seisan - Centro de Geociencias ::.. UNAM

seisan - Centro de Geociencias ::.. UNAM
SEISAN: THE EARTHQUAKE ANALYSIS SOFTWARE
FOR WINDOWS, SOLARIS, LINUX and MACOSX
Version 8.1
Jens Havskov(1) and Lars Ottemöller (2)
Editors
(1)Department of Earth Science
University of Bergen
Allégaten 41, 5007 Bergen, Norway
Phone +47 5558 3414
Fax +47 5558 3660
E-mail: [email protected], [email protected]
(2) British Geological Survey
Murchison House, West Mains Road
Edinburgh, UK
E-mail: mailto: [email protected]
mailto:
August 2005
ii
iii
List of contents
1. INTRODUCTION................................................................................................1
1.1 Latest Changes.................................................................................................................................................. 1
1.2 Information about SEISAN online ................................................................................................................... 3
2. STRUCTURE OF SEISAN.................................................................................4
2.1 Directories .......................................................................................................................................................... 4
2.2 The database...................................................................................................................................................... 4
2.2.1 Phase data and hypocenters, the REA directory ...................................................................................... 5
2.2.2 Waveform data, the WAV directory............................................................................................................ 6
2.2.2 Continuous waveform data......................................................................................................................... 8
2.3 File types used with SEISAN ........................................................................................................................... 8
2.4 Upper and lower case ....................................................................................................................................... 9
2.5 Moving data between Sun and PC.................................................................................................................. 9
3. INSTALLATION ...............................................................................................10
3.1 Unix (SOLARIS and Linux)............................................................................................................................. 10
3.2 MacOSX............................................................................................................................................................. 13
3.3 Windows (95, 98 and NT)................................................................................................................................ 14
3.4 Database directories for your own system, MAKEREA ............................................................................ 16
3.5 Default parameter files for the main programs........................................................................................... 16
3.6 Color settings................................................................................................................................................... 17
3.7 Compiling SEISAN programs ........................................................................................................................ 18
3.8 Program Version.............................................................................................................................................. 18
3.9 Setting general SEISAN defaults, SEISAN.DEF.......................................................................................... 18
3.10 Format description in parmater files .......................................................................................................... 21
3.11 Using SEISCONF to modify parameter files.............................................................................................. 21
4. USING SEISAN................................................................................................26
4.1 Short user guide .............................................................................................................................................. 26
4.2 Getting data into the database ...................................................................................................................... 29
4.2.1 System with digital data ............................................................................................................................ 30
4.2.2 System without digital data ....................................................................................................................... 30
4.2.3 Database security ..................................................................................................................................... 30
4.2.4 Data base tools, content and checks ....................................................................................................... 31
4.2.5 High accuracy in SEISAN......................................................................................................................... 31
4.3 Interactive work with earthquake locations, EEV command.................................................................... 32
4.4 How EEV works ............................................................................................................................................... 33
4.5.1 EEV driver program: JSEISAN ................................................................................................................... 42
4.5.2 EEV Windows driver program: SEISAN.................................................................................................... 51
4.6 System response............................................................................................................................................. 53
4.7 Working with catalogs .................................................................................................................................... 55
4.7.1 Explosions in SEISAN............................................................................................................................... 55
4.8 Printing.............................................................................................................................................................. 55
4.9. General Work with SEISAN........................................................................................................................... 56
4.10. Graphics in SEISAN ..................................................................................................................................... 56
iv
5. INTRODUCTORY TRAINING COURSE .........................................................57
6. DESCRIPTION OF PROGRAMS AND COMMANDS ....................................58
6.1 Hypocenter location programs: HYPOCENTER, HYPO71 and HYPOINVERSE .................................... 58
6.1.1 The hypocenter program, HYP................................................................................................................. 58
6.1.2 HYPO71 .................................................................................................................................................... 68
6.1.3 The Hypoinverse program, HYPINV (SUN and PC) ............................................................................... 71
6.1.3 HYP_ISC ................................................................................................................................................... 72
6.2 Trace plotting, phase picking and spectral analysis, MULPLT ................................................................ 72
6.2.1 MULPLT main functions ........................................................................................................................... 73
6.2.2 Use of MULPLT from EEV........................................................................................................................ 74
6.2.3 Continuous plotting ................................................................................................................................... 75
6.2.4 Commands in MULPLT, overview ........................................................................................................... 75
6.2.5 Registering new events into SEISAN....................................................................................................... 85
6.2.6 Phase picking, amplitude, weight and polarity......................................................................................... 86
6.2.7 Theoretical arrival times for global and local phases and location ........................................................ 87
6.2.8 Instrument correction and magnitudes Ml, mb and Ms ........................................................................... 88
6.2.9 Determining azimuth of arrival (3 comp or array) and doing component rotation.................................. 89
6.2.10 Data manipulation commands............................................................................................................... 90
6.2.11 Spectral analysis, ‘s’(Spec)..................................................................................................................... 94
6.2.12 The MULPLT.DEF and SEISAN.DEF file ............................................................................................ 103
6.2.13 Distance trace plot with GMT, TRACE_PLOT.....................................................................................106
6.3 Plotting epicenters, EPIMAP........................................................................................................................ 108
6.3.1 W_EMAP, Windows based map program............................................................................................... 115
6.4 Searching in the database, SELECT........................................................................................................... 116
6.5 Extracting events from the database, COLLECT...................................................................................... 121
6.6 Inserting events into the database, SPLIT................................................................................................. 121
6.7 Updating final locations in database, UPDATE and UPD....................................................................... 122
6.8 Using ‘filenr.lis’, DIRF and DELF................................................................................................................. 124
6.9 Making a bulletin, BUL.................................................................................................................................. 124
6.10 Reports and statistics................................................................................................................................. 125
6.11 Waveform file management tools ............................................................................................................. 129
6.12 File conversion and modification programs ........................................................................................... 136
6.13 Signal processing programs ..................................................................................................................... 149
6.13.1 PITSA .................................................................................................................................................... 149
6.13.2 SAC2000 ............................................................................................................................................... 150
6.13.3 DEGTRA A4.......................................................................................................................................... 150
6.14 Calculating b-value, BVALUE .................................................................................................................... 152
6.15 Automatic phase picking, AUTO,AUTOPIC,AUTOSIG...........................................................................155
6.16 Fault plane solution, FOCMEC .................................................................................................................. 160
6.17 Calculation of coda q, CODAQ.................................................................................................................. 165
6.18 Merge events near in time ASSOCI .......................................................................................................... 171
6.19 Making synthetic seismograms ................................................................................................................ 171
6.20 Calculation of travel times ......................................................................................................................... 179
6.20.1 IASPEI travel time software, program TTIM ........................................................................................ 179
6.20.2 Calculation of travel times for layer model and gradient model, TTLAYER ....................................... 179
6.20.3 IASP, travel times for MULPLT ............................................................................................................ 179
6.21 Inversion for QLg, QLG ................................................................................................................................ 180
6.22 Wadati ........................................................................................................................................................... 181
6.23 Calculating spectra, the SPEC program .................................................................................................. 183
6.24 Seismic risk related programs .................................................................................................................. 189
6.25 Magnitude relations, MAG.......................................................................................................................... 201
6.26 Explosion filtering, EXFILTER ................................................................................................................... 205
6.27 Inversion of travel time data and joint hypocenter determination.......................................................208
6.27.1 VELEST................................................................................................................................................. 208
6.27.2NOR2DD ................................................................................................................................................ 211
v
6.27.3 NOR2 JHD_PUJOL .............................................................................................................................. 211
6.28 Analysis of volcanic earthquakes............................................................................................................. 211
6.29 FK Analysis .................................................................................................................................................. 214
6.30 Surface wave analysis (SUN)..................................................................................................................... 217
6.31 Instrument response................................................................................................................................... 217
6.31.1 Create instrument response files, RESP ............................................................................................. 217
6.31.2 SEED response .................................................................................................................................... 224
6.31.3 SEED response to GSE, SEEDRESP2GSE....................................................................................... 225
6.32 Macroseismic data in SEISAN................................................................................................................... 225
6.33 Correlation of waveform signals, CORR.................................................................................................. 226
7. Programming in SEISAN and list of files in SEISAN
distribution ..............................................................................................229
7.1 Programmers guide and some test programs.......................................................................................... 229
7.2 CONTENTS OF PRO, LIB, INC, INF, COM, DAT and PIC DIRECTORIES.............................................. 230
8. Acknowledgments........................................................................................231
9. References ....................................................................................................232
Appendix 1 - The Nordic format......................................................................234
Appendix 2 - The Seisan waveform file format .............................................239
Appendix 3 - Response file formats used in SEISAN ..................................243
INDEX.................................................................................................................246
1
1. INTRODUCTION
The SEISAN seismic analysis system is a complete set of programs and a simple database for analyzing
earthquakes from analog and digital data. With SEISAN it is possible using local and global earthquakes
to enter phase readings manually or pick them with a cursor, locate events, edit events, determine
spectral parameters, seismic moment, azimuth of arrival from 3-component stations and plot epicenters.
The system consists of a set of programs tied to the same database. Using the search programs it is
possible to use different criteria to search the database for particular events and work with this subset
without extracting the events. Most of the programs can operate both in a conventional way (using a
single file with many events), or in a database manner. Additionally, SEISAN contains some integrated
research type programs like coda Q, synthetic modeling and a complete system for seismic hazard
calculation.
The data is organized in a database like structure using the file system. The smallest basic unit is a file
containing original phase readings (arrival times, amplitude, period, azimuth, and apparent velocity) for
one event. The name of that file is also the event ID, which is the key to all information about the event in
the database. Although the database in reality only consists of a large number of sub-directories and files
(all of which the user has access to), the intention is that by using the surrounding software, the user
should rarely need to access the files directly, but rather do all work from the user's own directory. Test
data and a tutorial (see section 5) are supplied with the system.
The programs are mostly written in Fortran, a few in C and almost all source codes is given, so the user
should be able to fix bugs and make modifications. The programs have been compiled and linked with
system compilers and linkers on SUN, GNU compiler on Linux and MaxOSX, and Digital Visual Fortran
6.0 under Windows. SEISAN runs under Sun Solaris, Linux, MacOSX, Windows95/98, and Windows
NT/2000. No format conversion is needed to move data files (binary and ASCII) between the systems if
one of the standard formats (SEISAN, GSE, SEED, SAC ASCII) is used.
This manual resides in the directory INF (see below), when the system has been implemented on your
computer. The file is called seisan_8.0.pdf (Adobe PDF).
The SEISAN system is built of programs made by many different individuals without who it would never
have been possible to make SEISAN. Acknowledgement is made throughout this manual where
appropriate or in the acknowledgement section at the end. SEISAN now contains so many programs that
when a new version is released, it is not possible to check all the options in all programs and we rely on
the user to help finding the bugs, please report!
SEISAN is freely available for all non-commercial use.
1.1 Latest changes
Version 8.1
SEED – read seed and miniseed data, write miniseed, read seed response files
Macromap- make a map with macroseismic data
Correlation program – tool to use cross-correlation to identify groups of similar events and to
determine relative phase arrivals
Seicut – tool to cut waveform files
Hyp_isc – ISC location program
Seedresp2gse – convert seed response file to GSE format
Rdseed_many- cut a large SEED volume into file of equal size
Nor2dd and nor2jhd_pujol – prepare input files for travel time inversion programs
Postscript output file name suffix is changed from plt to eps
2
Version 8.0
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Improved graphics: scaling, redraw after window was covered, cursor and font can be
selected, colors on Linux (Unix only).
More output for use with GMT (script gmtxy)
Older conversion programs removed, new have been made
A new Java interface JSEISAN doing much of the functions in EEV,SELECT and MULPLT
Change of NORDIC S-file format, signal to noise ratio has been replaced by angle of
incidence
Program COMPACT change to NORHEAD to avoid conflics on Windows
Program EXTRACT changed name to WAVETOOL to avoid conflict on Windows, the
program has also got new features
Original FOCMEC by Snoke is now used, FOCMEC is interface to FOCMEC_EXE, amplitude
ratios can be used
Improved handling of continuous data with MULPLT
Grid search added to spectral fitting in AUTOSIG
Java program SEISCONF to edit main parameter files
New program DEGTRAA4, application for earthquake engineering
SELECT now works on S-file database
Seisan compiled on Mac OSX (Dieter Stoll)
3
1.2 Information about SEISAN online
SEISAN homepage
The URL address where SEISAN and related software can be found is:
http://www.geo.uib.no/Seismologi/SOFTWARE/ (or 129.177.55.13 instead of www.geo.uib.no)
Here you can find information on the latest changes in SEISAN, access the online manual, download
the software and much more.
SEISAN anonymous ftp server
Seisan is available from the following ftp server:
ftp.geo.uib.no ( or 129.177.55.4)
Login: ftp
Password: <your email address>
The files are stored in the directory /pub/seismo/SOFTWARE/SEISAN_8.0.
SEISAN mailing lists
There are 2 mailing lists, which are set-up to improve the exchange of information and questions on
SEISAN. We strongly recommend that all users subscribe to the SEISAN list. The lists are:
[email protected]
[email protected]
The purpose of the lists is:
seisan
seisan-help
SEISAN experts who are willing to help
others
Who is on the
list?
SEISAN users
Purpose of the
list
Information on upgrades, bugs and
As SEISAN user, you can send your
other related topics will be sent to this
questions or problems to this list
list
To subscribe to one of the lists, send an email to: [email protected]
In the body of the email, not as subject, put:
subscribe list
where list is seisan or seisan-help
You can obtain help from the majordomo list-server by sending the message:
help
4
2. STRUCTURE OF SEISAN
2.1 Directories
The whole SEISAN system is located in subdirectories residing under the main directory SEISMO. For
more details, see section 3 on installation. The system contains the following main subdirectories:
REA:
WOR:
TMP
PRO:
LIB:
INC:
COM:
DAT:
WAV:
CAL:
INF:
SUP:
Earthquake readings and full epicenter solutions in a database
The users work directory, initially empty
Temporal storage of files, initially empty
Programs, source code and executables
Libraries and subroutines
Include files for programs and subroutines in PRO and LIB
Command procedures
Default and parameter files, e.g. station coordinates
Digital waveform data files
System calibration files
Documentation and information
Supplementary files and programs
In the following, the above subdirectories will mostly be called directories to avoid always referring to
SEISMO. All directories use capital letters, however this only makes a difference in the Unix
versions. The directory structure is used as a tree like structure for quick access to individual files in
the REA directory, which therefore will appear as a simple database to the user. The next section is a
description of the database directories; the other directories are described in section 7. Figure 1
shows the tree structure of SEISAN.
Figure 1. Structure of SEISAN. Note that BERGE under WAV is optional and DELET (not
shown) under REA has a similar directory structure as e.g. NAO.
SEISMO
REA
DAT
BERGE
CAL
NAO
KONO
WAV
BERGE
NAO
1996
CAT
1995
1996
01 02
01 02
12
01 02
12
12
2.2 The database
The database of SEISAN consists of the two directories REA and WAV. The REA directory and its
subdirectories contain readings and source information while all waveform data is normally in the directory
WAV (see 2.2.2) with no subdirectories. Optionally WAV can also be divided into a similar subdirectory
structure, see 2.2.2, which is useful when storing continuous data in particular. The DELET database
contains all events deleted from any of the databases (here BERGE/BER and NAO). Filenames are
5
identical between all platforms.
2.2.1 Phase data and hypocenters
The REA directory contains phase readings and derived source information like hypocenters, fault plane
solutions etc. The REA directory has one or several subdirectories corresponding to separate databases
(see Figure 1 for an example with two databases). The database names can have between 3 and 5
characters. If less than 5 characters are used, the character ‘_’ is added in the file system to make it 5.
The user does not have to put the ‘_’ when running a program, they will be added by the software. If a
directory is made manually, the ‘_‘ must be put in. It is assumed that a database is always present in the
system. The name of the default database is given by an environmental variable (see section 3.1),
however if not set, it will default to AGA for agency. Here, BER will be used as an example throughout the
manual. A database has a duplicate storage of the events. For quick reference and interactive work the
events are stored in single files (S-files) in yearly directories and monthly subdirectories. When new data
is entered into the database, it comes in as individual event files. However, once the interactive work has
finished, the single event files are overwritten with the final location and additionally stored in monthly files,
which are only changed when updating (UPDATE command, see section 6.7). The monthly files, called
CAT-files for catalog, are stored separately in the CAT directory and primarily used for quick searching
and backup for the single files. In addition to the event data, there is also a LOG directory in each
database to keep a log of the data processing, see 6.7.
S-file database structure
The structure for the single file storage is as follows (Windows example):
\REA\BER__\
\REA\BER__\1999\
\REA\BER__\1999\01\
Main readings directory, all data
Data for 1999
Data for January 1999, each event in one file
On Unix, the last line would have been /REA/BER__/1999/01.
Each event contains original phase readings in the Nordic format (Appendix 1) which includes file names
of all corresponding waveform files. One event is one file. Each event has an ID line. The ID line contains
a unique ID, which will follow the event through all COLLECT and SPLIT operations (see section 6.5 and
6.6). The ID line also contains status information about the event like last action, when it was updated etc.
The ID-number can be fixed, which is useful if data is taken out from the database, processed on another
computer and later put back into the database, since otherwise the ID of an event might be changed and
the existing file would not be overwritten. An example of an S-file name is
27-1112-11L.S199401.
The S-files are used as input for the location program and, when making a permanent update, also for
output, see 6.1. The letter in front of the "." indicates the event type and can be L, R or D for local, regional
or distant event respectively. It is the same indicator as given in the header line of the S-file, see the
Nordic format. The remaining numbers give (in order) day, hr, min, sec, year and month.
As mentioned above, the system can contain many other databases, which may function exactly like the
BER directory. A data base can be used to store a subset of data or data from different networks. Data
can be moved between databases or in and out of the databases, for details, see description on EEV (4.3
and 4.4).
Monthly location files, the CAT directory
Events located in monthly files are in a directory called /SEISMO/REA/BER__/CAT in addition to the
6
individual S-files. Additional databases like e.g. NAO will have epicenters stored under
/SEISMO/REA/NAO__/CAT.
The monthly epicenter files are called 199901.CAT for e.g. January 1999. Although the files generated by
SEISAN normally are monthly files, the CAT directory can also contain yearly files or any other time
interval. The only rule is that the name of the file must give the year and month of the first event in the file.
This is because the search program SELECT uses the file names to search requested time intervals. If a
user has a historical catalog, this can be added as an individual file. If the historical catalog starts in 1820,
the file name would be 182001.CAT. The files in CAT do not need to be continuous in time, but they must
not have overlaps in time and each file must have data in chronological order. The format of the CAT files
is the same as for the S-files. Additionally, CAT files can also be compact files, meaning just the header
lines of the S-files (see also section 2.3).
2.2.2 Waveform data and formats
SEISAN works with various waveform formats including SEISAN, GSE, SEED/MINISEED and SAC
binary and SAC ASCII. The SEISAN format is described in Appendix 2, while for a format description
of GSE and SAC the user is referred to GSETT-3 (1997) and Goldstein (1999), respectively. The
SEED format is described in IRIS Consortium (1993). The GSE reading routines are based on the
codeco routines written by Urs Kradolfer, Klaus Stammler and Karl Koch. The different formats can be
used in parallel by several programs. With MULPLT for example it is possible to plot data in the four
formats at the same time. Other formats can be added by adding reading routines and adding the
respective calls to LIB/wave.for. Note that SAC binary files can only be used on Solaris and Linux,
while SAC ASCII can be used on all platforms. To use other formats, a conversion program must be
used first, see section 6.12.
In general it is recommended to keep the waveform data in one format only, mainly for simplicity and
maintenance reasons. There may be different arguments for or against one or the other format
depending on the user’s preferences and requirements. SAC and GSE are widely used formats and
therefore may be attractive. SEISAN is a multi-trace binary format with direct read access to individual
traces. The SEISAN format is probably your best choice if your main processing system is SEISAN
and because it is easily used on all computer platforms. SAC is a single trace binary or ASCII format
with a large number of header parameters. The SAC format is widely used in research-oriented
programs. GSE is a multi-trace ASCII waveform format that includes various sub-formats. It is widely
used for data exchange. Although the GSE format can keep any number of traces, it is recommended
to include no more than 3 traces in a single file depending on the number of samples, since when
reading a particular trace, the whole file may have to be read.
For the future, the SEED/MINISEED format might be the best option since most data centers use it.
However, the SEISAN implementation should probably be tested a bit more. SEISAN cannot read
SEED files using all options possible in SEED, but data from the largest data centers as well as many
observatories have been used for testing. With respect to MINISEED, there are probably less
problems since MINISEED is simpler than SEED. SEISAN can also write MININSEED (program
WAVETOOL), but cannot write SEED (unless GSE2SEED is used). The SEISAN implementation is
described
in
http://www.geo.uib.no/Seismologi/SOFTWARE/SEISAN_8.1/SEED/Technical_18_seed.doc
The WAV directory contains files with digital waveform data. The directory normally has no subdirectories
or any other organization. However, in case of large databases, WAV can be subdivided, see below. In
addition any directory can contain waveform data, it has to be specified in SEISAN.DEF (section 3.8). The
amount of data that can be stored is only limited by the disk size. The analysis system will always look in
WAV for particular files if they are not in the user’s own directory. Waveform files will automatically be
transferred to WAV on initial registration into the database (see MULPLT). Registration is the process of
automatically creating an S-file in the database with the name of the waveform file and header
information. Phase pickings are done later. See section 6.2.
There is normally no requirement for particular filenames for the waveform files in WAV or elsewhere,
however many programs will make file names like:
7
yyyy-mm-dd-hhmm-ssT.NETWO_nnn e.g. 1995-01-23-1230-20T.BERGE_013
With the abbreviations yyyy: year, mm: month, dd: day, hh: hour, mm: minute, ss: second, T: file type
indicator (normally S), NETWO: maximum 5 letter network code and nnn: number of channels.
File type indicators are: S: Standard, R: Resampled, A: Appended
WAV database: In case a large number of waveform data is stored, it might be an advantage to also split
up the WAV directory in subdirectories. This is done in the same way as in the REA directory, e.g.
waveform files for BER from July 1994 would be found in WAV/BER__/1994/07. Programs that use
waveform files will automatically search, in order, the current directory, TMP, WAV and the monthly WAV
directory. When storing in the WAV database, it is a requirement that the waveform names start
with either yymm (like 9902) or yyyy-mm (like 1999-02). Waveform files created on Windows and
Linux version 7 or newer cannot be read on older SEISAN versions.
The SEISAN binary waveform format is explained in Appendix 2. The files are written and read with the
same Fortran statements on all platforms, however the internal structure and byte order are different. As
of SEISAN version 5.1, files written on either machine can be read on the other and there is no need for
any conversion when the binary waveform files are moved between Sun, Linux, MaxOSX and Windows.
Compression of waveform data
Waveform files can be stored in compressed format. The compression must be done by the user.
Programs that access the compressed waveform files copy the file to the TMP directory, and uncompress
there. The uncompressed file remains afterwards and will be found the next time one of the programs is
looking for the same waveform file. The content of the TMP directory has to be deleted manually. On Unix,
you may automatically delete the content of the TMP directory by a cronjob, see manual pages on
crontab. On Unix the compression formats supported include gzip, compress, bzip2 and zip. So far, no
automatic decompression is supported on Windows (will be put in). With the introduction of SEED format,
there is less need for external compression since the SEED data usually is compressed and therefore
decompressed on the fly when read.
Component codes
The SEISAN waveform format has 4 characters for the component code. The first character indicates the
type of sensor, for example ‘B’ for broadband, ‘S’ for short-period or ‘L’ for long-period. For acceleration
data the first character has to be ‘A’ because SEISAN assumes that the corresponding response has
been given as acceleration response. The fourth character has to give the channel orientation, ‘Z’ is used
for vertical, ‘E’ for east-west and ‘N’ for north-south. Other orientation of the horizontal components is
possible in GSE, SEED and SEISAN are not understood by SEISAN. If data are rotated, ‘T’ is used for
transverse and ‘R’ for radial. The second and third characters can be chosen by the user. The component
code is part of the response filename and is used to find the response corresponding to a given station
and component. Program WAVFIX can be used to change station and/or component codes.
The Nordic format only has space for two characters for the component code. The definition in SEISAN is
that these are the first and fourth character of the waveform component code. This means that the relation
between the component code in the Nordic file and the waveform data is non-unique.
The GSE and SEED waveform formats have three characters for the channel code, see GSETT (1997)
and IRIS Consortium (1993) for the detailed definition of the component codes. SEISAN, when reading
waveform data in either GSE or SEED format internally keeps the first two characters and moves the third
to fourth, so for example ‘BHZ’ becomes ‘BH Z’. GSE response filenames contain four characters, while it
has three characters inside, the same conversion as for waveform data is applied. SEED response
filenames have only three characters, so matching to find the correct response file is done between the
first two characters and internal character four with character three in the filename. When converting from
SEISAN to GSE or SEED, the third character of the channel code is omitted to shrink to three characters.
SAC has more than four characters for the component code and sacsei.def has to be used to define the
8
conversion. However, normally SAC data will have three character component codes as well. Conversion
of component codes from SEISAN to SAC is also defined in sacsei.def.
Data streams in SEED also have a location code, which allows to distinguish for example between two
‘BHZ’ components (for example a 30 second 120 second sensor with the same sampling rate and high
gain) at the same site. SEISAN does not use the location code. This can be a problem when using
instrument response that is different for the two location codes. In this case the user has to ensure that
only the instrument response file for the corresponding data to use is stored in the CAL directory.
When converting between SEISAN and other waveform formats, component conversion is defined in the
respective definition files, see section on conversion programs.
2.2.2 Continuous waveform data
In SEISAN continuous data has no special format. Continuous data is simply ordinary waveform files that
follow each other in time. In order to treat the data as continuous, the data can be put into a SEISAN
continuous data base. Such a data base is made as follows:
-
For each waveform file from a station or network, an S-file is created. The S-files only contain
reference to the waveform file. Program AUTOREG can be used.
The waveform data is put into the corresponding waveform station directories.
The continuous databases are defined in SEISAN.DEF in DAT.
If e.g. data is to be stored from 3 different stations, create 3 databases under WAV and REA with the
name of the stations (program MAKEREA). If the continuous data consist of 20-minute files, this
would mean about 2200 files pr month, which is a reasonable number. It is now possible for some
programs (MULPLT, WAVETOOL) to get access to any or all of the traces in the continuous data
base and plot and extract data.
A simpler way to use smaller quantities of continuous data is to make a list of these files with DIRF and
an application program can then use that list to work with the data. Currently two programs have special
options for this kind of continuous data. The MULPLT program will plot data from several files as if it was
one file in one continues trace the RESAMP program will resample the data from several files and put it
into one output file.
2.3 File types used with SEISAN
A description of the different file types is given below with typical names. Most names must be exactly as
specified, others can be given names. However it is VERY important that no name including full path is
more than 80 characters long. Until now this has not been a problem, however it has to be considered
when SEISAN is installed.
The basic unit is a file in the Nordic format, (see Appendix 1). For practical purposes 3 descriptive names
are used for Nordic files:
S-file: Single event file with phase readings, with or without source parameters such as location and
magnitude. In the database these files are named with the extension: .Syyymm This is the standard type
of file in e.g. the BER__/1998/08/. An example is 11-1234-11L.S199808.
CAT-file: A catalog file containing many S-files with location or just a catalog of hypocenters, a compact
file, see below. This is the standard type of file in e.g. the /REA/BER/CAT directory. An example is
199801.CAT. This file format is also output from several programs like SELECT and COLLECT. There is
a blank line between events.
Compact file: This is a CAT-file with only the source information. One event is represented with one line,
(the header line in the S-file). There is no blank line between events. A compact file can be generated by
either COLLECT or NORHEAD ( earlier called COMPACT).
9
In addition there are the following types of files:
SEISAN waveform file: Waveform data can be stored in SEISAN, GSE and SAC format, see section
2.2.2. An example of a name is 1992-01-11-2233-22S.BERGE_011.
Response file: File giving the response of a given channel at a given station. They are generated with the
RESP program, see description of CAL directory, section 6.31. This is the standard type of file in the CAL
directory. An example of a name is ODDA_S__Z. 1999-05-01-0000_SEI.
File listing: This is just a file with a list of numbered files. The file name is ALWAYS filenr.lis, and it is
generated with the DIRF program, see 6.8.
Index file: This file contains a listing of absolute paths to a series of S-files. The index file can be used as
input instead of the CAT-files to several programs. Several programs generate index files as e.g. select
and eev. The index file has the same format as the filenr.lis files described above and can be generated
with the dirf command using S-files. The index file name must contain a ‘.’. An example is shown below:
1 \SEISMO\REA\TEST_\1993\09\29-2228-26D.S199309
2 \SEISMO\REA\TEST_\1994\06\16-1841-57D.S199406
3 \SEISMO\REA\TEST_\1996\06\03-1955-40D.S199606
2.4 Upper and lower case
Upper and lower case file names only makes a difference on SUN, Linux and MacOSX. The intention is
that all permanent data file names used by SEISAN should be in upper case (e.g. S-files, crustal model
file, directories (e.g. REA) while temporary files should be in lower case (e.g. print.out). Programs are also
in lower case. It should then be a bit more difficult to delete the permanent files. NOTE THAT
THROUGHOUT THIS MANUAL, PROGRAM NAMES ARE GIVEN IN UPPER CASE TO INDICATE
THAT THEY ARE NAMES, HOWEVER WHEN USING THE PROGRAMS, LOWER CASE MUST BE
USED ON SUN. In program MULPLT, commands are case dependent.
2.5 Moving data between Sun, Linux, MacOSX and Windows
All S-files and file names are identical on the three platforms. To move many events (S-files) from one
system to another, make a COLLECT (section 6.5) on the original system and a SPLIT (section 6.6) on
the receiving system. As mentioned in section 2.3, the SEISAN binary waveform files have different
internal structure if written on Sun, Linux, MacOSX or Windows, but this is corrected for in the reading
routine, so files can be copied directly. GSE files can be copied directly since they are ASCII files. SAC
binary files are different between Linux and Solaris; SEISAN can only read files that were written on the
same platform. SEED/MINISEED files can be used directly on all platforms.
The only other files that are different are the binary earth model files IASP91_platform.HED and
IASP91_platform.TBL in the DAT directory (where platform is either sun, linux, macosx or windows). The
platform is included in the filename so that possibly SEISAN can be used on different platforms with only
one file system. For example the data may be kept on a Sun file system, but you also want to share the
disks from a Windows system and process the data using the Windows version of Seisan. Otherwise, the
files cannot be moved, but are easily regenerated with the IASP91 programs, see section 6.20 and 8.8 in
the Hypocenter manual.
10
3. INSTALLATION
SEISAN has been tested and compiled for Windows 98/NT/2000/XP, Solaris 2, Redhat Linux and
MacOSX.
Upgrade from version 7.0 or higher
Before you start, take a backup copy of your DAT directory. Note that when you upgrade, many
parameter files will be overwritten so make sure old parameter files are copied before putting in a new
version of SEISAN. The most important are in DAT: STATION0.HYP, SEISAN.DEF, MULPLT.DEF. Also
the Unix setup file .SEISAN is overwritten. You may also want to keep copies of PRO, LIB and INC to
keep a copy of the old source code, especially if you have done any modifications to the code. You can
keep almost all of your parameter files, only SEISAN.DEF has been changed. Check this file and change
to your system. Some individual program parameter files like for SPEC have changed.
How to get SEISAN
SEISAN can be copied from ftp.geo.uib.no (129.177.55.4), login is ftp and password is your email
address or http://www.geo.uib.no/seismo/software/seisan.html (or 129.177.55.5 instead of
www.geo.uib.no). On the AFTP server go to /pub/seismo/SOFTWARE/SEISAN_8.0. Use binary mode
for the compressed files (tar and zip). Before copying, check the readme file for latest updates,
changes and current content of the directory. The directory will at least contain the following files:
seisan_X.Y_UNIX.tar.Z: a compressed tar file, whole distribution with executables and test data, X.Y
stands for the latest distribution number and UNIX for the respective
Unix system
testdata_X.Y.tar.Z
SEISAN testdata
seisan_short_X.Y_UNIX.tar.Z: a compressed tar file, distribution without executables and test data,
must be compiled
seisan._X.Y.zip:
Windows distribution as zip file (WinZip)
seisan_X.Y.pdf: the SEISAN manual, Adobe PDF
seitrain.pdf
SEISAN training course
Alternatively SEISAN might be obtained on a CD with the same content as above (write to
[email protected]).
Section 3.6 gives additional information about modifications and recompilation.
3.1 Unix (SOLARIS and Linux)
Solaris: The SEISAN programs have been compiled on Solaris 7 using Sun Workshop 5, which means
you have to recompile if you use an earlier version of the operating system or compiler. If you can
recompile on Solaris, please do so! The programs on Solaris are compiled dynamically, which means not
all system and compiler libraries are included in the executables. If you are running Solaris, the system
libraries are normally installed, but the Sun system compilers might not be installed. If the compilers are
not installed, you have the following options: (1) you install the Sun workshop compilers, license is not
needed, since only the libraries are required; (2) you install the required libraries, which are part of the
Solaris SEISAN distribution (instructions below).
Linux: The programs have been compiled under Redhat Linux7.2 using the GNU compilers gcc and g77.
It is recommended to recompile the programs, since otherwise the programs might not run on your Linux
distribution. In the Redhat distribution of Linux the Fortran compiler is not part of the standard distribution,
it has to be installed (see your Linux manual for instructions). THE USER ACCOUNT MUST BE SET UP
TO USE csh or tcsh in order for the SEISAN scripts (e.g. .SEISAN) to work, like. Otherwise the scripts
11
need to be adopted to the shell used.
Instructions
The first step is to install the distribution, the procedure is the same for all Unix platforms.
1. Get tar file
Copy the distribution file for your platform from CD or transfer it through FTP or from the web site to the
SEISAN top directory, this could be a directory ‘seismo’ under the home directory.
2. Decompress
uncompress seisan_X.Y_UNIX.tar.Z
or (if suffix .gz)
gunzip seisan_X.Y_UNIX.tar.gz
There should now be a file called seisan_X.Y_UNIX.tar in your directory.
3. Install SEISAN
tar xvf seisan.tar
Check that the SEISAN directories have been created.
If SEISAN has been installed without executable files, they can all be generated with the command ‘make
all’ from the PRO directory. On Sun this requires that the Sun compilers be installed, on Linux it requires
the GNU Fortran compilers. See also section on compilation (3.6).
Install Workshop libraries
In the SUP directory of the Solaris distribution the file sun_ws_lib.tar.Z includes the libraries that are
needed to run SEISAN on Solaris in case the compilers are not installed. The file is a compressed tar file.
The files can be extracted with ‘uncompress sun_ws_lib.tar.Z’ and then ‘tar xvf sun_ws_lib.tar’. The library
files can be stored in any directory in the system, but the environmental variable ‘LD_LIBRARY_PATH’
has to be set accordingly. If you are using the C-shell, this can be done by adding to the .cshrc file the line
‘setenv LD_LIBRARY_PATH /path/:$LD_LIBRARY_PATH’. This would add /path/ (which is the path to
where the libraries are) to LD_LIBRARY_PATH, which normally is already defined.
4. Set system parameters
If you are doing an update, some of the following settings can be skipped.
Activate SEISAN:
In your .cshrc file, the aliases and paths used by SEISAN are defined by adding the line ‘source
..seismo/COM/.SEISAN’ where ‘..seismo’ is the directory below which SEISAN has been installed. The
.SEISAN script file assumes that you are running either csh or tcsh as your shell. If you are using another
shell (e.g. bash on Linux), you need to modify the script accordingly or change the shell. It is assumed that
X-windows is installed.
SEISAN path for programs:
In order for programs and subroutines to know the path to the SEISAN program directory, this must be
defined in the file .SEISAN in COM. Edit that file and set the environmental variable SEISAN_RES to the
name of the top directory, meaning the directory structure below and including seismo e.g.
/top/users/seismo. This variable is then used to set the path to SEISAN directories.
12
Search path for libraries:
To run the NANSEI conversion program under Solaris, the SEISAN LIB directory needs to be included in
the environmental variable LD_LIBRARY_PATH. The LIB directory as default is already added to the
library search path in the .SEISAN file.
SEISAN path for databases, parameter files etc:
The SEISAN database can be under the same top directory as programs, however it can also be different.
This is practical if several users have their own databases, but use the same software. Set environmental
variable SEISAN_TOP to top directory e.g. /top/users/seismo.
Java settings:
Aliases are defined to run the Java tools jseisan, seisconf and sformat. The setup in the .SEISAN file as
default is:
alias jseisan 'java -DSEISAN_TOP=$SEISAN_RES -classpath $SEISAN_TOP/PRO/jseisan.jar SEISAN'
alias seisconf 'java -DSEISAN_TOP=$SEISAN_RES -classpath $SEISAN_TOP/PRO/jseisan.jar SEISCONF'
alias sformat 'java -DSEISAN_TOP=$SEISAN_RES -classpath $SEISAN_TOP/PRO/sformat.jar Sformat'
These settings may have to be changed in case you want to keep the jar files in some other directory. The
CLASSPATH (which is used to search for classes) is also modified to include PRO and PRO/jseisan.jar,
although this is not needed when using the aliases.
SEISAN agency:
In .SEISAN also set the environmental variable AGENCY (upper case) to your 3-letter agency code
(upper case). This variable is only used by program MACROIN from EEV in connection with entering
macroseismic data so for most users ignore this setting.
SEISAN default database:
To locate the default database directory (here BER) set environmental variable DEF_BASE in .SEISAN. If
not set, the name AGA is used. The data bases are found under SEISAN_TOP.
SEISAN editor used in EEV:
The default editor is vi, any other editor can be set with the environmental variable SEISAN_EDITOR.
SEISAN calibration file directory:
By default, calibration files are in CAL, but they can be in a directory set with variable LOCAL_CAL. The
directory name must be complete like /home/users/calibration/
SEISARCH
Gives the architecture, can be either sun, linux or windows. Used in Makefile when compiling.
SACAUX
Path to SAC aux directory, required by the SAC routines for reading and writing, although not really used.
SACLIB
Specify path and filename to SAC libraries, only needed when you compile programs (Unix) and you have
the libraries installed on your system.
Printer for Postscript plots:
The hard copy files from programs are sent to the printer from within the programs using the standard lpr
command. In the .SEISAN file, define lpr using the standard environmental variable PRINTER.
Remember that the printer must accept Postscript. PostScript files can also be viewed and printed on
most printers outside SEISAN using GhostView, however in that cases files cannot be printed from within
a program.
Scaling for Postscript plots:
By default, plots will be in A4 size. This can be changed by setting the environmental variables
SEISAN_PSSCALE_X and SEISAN_PSSCALE_Y. The default for A4 size is 1.0 for both variables. For
Letter size the Y-scaling can be set to 0.9.
Seisan Extension:
User specific code
can
be
implemented
by
making
use
of
the
environmental
variable
13
SEISAN_EXTENSION. The idea is that programs read this variable, if set to the user specific string, the
user’s source code will be used instead of the default. An example could be the computation of error
ellipses. Currently used codes are: BGS.
4. Testdata
The testdata set can be extracted from the file testdata_X.Y.tar.Z. Use programs uncompress and tar to
extract the data in the SEISAN top directory (keep subdirectory structure).
Dimensions
Most dimensions are set in file seidim.inc in the INC directory. In order to change dimensions, first change
in the include file and then recompile the whole SEISAN distribution. The most important dimensions are:
Number of points in one trace
Number of lines in NORDIC format file
Maximum number of traces in one plot
Maximum number of events in one month
Maximum number of calibration files
1 200 000
4 000
1 000
90 000
1 500
The maximum number of points used by the SEED reading routine are set in seed_internal.in.f. Currently
its is also set to 1 200 000.
SEISAN has been tested with much larger dimensions, like 10 000 000 for number of points in one trace,
however large dimensions might slow down the speed due to swapping (particularly if memory is not
large) so a smaller dimension has been chosen. 1 200 000 samples corresponds to almost 23 hours of
100 Hz data. There is rarely a need for larger dimensions. For continuous data, SEISAN works with many
files so smaller dimensions can be used.
Note:
In case programs don’t work, you might have to recompile, see section 3.6.
Graphics problem: .
On Solaris, if no colors, make sure color setting is 8 bit. Can be set with command m64config –depth 8.
See Solaris manual.
3.2 MacOSX
The MacOSX version does not come pre-compiled, and will have to be compiled by the user. The
installation is basically the same as for Solaris/Linux (see previous section), but compilation needs to
be done, see section 3.7. The MacOSX is basically the same as the Sun distribution. However, there
are some differences in the Makefiles, since GNU compilers are used the compilation is similar to
Linux.
Additional hints on MacOSX 10
By Dieter Stoll, Lennartz electronic GmbH, email: [email protected]
At the time of this writing (September 2003), 10.2.6 was the most current version of OS X, with the
next major release (codename "Panther") already out as a beta. The following hints refer to 10.2.x
versions; chances are that by the time you read this, 10.x (x > 2) is already out, and some of the links,
hints, and methods discussed below may be obsolete.
There are two prerequisites for using Seisan on OS X. The first prerequisite concerns compiling
Seisan from the distributed sources; the second prerequisite concerns the actual use of Seisan. On
top of these technical prerequisites there is a "human factor" - you need to be at least superficially
familiar with command line operation (Terminal.app or xterm). In order to compile and link Seisan off
the source distribution, you need to have the Apple Developer Tools installed. These come on an
extra CD together with the OS. If you don't have access to that extra CD, DevTools can be
downloaded from Apple (it's a fairly major download, around 600 MB), but you need to go through a
14
registration process before. The steps are:
1. Go to http://developer.apple.com
2. Click the Join Now link in the Partner with Apple section and follow the instructions to join
Apple Developer Connection.
3. Once your Apple Developer Membership is confirmed, log in and click the Download
Software link. The Mac OS X 10.0 Developer Tools are available for download there.
In order to actually use Seisan you need to have the X Windows system installed (as opposed to OS
X's native "Aqua" GUI). X11 runs under Aqua, it is not a mutually exclusive alternative. X11, ready to
use
(no
compilation/linking
required)
for
OS
X
is
a
free
download
from
http://www.apple.com/macosx/x11.
3.3 Windows (95, 98, 2000, XP and NT)
It is assumed that you transferred the Windows distribution, copied from CD or decompressing directly
from CD. In the following, it is assumed that you install on disk drive C. The windows graphics work with
any resolution, small fonts should be used, however, SEISAN is designed to work with 1024x768. This
version of SEISAN has not been tested on 95, 98 or NT.
Using install script
Click on seisan81.exe and follow the instructions, use all defaults. If you later want to change some
environmental variables, see instructions below. If SEISAN already is installed, a window comes up for a
possible upgrade or removal of SEISAN.
Using the seisan_8.1.zip file
1.Use WinZip to unpack the compressed archive file seisan_X.Y.zip to the directory ‘seismo’ either
from the CD or a file copied to the hard disk. The old pkunzip can no longer be used, since it does not
support long file names.
2. Changes to the environmental settings: There are two options (1) Edit the autoexec.bat file in the
root directory or (2) set the variables directly (Windows 2000/XP). The Autoexec.bat works on all
systems.
Option 1:
Edit autoexec.bat in root directory:
-
Add C:\SEISMO\COM and C:\SEISMO\PRO to path
path c:\seismo\pro; c:\seismo\com, or set path=
Add line
set SEISAN_TOP=c:\seismo
This example assumes that SEISAN is installed under c:\seismo, not correct if SEISAN is
installed differently. Note, from SEISAN 8.0, SEISAN_TOP must be set in order for java programs
and commands like wo, da etc to work.
-
OPTIONAL: Add environmental variables
SEISAN_EDITOR, SEISAN_PSSCALE_X,
SEISAN_PSSCALE_Y and DEF_BASE (see UNIX section for definition). If not added to
autoexec.bat, the defaults \SEISMO, edit, A4 and AGA are used. SEISAN_TOP would be set as ‘set
SEISAN_TOP=\SEISMO ’, but could also be e.g. ‘\test\best\analysis ’ or ‘d:\seisan ’. Note there is one
blank character at the end and the first character MUST be "\" or the second ":". If no seismo
directory, it could e.g. be just E:.The default editor is notepad. The Unix variable AGENCY is not used
on the PC.
-
Install doskey for convenience when working in DOS windows. Add line
autoexec.bat (not needed on Windows2000 and XP, already there)
doskey /insert to
15
If installed on W95 or W98, the commands wo, da etc will not work and the files wo.bat etc in COM must
be modified. Just replace the /d %seisan_top with the true name like e.g. cd c:\seismo\wor
Example file
Set path=c:\seismo\com;c:\seismo\pro
Set def_base=test_
Set seisan_top=c:\seismo
Option 2
- Go to control panel – system - advanced and select environmental variables. Here all the above
variables can be set.
3. Optional: If you want to recompile, you need the Fortran compilers installed.
4. Printer:
It is assumed that a Postscript printer is connected to the PC, either directly or by a network. When a
program sends a plot to the printer, it issues the command SEISANPR. In the COM directory, there is a
file called SEISANPR.BAT containing the print command. Several suggestions are made in that file, the
default is to use ‘COPY filename PRN. If you do not have a PostScript printer, it is possible to view and
print PostScript files with GhostViev, a public domain software available from many sites, e.g.
http://www.seas.ucla.edu/~ee5cta/ghostView/
http://www.cs.wisc.edu/~ghost/gsview/get47.htm
http://www.geo.uib.no/Seismologi/SOFTWARE/
5. Testdata
The testdata set can be extracted from the file testdata_X.Y.tar.Z with for example the WinZIP program.
Extract the data to the SEISAN top directory. To use the test data base as the default, add line set
def_base=test_ in autoexec.bat file.
Installing Windows SEISAN interface
The program is called SEISAN.EXE and located in the PRO directory. To install it on the desktop, do the
following steps:
1: Right mouse click with cursor on the desktop
2: Select New/Shortcut
3: On the command line box, type c:\seismo\pro\seisan.exe, press Next
4: Type a name for the shortcut, e.g. SEISAN
5: Press finish button
In this installation example, SEISAN is installed under c:\seismo.
During installation, a file seisan.ini is created in the [c:\] directory (if, it does not exist) and thereby the
configuration variables are set from this file. If you want to work with different configuration, the file
has to be changed. In particular if you want to work with a data base on a different disk drive.
Example of seisan.ini file :
[Seisan]
SeisanDirectory=c:\seismo\
WorkDirectory=c:\seismo\work\
ProgDirectory=c:\SEISMO\PRO\
DatabaseDirectory=c:\seismo\REA\
[Database]
16
LastDatabase=CUB2_
LastYear=1998
LastMonth=04
IndexFile=[None]
Running SEISAN on a PC with data on a Unix system, or the other way around.
With SAMBA, a disk can be mounted between different platforms, for example a Unix file system under
Windows. The SEISAN_TOP must then be set to the name of the Unix disk a seen on the PC. All files can
be read directly from the Unix disk, however the files IASP91_platform.HED and IASP91_platform.TBL
must be present. SEISAN works out which of these files to use. With SAMBA, PC users can then access
a Unix SEISAN data base directly using the programs on the PC.
Potential problems
Program takes a long time to load: If the program is large, it might use disk swap files, which can take a lot
of time. Reduce array size (seidim.inc in INC) and recompile.
Commands like P or L do not work in EEV. If swapping takes place, some damaged swap files or other
files might remain which prevents the program from starting. Clean disk with command SCANDISK.
Digital Fortran seems to be very sensitive to viruses. If you get memory problems or programs will not run,
particularly under EEV, check for virus.
Graphics problem: Graphics programs write text to a console window. If there is a program crash, the
error message goes to the console window, which then disappears.
3.4 Database directories for your own system, MAKEREA
When you want to use SEISAN with your own data, the necessary directory structure must first be created
and your own station coordinates and crustal model must be defined. Both REA and WAV directory
structures are created with the program MAKEREA. The program asks for database name (must be 1-5
letters and UPPER CASE), start time (year and month), end time (year and month) and the structure to
create (REA or WAV). An output will then show the names of directories created. If a directory already
exists, a message will be given. It is possible to create both structures at a time, since the program after
finishing asks if the respective other structure should be created.
3.5 Default parameter files for the main programs
Once you start locating events, the location program HYP will look for an input file with station
coordinates, location parameters and crustal model. This file is located in the DAT directory and called
STATION0.HYP. To edit the file, move to the DAT directory by using command DA and edit the file. For
more information about this file, see section 6.1. If you want to keep the original file for test purposes, first
copy it to a file with another name.
When plotting epicenters (command EPIMAP) input files with map-contours for the EPIMAP program are
used (type *.MAP). These files are also located in the DAT directory. If you want to use more detailed
map contours, you must get hold of your own data and put them into file called e.g. MYMAP.MAP. In the
DAT directory there are two sets of contours: WORLD.MAP and EUROPE.MAP, a more detailed
European map than found in WORLD.MAP. Detailed map files in SEISAN format for the whole world is
found at the SEISAN ftp site and the SEISAN CD.
The plotting program MULPLT can use a default file for those stations, which are to be routinely plotted,
as well as other default information. This is defined in MULPLT.DEF in DAT, see example file.
MULPLT.DEF also defines which keys are assigned to which phases and what character is used for the
17
mouse. An example is given in DAT, see also example in 6.2.
Both the MULPLT.DEF and STATION0.HYP can also be in the working directory. Programs always look
there first and on a multi-user system, this enables different users to have their own setup. It also makes it
possible to work with different setups by just changing directory.
The AUTOPIC PROGRAM requires AUTOPIC.INP in the DAT directory.
Most SEISAN program use the SEISAN.DEF file (in DAT, see section 3.8) where many general
parameters are set.
The bulletin program requires a front page with whatever you like and some set for fonts can be made.
The text of this page is located in the BUL.INP file in the DAT directory.
3.6 Color settings
All programs using color can use a color definition file called COLOR.DEF. The file can be located in the
current working directory or in DAT. Programs will first look in the working directory, then in DAT. If no
COLOR.DEF file is given, default is used. Be careful with color setting, unexpected results might occur,
like getting a blank screen when plotting white traces on a white background. Several color settings are
universal like setting colors for titles, however several color settings are specific for particular programs,
see example file below. Since colors can also be used for the Postscript file, there is an option for using
color on the screen but not for Postscript, since it is more likely that the user will have a color screen than
a color Postscript plotter. Sending a color plot file to a black and white laser printer results in a plot with
gray tones.
The COLOR.DEF file:
This file is for defining SEISAN color objects. If no file available, default colors, as given at end of line
below, will be used. The entries can come in any order, however the keyword MUST appear as shown
below. If an object color is not defined, it will retain its default value. If a black or white background and
no colors are chosen, all colors will be reset accordingly.
The color codes are:
1: blue
2: green
color_screen
color_hard_copy
color_back
color_trace
color_pic
color_zoom
color_def
color_frame
color_title
color_spec
color_axis_not
color_epi
color_station
color_map_contour
color_map_grid
color_label_grid
color_symbol_key
color_prompt
color_section
color_bval_np
color_bval_ac
color_bval_line
color_box
color_box_letter
color_foc_dilat
color_foc_comp
color_foc_p
color_foc_t
color_foc_plane
color_syn
3: red
0
0
5
6
3
2
6
1
6
1
3
3
3
1
6
6
6
1
3
2
1
1
5
6
3
1
2
3
4
1
4: yellow
5: white
6: black
!0: no colors, 1: colors
! ----------------------! background color
! seismic traces, map contours,white
! phase picking,yellow
! zoo lines in mulplt, red
! default color, white
! frames like epimap map frames, mulplt, green
! titles on top of plots, yellow
! spectras, red
! axis notations, yellow
! epicenters, red
! seismic stations, blue
! epimap contours, yellow
! Lat-long/(x,y) grid, green
! Grid labels for map, green
! Diagram key, red
! Prompt text., black
! section outline in epimap,green
! bvalue, number of events, green
! ------, accumulated ---, blue
! ------, lsq line,blue
! box for interactive input,yellow
! letters in --------------,black
! focmec dilatation, red
! focmec compression, green
! focmec P-axis, red
! focmec T-axis, blue,
! focmec fault planes, blue
! synthetic picks, blue
18
3.7 Compiling SEISAN programs
The SEISAN distribution for all platforms includes the executables. Therefore in general it is not necessary
to recompile. However, you may have the source distribution or you might want to modify some of the
programs for your own needs or remove bugs and will have to compile programs.
The SEISAN programs on all platforms can be compiled using the make utility, which is make on Unix
(Solaris, Linux and MacOSX) and nmake in DIGITAL Visual Fortran under Windows. On all platforms
there is a ‘Makefile’ in both the PRO and LIB directories. The files should not need any modification. If you
have the SAC libraries installed on your system, you need to set the SACLIB environmental variable in
.SEISAN, otherwise SEISAN is compiled with SAC dummy routines and you will not be able to read or
write SAC files. The default setup in makefile is sac_dummy.o.
It is possible to move SEISAN between all Unix platforms without changing the Makefiles. You can
compile SEISAN from different operating systems under one file system by defining SEISARCH
(environmental variable), which can be solaris, linux or MacOSX. (Note: Without setting SEISARCH, the
compilation will not work on Unix, since make will not know what SEISARCH is.) The PC distribution is not
the same as for Unix, which means it is not possible to compile the Unix distribution under Windows or
vice versa. Most of the programs are the same on Unix and PC, but not all, which means the Makefiles
are different. See section 7 for differences between the platforms.
After you have changed the Makefiles according to your system, the compilation can be started from the
PRO directory by starting ‘make all’ on Unix and ‘nmake’ on Windows. From the Makefile in the PRO
directory, the Makefile in the LIB directory is started to create the object libraries. This does not work
under Windows, so you first need to start ‘nmake’ in the LIB directory, and afterwards in the PRO
directory. Then a SEISAN archive is created, ‘seisan.a’ on Unix and ‘seisan.lib’ on Windows. The archive
contains all library subroutines, and you can easily link to the archive if you want to use SEISAN
subroutines in your own programs. Finally all programs are compiled. Single programs can be recompiled
by starting ‘make program’ on Unix and ‘nmake program’ on Windows.
If you do changes in the LIB directory you need to compile using ‘make all’/’nmake all’, which will also
create the archive file. Then you can recompile individual programs in PRO as explained above.
When compiling SEISAN on SUSE Linux it has been suggested to use the following compiler options ‘malign-double -finit-local-zero’ in addition to the ones already used. Testing this on Redhat Linux
produced very large executables, but it may be worth trying on SUSE.
If graphics programs do no link, check that you have X11 libraries in $LD_LIBRARY_PATH. You can
check what is there by command echo $LD_LIBRARY_PATH.
3.8 Program Version
The version number of this SEISAN distribution is 8.0. New from version 7.0 is that all individual programs
have a version number and a date of the last changes. This will help the users and developers to find out
which version is in use and from which date. This is particular useful when bugs are reported. In case
major bugs have been fixed, the new versions (same version number, different date) of the individual
program will be available on the anonymous ftp site and a message will be sent to the seisan mailing list.
In all programs you can obtain the version number by starting a program with ‘-version’ as argument. For
example ‘collect –version’ will show you the version and date of your COLLECT program.
3.9 Setting general SEISAN defaults, SEISAN.DEF
There are several parameters which are common for several programs which are set in the file
SEISAN.DEF. This file can be in your local directory or the DAT directory. The system will first look in
the local directory, but normally the file is in the DAT directory. The file does not have to be there
since there are defaults for all parameters, but the parameters might need to be set.
19
The format of the file is as shown below:
This file is for defaults for SEISAN and called SEISAN.DEF. The name must
be in upper case on Sun. The following shows the parameters which can be set. The file can
contain any lines in any order, only the lines with
recognized keywords and a non blank field under Par 1 will be read. The
comments have no importance. Numbers are real numbers.
KEYWORD............Comments.............Par 1.....Par 2
CONFIRMATION
level 0 or 1
0.
CONT_BEFORE
start min before
20.
CONT_AFTER
start min after
1.
CONT_BASE
REA continuous bas
CBER
CONT_BASE
REA continuous bas
CBER
COPY_WAV_DIR
data base copy reg. BER
CURSOR
0,1,2=poi,cr,crhair 2
EPIMAP_STATIONS
pl. st. a for all
EPIMAP_MAP_FILE
name of map
EUROPE
EPIMAP_PROJECTION real number
3.
FOCMEC MAXSOL
100.
FONT
graphics font
HERKIJ_DISTANCE
100.0
HIGH_ACCURACY
high accuracy (0/1)
1.
HYPO71_OFFSET
offset in degree
0.0
INIT_IMGMAP_FILE
path to fi
c:\seismo\dat\IMGWORLD.GIF
MAP_LAT_BORDER
dist from center
3.0
MAP_LON_BORDER
"
6.0
MAP_SERVER
2
MERGE_WAVEFORM
Code for merging wa NSN
SPECTRAL GEO_DEPTHS
10.0
14.0
REG_KEEP_AUTO
keep phases when reg. 1.0
TEXT_PRINT
printer command EEV nenscript -Psps
WAVEFORM_BASE
Waveform base name
AGA
WAVEFORM_BASE
Waveform base name
LOF
WAVEFORM_BASE
Waveform base name
BER
WAVEFORM_DIRS
Waveform directory
/net/seismo/seismo/WOR/seisnet
OUTPUT_DIR
./
INIT_IMGMAP_FILE
/net/seismo/DAT/IMGWORLD.gif
MAP_SERVER
0
IMGMAP_PATH
/net/seismo/DAT/MAP
INIT_MAP_LOWER_LATITUDE
-90.0
INIT_MAP_UPPER_LATITUDE
90.0
INIT_MAP_LEFT_LONGITUDE
-180.0
INIT_MAP_RIGHT_LONGITUDE
180.0
INTERNET_BROWSER
/prog/netscape
HELP_DIR
/net/seismo/INF
The parameters are:
Confirmation: Level of confirmation required for example when deleting files, 0.=no confirmation and
1.=always confirm. (at the moment only used in eev)
Cont_base: Waveform database to be searched (there can be several)
Cont_before: Minutes to read into the memory before the required start time (must
be at least the length of a waveform file)
Cont_after: Minutes to read into the memory after the data that is plotted
Copy_wav_dir: Normally, waveform files are copied to WAV when registering an event. If this parameter is
set, waveform files can be copied to the WAV data base specified. Max 5 characters.
CURSOR: Select cursor: 0. for pointer, 1 for cross and 2 for crosshair (Unix only, on PC only pointer
is available)).
Epimap_map_file: The map coordinates file to use with MAP option in EEV
Epimap projection. The projection number used by EPIMAP, see EPIMAP program for choices.
Epimap_stations: One letter indicator for plotting stations, in first column. See EPIMAP program for
codes. A is all.
FOCMEC MAXSOL: Maximum number of solutions in FOCMEC grid search, default is 100, however,
20
it may be required to allow for more solutions.
FONT: Select font available on X system. To check available fonts use command ‘xlsfonts’ or
‘xfontsel’ and ‘xfd –fn font’ to display a font. (Unix only)
High_accuracy: Setting it to 1.0 enables high accuracy operation. This parameter affects programs
MULPLT, HYP and UPDATE
HYPO71_OFFSET: Apply offset in degree to station and epicenter locations, required for example when
not all stations are either east or west of 0 longitude.
Map_lat_border, map_lon_border: These parameters are used with command MAP and GMTMAP in
EEV, which plot a map centered on current epicenter. The two parameters give the distance in
degrees from the epicenter that the map should be plotted. If both set to ‘0.’ EEV will ask for the
parameters.
Merge_waveform: The network code given to waveform files merged with MULPLT when running
from EEV. See MULPLT and EEV. Also used in WAVETOOL and SEISEI. Max 5 characters.
Spectral_geo_depths and herkij_distance. See MULPLT spectral section for explanation. Parameters
used to calculate geometrical spreading.
Reg_keep_auto: If flag set to 1.0, keep automatic pics when registering event from EEV.
Text_print: Printer command used to print an S-file from EEV.
Waveform base: Name of waveform data base to be searched. Normally this is a 1-5 letter data base
name. The name must be written as shown above under Par1. Not needed for the default data base.
Waveform_dirs: The complete path to directories where the system should look for waveform files.
Output_dir: Output Directory for SEISAN commands results. Default “./”
Init_imgmap_file: File name of the
“/seismo/DAT/IMGWORLD.gif”, for JSEISAN.
initial
map
represented
as
image.
Default
Map_server: Type of map retrieved from Internet or locally; 0: Static local image (country boundaries),
1: Static remote image (country boundaries), 2,3,4,5: Dynamic Remote image (2:country boundaries,
3:Relief from GTOPO30 only land, 4:Two minute shaded relief, 5: Combine 3 and 4). Default 0, for
JSEISAN.
Imgmap_path: PATH for the static local maps (images) stored in the local hard disk used for zooming.
Default “/seismo/DAT/IMGMAP”, for JSEISAN.
Init_map_lower_latitude: Lower latitude of the initial map. Default –90.0, for JSEISAN.
Init_map_upper_latitude: Upper latitude of the initial map. Default 90.0, for JSEISAN.
Init_map_left_longitude: Lower longitude of the initial map. Default –180.0, for JSEISAN.
Init_map_right_longitude: Upper longitude of the initial map. Default 180.0, for JSEISAN.
Acrobat_reader: Path for the Adobe Acrobat Reader (needed for the help files). Default
“/prog/acroread” , for JSEISAN in Unix.
INTERNET_BROWSER: Location and place of browser
HELP_DIR: Directory of help files, usually INF
21
3.10 Format description in parmater files
Since most SEISAN programs are written in Fortran, the format descriptions follow the Fortran
convention. The following are the main format descriptors:
In: Integer format: E.g. I5 means an integer with 5 digits normally right justified. A completely blank
field will be read as zero. Examples:
123456..
1100
11
(Position)
read as 1100
read as 11
read as 0
Fn.k: Real number format: E.g. f7.3 is a real number occupying 7 places like 111.111 and the decimal
point is 3 places from the right. Any real number can occupy the 7 places like
1234567
(Position)
1.1
1.1
1.1
All of these will be read as 1.1. If there is no decimal place given, it is automatically put k places from
the right. Like the number 1234 read with f4.2 will be read as 12.34.
nX . Spaces. Like 5x means 5 spaces.
An: Character format: Like A5 means reading 5 characters
Combining format specifications, example
1234567890123456789
position
2f5.1,1x,a4,2x,i2
11.1 12.1 text 12
Do not use tabulator instead of blanks
3.11 Using SEISCONF to modify parameter files
By Bladimir Moreno
Introduction
The SEISCONF program is a JAVA application for editing some of the configuration files defined in
the SEISAN earthquake analysis software. The editing is based on a simple Graphical User
Interfaces composed of dialog-boxes where the input data is validated and an on-line help about the
parameter being configured is given. The user-friendly interface allows to minimize the possible error
in the configuration file format as well as to avoid unrealistic values for a particular parameter.
SEISCONF was developed with Visual Cafe 4 standard edition. The software operates on Unix and
Windows.
When all settings are done as described under Installation, you can start the program from the prompt
line as with the command ‘seisconf’.
Program options
The program starts with a main window (Fig. 1) with three options: (1) Edit the general SEISAN
configuration parameters (SEISAN.DEF), (2) Edit the parameters for the hypocenter determination
(STATION?.HYP) and (3) Edit the parameters for the waveforms analysis tool (MULPLT.DEF). An
option can be selected by clicking the associated button. Normally these files are expected to be in
the DAT directory. If the selected configuration file is found in the current directory, the user is
prompted for choosing which file, either the local or the global.
22
Figure 2. Main Window of the program.
Changing SEISAN.DEF
The editing process is based on a list made up with the configuration parameters. The parameters
which are present in the configuration file are marked with “*” (Figure 2). Parameters that are not
found in the file, are set to defaults values. These default values are taken from the file
“SEISANDEF.INP”. The user navigates through the list and can edit the associated values by using
the edit-boxes labelled as VALUE1 and VALUE2. The entered data is validated according to the type
and allowed range of values. The check-box labelled as “Use” permits to select/unselect the selected
parameter to be saved in the configuration file. The button <Add New> allows add a new parameter,
same type of the one selected in the list, if possible. A brief description about the meaning of the
selected parameter is given at the bottom of the window.
23
Figure 3. Edit Window of the configuration file: MULPLT.DEF
Changing STATION?.HYP
The editing window (Figure 3) is composed of four blocks: (1) A list with the general parameters
(RESET TEST) of the HYPOCENTER program, (2) The seismic stations list, (3) The velocity model
and (4) The control line. In addition there is a edit-box for the reporting agency and a combo-box
labelled as “Model” for selecting which STATION?.HYP file is being edited. The question mark (?) in
the file name takes the value selected in the combo-box. By default, STATION0.HYP is selected. The
combo-box is made up from the set of station files (STATION?.HYP) found in the DAT directory or in
the current local directory. There is two buttons <add> and <remove> in the station list block and
velocity model block. Their function is for adding or removing items from the list. If you want to add a
new item and there is one already selected (highlighted), you must click the button <add> and then
change the value with the new one. Do not try to enter the data before clicking the button <add>
because that will modify the current value of the selected item. The editing process is similar as it was
explained in the previous section, except that the data for validation and on-line help is taken from the
file “STATIONDEF.INP”.
24
Figure 4. Edit Window of the configuration file: STATION0.HYP
Changing MULPLT.DEF
The changing is the similar to SEISAN.DEF. The data for validation and on-line help is taken from
“MULPLTDEF.INP”
Input definition file format
The input files SEISANDEF.INP, MULPLTDEF.INP and STATIONDEF.INP are used for data
validation purpose and user’s help support. Normally the user will not edit these files. The first two
lines of the files are comments followed by three or more lines for each parameter: (1) The data
numerical description, (2) The name and a short description of the parameter shown on the list and
(3) A longer description (help) about the meaning of the parameter(s). This can be several lines. The
data numerical description has the following format:
<Keyword>:<default>,<range>,<value1>,<value2>,…;<type>,<column>,<width>,<dec>|…
<keyword>: Identify the name of the parameter used in the configuration file.
<default>: Default value used when the parameter is not found in the configuration file.
<range>: A character for specifying when the parameter is enclosed by a range of values or is
allowed to take a set of values. It is set with “R” for range or “U” for a set (see example below).
<value1>,<value2>: Represent the lower and upper limit of the range of values which are allow to be
25
taken. It is make sense when “R” is used in the <range> option. When using “U”, then
<value1>,<value2>,..,<valueN> are the set of permitted values.
<type>: A character for identifying the type of data: “F” float,”I” integer,”S” string of characters.
<column>: The position (column) of the parameter in the configuration file.
<width>: The number of characters occupied in the configuration file.
<dec>: The number of digits after the point for float values.
When more than one value is taken for a particular parameter, then the character “|” is used for
dividing the two formats.
Examples:
SPECTRAL F-BAND:0.05,R,0,20 ; F,40,5,2 | 20.0,R,0,20 ; F,50,5,2
The keyword “SPECTRL F-BAND” has two values. The first one is set by default to 0.05 Hz and can
take values between 0 and 20 Hz. It is a float value found in the column 40 of the configuration file
and occupies 5 characters with 2 decimal digits. The second value is set by default to 20 Hz and has
the same numerical format as the first one, except that it is found in column 50 of the configuration
file.
MAP_SERVER:0,U,0,1,2,3,4,5 ; I,40,1
The keyword ”MAP_SERVER” is set by default to 0 and can only take the values 0,1,2,3,4 and 5. It is
an integer data found in column 40 and occupies a 1 character position.
WAVEFORM_BASE:BER ; S,40,5
The keyword “WAVEFORM_BASE” is set by default to “BER”. It is a string with a maximum of 5
characters and it is found in the column 40 of the configuration file.
Below is shown part of the content of MULPLTDEF.INP
# KEYWORD:D,R,V1,V2,..;T,C,W,X|D,R,V1,V2,..;T,C,W,X
# D:default,R:Range(R,U),V1:value;T:type(F,I,S),C:column,W:width,X:decimals
X_SCREEN_SIZE:90.0,R,10,100;F,40,4,0
Size of initial X-window in % of total screen
Size of initial X-window in % of total screen
SPECTRAL Q0:440.0,R,0,1000;F,40,6,1
Q0 for spectral analysis
Q is defined as q0*f**qalpha, default 0 meaning no Q-correction
SPECTRAL KAPPA:0.0,R,0,10;F,40,5,2
Spectral kappa - near surface attenuation
Spectral kappa - near surface attenuation, default 0 meaning no attenuation
………………………………………..
26
4. USING SEISAN
Once the system has been installed, it is ready to use. Usually all work should be done in the WOR
directory or on a multi user system from your own directory. To move to WOR, type WO. Unless you have
to do system work, it will not be necessary to move to any other directories. However to do so, just type
the first two letters of the directory name like DA to move to the DAT directory. On a PC the Edit editor is
default (invoked with command edit), and on SUN the vi editor.
The system has two basic modes of operation. The first is to work interactively with the database. That
means jumping around from event to event, plotting, interactive phase picking, locating, deleting, typing,
editing or appending events (S-files). This mode is invoked with the command EEV, which uses several
programs, controlled by a driver program and is intended for testing and editing of single events. Once
the input data seems OK, the second mode of operation can be used.
On Windows95/98/2000/NT, program SEISAN is equivalent to EEV and whenever EEV is mentioned, this
is meant to also include W95/98/NT, SEISAN, see section 4.5.2.
The second mode is more like traditional data analysis where single programs are made to work on the
whole or part of the database. In this mode the updated S-files and CAT-files are created. Examples are
also plotting of epicenters, waveform data or searching for data fulfilling certain criteria.
The system comes with a test data set from different networks, mainly the Norwegian National Network
for the time periods 199309 to 200002. The data has waveform data in different formats. The data set
includes events from both local and teleseismic distances. The installation of test data is separate from
installation of SEISAN.
If you want to try the system, go directly to section 4.3 to get a feeling for how the system works.
SEISAN problems: Some of the most common problems have been collected in the index under the
header "Problem".
4.1 Short user guide
The SEISAN manual has been divided into sections describing the individual programs. However, many
tasks require the use of many programs and it is not always easy to find what can be done and which
programs to use. The following section intends to give an overview of some general problems that
SEISAN can work with and a list of programs to use. The following tasks have been identified:
- Routine processing: Phase picking, hypocenter location and magnitudes
- Determination of source parameters: Fault plane solution, stress drop, etc
- Crustal structure: Velocities, layer thickness and attenuation
- Seismic catalogs: ISC data, database management, completeness, statistics, etc
- Seismic hazard: Attenuation, catalogs, and soil response
Routine processing
The main work of a seismic observatory is to quickly process and organize incoming data from
different sources. SEISAN has a simple time ordered database (see later section) and a set of
programs for these tasks. The most important programs are:
EEV: The EEV program is the interactive program for working with single events in the database. The
program is used to navigate in the database to find a given event as well as for housekeeping
27
(splitting, merging and deleting events). Once an event has been selected, a large number of options
are available like phase picking, earthquake location, fault plane solution, macroseismic information
etc. All results of the interactive processing are stored in the database (S-files).
MULPLT: This is the general plotting and signal analysis program and can be used to pick phases
and amplitudes, correct for instrument response, produce Wood-Anderson seismograms for
determining Ml, simulate WWSSN SP and LP records, determine azimuth of arrival for 3 component
stations, rotate seismograms, display theoretical arrival times for IASP91 phases to help identifying
global phases and do spectral analysis. MULPLT can be used from EEV or as a stand-alone
program.
HYP: This is the general program for hypocenter location and is based on HYPOCENTER (Lienert et
al., 1986, Lienert and Havskov, 1995). The program can use nearly all common crustal and global
phases (8 character ISC codes), locate teleseismic events using the IASP91 model and use observed
azimuth and apparent velocity. The program can therefore be used with all types of input data
whether from single stations or arrays. HYP can be used from EEV or as a stand-alone program.
Apparent velocity is currently only used for starting location.
EPIMAP: This is the general hypocenter plotting program for making epicenter maps and hypocenter
profiles. The hypocenters can be plotted with elliptical error ellipses and EPIMAP can also be used
for interactive selection of events in polygon areas. For plotting hypocenters, there is also an
interface to GMT.
BUL: The function of this program is to produce a bulletin. The user can tailor the appearance to local
needs and the program can produce bulletins of hypocenters only or both hypocenters and phase
readings.
In addition to the above programs, several programs are available for database creation, input and
output of large data sets and conversion and manipulation of waveform data.
In order to get an idea of how routine processing works, some examples of routine processing will be
given below.
Case A: Telemetry network with 32 channel central recording
The network generates waveform event files, which are transferred to SEISAN. The tasks are:
1: Convert waveform files to SEISAN format (many events can be converted in one operation).
Inspect events with MULPLT. From MULPLT, false triggers are deleted and real events are put into
the database. Events are at this stage identified as local, regional or distant. Phase picks can be done
at this stage, but is usually done later.
2: Interactive phase picking, earthquake location, magnitude etc done with EEV. Automatic phase
picking is also possible at this stage.
3: Database is updated (UPDATE) once a suitable interval has been processed interactively, usually
a month. Updating means permanently storing the hypocenters etc in the database.
4: Make hypocenter maps with EPIMAP.
5: Produce a bulletin with BUL.
Case B: 3 telemetry networks and one broad band station
The routine is the same as above except for one additional step between 1 and 2. Since several data
28
sets are available, some of the detections from different networks or the broad band station might
correspond to the same event. There are now two options. The first is to merge the waveform files for
corresponding events and then put the events into the database. The second option is to put all real
events into the database and then do the merging from EEV.
Case C: A mix of stations and networks and additional phase readings
The steps are as in case B except that before step 2, the additional phase data is put into the
database. In this case the merging of events must be done with EEV.
It should be noted that data collection and step 1 to 3 is fully automated using SEISNET (Ottemöller
and Havskov, 1999).
Example of using EEV for interactive processing:
Find event in default database nearest the given date and time: EEV 1999020303
Once EEV is started, an EEV prompt is given and different EEV options are available. Examples are:
E: Edit event, P: Plot event, L: Locate event, F: Make fault plane solution, d2201: Find event nearest
day 22 at 01 hour, MAP: Start EPIMAP to show earthquake location and SAC: Start SAC processing
of event using all parameter and waveform data from SEISAN database.
The above examples have mostly described the interactive processing of single events. However,
once the data is in the database, operations can be done on the whole database, for any time interval
or for events fulfilling certain criteria (like magnitude, area etc). Examples are relocating events,
extracting data and determining coda Q.
Source parameters
The routine processing normally produces magnitudes and hypocenters. The fault plane solution can
be determined using polarities and one event (Snoke et. al., 1984). Composite fault plane solutions
can also be made. A second way of determining fault plane solution is to synthetically model the
waveforms using the modeling programs. In addition, seismic moment, stress drop and seismic
source radius can be determined by doing spectral analysis or spectral modeling. This can also be
done automatically with AUTOSIG. The moment tensor of local earthquakes can be determined by
inverting the amplitudes of the Pg and Sg waves (Ebel and Bonjer, 1990)
The full wave modeling programs integrated with SEISAN, are written by Bouchon (1981) and
Herrmann (Herrmann,1996). The ray-tracing program is based on WKBJ and written by Chapman et
al (1988) and integrated with SEISAN by Valerie Maupin.
All the above programs are executed from EEV in order to use known source parameters.
Crustal structure and Q
A large database can be a good source of information for determining structural parameters and
SEISAN provides several programs to determine the crustal structure and Q. Using seismic arrival
times, it is possible to invert for the crustal structure using the VELEST program (Kissling et al, 1994).
It is also possible to do forward modeling using the location program for a large number earthquakes,
since it at the end of a run, a summary of average station travel time residuals and event RMS is
given. A special option of HYP is to locate a data set with all permutation of a given range of models
in order to find the model giving the lowest RMS.
Deep earthquakes under a local network produce clear phase conversion at crustal interfaces (Chiu
et al, 1986). They can be modeled with one of the full wave modeling programs both with respect to
amplitude and arrival time.
29
SEISAN can, when displaying surface waves, make spectral files ready to be processed for surface
wave dispersion with Herrmann's programs (Herrmann, 1996).
Attenuation can be determined using the coda Q method for local earthquakes (CODAQ). The coda Q
program will calculate q for a series of events and stations at given frequencies. Average values are
calculated and a q vs f curve is fitted to the calculated values. The principle for calculation is the
standard coda q method, whereby a coda window is bandpass filtered, an envelope fitted and the
coda q at the corresponding frequency calculated (Havskov et al, 1989). The SPEC program will
determine Q by calculating spectral ratios or the near surface attenuation using the spectral decay
method. An alternative is to use spectral modeling where Q, stress drop and seismic moment are
modeled simultaneously.
Catalog and database work
Once a large database has been created, several programs are used to manipulate and analyze the
data. The catalog can be searched for a large number of parameters. Selection criteria are:
Magnitude range, magnitude types, event types (e.g. local, distant, volcanic, explosion), latitude,
longitude and depth range, RMS of travel time residuals, number of stations used in the location, felt
events, number of polarities, presence of certain stations etc. Events can also be selected in an area
with the program used for hypocentral plots.
A very useful source of data is the ISC. Data from ISC CD ROM's can be read and converted to
SEISAN format (hypocenters and phase data) and put into a database. The data can then be used for
e.g. seismic hazard, fault plane solution or it can be relocated. A general task with catalogs is to
homogenize magnitudes. Magnitude relations between e.g. Mb and Ms or Ms from one agency to Ms
from another agency can be done with the program MAG. The program will also convert one
magnitude to another once the linear regression has been determined. Event statistics can be made
with STATIS and b-values calculated with BVALUE. The number of events as a function of time is
plotted with CATSTAT.
Seismic hazard
Probabilistic earthquake hazard computations is done, using the EQRISK program (McGuire, 1976)
or the CRISIS99 program (Ordaz, 1991; Ordaz, 1999). EQRISK computes seismic hazard in terms of
probabilities of exceedence vs earthquake intensity measures such as peak ground acceleration
(PGA), for a given site or a grid of sites for up to eight different return periods.
The site amplification is calculated with the SPEC program. This is used for making spectra of many
seismic signals in a semiautomatic manner. The program is intended for two purposes: (1) making
relative spectra for a series of pairs of stations terminated by the average spectra, (2) Making a series
of spectra for a number of stations and events. The spectra can be corrected for distance, q, and
instrument response.
This section involves a large number of programs and a more detailed description is given in section 6.24.
4.2 Getting data into the database
The first requirement for interactive work with the event editor EEV is to get the data into the database.
There are two ways to get data into the database, as described in section 4.2.1 and 4.2.2. It is of course
possible to make the individual S-files directly in the REA directories with the editor. This would be rather
slow, and be against the philosophy of the system. However, it is mentioned in order to point out how
simple the database structure is.
The SEISAN system can be used with or without digital data, the only difference to the directory structure
is that the WAV and CAL directories are present when using digital waveform data. However the way of
getting data into the database differs in the two cases and will be described separately.
30
4.2.1 System with digital data
This means that the original data is individual digital event waveform files generated by some data
acquisition system. The waveform data can be stored in SEISAN, GSE and SAC format as single or multi
trace files. The files that are used in conjunction with the database are normally stored in WAV but can
also be in the user's directory, e.g. WOR. The normal scenario would be that multiplexed files would be
transferred from a digital field station, demultiplexed and converted to SEISAN waveform format.
Programs are provided to convert from most of the popular waveform formats like MINISEED, GSE,
PCSUDS and from commercial recorders. It is most practical to initially put the files in WOR, check the
events for false triggers, save the true events in WAV, make the corresponding S-file and a hardcopy of
the digital data.
All of this can be done with the program MULPLT. The program plots channels from a single waveform
file. The user can then interactively decide if this is an event to keep, in which case an S-file is created in
the database and the event is moved to WAV.
Alternatively, all new waveform files can be auto-registered into the database (AUTOREG) and all
checking takes place from EEV.
When digital data is the input to the analysis system, MULPLT is the program to use to get data into the
database. From there on further analysis can be done with EEV (picking phases, locating and editing).
MULPLT is also the program used with EEV. For more details on MULPLT, see detailed description in
section 6.2.
4.2.2 System without digital data
In this case the user would get phase data from other sources, e.g. analog seismograms or files with
readings from other stations and agencies. These files are assumed to be written in Nordic Format.
Conversion can be done from other formats like ISC, NEIC and HYPO71.
If a user already has a file with one or several events in Nordic Format, this file can be split up into single
files which are copied (from any directory) into the database by using the command SPLIT. Creating a
new file in Nordic Format can also be done with the program NEWEVE (use command NEWEVE).
The SPLIT program then reads the NEWEVE output file and writes out single S-files with correct names
either in the current directory (default) or in the database specified (BER or another). The reason that the
database specifically must be given is that the user should not accidentally put data into the database
(see section 6.6).
4.2.3 Database security
Duplicate ID:
Since the database consists of single files with names corresponding to time down to the second as well
as the event type (L, R or D) it will sometimes happen that two events will get the same name. Thus
copying in a new event with the same name could overwrite the existing event, and the user would never
know. In SEISAN, from version 5.0, some security has been put in. New data can enter the database with
4 programs: SPLIT, EEV, MULPLT and AUTOREG. With all programs, the user will be prompted if a new
event is about to overwrite an existing event. Both SPLIT and EEV have the possibility to create
alternative ID's if the user wants both the new and old event, while MULPLT and AUTOREG just offers
the possibility to skip a double event. If a new ID is created, an attempt will be made to use a time one
second later. If that also corresponds to an existing event, the next second is attempted etc. This allows
for 60 events to be registered in the database with the same minute and event type. If an event has got
the ID changed, the header line in the file is NOT changed, however the ID line is of course changed. This
31
will be indicated on the ID line with a ’d’ at the end of the ID number.
Deleting events:
Event here means S-file in the database. Events are only deleted when using EEV, either with the EEV
delete command D or the EEV append command A. In both cases, the deleted event is stored in the
DELET database before being deleted from whatever database. Even if the system contains many
databases, there is only one DELET database. This means that deleted events from different databases
are mixed in DELET. In order to restore an event, enter DELET database with EEV and copy the deleted
event back with the C command. It is up to the user to manually clean up the DELET database.
There is one more final security. If an event has been deleted from a database, but an UPDATE has not
yet been made, the event might be in the CAT part of the database and can be extracted by SELECT or
the editor.
4.2.4 Data base tools, content and checks
Content of data bases, program BASE:
In the REA directory, a binary file called REA.LOG contains information about number of events in all data
base. Initially the file has no information, but each time programs EEV, HYP, UPD, CHECK_BASE or
COLLECT are executed, the information is updated for the months accessed. The information can be
displayed with program BASE, which first shows available data bases and the user, can then select one to
get info for particular months. Make sure to use right case for data base names, always in upper case on
Unix systems. The program is still a bit experimental !!
Check content of S-files for magnitudes and residuals etc, program CHECKRE:
The program can read data bases or CAT files and check events for large residuals, abnormal depths
etch. The program is intended for quality control, the parameters hardwired in the program might not suit
all. Check program source listing.
Check for data base related errors, program CHECK_BASE
The data base depends on error free S-files and that there is a correspondence between the S-file name
and the event ID. This should normally be ok, however errors can occur during editing or there can be
program crashed producing errors. The program reads the data base and checks for:
Missing ID lines: If ID line is missing, it can be put in manually or doing an UPDATE.
No correspondence between ID line and S-file name: A serious error has occurred. try to find out what is
correct, the ID or the file name. An UPDATE cures the problem, however data might be lost.
Error in S-file: All parameters are checked and files with non standard parameters are indicated. The error
can be a number in a wrong position. The errors should be corrected.
For all the above 3 cases, an index file is generated with bad S-files and EEV can the be used directly
with the index file to access the bad S-files. THIS ONLY WORKS WITH ONE DATA BASE AT A TIME.
It is recommended to run check_base in case of system crash or as a security, just before an UPDATE.
4.2.5 High accuracy in SEISAN
SEISAN can use higher accuracy than the default. The goal is to have an accuracy of 1 ms in time
32
and 1 m in location.
In order to write out the high accuracy numbers, a new parameter has been added to SEISAN.DEF.
The parameter is HIGH_ACCURACY. Setting it to 1.0 enables high accuracy operation. This
parameter affects the programs MULPLT, FK, HYP and UPDATE.
Station locations: The station file looks like before except that in order to get higher accuracy of
station locations, the minutes of latitude and longitude are specified without the point. E.g. the
minutes 22.122 can now be written as 22122 in the same columns as before while if the point is
given, only 2 decimals can be used as 22.12. This changes do not affect any old station coordinates.
Programs reading station coordinates, will use high accuracy input if available.
EPIMAP will always read in high accuracy mode, if any high accuracy data is present, whether station
locations or hypocenters.
FK will always read high accuracy station coordinates, if available and FK can therefore now be used
with very small arrays.
Programs with output affected by high accuracy mode:
MULPLT will write the phase readings as f6.3 instead of f5.2 like e.g. 11.234 instead of 11.23. For
normal use, this is not needed and the files look better if high accuracy mode is not used.
HYP and UPDATE writes an extra high accuracy hypocenter line which has been given type H. An
example is
1996
1996
6 3 2006 35.5 D
6 3 2006 35.511
The format
Column
1 -15
16
17
23
24-32
33
34-44
44
45-52
53
54-59
60-79
80
46.787 153.722 33.0
46.78711 153.72245
TES 15 1.9 3.4STES 5.8BTES 5.6BPDE1
33.011 1.923
H
is
As type 1 line
Free
Seconds, f6.3
Free
Latitude, f9.5
Free
Longitude, f10.5
Free
Depth, f8.3
Free
RMS, f6.3
Free
H
4.3 Interactive work with earthquake locations, EEV command
The idea of SEISAN for interactive work is that the user should be able to easily jump from event to event
and run several different programs with one event without restarting every time. This is done with the
command EEV (see below). In this interactive mode, events are picked, edited, located, moved, deleted
etc. until a satisfactory solution is found. In the interactive mode, NO UPDATING of the location in the Sfile or the permanent output CAT directory is done since it is too easy in interactive mode to accidentally
change something. The permanent updating of S-files and CAT directories can only be done for one or
several months at a time (see UPDATE command) in order to ensure that nothing is forgotten within a
month.
Once the events have been updated, further work can be done (like searching for specific events or
making a bulletin) using single programs which read directly from the database. Most of the analysis
programs will also work without using the database structure that is e.g. searching in single file with many
events. For more details of the analysis programs, see section 6.
33
4.4 How EEV works
It is now assumed that data has been entered into the database. The fundamental tool for the database is
then the EEV program, which mostly works within the limits of one month in the standard database or with
whatever the user has of S-files in his own directory. Optionally, EEV can also work with several months.
A special option is to use a list of files in an INDEX file, see end of this section and SELECT program.
Some of the commands available within EEV are also available within programs. See below for more
details on EEV.
The EEV program reads the file names of all S-files in the database monthly directory (or local directory or
index file), positions the pointer at the first event and asks for a command to be performed for the current
event or to find another event. If the command is to use a program, control is handed over to that
program, which on completion hands control back to EEV. In this way, many different independent
programs can be used from within EEV, e.g. several different location programs can be installed.
EEV can be started in several ways:
EEV with one month in default database: EEV yyyymm.
E.g. EEV 199201 would work on January 1992 on the standard BER database. It is here also possible to
give a more precise start time like EEV 1992011520 to start with the first event at or after January 15 at 20
hrs.
EEV with one month in alternative database: EEV yyyymm BASE.
BASE is the database. To work on the NAO base, the command would be EEV 199201 NAO.
EEV with several months in default database: EEV yyyymm YYYYMM
yyyymm is start year and month and YYYYMM is end year and month.
EEV with several months in alternative database: EEV yyyymm YYYYMM BASE
yyyymm is start year and month and YYYYMM is end year and month.
EEV to work with events is local directory: EEV
Only the S-files in local directory will be used.
EEV to work with an index file: EEV index.out
EEV can work with an index file and the command would be EEV index.out, where index.out is the index
file name (can have any name as long as it contains a ‘.’ except when used with HYP). For information on
index files, see 6.4.
Databases can have 1-5 letter names and the user specify 1-5 letters. The real names in the directory
structure are always 5 letters so if the user specifies e.g. a base name of BA, the real name will be BA___
. The full 5-letter name can also be used.
The commands in EEV mainly use only one letter unless a date or a number has to be given. To get a
short explanation, type ? and you will get:
34
----------------------------------------------------------------------Help on EEV
----------------------------------------------------------------------?
#xx:
Axx:
Print this help file
Go to event # xx, also works without the #
Append event # xx to current event, original event
remains
AA:
Append current event to next event, original event remains
AUTOSIG: Automatic processing
B:
Back one event
BOUCH:
Run Bouchon's modeling program
BOUSEI:
Make SEISAN file from Bouchon synthetic file
C:
Copy event to another data base or to current directory
COMMENT: Write comment lines in S-file
D:
Delete current event, you will be prompted to confirm
DUP:
Duplicate current event in data base, different id
Dxxxxxx: Go to first event on date xx, hour xx, min xx
E:
Edit
EXP:
Enter explosion information
Eyyyymm: Let EEV session end with year yyyy and month mm
F:
Make a fault plane solution
FC:
Accumulate data for composite fault plane solution
GRID:
Locate by grid search
GMTMAP:
Makes a GMT map like the MAP commnad
HERRMANN: Run Herrmann's modelling program (not PC)
HERSEI:
Make SEISAN file from Herrmann synthetic file (not PC)
H:
Locate with Hypoinverse
HYPO71:
Locate with Hypo71
IASP:
Generate arrival times using IASPEI91 tables
IL:
ISC location program (unix only)
INPUTEPI: Input hypocenter and origin time in S-file from st. input
INPUTONE: Input an additional type one line (hypocenter line)
INPUTX:
Input of xnear and xfar.
INVRAD:
Make moment ternsor inversion with Invrad
Jyyyymm BAS:Jump to year yy and month mm in base BAS
L:
Locate event, will also calculate magnitude if not
locatable but distance is present, Hypocenter
LL:
Locate current and next event together
Lxx:
Locate current and event xx together
MAC:
Input macroseismic information
MACROMAP: Make a GMT based map of felt information, Unix only
MAP:
Make a map of current location
MODELS:
List MODEL.DEF file in DAT
MOM:
Make moment ternsor inversion with Invrad
NEW:
Add a new event to data base
WKBJ:
Make synthetic seismograms with WKBJ program
O:
Operating system command, e.g. ols is ls, ocd test is cd test
do not currently work on command with prompt input like epimap and
collect
P:
Plot event, also make hard copies and pic phases
PO:
Plot event with defaults
PITSA:
Start Pitsa program (not on PC)
PRINT:
Print S-file on printer
PMAC:
Macroseismic Windows program
PUT:
Register event
Q:
Quit EEV
R:
Rename event type, must be L, R or D
REG:
Register event
RMSDEP:
Calculates and plots RMS as a function of depth
SAC:
Run SAC
Sxxxxxx: Search for next two events which are within xxxxxx seconds. If time
blank, a default of 180 secs is used
SS:
Find next unprocessed event in base
SYNT:
Make parameters for synthetic modelling
T:
Type event
TT:
Type only header line of event
UPDATE:
Updates S-file with hypocenter etc.
USERCOM: Start user defined command as 'usercom -sfile <sfile-name> '
WAD:
Make a wadati diagram
U:
Update list of S-files
W:
Show location of waveform files
Z:
Auto pic current event, if readings available, new pics
will be added with a flag
-----------------------------------------------------------------------
35
Note: Command letters can be upper or lower case.
Comments to commands:
#XXX : Go to event by number. When giving a number, only give the number of digits needed, no
formatting. Thus e.g. to find event 7 or 777, write 7 or 777 respectively. If there is not an event
corresponding to the parameter specified, EEV will go back to event #1. In the number command, # can
be omitted.
Axxx: Append another event to current event.
The event specified is appended to current event. All header and lines in both files are saved and put in
order in the current event. The main first header is from the current event. The ID line for the appended
event is saved as a comment line. The user will be questioned if the appended event is to be deleted.
AA: Same as above using next event.
AUTOSIG: Automatic processing with autosig program.
B: Back one event
BOUCH:
Run Bouchon's modeling program
BOUSEI: Make SEISAN file from Bouchon synthetic file
C: Copy events
There are two options, copy the event to another database given by a 1-5 letter name (upper case) or to a
file EEV.OUT in your working directory. Several files can be extracted within one EEV session to the
same EEV.OUT file. A new EEV session deletes the previous eev.out file. The C option can be used to
recover files from the DELET database of deleted events. In addition to making the EEV.OUT file, an
index file is also made called indexeev.out. THIS FILE IS NOT DELETED WHEN EEV STARTS UP since
the intention is to be able to use EEV to make an index file of interesting events from several months. You
can then start eev with the selected events with command EEV eevindex.out. Note: The other data base
can also be a local data base “,,” in which case EEV should not operate on the same local data base.
COMMENT: Comment are written into S-file, terminated by a blank line.
DXXXXX; The D-command is used to jump to another event at a given date and time, normally only day is
used: The hour can optionally be specified. E.g. d2205 will find the event nearest in time after day 22 at 05
hours. If both day and hour is used 4 digits MUST be given e.g. 0708.
Highest accuracy is the nearest minute.
D: Delete event
You are asked for confirmation. After the event has been deleted, all S-file names are read in again and
all event numbers after the deleted event are therefore changed. The deleted event is automatically
saved in the DELET database. If the event is present in the CAT file, it remains there until the next
update is done, see UPDATE command in 6.7.
DUP: Duplicates an event in the database. The duplicated event has an ID, which is one second different
from the original event. The command can be used to split an event in two and then manually deleting
phase lines in each.
E: Edit the event. As default on SUN vi is used and on PC edit is used. The editor can be changed, see
section 3. When control goes back to EEV, the file is checked for possible typing errors or other format
problems. If a problem is encountered, the line with the problem is displayed with an indication of where
the mistake might be, and the user is returned to the editor. Alternatively the error can be ignored. The file
is also checked for missing iD and consistency between file name and ID. Problem: Some editors will
36
keep a backup copy of the original file so 2 files might be present with one e.g. with the additional
extension .BAK. EEV (from version 7.2) will only use the original file, but there is no check on what backup
files might accumulate.
Eyyyymm: Giving this command will make the current EEV session end with year yyyy and month mm
within the same data base. When EEV gets to the end of the month, pressing return will move EEV to the
first event of the following month instead of to the first event of the same month.
EXP: Input of explosion information. This command creates 3 new lines (see format description in
Appendix 1) and changes the main header line event type to explosion (E). The user is asked for
location, time, charge and comments. The explosion agency is used to classify types of sites and can be
used by SELECT for searching.
F: Make a fault plane solution.
The program uses polarities. See section 6.16 for more details.
FC: This command accumulates data in a file focmec.inp to be used for a composite fault plane solution.
For each event to be used, type command FC. See 6.16 for more details.
GMTMAP: Start ‘gmtmap.exp’ program (not included in SEISAN) to plot GMT map. GMTMAP
automatically creates a map using GMT. (UNIX only)
GRID:. Hypocenter is started up and will ask for the grid: Latitude and longitude range and grid
spacing. A maximum of 71 points can be used in each direction. The point with the lowest RMS is
found and the corresponding location and residual is printed on the screen. It is now optionally
possible to plot the contours on the screen. The map coordinates used are as defined in
SEISAN.DEF. Note that the grid search is using exactly the same parameters as Hypocenter. This
includes all weights and phase types. The depth is fixed to the depth given in the S-file header line.
For more details and an example, see application note 'epi.pdf' in INF.
H: Locate with Hypoinverse, no database update is made, no Nordic output format file.
HERRMAN: Herrmann’s modeling programs, only on Sun, might work on Linux, not tested.
HERSEI: Make a SEISAN waveform file from output of Herrmann modeling, only tested on Sun.
HYPO71: Locate with HYPO71. The database is not updated (not well tested on PC).
IASP: Generate a file with theoretical arrival times for the current event. The command will only work if the
event has an epicenter and origin time in header line or a subsequent type 1 line, see also INPUTEPI and
INPUTONE. These theoretical times will then be displayed with mulplt, the next time command P is used
in EEV. The theoretical times are listed in file iasp.out. See section 6.20.3 for more information. The
command can also be used directly from MULPLT.
IL: Makes a location with the ISC location program. For more info, see section 6.1.4:
::
INPUTONE: Makes an additional type one line (hypocenter line) in the file. Enter the data exactly under
the columns indicated. The line will be entered exactly as written, so it is possible to enter any part of the
information.
INPUTEPI: Works like INPUTONE, except that it overwrites information on the first header line if nonblank information is given. Use INPUTEPI to add information to the first header line like e.g. the depth. If
existing nonblank characters on the line are to be replaced by blanks (e.g. remove a magnitude), use
underscore "_".
INPUTX: Makes a comment line with xnear and xfar values.
INVRAD: Runs the moment tensor inversion program, see description below on moment tensor
37
inversion.
Jyyyymm BAS: This command makes it possible to change month and database during an EEV session
by giving a new year yyyy and month mm and optionally a new database BAS. If no database is given, the
same database is assumed.
L: Locate event with HYPOCENTER (same as HYP). The location does not update the S-file.
Lxx: Locate current event with event number xx. This is used to check if two events belong together.
LL: Locate current and following event together.
MAC: Enter macroseismic information, you will be prompted for all information. For details of the type of
information, see definition of Nordic format, Appendix 1. See also command PMAC.
MACROMAP: Felt information is read from a file with macroseismic information and plotted with GMT.
The file name of the file with macroseimic observations is given in the S-file.
MAP: Start EPIMAP program to produce a map of current location. If a location is given in the S-file, this
location is plotted, otherwise the event is located if possible and the resulting location used for plotting.
The parameters for generating the map are set in the SEISAN.DEF file (see 3.8).
MODELS: Lists MODEL.DEF file in DAT that assigns names to single characters in STATIONx.HYP file.
NEW: Creates a new event in the database. The user is asked to give date and time and the event is
created in the current monthly database.
O: Give a command to the operating system. This is a very useful command, since it is possible to do
almost anything without leaving EEV, including starting a new session of EEV !! E.g. the command ols on
Sun and odir on PC would make a directory listing.
PUT: Register event. This option is mainly meant to be used with the SEISNET data collection
system. The command cleans up the S-file for all SEISNET operations. It removes commented out
ID-lines and copies the waveform files given for the event from the current directory to WAV. The
command is equivalent to the register command in MULPLT. If events are auto registered with
AUTOREG, the command can be used to clean up and inspect incoming data without using MULPLT
directly.
PMAC: Windows only program PROMAC for processing macroseismic information to calculate
intensities from felt information and model the macroseismic intensities. The program can also plot
associated pictures (in directory PIC). All information is stored in the S-file. The program was written
by Bladimir Moreno, and has a separate manual, see INF directory. Program must be installed
separately, zip file in SUP.
P: Plot event with MULPLT
PO: Use MULPLT with defaults. This means that no questions will be asked and the plot appears in
multi trace mode with default channels and default filters as given in the MULPLT.DEF file in DAT.
Useful option for routine inspection of raw data.
PITSA: Run the Pitsa program, see section 6.13. (Not on PC)
PRINT: The current S-file is printed on the default printer, to set up printer command, see SEISAN.DEF
(section3.8).
Q: Quit EEV
REG: Same as PUT.
38
R: Rename event type
Giving an event a new type requires changing the header in the S-file and the S-file name. All this is done
with R-command. You are prompted for a new type (can be the same in which case nothing is done). A
new S-file is made and the old deleted. The CAT-file is NOT changed so if no UPDATE is done, the event
there will remain with wrong type. Event types are L: Local event, R: Regional event and D: Distant event.
RMSDEP: Calculates and plots RMS as a function of depth for current event. RMSDEP also operates as
a freestanding program with additional capabilities, see description in program.
SAC: Convert all data to SAC format and starts the SAC processing system ( not distributed with SEISAN,
must be obtained separately), not on PC.
Sxxxxxx: Search for next pairs of events separated in time by xxxxxx secs (max 999999). If no value is
given, 180 secs is used. The command is intended for finding events to be merged after putting together
two different data sets with SPLIT. If a new time instead of the default 180 is entered, it will remain in
effect for the whole
EEV session. NOTE, that the search starts with the current event, so after using S, one return to go to
the next event must be given to start a new search.
SS: Find next unprocessed event in database. Events, which have status in ID line as follows: SPL:
split with SPLIT program, HYP: auto-located with HYP, NEW: new event from EEV or ARG:
registered by AUTOREG. The idea is that when new unprocessed data have entered the
database by one of these programs, it should be easy for the operator to find the event. In
EEV, an N near the end of the prompt line indicates an event with this status.
T: Type event.
TT: Type only header of event.
UPDATE: Updates (overwrite) S-file with hypocenter, magnitudes, residuals etc. Note that the CAT file IS
NOT UPDATED. This can only be done with stand-alone command UPDATE, see section 6.7.
U: Update EEV event list. All S-file names are read in again. Is useful if data arrives during an EEV
session, like when using Copy command from another data base.
USERCOM: Starts user defined program with command ‘usercom –sfile <sfile-name>’, where usercom is
the command name. This command is useful for example if you want to start your program to create a
report based on the S-file, from EEV. Note: the usercom is not a SEISAN program.
W: Check if event has waveform files. If so, check in which directory they are if present on the
system. The search will start in current directory, then WAV followed by all directories defined with
keyword WAVEFORM_BASE in SEISAN.DEF in DAT.
WAD: The program reads the data for the event and then asks if all phases are going to be used or
only phases of the same type like Pg and Sg. Ideally, only phases of the same type should be used,
however in practice it might be interesting so see all data, it might give an idea about phase
identification. The Wadati parameters will now be calculated and shown on the screen. Optionally a
plot can now be made. The plot shows the Wadati diagram. On the left is shown all stations with
corresponding S-P times. Any station on the plot can be identified with the cursor. Point the cursor
near a symbol and click and the station data will be shown in the upper right hand corner. This facility
is used to identify bad picks.
The plot output file is called wad_plot.eps.
Z: Automatic phase picking. A waveform file must be present. See also the AUTO program section 6.15.
Below is shown a session with EEV on PC.
39
Example of using EEV for November 1993
eev 199311
1993 11 Reading events for base
#
1 2 Nov 1993 17:06 48 L
#
2 5 Nov 1993 22:37 21 D
#
3 5 Nov 1993 22:37 23 D
#
4 5 Nov
93 22:39 2 L
#
5 5 Nov
93 22:40 58 L
#
6 7 Nov 1993 23:40 43 L
#
7 7 Nov 1993 23:43 17 L
#
17 19 Nov 1993 01:45 29 D
AGA
60.443
18
4.512
2.0
67.837
20.059 15.0
66.307
6.919 31.0
70.069 -139.780
.1
1.5 N 1.8CBER
6
1
1
0.7
1.4
0.1
7
8
7
2.5CBER
3.1CBER
?
?
?
?
?
?
? 17
? t
File name: \seismo\REA\AGA__\1993\11\19-0145-29D.S199302
1993 1119 0145 29.0 D 70.069-139.780
.1 BER 7 .1
1
.19
999.9
821.9999.9 -.3206E+06 -.2536E+07
.2639E+08E
ACTION:UPD 97-03-25 21:28 OP:jh
STATUS:
ID:19931119014529
I
93111901.K41
6
93 1119 153 6.5 D
1
9311-19-0153-06S.NSN_09
6
STAT SP IPHASW D HRMM SECON CODA AMPLIT PERI AZIMU VELO SNR AR TRES W DIS CAZ7
KBS SZ EP
151 54.8
13.4 0 3365 161
TRO SZ EP
153 03.0
.010 4420 169
MOL SZ EP
153 50.51
.010 5070 165
ASK SZ EP
154 04.0
.010 5262 164
BER SZ EP
154 05.0
.110 5274 165
EGD SZ EP
154 05.5
-.110 5285 165
KONO BZ EP
9
153 49.21
-25.5 0 5413 167
#
17 19 Nov 1993 01:45 29 D
70.069 -139.780
.1
#
18 21 Nov 1993 01:53 56 L
60.184
4.965 15.0
1993 11 Reading events for base AGA
18
1 2 Nov 1993 17:06 48 L
60.443
4.512 2.0
N 0.5
2.6CBER
7
11
2.2
1.8CBER
6
?
?
? q
In the above example (PC), the month has 18 events. For each event, vital information is displayed: Date,
type, hypocenter, RMS, first magnitude and number of stations (number in S-file which might be larger
than number used for location as given in S-file header line after a location). In this way the user can
quickly search for events wanted and get important information without looking at all the details. The first
event in the list is newly entered into the database as indicated with the N near the end of the line. In the
above example, a return was made to go to next event until event #7 after which a jump was made to
event 17. For this event, all parameter data was displayed with the 't' command. A return was made to
event 18, another return and the event list was read in again and event #1 again became the current
event. Note that not all events had a location.
Below are shown examples of the commands (C)opy, (D)ate, a(S)sociate and (A)ppend. Comment are
preceded by '!' and written in bold. The database is EAF.
EEV 199405 EAF
1994 5 Reading events for base EAF
#
1 1 May 1994 1:18 8 D
#
2 1 May 1994 11:37 6 L
#
3 1 May 1994 12:00 33 D
36.607
# 366 20 May 1994 5: 2 8 R
! the month has 613 events
?
?
68.449 15.0 2.4
! go to day 20 ? d20
? c
! copy an event to working dir.
Copy event: Other database, give 1-5 letter name
Working directory in file eev.out: return
#
#
#
366 20 May 1994 5: 2 8 R
367 20 May 1994 10:59 32 D
530 26 May 1994 8:55 11 D
549 27 May 1994
548 27 May 1994
9:27 41 L
9:27 1 L
613
! jump to 530
! look for time association
! append to next event
?
? 530
? s
Associated
? aa
40
Event # 549 appended to event # 548
Appended event still present
Do you want to delete appended event(y/n=return)y
! delete appended event
Backup copy saved as: \seismo\REA\DELET\1994\05\27-0927-41L.S199405 ! del. ev. save
Deleted file
\seismo\REA\EAF__\1994\05\27-0927-41L.S199405 ! app.ev. del.
1994 05 Reading events for base EAF
612
! event list updated
# 548 27 May 1994 9:27 1 L
! jump to 222
? 222
# 222 12 May 1994 23:28 10 L
! change event type
? r
Change event type to L,R or D ?r
New file
\seismo\REA\EAF__\1994\05\12-2328-10R.S199405
Deleted file: \seismo\REA\EAF__\1994\05\12-2328-10L.S199405
Reading events for base EAF
612
# 222 12 May 1994 23:28 10 R
# 223 13 May 1994 1: 1 37 L
# 224 13 May 1994 1:16 44 L
?
?
? q
Stop - Program terminated.
***************************************************************************
When the interactive location is finished, the database should be updated, see section 6.1.
***************************************************************************
Using EEV on a subset of events or using alternative databases:
Since the EEV procedure or the HYP program will work on an index file, the user can create a subset
of his own interesting events to work with by creating his own index file with just these events. The
index file can be created by searching through the database using SELECT or it can be created
manually with the C-command in EEV.
Local database:
If data is extracted by using the COLLECT or SELECT and then split up again using SPLIT, it is
possible to keep all files in a working directory by not specifying database when splitting up. Another
simple way is to use the Copy function in EEV and copy directly from a named data base to the local
data base. Programs will then look for S-files in the current directory instead of in the database.
In addition to working with index files, there is also the possibility of storing data in different
databases. By default, the data is always stored in BER. However, the user can also create another
database structure (file structure) with another name and programs and procedures will work on that
database too. There are some restrictions: The new database, which is a subdirectory under
SEISMO/REA, just like BER, MUST have a 1-5 letter name. Currently, the alternative database is
used in our Institute to store data from other agencies like NAO, which in some cases are copied to
our own database (C-command under EEV).
The name DELET is reserved for the DELET database, which is always present.
Moment tensor inversion program, INVRAD
The program is written by John Ebel (Ebel and Bonjer, 1990) for moment tensor inversion for very
local events. The program uses instrument-corrected amplitudes of the direct (upgoing) phases of P,
SV and SH phases and makes a linear inversion for the moment tensor. The program then finds the
largest double couple component of the traceless moment tensor. For more details see file invrad.txt
in the INF directory.
The original program has been slightly modified in input and output to be integrated with EEV in
SEISAN. The steps to get the fault plane solution are:
41
Select the event from EEV
1: Plot each trace and select preferably the first clear amplitude of the direct wave. Mark the
amplitude as usual and associate the amplitude with amplitude phases AMPG or AMSG (direct
phases). This will create a separate line with amplitude readings only. The polarity must also be
indicated on a separate phase since the inversion program uses the polarity of the amplitude. The
amplitudes MUST be picked on instrument corrected traces if all instruments do not have the same
response function.
At least 5 amplitudes must be selected. S phases picked on vertical or radial components will be
considered SV while S-amplitudes picked on transverse components will be considered SH. Phases
picked on NS or EW component cannot be used. If these new phases are not to be used for location,
they can be weighted out.
2. Update event with command update to make distance and azimuths available.
3. Use command INVRAD to do the inversion. This command does several things hidden for the user:
- Creates the model input file for INVRAD called invrad.mod. This file is created from the
STATION0.HYP file, either from the current directory or DAT.
- Creates the data input file for INVRAD called invrad.inp. This file is made from the current database
file (S-file) by extracting all amplitudes associated with Pg and Sg amplitudes and converts to P, SV
or SH amplitudes in microns.
The depth of the event is taken from the S-file header and the estimated error is fixed to 0.1 micron.
- Runs the INVRAD program which produces the invrad.out file
- Reads the invrad.out file to get the fault plane solutions which overwrite the current fault plane
solution in the S-file. If you do not want to get the current solution overwritten, put a character in
column 79 on the solution, see also focmec program.
The fault plane solution can then be plotted with FOCMEC.
42
4.5.1 EEV driver program: JSEISAN
Software and manual by Bladimir Moreno
JSEISAN is a JAVA application for providing a user-friendly graphical interface of some of the
functions of the SEISAN earthquake analysis system. The program uses the functions implemented in
SEISAN by executing external commands, which mean that most of the data processing is performed
with the SEISAN software. JSEISAN is mainly a tool for formatting the input data and display the
results. The program was developed with Visual Cafe 4 Standard Edition. The software operates in
Unix and Windows environments.
Using JSEISAN
The program is started by giving command jseisan and the window (Figure 5) below will appear.
Figure 5. Main window of JSEISAN.
The user interaction with JSEISAN can be classified as: (1) Searching for earthquakes and (2)
working with a particular earthquake. It is assume that a standard SEISAN data base is available with
S-files and CAT files in the CAT directories (see SEISAN manual). The interface has a main window
where the user can search earthquakes (<Search> button) set up by a range of values (Figure 6). As
a result of the searching process, an earthquake list is shown on the screen and a set of functions is
offered. In addition, the epicenters of the earthquake list are plotted on the map. Because there is a
direct link between the map and the list, the selected earthquake is highlight on both the map and the
list when the user is navigating either through the list or through the map (see Interactive mapping
tool).
43
Configuring JSEISAN
JSEISAN has several configuration variables to be set up by the user. In order to configure the
variables, you have to click the button <Config> (Figure 5). The configuration parameters are stored
in the file “SEISAN.DEF”. The editing process of this file is performed through the JAVA program
“SEISCONF.class” (see SEISCONF). A complete description of the configuration parameters is
given in at the end of this section.
Selecting the database to work with
The desired SEISAN database is selected from the combo-box labeled as “DATA BASE” located in
the left upper part of the main window (Figure 5). The last item in the combo-box corresponds to a
SEISAN CAT file labeled as “USER FILE”. When this item is selected, the user can load a file, for
example “collect.out”, to work with.
Searching
There are two windows for setting the parameters used in the searching process: (1) A main window,
which contains the most common parameters (Figure 5) and (2) A second window (Figure 6) which
contain the rest. This second window is opened by clicking the button <More Param.>. In order to
perform the search you have to:
1. Set up the parameters (period of time, magnitude range, geographic
area, etc.) needed in the searching process.
2. Click the button <More Param.> for refining the searching process (if
needed). A dialog-box (Figure 6) for setting the additional parameters is
shown.
3. Press the <Search> button.
Figure 6. Dialog-window for setting other parameters used during search.
44
Interactive Mapping Tool
Figure 5 shows a map with the locations of the earthquakes obtained during the searching process.
The map has an equidistant cylindrical projection, which means that one unit of longitude has the
same distance to one unit of latitude. This map can be used as interactive mapping tool. For this
purpose, 4 options are included in the combo-box labeled as “Map Click Action” (Figure 5). The
options are “Select”, “Zoom in”, “Zoom out” and “Polygon”. The function of each radio-button is as
follows:
“Select”: When it is active, the user can click over the circle identifying the location of the earthquake.
The associated earthquake in the list will be highlighted.
“Zoom in”: When it is active, the user can zoom in the map by a factor of 2,4,8,16 or 32. The zoom
factor is selected from the combo-box labeled as “Zoom by”. The clicked point on the map will be the
center of the new map.
“Zoom out”: Similar to “Zoom in” but with reverse effect.
“Polygon”: When it is active, the user can define the points of a polygon by clicking on the map. The
polygon is closed when a click over the starting point is made. Once the polygon is defined, the user
can edit its points by using the refined search window, which is reach by clicking the <More Param.>
button. Epicenters inside the polygon can now be selected with the search function.
The zoomed map is taken either from Internet or locally. In the case of being retrieved locally, the
images are taken from the directory specified by the configuration variable IMGMAP_PATH. The
images are saved in sub-directories according with the level of zoom. For example in case of 6 levels
of zoom you will find the directories ZOOM0, ZOOM1, ZOOM2, ZOOM3, ZOOM4, ZOOM5 and
ZOOM6. ZOOM0 only has one image, the whole world. ZOOM1 contain 9 images, the world is
divided into 9 equal areas. ZOOM2 contain 81 images, the world is divided into 81 areas and the last
6
level of zoom (ZOOM6) contains 9 images. The image files are numbered as IMG#.gif from left-up to
right-down, like a matrix but using only one index. When the map is being retrieved from Internet
there is no zooming limitation. You can go as deep as you want. The configuration variable
MAP_SERVER controls where to retrieve the image from.
Making an epicenter map from the earthquake list
An epicenter map can be made through the SEISAN command “EPIMAP”. The steps are:
1. Press the <Epimap> button
2. A new dialog-box for the configuration of the map is shown (Figure 7), then click the <Ok> button.
45
Figure 7. Dialog-box for setting the EPIMAP input data. Only the most common EPIMAP
parameters are set.
Working with a particular earthquake
There are four available options for working with a particular earthquake of the list: (1): Plot
waveforms, (2) Edit phases, (3) Locate and (4) Update. The associated buttons are located on the
right side of the earthquake list (Figure 5).
Plotting and processing traces
Any number of channels, up to 75, can be plotted. In order to visualize a large number of channels, a
page system is implemented, when more than 12 channels are selected. For practical reason, two
operation modes are given in the plotting process: “multi-traces mode” (Figure 8) and “single-trace
mode” (Figure 9). The distinction is made mainly for choosing which channels will be plotted together
(multi-trace mode with select/unselect) and to navigate through the selected channels in more detail
(single-trace mode). This is similar to MULPLT in SEISAN. In both operational modes, picking
phases, zooming and filtering can be done. For “multi-trace mode”, a set of additional options are
available, e.g. save traces, which is applied to the plotted channels on the screen within a selected
time window.
46
Figure 8. Plotting window. Multi-Trace mode
Figure 9. Plotting window. Single-trace mode.
The steps to plot a particular earthquake selected from the list are:
1. Select a desired earthquake from the list by clicking the left button of the mouse on it (Figure 5).
2. Press the <Plot> button.
3. A new dialog-box for the selection of the channels to be plotted is shown (Figure 10), then click the
47
<Plot> button.
Figure 10. Selection of channels.
Filtering one/several channels
1. Select the low corner frequency of the Band-Pass filter (select “none” if a High-Pass filter is
desired) from the combo-box labeled as “Band-Pass filter” (Figure 8).
2. Select the high corner frequency of the Band-Pass filter (select “none” if a Low-Pass filter is
desired).
3. Make a zoom over the traces (see Zooming section) or click the button <Replot - Apply>
Instrument corrections
Converting the amplitude of one/several traces into ground velocity, displacement or acceleration and
simulating Wood-And., Mb and Ms amplitude.
1. Select the desire amplitude from the combo-box labeled “Amplitude Scale” (Figure 8).
2. Make a zoom over the traces (see Zooming section) or click the button <Replot - Apply>
Picking and removing phases
1. Press the <+>/<-> button (Figure 8) in order to select/unselect the polarity (Compression/
Dilatation) of the first impulse (if needed)
2. Press the <I>/<E> button in order to select/unselect the shape (Impulsive/Emergent ) of the first
impulse (if needed)
3. Press the button identifying the phase to be picked/removed (<P>/<S>,…)
4. Press the <Pick>/<Del> button to active the picking/deleting process
5. Move the mouse pointer to the position of the first impulse
6. Click the left button
* The previous buttons will be kept active. Furthermore you can pick the selected phase for the rest of
the channels without going through the first 4 steps.
Reading amplitude
1. Press the <Amp> button (Figure 8) to active the reading process
2. Move the mouse pointer to the position of one extreme (upper or lower) of the signal
3. Click the left button (a horizontal line will be drawn)
4. Move the mouse pointer to the position of the opposite extreme of the signal one half period away.
5. Click the left button
48
* The <Amp> button will be kept active. Furthermore you can continue reading amplitudes for the rest
of the channels without going through the first step.
Zooming
1. Move the mouse pointer to the initial time of the zooming window (the time is shown above the
right-upper-corner when you move the mouse)
2. Click the left button
3. Move the mouse pointer to the final time of the zooming window
4. Click the left button
Extracting traces from a time window
You can extract either raw or processed traces from the plotting window (Figure 8). The steps are:
1. Select/unselect (see “Other options” section) the channels to be downloaded
2. Make any desire processing (zoom, filter, etc.) if you want to save processed traces or a data
selection of the original traces.
3. Press the <DownLoad> button.
The data will now be stored in your working directory in SEISAN format in a file with name specified
by the user. Since an integer format is used, processed data with amplitudes less than 1 will have
amplitude values 0.
Other options
Switching between “single” and “multi” traces mode
1. Press the <single-trace mode>/<multi-traces mode> button (Figure 8 and Figure 9).
Show/hide channels to be plotter together on the screen
1. Press the <Replot - Reset> button (Figure 8)
2. Click the left button of the mouse over the check-box identifying each channel in order to
select/unselect the channel
3. Press the <Replot - Apply> button
Navigating through individual channels
1. Go to the “single-trace” mode (Figure 9).
2. Press the <Next Channel> button to go forward or <Previous Channel> to go backward
Navigating through pages of channels (12 channels as maximum on one page)
1. Go to the “multi-trace” mode (Figure 8)
2. Press the <Next Page> button to go forward or <Previous Page> to go backward
Editing S-files
The editing window (Figure 11) has 3 buttons for closing the window. When you click the button
<Save & Exit> the S-file is physically modified on the hard disk. The <close> button keeps the
changes in memory until another earthquake is selected on the list. So, you can still perform full
processing without making any real change in the S-file. The steps to edit the S-file are:
1. Select a desired earthquake from the list clicking the left button of the mouse on it
(Figure 5)
2. Click the <Edit> button
3. An editing window with the information is shown (Figure 11).
49
Figure 11. Editing window.
Note that this editing function do not yet, as in SEISAN EEV, check the file for correct editing.
Determining hypocenter and magnitude
1. Select a desired earthquake from the list clicking the left button of the mouse on it
2. Press the <Locate> button
3. A new window with the results is shown (Figure 12)
Figure 12. Results of the hypocenter determination.
The usual SEISAN files print.out and hyp.out are generated (see section 6.1.1).
Updating the earthquake in the database
50
1. Select a desire earthquake from the list by clicking the left button of the mouse.
2. Press the <Update> button (Figure 5)
Note: This updates the S-file, but the original CAT file in the database is NOT updated. In order to do
so, an update outside JSEISAN must be done (see SEISAN manual).
Help Function
The help system implementation is based on PDF files. When the user clicks the <HELP> button, a
list of PDF files is shown (Figure 13). The list is stored in the file “helpmenu.inp”, which is found in the
DAT directory. You can edit this file for adding new PDF files to the help system
Figure 13. Example list of PDF files used for the help system, normally in DAT.
Configuration parameters in JSEISAN
The following parameters used by JSEISAN are stored in the configuration file “SEISAN.DEF”. They
can be edited through the program SEISCONF. Most of them are defined to be used only with
JSEISAN, but some are already defined in the SEISAN configuration file: “SEISAN.DEF” (
EPIMAP_PROJECTION and EPIMAP_MAP_FILE)
OUTPUT_DIR: Output Directory for SEISAN commands results. Default “./”
EPIMAP_MAP_FILE: Name of the map file for EPIMAP command
EPIMAP_PROJECTION: Number of the projection for EPIMAP command (see SEISAN Manual)
INIT_IMGMAP_FILE: File name of the initial map represented as image. Default
“/seismo/DAT/IMGWORLD.gif”
MAP_SERVER: Type of map retrieved from Internet or locally; 0: Static local image (country
boundaries), 1: Static remote image (country boundaries), 2,3,4,5: Dynamic Remote image (2:country
boundaries, 3:Relief from GTOPO30 only land, 4:Two minute shaded relief, 5: Combine 3 and 4).
Default 0.
IMGMAP_PATH: PATH for the static local maps (images) stored in the local hard disk used for
zooming. Default “/seismo/DAT/IMGMAP”
INIT_MAP_LOWER_LATITUDE: Lower latitude of the initial map. Default –90.0
INIT_MAP_UPPER_LATITUDE: Upper latitude of the initial map. Default 90.0
51
INIT_MAP_LEFT_LONGITUDE: Lower longitude of the initial map. Default –180.0
INIT_MAP_RIGHT_LONGITUDE: Upper longitude of the initial map. Default 180.0
ACROBAT_READER: Path for the Adobe Acrobat Reader (needed for the help files). Default
“/prog/acroread” for Unix.
INTERNET_BROWSER: Location and place of browser
HELP_DIR: Directory of help files, usually INF
4.5.2 EEV Windows driver program: SEISAN
The program is an alternative to the standard EEV and it has all the functions of EEV. The main difference
compared to EEV is that is has a 'Windows type' selection of events in the database and that the most
used commands in EEV and SEISAN can be executed by pressing a Button. The intention is that the
majority of routine tasks in SEISAN can be done within the W95 interface without learning all the SEISAN
prompt line commands.
Starting SEISAN Windows
When Windows is running, SEISAN can be started by clicking on the SEISAN icon if installed (see section
3) or writing SEISAN on the prompt line. SEISAN will start up and show a figure as shown in Figure 14. In
addition to the main SEISAN window, there will also be a console window used for input and output since
all underlying programs are started from the prompt line.
Working directory
Most programs read and write to the current working directory. The name of the working directory is
displayed on the bottom of the screen. To change the working directory, press file selection at the top lefthand corner.
52
Figure 14. Windows SEISAN display
Database selection
When SEISAN starts up, it will start with the database used when it last was closed. Other databases can
be selected with the 'Database Selection' menu, which also displays the current database. The choices
are
1: One of the 1-5 letters databases already in existence.
New databases are created as usual with MAKEREA on the prompt line.
2: A local database in the current working directory. The current working directory is displayed on the
bottom of the SEISAN screen. To change the working directory use file selection at the top left hand
corner.
3: An index file. The file name is selected on the menu 'Selected Index File'
Year and month selection: If a 1-3 letter database has been selected, the years and months available are
seen under year and month buttons and can be selected there.
Selecting an event
Once the database has been selected, SEISAN will work much like EEV. The event window will show 12
events with the same information as seen in EEV. The total number of events for the month is shown
above the event selection box. The first event in the list will be the current event. Any other event on the
list can become the active event by clicking on it and it will be highlighted. Events outside the window can
53
be displayed using the scroll bar. In addition, all EEV commands can be used including event selection
commands. This works exactly as in EEV. Write e.g. 22 anywhere on the screen, press return (or click on
execute button) and event 22 will be highlighted.
Commands
All commands from EEV can be used and they are used like in EEV. Typing e.g. 'l' and return (or click on
execute button), will locate the event. While typing the command, it will appear in the 'EEV Command
Line' window. The command can be edited and a command can be repeated by just pressing the
'Execute' button or hitting 'return' again. However, the 4 most used EEV commands can also be executed
by clicking a command button:
Type: Will display the content of the S-file, same as EEV 'T'
Edit: Edit the S-file, same as EEV command 'E'
Plot: Plot the traces, same as EEV command 'P'.
Locate: Hypocenter location, same as EEV command 'L'
In addition it is possible to display the S-file header line by double clicking on the active event. This
corresponds to EEV command 'TT'.
Program output and interaction
Since all programs started by SEISAN are console based programs, the screen output and input will
appear on the console window. The console window will come in the foreground if data is output or input
is required. As soon as the action stops, the SEISAN window comes back to the foreground. With a large
screen resolution, it is possible to see both windows at the same time. It is also possible to switch
between the two windows by clicking on the 'View Console' button.
Access to the DOS prompt
Since all programs under SEISAN run in the prompt mode, it is often practical to get a DOS window on
the screen. The 'DOS Prompt' button will open a DOS window in the current working directory from which
SEISAN or other program can be executed. On NT, the equivalent is a console window.
Other programs
UPDATE, SELECT, PROMAC and EPIMAP can also be started from SEISAN by clicking a button. These
programs have been selected since they are often used in routine operation.
4.6 System response
The instrument response can be defined for each channel of digital data in either SEISAN, GSE or SEED
response format. There are three places in the system where it can be stored. Often the instrument
response is part of each channel header in the digital waveform file in SEISAN waveform format (see the
Appendix 2 for format description). However, the instrument response is often not available at the moment
the data arrives, or it is later discovered that the response given in the waveform file is wrong. There is
therefore by default a directory CAL that contains one response file for each channel and for each date
from which it is valid. Since the filenames contain the date from which a change in the response was
made and the channel code and component code, a directory listing of CAL will give the history in
chronological order of the response of a given channel. This is the most common way to use the response
information in SEISAN.
Response information can also be kept in any other directory specified with the environmental variable
LOCAL_CAL. The variable must be set with the full path to the directory e.g. /home/seismo/WOR/test/, or
54
on PC, \seismo\new\cal\. On Sun it can be set in the .SEISAN file and on PC in the autoexec.bat. The
variable can also be set from the keyboard (Sun: 'setenv LOCAL_CAL directory', PC, 'set
LOCAL_CAL=directory'. This is a useful option when testing response files.
The response information gives the gain of a channel in counts/m and to get the correct ground
displacement, the count values must be divided by the response values.
In the current SEISAN system, only the programs MULPLT, WAVETOOL and SPEC use the response
information when doing spectral analysis, generating Wood Anderson or ground motion traces. The
programs will look first in the CAL (or alternative) directory for a valid response file and if not found there
use the header information in the waveform file. A message will be given if the file header information is
used.
If waveform files are generated on the SEISAN system from raw field station files or other input files
without response information, the conversion programs (e.g. QNXSEI from a SEISLOG QNX system) will
look in the CAL (or alternative) directory to find the response information to include with the Seisan
waveform file. The response will be only put into the SEISAN waveform file, if the response is stored in
SEISAN format. The response files are generated with RESP, see below and Appendix 3.
The RESP program (section 6.31) can be used to generate the response files. The filenames for the
response files are STATTCOMP.YYYY-MM-DD-hhmm_FOR where STATT is station code, COMP is
component, YYYY is year, MM is month, DD is day, hh is hour, mm is minute and FOR is the format
indicator which can be SEI or GSE. If FOR is not given, the format is SEISAN. An example is
BER__S__Z.1999-05-05-1244.. You should take a backup of the response files before you run the
program (see section 3).
The response files can be located in CAL, or, if many files are available optionally also in a subdirectory
structure. This optional structure simply consist of a subdirectory for each station and the subdirectory
name must have 5 letters so base BER would have the name BER__. The system automatically locates
the response files whether all are in CAL or in the subdirectory structure.
The response file can store the response in different ways:
1 SEISAN format:
a: Parameters used for calculating the response: Generator constant, filters etc. In addition, the response
(amplitude and phase) at 30 frequencies is listed. In this case the response is calculated from the
parameters.
b: Incomplete set of parameters or no parameters and the response at 30 frequencies. In this case the
response is calculated by interpolation of the 30 values.
c: Poles and zeros: No discrete values are given and the response is calculated directly from the poles
and zeros. The number of poles and zeros in the SEISAN format is limited.
2 GSE CAL2 format:
a: Poles and zeros, number is unlimited, the response is directly calculated from the poles and zeros.
b: Pairs of frequency amplitude phase, number of pairs is unlimited, the response is calculated by
interpolation.
3 SEED
Poles and zeros (only Transfer function type A (Laplace Transform (Rad/sec)), number of poles and zeros
are unlimited, the response is directly calculated from the poles and zeros. Only reads SEED response in
ASCII format as written out by rdseed, not dataless SEED volumes. Standard filename such as
‘RESP.IU.TRIS.10.BHZ’ are understood. Files are read from CAL directory.
When rotating signals, it is assumed that the response is the same on all 3 channels. !!!!
Response files can be plotted from MULPLT showing the actual response information that is used with a
given trace. Response files can also be plotted directly with program PRESP, see below.
55
All or a subset of the response files can be printed out in a table with program PR_RESP. The program
must be executed from the directory with the response files. Make a listing (file filenr.lis) of files to print out
with DIRF and run the program. It will produce an output file ready for printing.
A response file can be plotted with the program PRESP. The program is started with command presp
filename, where filename is the response file name. If no file name is given, the program asks for a
filename or number. If a DIRF has been made and the list of files in filenr.lis is available, a response file
can then be selected with a number. The program produces a PostScript output file with name presp.eps.
4.7 Working with catalogs
It is often convenient to have multiple solutions of hypocenters in the database S-files or the CAT-files.
Typically data has been entered from different sources and merged to form a single catalog. The first
hypocenter line in the file is then considered the prime hypocenter estimate and this is the one used by
e.g. EPIMAP to plot the hypocenters. The order of the hypocenters can be rearranged by CAT_AGA.
Several programs use all the hypocenter lines. The magnitude correlation program will search any
hypocenter line and the database selection program SELECT will optionally also use all the hypocenter
lines. When the data base is updated with a new location and magnitude (UPDATE, section 6.7), it is only
the first hypocenter line which is overwritten. If there is a magnitude in the 3 position, it is left unchanged
unless it has the same agency as used for updating. This is useful in normal observatory practice, where it
is common to put in some external agency magnitude which then must be left unchanged. If more
magnitudes than 3 are calculated, they will be placed on a subsequent hypocenter line identified by
having the same year, month, day and hypocenter agency as given on the first line.
In order to merge different catalogs, it might be an advantage to put all the data into a complete database
where each event is one file, even when only hypocenters are available. This is done by first splitting up
the catalogs with SPLIT and then using EEV to merge the events. Since there is no requirement for
monthly directories to have data, this methodology can also work for historical catalogs. The data can then
subsequently be put into the CAT database without relocation using the UPD command.
4.7.1 Explosions in SEISAN
Many catalogs are contaminated by explosions and in SEISAN, explosions can be dealt with in
several ways. In the data base, confirmed explosion are marked with E and probable explosions with
P. These indicators are mostly put in when the operator first registers the event. However, there is
also a possibility to automatically identify events which are probable explosions. This is done with
program EXFILTER (section 6.26). In the data base S-files, there is a special format for recording
explosion information (command EXP in EEV). The explosion site there can be assigned a three letter
code, which can be used by SELECT to find explosions from specific sites. In this format it is also
possible to store the explosion charge and explosion location and origin time separately from the
calculated location and origin time.
4.8 Printing
All SEISAN programs, that produce graphical output, also generate Postscript files with the file suffix eps
(note this was plt before version 8.1). These can be directly sent to a Postscript printer. It seems that
programs like Microsoft Word don’t like the SEISAN Postscript and you will need to convert your files to
another Postscript, this can be done for example with the program ghostscript using pswrite as output
device.
Note: On Solaris 7, both the lpr and the lp command for sending files to the printer, don’t create a copy of
the file before sending it (bug in Solaris). This means that a plot file can be overwritten before being sent
to the printer. Therefore when SEISAN on Unix is sending plots, the system waits for 5 seconds after a file
is sent to the plotter before continuing. This is most important when plotting continuous data or a large
number of files with MULPLT.
56
4.9 General Work with SEISAN
Once data is in the database and the routine analysis has been finished by running UPDATE (final
epicenters recorded in CAT and the S-files), it is possible to go on with general work with the data. This
means searching the database, making a bulletin or plotting the epicenters. It is also possible to use some
of the more specialized tools of SEISAN which include working on subsets of data or creating other
databases, see 4.4.
For general use, the basic philosophy is that the user should not enter the REA directories. All commands
and programs should be used from the user's own directory or the WOR directory. To access part of the
main database, the programs always ask for start and end date as follows:
19880602011001
198806020110
1988060201
19880602
198806
1988
BLANK
:
:
:
:
:
:
:
including
including
including
including
including
including
only used
from or to the second
from or to the minute
from or to the hour
from or to the day
from or to the month
from or to the year
as end date, means to end of month
Note that the end time is inclusive, this means that e.g. 198806 includes all of June 1988.
Thus most programs will work from any given date-time to any other given date-time. Programs that work
directly on the S-files in the database (e.g. COLLECT) can work with any time interval in which the
database structure has been created. THERE IS NO REQUIREMENT THAT THERE IS DATA IN THE
INDIVIDUAL MONTHLY DIRECTORIES, ONLY THAT THEY EXIST.
There are usually 4 options for database, either the standard base (often by default), the user's own
subset of the standard base (an INDEX file or S-files in local directory) or another database. If the user
has his own database specified by an INDEX file, the event ID's must be in that INDEX file. Since the
index file gives complete file name of event files, the index file can work on a subset of the main database.
Note that most of the programs are used as stand-alone programs, disregarding the database structure. If
one for example prefers to have all events gathered in one file rather than split into many files and
directories most programs will therefore work.
4.10 Graphics in SEISAN
Most programs in SEISAN producing graphics on the screen use the SEISAN graphics system (see
also section 7). This produces fast and low quality graphics both on the screen and similar PostScript
output files. Most of these plots are not suitable for publication and many programs therefore also
create output ASCII files of the main results, which then can be put into more professional plotting
routines. The GMT (Generic mapping Tools) system is one of the more widely used plotting systems
used in seismology. Several programs in SEISAN therefore produce output that can be used with
GMT or makes plots directly in GMT. From version 8.0 of SEISAN, a script is included (GMTXY,
manual in INF) which will produce nice xy-graphics from specially made output files. So far only
programs SPEC, CATSTAT and LSQ produce these output files, the intention is to include this
feature in more SEISAN programs.
If there is a need to produce better quality graphics there are several possibilities:
Maps: GMTMAP, Unix (must be installed separately, on CD
W_EMAP: Windows based mapping system
Seismograms: TRACEPLOT (GMT based)
XY-plots:
GMTXY (GMT based)
Maps: create GMT input files with SEIGMT
Volcanic event distribution: VOLCSTAT
57
5. INTRODUCTORY TRAINING COURSE
The document ‘SEISAN Version 8.0 - introductory training’ which is a tutorial for new users, is
included in the distribution. The testdata used in the exercises need to be installed, see section 3.
Going through the exercises of the tutorial might be the best way to learn SEISAN. The exercises are
based on the testdata that are distributed with the full version of SEISAN. The document is given as
PDF file (seitrain.pdf) in the INF directory.
The main goal of the introductory training course is to become familiar with the database program
EEV, the plotting program MULPLT and the location program HYP. Of course additional reading of
relevant sections in this manual is required. The exercises can be completed within one or two days.
Overview of the Exercises:
1) Interactive work with the database using EEV
2) Selecting data from the database and making an epicenter map
3) Putting new data into the database
4) Plotting digital data
5) Phase picking
6) Magnitude determination
7) Fault plane solution
8) Spectral analysis
9) Reading global phases using IASPEI91 tables
10) Waveform formats in SEISAN
11) Response in SEISAN
12) Additional exercises in earthquake location
58
6. DESCRIPTION OF PROGRAMS AND COMMANDS
This section gives user manuals for programs and command procedures used with SEISAN. Not all are
as detailed as one could want, however many questions from programs should be self-explanatory. Most
programs will produce output files with the extension .out and proceeding by the name of the program.
E.g. output from collect, will be collect.out. Running a program twice will erase the earlier output files. If
these files are to be used later, remember to rename them before running a program again. There are
several programs, which have separate manuals in the INF directory.
6.1 Hypocenter location programs: HYPOCENTER, HYPO71 and HYPOINVERSE
6.1.1 The hypocenter program, HYP
The hypocenter program is a modified version of HYPOCENTER (Lienert et al., 1986, Lienert, 1991,
Lienert and Havskov, 1995). The main modifications are that it can accept more phases, locate
teleseismic events and use input in Nordic format directly from the database. A detailed manual (earlier
version, hypocent.pdf) and some of the later changes (hypocent_latest.pdf) is given in INF directory. The
input parameter file with station coordinates, model etc is STATION0.HYP, see later.
Local crustal phases:
The program will accept P, Pg, Pn, S, Sg, Sn, Pb, Sb, Rg, T and Lg phases and when locating teleseismic
events most of the IASPEI phases (see below). If only P or S is given, the fastest phase is used as in the
original version of the program.
Azimuth:
The program also uses observed station azimuths as given in the Nordic Format. Station azimuths can be
obtained with either 3-component stations or array stations. This means that the program can locate with
one station if it has P, S and azimuth. Azimuth residuals contribute to the overall rms, see TEST(52) and
section on weight. In order to locate with one station, azimuth and P and S, TEST(56) MUST be set to 1.
Know problem: When locating with one station and very short distance, the location does not work if
hypocentral depth is zero, fix to some non-zero depth.
Magnitudes:
Magnitudes are calculated using coda, amplitude and spectral level. Parameters are given in the station
file using the RESET TEST variables. For magnitude based on amplitude, the amplitude must be given in
nanometers in the input file (SEISAN standard).
The formula used to calculate local magnitude is
Ml = a * log(amp) + b*log(dist) + c*dist + d
where a,b,c,d are constants, log is logarithm to the base 10, amp is maximum ground amplitude (zeropeak) in nm and dist is hypocentral distance in km (RESET TEST 75-78). The default constants are for
California (Hutton and Boore, 1987) which gives the following relation
Ml = log(amp) + 1.11 log(dist) + 0.00189 dist - 2.09
Local magnitudes are only calculated for events with epicentral distance LESS THAN TEST(57)
(default 1500 km) and if the period is less than 5.0 secs. All amplitudes for the phases L, ‘S ‘ Sg, SG,
59
AMP, and AML, AMPL or blank are used. This means that if an amplitude is picked on both Lg and Sg,
both will be used. The period is not used. The many possible phase names is a result of changes over
time and thus to ensure that Ml is calculated correctly with older data. The current version of MULPLT
produces the standard name AML.
The coda magnitude is calculated using
Mc = a * log(coda) + b * dist + c
where coda is coda length in secs and a,b and are constants (RESET TEST 7-9). If a is given as a
negative number, the following formula will be used
Mc = abs(a)*log(coda)*log(coda) + b * dist + c
Coda magnitude is only calculated if the epicentral distance is less than TEST(57).
Ms is calculated using the standard
Ms = log(amp/T)+1.66log(dist)+3.3
where T is period. Amplitude is in micrometer and distance in degrees, however in the Nordic format nm
and km are used and the program converts. Ms is only calculated if the period is larger than 10.0 seconds
in which case the program automatically assumes that Ms is the wanted magnitude. The phase used can
be AMS, AMP or blank. The current version of MULPLT produces the standard name AMS.
Mb is calculated using
Mb = log(amp/T) + Q(dist,depth)
where Q is a hardwired function of distance and depth (Veith and Clawson, 1972). Mb is only calculated
if the epicentral distance is less than 100 degrees and the period is in the range 0.2 - 5 secs and
either the phase is P, AMP, AMb or the phase is blank and the epicentral distance is larger than
TEST(57). The current version of MULPLT produces the standard name AMb
Mw is calculated as
Mw = 2/3 * log10(moment) - 6.06
where moment is in Nm (see also section 6.2.11). When an event is relocated, the moment is also
recalculated according to revised hypocentral distance
NOTE: If an amplitude has a given period between 5 and 10 secs, it is not used for magnitude calculation,
see above. If an event is not located, there will normally be no magnitude calculation and all magnitude
and distance information is deleted from the output S-file (hyp.out) except, the magnitude in the 3 position
on the header line if it has an agency different from the default agency. The only exception is that if a coda
is given, the epicentral distance is retained and coda magnitude will therefore be calculated. This means
that for events, which cannot be located, it is still possible to calculate coda magnitudes by manually
entering the epicentral distance on the line containing the coda length.
On the first header line, there is room for 3 magnitudes. If there is a magnitude in the 3 position, it is not
overwritten, so there will only be room for 2 magnitudes on the first header line. If more magnitudes are
calculated, they will be written on a subsequent hypocenter line, which is identified by having the same
year, month, day and hypocenter agency as the first header line. This means that there is room for a total
of 6 magnitudes, which can each, be updated when relocating. Hypocenter info and all 6 magnitudes can
be printed out on one line with program REPORT.
All magnitudes can have a station dependent correction given in the station file. This correction does not
affect the Mc in the print.out file.
Only calculate magnitude: If TEST(106) is set to 1.0, only magnitudes are calculated, provided a distance
60
is given.
Use of S-P and L-S differences:
Uncertainty in absolute times often makes it necessary to be able to use the difference in time between
two arrivals such as P and S or P and L. If no absolute times are available, the calculated origin time will
be close to that at the first arrival station and is of course meaningless. However, a perfectly good
epicenter and depth can still be obtained from P-S or P-L differences alone. To enable this feature, set the
weight for the P phase input record to 9. This P is then assigned a weight of 0, effectively disabling its use.
However, a time residual and azimuth, etc., will still be calculated for it, enabling an assessment to be
made of its absolute time. A search will then be made of the entire input phase set for an S or L phase at
the same station. If such a phase is found, its variables are used to store the observed and calculated
difference times and their derivatives, and it’s weight (0-4) is used for the difference phase. DON'T SET IT
TO 9!! If two or more such phases (e.g., SN, SG, LG, etc.) are found, all their differences with the P time
will be used instead of their absolute times. Blanks will appear beneath 'hrmn' in the residual summary for
all such phases, while the observed and calculated difference times with the first P will appear beneath
't-obs' and 't-cal'.
NB. There must be at least one phase with absolute time to get a location.
Global event location:
When locating globally, the program uses the IASPEI91 travel time software described by Buland and
Chapman (1983) and Kennett and Engdahl (1991).
HYP evaluates all the IASPEI91 phases (up to 60) at each delta, and searches for the phase specified in
the 4-character phase identifier. If no phase is found, the phase is given a weight of -1, which effectively
removes it from the phase set. If a phase is labeled as 'P ', 'S ', 'PKP ' or 'SKS ', and this phase is not in
the IASPEI91 list, the first arrival phase having P or S as its first letter is used, or PKP, SKS as its first 3
letters. In addition, include the PKiK phases in this search for 'PKP ' and 'SKiK' phases in the search for
'SKP '.
The IASPEI91 phase set currently includes: P, Pdiff, PKP, PKiKP, pP, pPdiff, pPKP, pPKiKP, Sp, sPdiff,
sPKP, sPKiKP, PP, P'P', S, Sdiff, SKS, sP, pSdiff, pSKS, Ss, sSdiff, sSKS, SS, S'S', PS, PKS, SP, SKP,
SKiKP, PcP, PcS, ScP, ScS, PKKP, PKKS,
SKKP, and SKKS.
Long phase names:
Normally SEISAN and the Nordic format assume up to 4 character phase names. However, when
working with global phases, the phase name length can in a few cases be up to the ISC standard of 8
characters. The program then uses column 9 for weight (normally blank) and column 11-18 for the phase.
In this case it is not possible to give a polarity.
Criteria for a Solution:
The cases where a solution will not be attempted are as follows:
1. Multiple phases at two stations, but no azimuths. This is a non-unique case, even though four different
arrivals are present.
2. Less than three phases from three different stations and no azimuths.
3. A single phase at one station with an azimuth.
Note that if phases are weighted out due to large distance or a bad fit during the first iteration, there might
not be a location even if more than 3 stations are available.
Weighting:
61
A number of different weights may be used to calculate the solution.
1. User specified weights: These are calculated using the HYPO71 style weight number 0 to 4, read with
each phase, where 0 corresponds to w1=1.0, 1 to w1=0.75, 2 to w1=0.5, 3 to w1=0.25 and 4 to w1=0.
Uncertain time is 9 meaning that absolute time is not used, see also use of S-P times on previous page.
2. Distance weighting: This is given by the formula w2=(xfar-delta)(xfar-xnear) where delta is the distance
(km) of the event from the station and xnear and xfar are read from the station file, STATION0.HYP.
3. Bisquare weighting: This scheme, described by Anderson (1982) calculates residual weights, see
details in HYP manual. Used for distant events.
4. Azimuth weighting: Azimuth residuals are divided by test(52), which is the error in azimuth that
corresponds to a one-second error in arrival time. For example, if test(52)=5 (default), a phase residual of
5 degrees will become a residual of 1 (5/test(52)) in the parameter corrections and rms calculation.
All the above weights are multiplied together to calculate the weight used in the inversion. If the userspecified weight, w1, is changed by (2) or (3) above, changed to zero by the consistency check, or set to
-1 because the phase is not recognized, an asterisk will appear after the final weight in the residual
printout.
Determining which travel time software is used:
The parameter test(57) is used to determine whether a layered model or IASPEI91 software is used to
calculate the travel times and their derivatives. For the initial starting location, the distances from each
station are calculated and IASPEI91 is used if any of them exceed test(57). However, this can be
overridden by the distance indicator in column 22 of the Nordic header record. If this is L, a crustal model
is used regardless of distance, whereas if it is D, IASPEI91 is used, while R has no effect i.e., test(57) is
still used. So if either a crustal model or IASPEI91 tables are wanted, use either L or D respectively.
Starting epicenter location:
The program uses a starting location algorithm (reset test(56)) which tests the rms of all starting locations
and select the minimum rms solution, see HYP manual.
User defined start location: If an S is written in the input S-file at column 45 of the epicenter line, the
location starts at the location (epicenter) given on the header line. If an S is written in column 44 on
header line, the depth iteration will start at depth given on the header line. If N is written in column 45, the
nearest station will be used irrespective of global settings.
Starting depth:
If no event specific start depth is given in S-file, the starting depth is taken from the first number on the
control line (see later) in the HYPO71 style. However, there is often problems obtaining a reliable depth
due to local minima. This can be manually checked with program RMSDEP from EEV. HYP can also be
set up to locate the same event starting with a range of different start depths, and then choose the one
with the lowest RMS. This can significantly improve the reliability of depth determination. Selecting 3 to 5
different start depth is often enough. This option is set on the control line in the station file.
Fixing location:
Using F instead of S, fixes the position (depth and location)..
Do not locate event:
62
If a * is written in column 45, the event is not located, can be used if an external location is to be kept
unchanged
Fixing origin time:
Using an F in column 11 of header line will fix the origin time given on the header line.
If both depth and location are fixed, but not the origin time, new origin time and residuals will be
calculated. This can be useful when working with readings from a few stations which should be checked
against known locations. If e.g. distant events are read, it is often the practice to put in the PDE location on
the header line and calculate residuals relative to the observations. When the UPDATE is made, the
agency of the location is NOT changed, assuming that if both depth and epicenter are fixed, the
hypocenter must come from an external agency.
Alternative model:
By default, an event is located using the STATION0.HYP input file. However, each event can use its own
model (with all the location parameters) which is specified with one character in column 21 on the Nordic
input file header line. The model then has a corresponding name. If e.g. the model is called W, the
corresponding input station file will be called STATIONW.HYP. It is therefore possible to have as many
different station files, as there are printable characters. Note that if a different model x has been specified
and is not present, the program will stop with the message "STATIONx.HYP does not exist". The file
MODEL.DEF in DAT can be used to assign the single character a name, which can be listed from EEV.
The format in MODEL.DEF is one line per model, the model indicator is given in column 1, column 2 is
blank and the model name is given in columns 3 to 80.
Using HYP to determine crustal structure
HYP has an option to locate a data set for a large number of different models and then determined which
model gives the lowest average RMS for the data set. This might be a useful option, particularly when a
sparse data set is available. In order to use this option, an additional input parameter file h_model.par is
given. When this file is in the working directory, HYP will switch to multiple model mode SO ONLY HAVE
THIS FILE IN WORKING DIRECTORY IF MULTIPLE MODEL MODE IS INTENDED. When using this
option, all events must use the same STATIONx.HYP file, otherwise the program fails. The input
MUST be from a single file, NOT from the data base. Below is an example of an input file
layer #
1
2
3
4
4
start vp
4.55
6.3
6.8
7.90
8.05
delta vp
0.1
0.1
0.1
0.05
0.05
# delta
5
5
5
3
4
start h
0.0
4.0
22.5
32.5
40.0
delta h
1.0
1.0
1.0
1.0
1.0
# delta
1
1
1
1
1
The first line is info only. Layer # is also only for information. For each layer, there is a start P-velocity
(start vp), increment in velocity (delta vp) and number of increments (# delta). The following inputs are
then the same for layer depths. There must be an entry for each layer even if no variation is used. In
the above example, no variation in layer thickness is tested for. An example input file is given in DAT.
The parameters for location not set in h_model.par like Vp/Vs, Lg velocity etc remain unchanged.
When HYP starts up, it will print out how many permutations are required. If more than a few
thousand, reduce the number of models. In any case it is an advantage to first try with just a few
models to get a feeling for how sensitive the data is for model changes.
An output file h_models.out is generated, see example below. For each model tested, one output line
is given with the RMS and the model. In the example below only the last 5 models are shown. Since
many models can have very similar average RMS, the best 10 models are printed at the end.
63
0.946
2.607
0.934
0.994
2.677
4.95
4.95
4.95
4.95
4.95
Minimum rms
0.00
0.00
0.00
0.00
0.00
6.70
6.70
6.70
6.70
6.70
4.00
4.00
4.00
4.00
4.00
7.20
7.20
7.20
7.20
7.20
24.50
24.50
24.50
24.50
24.50
7.90
7.90
8.00
8.00
8.00
32.00
32.00
32.00
32.00
32.00
8.20
8.30
8.10
8.20
8.30
40.00
40.00
40.00
40.00
40.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
7.10
7.00
7.00
7.10
7.10
7.20
6.90
7.00
7.00
7.00
24.50
23.50
24.50
23.50
24.50
24.50
22.50
22.50
23.50
24.50
7.80
7.80
7.80
7.80
7.80
7.80
7.80
7.80
7.80
7.80
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
8.10
8.10
8.10
8.10
8.10
8.10
8.10
8.10
8.10
8.10
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
0.764057
The best models
0.771
0.766
0.767
0.769
0.766
0.772
0.771
0.771
0.770
0.771
4.95
4.85
4.85
4.85
4.85
4.85
4.95
4.95
4.95
4.95
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
6.50
6.50
6.50
6.50
6.50
6.50
6.50
6.50
6.50
6.50
Running HYP:
The program is started with command HYP from the prompt line (interactive mode) or with 'L' in EEV.
HYP can also be started with an argument like ‘hyp input.dat’, where input .dat is an S-file. The first event
in the S-file will then be located without further user interaction. Below follows an example of running
outside EEV, explanations are in lower case. Note that the STATION0.HYP file MUST be present in the
DAT directory for HYP to know that it is working with a SEISAN database. If not present, HYP will only ask
for an input file name, see HYP manual.
HYP
Arrival time data input, select one:
SEISAN database or
: RETURN
Alternative database, give 1-5 letter code
:
Local index file, name must start with index or :
Local database, write ,, or
:
File name for one file in NORDIC format
:
Your answer here determines the input
source. A return means that you work directly on the BER database. A 1-5 letter
code gives name of database, e.g. NAO. An index file or the name of a readings
file is used when you want to work on specific subsets.
Local database is S-files in local directory.
Start Time
(YYYYMMDDHHMMSS) : 199012
End Time, RETURN is to end of month
: 19901205
Standard formatted time input.
Interactive operation (N/Y=return)
If N, whole time interval or file is located, one line output
pr event.
# 1
1992 12 3 0137 40.3 NPHS=
12 T Q L #XXX
# 2
1992 12 3 0237 43.3 NPHS=
14 T Q L #XXX l
! now locate
here comes location, see HYP manual*****************************
# 2
1992 12 3 0237 43.3 NPHS=
14 T Q L #XXX q
! stop
PRINT OUTPUT IN FILE print.out
CAT-FILE IN FILE hyp.out
Summary file in hypsum.out
In interactive mode, as shown above, event date is printed out for each event and action is taken as in
EEV for the options available. If HYP run on a single file, the options above are available meaning that
HYP can select and locate different events in a single file using the event number. If HYP runs on a
64
database, the EEV options D and B are also available, but not shown. If the option of no interactive input
is chosen, the program will locate from beginning to end without any more user interaction. This is a
useful option for testing a subset of the database with different models etc. without changing the
database. Note that the input file or database is never overwritten by HYP.
ALL TYPE ONE LINES WITH SAME AGENCY AS GIVEN IN STATIONX.HYP FILE WILL BE DELETED
SO THERE WILL NEVER BE MORE THAN ONE TYPE 1 LINE IN OUTPUT WITH CURRENT AGENCY
(except possibly a second magnitude line with a different type magnitude as given on main header line).
Problems: Sometimes HYP will not locate an event, look in the print.out file to see what happened. In
some cases, the initial location was put beyond the limits set by the parameters. If e.g. an event is defined
as a local event and no readings are to be used further away than 2000 km (distance weighting, see
following table or TEST(41)) then no location will be attempted. Try to change the event type to D and see
if the event locates. In a few other cases it might be an advantage to use a starting location.
Station and model files:
Station input is given in near standard HYPO71 format in the file STATION0.HYP in directory DAT. If
however the user wants to try a different model without changing the standard model in DAT, this is
possible by having a STATION0.HYP file in the working directory, since the program always looks there
first for the STATION0.HYP file (see example at end of this section). Another possibility is to use another
model for just one event by setting a flag in the phase input file, see below.
Below is an example of a STATION0.HYP file. The format is close to the HYPO71 format with one extra
line at the bottom. The test parameters 2-13 are as in HYPO71, see also HYPOCENTER manual section
4.1.2.
Comments are given after !'s
RESET
RESET
RESET
RESET
RESET
RESET
RESET
RESET
RESET
TEST(01)=0.3
TEST(03)=0.6
TEST(06)=0.1
TEST(07)=-3.0
TEST(08)=2.6
TEST(09)=0.001
TEST(11)=50.0
TEST(13)=5.0
TEST(50)=1.0
UPP 5951.50N
COP 5541.00N
KBS 7855.08N
EBH 5614890N
OSG 6029.80N
01A06049.43N
BERGE6057.12N
-BEBGE6157.12N
...
1737.60E 14
1226.00E 13
1155.44E 46
330490W 375
252.55E-100
1049.95E 426
1133.15E 100
1133.15E1100
6.2
0.0
6.6
12.0
7.1
23.0
8.05
31.0
N
8.25
50.0
8.5
80.0
15. 600. 1300. 1.73
BER
!
! one and only one blank line here
! station lines
! high accuracy lat-lon
! 5 char station name
! 5 char station name and at –1100 m
! one and only one blank line here
! model lines
3.8
2.2
200.0
300.0 **
! N indicates location of Moho
!
5 5.0 10.0
! control parameters
! Reporting agency
(a3)
Format of the station line is 2x,a4,i2,f5.3,a1,i3,f5.3,a1,i4,f6.2,5f5.2,9f6.2 or 1x,a5 .... if the station has 5
characters. The content is:
station code 4-5 chars (see above)
latitude in degrees
latitude in min
north or south (N or S)
65
longitude in degrees
longitude in minutes
east or west (E or W)
altitude in m, in some rare cases, the station is deeper than 1000 m in which case the minus sign has to
be put in column 1
P-delay in secs, S-delay is the same multiplied by Vp/Vs as given below
Magnitude corrections for 5 magnitudes: Mc, Ml, Mb, Ms and Mw
Spherical harmonic station corrections
The magnitude residuals are added to magnitudes calculated for each station but the result is only seen in
the final average magnitude. If the magnitude correction is set to 99.0, the magnitude is not used in the
average.
Format of model line: 3f7.3,a1. The information is:
P- velocity (km/sec)
Depth to interface (km)
S- velocity (not needed)
Interface indicator: N: Moho, B: Conrad
NB: Moho cannot be the last layer, there MUST be one layer below interface marked with N.
The line with ** indicates optional Vs, density, Qp and Qs. This is information only used with modeling,
see section 6.19. Format for additional info is 25x,4f10.1.
Format of control line: 3f5.0,f5.2,i5,2f5.1 Information is:
start depth in km, used if no range of start depths specified (see below)
xnear: distance at which distance weighting start
xfar: distance at which distance weighting is zero, beyond xfar, the phase is not used (local events only)
Vp/Vs ratio
number of start depths
start depth of range of start depths
increment in start depths
NB: If these parameters are used, the fixed initial start depth is not used
The input at the bottom is reporting agency used for both hypocenter and magnitudes.
Since the program locates distant events, max distance, reset test(41) must be set to a large value. To
avoid that local events move out in the blue, the parameters xnear and xfar must be set not larger than
2000 to 3000 km. Xnear and xfar are only used for local events (flag L) and regional events if the local
crustal model is used.
RESET TEST parameters:
HYP will assign reasonable default values for RESET TEST parameter. Below is shown a summary. For
full details see HYP manual. The number to the left is the control parameter and D indicates the default
value.
2:
7-9:
11:
13:
30:
Step length damping control, D: 500.0.
Duration magnitude coefficients used for calculating the coda magnitude, as MAG = TEST(7) +
TEST(8) * LOG(T) + TEST(9) * DELTA where
T is the coda length in seconds, DELTA is the hypocentral distance in km.
D: 7: 0.087, 8: 2.0, 9: 0.0035 (Lee, 1972)
If test(8) is negative, its positive value will be used and log(T) will be squared. Note however, that
the individual stations magnitude values printed out during the run of HYP still will be using the
unsquared log(T).
Maximum no of iterations in the least-squares rms minimization, D: 99.0
Increment in km for auxiliary rms, D: 20.0 km.
Initial damping factor, D: 0.005
66
31:
32:
34:
35:
36:
37:
38:
39:
40:
41:
43:
44:
45:
46:
47
49:
50:
51:
52:
53:
56:
57:
58:
59:
60:
61:
62:
63:
64:
65:
66:
67:
68:
69:
70:
71:
72:
73:
Max degs of freedom: Set to 3 for determining origin time and hypocenter, set to 2 for fixed depth
solution (depth on phase headers), -2 fix all events to starting depth in STATION0.HYP, 1 to fix all
hypocenters to value on phase headers, 0 to fix hypocenters and origin times to values on phase
headers. D:3.0
Magnitude of parameter change for convergence, D: 0.05
Minimum spread to normalize residuals, D: 0.1
Bisquare weighting width, D: 4.685
RMS residual low limit for bisquare weighting, D: 0.0
Maximum number of increases in damping before fixing depth, D: 10.0
Least squares errors (0.0), damped least squares errors (1.0), D: 0.0
Factor by which damping is increased when RMS increases, D: 4.0
Depth origin of coordinate system, 0: sea level, 1:maximum elevation station in station list, D: 0.0
Maximum distance (km) from nearest station at which hypocentral solutions will be generated, D:
20000.
Minimum rms for residuals to be used in average station residual calculation - doesn't affect the
final hypocenter solution, D:1.5
Rg phase velocity in km/sec, D: 3.0
Minimum rms difference between two location to use for average, D: 50.0
Minimum number of phases for average, D: 3.0
Prevent depth to go below Moho and Conrad for n and b phases respectively, 1: enabled, 0:
disabled, D: 0.0
T-phase velocity, D: 1.48 km/sec
Flag for using azimuth phases, 0 disables. Disabling the azimuths also means that they are not
used for a starting location. A better solution will often be to set the azimuth error, TEST(52) to a
large value, effectively disabling them.D: 1.0 (enabled).
Lg phase velocity in km/sec, D: 3.5.
Relative weighting of error in azimuth used in azimuth inversion (degrees). An error of test(52)
degrees will give the same contribution to the rms residual as a travel time error of 1 sec, D: 5.0
Critical distance phases moved to by start loc. if Pn or Sn, D: 130.0 km
A value of 1.0 enables the starting location algorithm, STARTLOC. Estimates are then obtained
from apparent velocity, distance, azimuths, etc. If test(56)=0.0 epicenter is taken 0.2 km from the
first arrival station. D: 1.0
MUST BE SET TO 1.0 TO LOCATE WITH ONE STATION ONLY.
Distance (geocentric km) beyond which IASPEI91 tables are used to calculate travel times. Can
be overridden by the distance letter L in the Nordic format. D: 1500 km
Maximum apparent velocity (km/sec) for phase data to be used. This option was added to
selectively disable some of the PKP phases, which have large errors due to their steep angle of
incidence. Their velocities were almost always > 25 km/s, D: 100.0 (effectively disabled)
Critical distance for PKP core phases, D: 13000 km
Seconds by which the arrival time difference between two adjacent stations can exceed the travel
time between them. Setting this to 0 disables the initial consistency check. D: 5.0
Multiple of apparent velocity regression residual rms at which arrival times are weighted to zero
during start location determination. Reducing this value will cause arrivals to be rejected when
they do not conform to the plane wave set of arrivals which is characteristic of distant events.
Unless you are getting a lot of messages ' xxx removed: Apparent velocity deviation =..', in the
output, it is recommend against changing this default value. However, you can disable this
feature by setting test(61)=0.0, D: 2.0
Use of IASP91 phases.0: Only calculate ‘basic’ phases, 1: calculate all, D: 1.0
Types of phases used when calculating travel time, D: 0.0
Allow temporary increase in RMS by this factor, D: 2.0
Number of iterations for which increased rms is allowed, D: 3.0
Print out of travel time calculation errors (1=y,0=n), D: 0.0
Recognize blank phases as P (y=1,n=0), D: 0.0
Apparent P-velocity(km/sec) to calculate start depth from pP-p, D: 5.0
Distance (deg) beyond which PKiKP or PKP is used as first arrival, D: 110.0
Maximum depth that the hypocenter is allowed to move to, D: 700 km
Sort output in distance,(y=1,n=0), D: 1.0
Auto phase identification for distant events (y=1,n=0), D: 0.0
Number of iterations with first P's before autophase id., D: 3.0
67
74:
75-78
79:
80:
81:
82:
83:
84:
85:
86:
87:
88:
89:
90:
91:
Print input phase data in print.out (y=1,n=0), 0.0
Ml magnitude coefficients. Ml = TEST(75)*log10(amp) + TEST(76)*log10(dist) + TEST(77)*dist +
TEST(78) where amp is amplitude in nm and dist hypocentral distance in km. The defaults are Ml
= 1.0 * log10(amp) + 1.11*log10(dist) + 0.00189*dist - 2.09 which is close to the original Richter
definition (Hutton and Boore, 1987).
Minimum number of stations to attempt a solution,D: 1.0
Minimum number of phases (azimuth is counted as a phase) to attempt a solution, D:
3.0
Disable location of local events if 0.0, D: 1.0
Disable location of regional events if 0.0, D: 1.0
Disable location of distant events if 0.0, D: 1.0
Disable ellipticity correction for distant events if 0.0, D: 1.0
A priori error(sec) of local events. This affects the error estimates, particularly when few stations
are present. D: 0.1
Number of degrees of freedom in estimating test(85) for loc. ev., D: 8.0
Confidence level, D: 0.1
RMS residual(sec) at which residual weighting is applied for distant events.
D: 10000.0
Use depth phases (y=1,n=0), D: 1.0
Use of core phases (y=1,n=0), D: 1.0
Same as TEST(85) for distant events,D 1.0
NOTE: In the Hypocenter manual, TEST(85) and TEST(91) have been interchanged.
92:
Number of degrees of freedom for test(92), D: 8.0
93:
Output longitude to always be positive (y=1,n=0), 0.0
94:
Value of residual below which zero weight phases (w=4) is used again, D. 0.0
95:
Disable use of core phases between 135 and 150 deg, 1: disabled, 0: enabled, D: 0.0
96:
Variation of depth to find minimum rms, 1: enabled, 0: disabled, D: 0.0
97:
Minute error correction 1: enabled, 0: disabled, D: 0.0
98:
Enable spherical harmonic station corrections, 1: enabled, 0: disabled, D:0.0
99-101: Lg, Rg and T weights put in permanently: D: 1.0,1.0,0.0
103:
Minimum number of depth phases for starting depth, D: 1.0
104:
Minimum distance of epicenter from array for distant events, D: 30.0 deg.
105:
Enable gradient model, not yet implemented
106:
Only calculate magnitudes and update spectral values, 1: enabled, 0: disabled, D: 0.0
107:
Use xnear and xfar from sfile, 0: disabled (xnear and xfar from STATION0.HYP file), 1 enabled,
D:0.0
The test parameter defaults are set in file hyposub1.for in LIB.
HYP output:
Output from the program is a CAT-file (hyp.out) and the original HYPOCENTER print file (print.out) with
more detailed information. The hyp.out file can be plotted directly using EPIMAP. In addition, there is also
the HYPO71 style summary file, hypsum.out. NOTE: In print.out and hypsum.out, year is only given with 2
digits. Magnitude in hypsum.out and print.out are only coda magnitude and will be different from same
magnitude in hyp.out if a magnitude correction has been used.
When HYP is executed from EEV, the print.out file has no station listing. In all other cases, there is a
station listing.
Some explanation is given below, for details see HYP manual
The output in print.out first shows the content of the TEST parameters in the STATION0.HYP file. After
that comes some routine output from the starting location algorithm. Then follows the output from the
iterations, which should be self-explanatory. The location is then given on one line containing origin time,
latitude longitude (deg min), depth, number of phases, ?, rms damping and errors, error estimates,
68
resolution matrix. Last are the station lines with the following abbreviations:
stn :
dist :
azm :
ain :
phs :
w
:
hrmn :
t-sec:
t-obs:
t-cal:
res :
wt
:
m
:
no
:
Station
Distance in km
Azimuth at the source
Angle of incidence at the source
Phase
Input weight
Hour minute
Arrival time sec
Observed travel time
Calculated travel time
Residual
Weight used, normalized to 1.0
The number of degrees of freedom in the spatial solution (maximum 3)
The number of non-zero weight valid phases, including azimuths, used in
the solution
A station weight wt=-1 means that the phase travel time could not be calculated.
Any change in the input phase ID is signified by an asterisk (*) before the phase ID.
If amplitudes are available, Ml, Mb. Mw or Ms will be calculated, and all stations calculating Ml, Mb, Mw or
MS will additionally be displayed at the end of the interactive printout.
Change of day:
If the origin time of the located event occur on the day before the time in the header line, the time in the
header line is changed to the previous day and all phase arrivals are changed accordingly. This means
that some hour values will be more than 23 since phase arrival times refer to the main header.
Seismic moments etc: After locating an event, HYP will check if there is spectral information (Moment etc,
see MULPLT) available in the S-file and average values will be calculated and written into the output file.
Problems with ”
If no location or an obviously wrong location is obtained, check print.out. Common problems are:
- Wrong location: Program gets into a local minimum. Use the individual event start location option (‘S’
in header line). If the problem happen often, try either option for start location (nearest station or start
location routine). If dept is the problem, try a range of start depths (set in STATION0.HYP)
- No location: The program iterates outside the maximum distance set for a local or regional event
(RESET TEST 57) or the initial start location is outside limits. Use a fixed start location or check
readings to get a better start location.
6.1.2 HYPO71 (Sun only)
By Brian Baptie, BGS
HYPO71 is a computer program for determining hypocenter, magnitude and first motion pattern of
local earthquakes written by Lee and Lahr (1972) using a stepwise statistical regression procedure
outlined in Draper and Smith (1966). The user’s manuals were originally released by the authors as a
series of open-file reports of the U.S. Geological Survey and contain a full description of input and
output parameters and usage. The SEISAN version of the program is essentially the same as the
original, the only differences being in the input and output facility. Input data required are phase
arrival times, station co-ordinates and a crustal velocity model. SEISAN extracts the arrival
information from a Nordic format phase readings file and the station and velocity information from the
station input file STATION0.HYP, found either in the SEISAN data directory, DAT, or the local
directory. The format of the STATION0.HYP file is described in this manual in the section on the
HYPOCENTER algorithm (6.1.1). HYPO71 supports 13 test variables that influence how the program
goes about locating the earthquakes. The default values for these variables were developed for the
large and closely spaced networks in central California. These variables are defined at the start of the
STATION0.HYP file by the values of TEST(01) to TEST(13). Brief definitions for each of these
69
variables can be found below and full definitions can be found in the HYPO71 manual.
SEISAN constructs a HYPO71 format input file called hypo71.input, containing the station coordinates, thickness and velocity for each layer of the crustal model and phase arrival times, then
runs the HYPO71 algorithm. The HYPO71 program generates a single output file called
hypo71.output. SEISAN reads the information contained in this output file to create two further output
files: hypo71.out, a Nordic format phase readings file containing the calculated location; and
hypo71.brief, a summary file containing origin time, epicenter, depth, magnitude and station residuals.
There are a number of limitations to the current version.
• The program is designed to run from eev and can only be used for one event at a time; there is
no facility for multiple event or batch location/relocation.
• HYPO71 is not included with the UPDATE command, so the database cannot be updated.
• Errors will result if the input phase readings contain arrivals from two different days, i.e. either
side of midnight
• All stations must have the same sign of latitude or longitude, so if stations extend across the
Greenwich meridian and/or the equator and an offset should be added to allow for this.
Running the program
HYPO71 is run from within eev by typing hypo71, at the command line. On successful completion, the
information from the hypo71.brief file is displayed on the screen. Below is an example of the screen
output.
EXAMPLE RUN
#
#
29 13 May 2001 8:26 59 L 55.1020 -3.6388 12.3 H
2.9L
16
29 13 May 2001 8:26 59 L 55.1020 -3.6388 12.3 H
2.9L
16
HYPO71 completed successfully
Date
: 13/05/01
Origin time : 8:26:59.78
Epicentre
: 55- 5.45 deg N
3-37.11 deg W
( 55.0908 /
-3.6185 )
Grid Ref
: 296.716 East / 578.581 North
Depth
:
2.13 Quality
: B B*B
Statistics
Magnitude
Magnitude
STN
BWH
BHH
BCC
GCD
ECK
ESK
BBH
BBO
BDL
BTA
CKE
XAL
EDI
GIM
WCB
CWF
DIST
9.7
25.6
26.8
32.7
32.8
36.4
44.2
46.1
54.1
63.2
65.0
93.4
96.6
104.4
200.0
302.7
?
? hypo71
NO DM GAP M RMS ERH ERZ
: 22 10 127 1 0.29 0.8 1.8
: 2.8 ML
(from 6 readings)
: No valid coda readings
AZM AIN
346 43
89 43
108 43
219 43
72 43
46 43
84 43
149 43
126 38
109 38
150 38
106 38
16 38
212 38
198 30
150 30
P-RES
0.08
-0.03
0.26
-0.40
-0.28
-0.39
-0.29
0.20
0.34
0.42
0.27
0.16
0.33
0.02
-0.91
-0.08
P-WT
7.22
1.03
0.71
1.75
1.02
1.00
0.77
0.54
0.70
0.69
0.54
0.54
1.00
1.35
0.45
0.24
S-RES
S-WT
-0.29
0.01
0.51
0.36
-0.91
0.26
0.01
0.36
1.04
0.09
0.36
0.87
Phase names
Only single character phase names are supported, denoted by P or S.
Weighting
Two weighting options may be used.
1. User specified weights assigned by a single integer value in the range 0 to 4 for a given phase.
These will assign a weighting factor of 1, 0.75, 0.5,0.25 or 0.0 to that phase. Also, a weighting of
9 will assign the absolute time a weighting of 0.0 but will allow the use of relative times if a valid
S-arrival is found for that station. The relative arrival time will be assigned the weight of the S-
70
phase.
2. Distance weighting as given by the relationship w = (xfar - Δ) / (xfar –xnear). By default the
parameters xnear and xfar are read from the STATION0.HYP file. If TEST(107) is set to 1., xnear
and xfar are read from s-file if there.
Using a starting location
The user can specify the use of a starting depth and epicenter by entering the character ‘S’ in
columns 44 and/or 45 respectively, in the header line of the input readings file. The starting depth and
epicenter are given by the values in the header line of the readings files. Otherwise, the starting
epicenter is set to be the latitude and longitude of the station with the earliest P-arrival.
Fixing the location
Using the character ‘F’ instead of ‘S’ in columns 44 and 45 of the header line fixes the depth and/or
epicenter to the values given in the header line.
Errors
The standard error output from the HYPO71 program is contained in an additional line in the Nordic
format readings output, hypo71.out, defined by the characters ‘83’ in columns 79 and 80.
The HYPO71 error line format is defined as follows:
Columns
2-14
19
21-23
25-27
28-30
32-34
36
38-41
43-46
48-51
79-80
Format
A13
A1
A1*A1
I3
I3
I3
A1
F4.2
F4.1
F4.1
A2
Description
‘HYPO71 errors’
Location quality, Q
QS and QD rating
Number phases used
Distance to closest station
Azimuthal gap
‘1’. (Always output?)
RMS
ERH (km)
ERZ (km)
‘83’
2
th
RMS is defined as √(∑Ri / N) where RI is the time residual at the i station. ERH is the standard error
2
2
in the epicenter in km given by √(SDX + SDY ), where SDX and SDY are the standard errors in
latitude and longitude. ERZ is the standard error in the focal depth in km. The location quality, Q, is a
measure intended to indicate the general quality of the solution and is defined by a single character.
Epicenter
excellent
good
fair
poor
Q
A
B
C
D
Focal Depth
good
fair
poor
poor
Q is taken as the average of QS and QD, where QS is a statistical measure of the solution and QD is
rated according to the station distribution.
QS
A
B
C
D
RMS (S)
< 0.15
<0.30
< 0.50
Other
ERH (km)
≤ 1.0
≤ 2.5
≤ 5.0
ERZ (km)
≤ 2.0
≤ 5.0
QD
A
N
≥6
Gap
≤ 90
DMIN
≤ Depth or 5 km
71
B
C
D
≥6
≥6
Other
≤ 135
≤ 180
≤ 2*Depth or 10 km
≤ 50 km
Magnitude
Both duration and amplitude can be used to calculate magnitudes as with HYPOCENTER (see above
for details). Duration, amplitude and period for each station are used to give a magnitude value for
each station. These values are averaged to give the event magnitudes.
The test variables
Test Variable
TEST(01)
Default Value
0.1 S
TEST(02)
10 km
TEST(03)
2.
TEST(04)
0.05 km
TEST(05)
5.0 km
TEST(06)
4.
TEST(07)
-0.87
TEST(08)
TEST(09)
TEST(10)
2.0
0.0035
100 km
TEST(11)
8.0
TEST(12)
0.5
TEST(13)
1.0 km
Definition
TEST(01) is the cut-off value below which Jeffreys’
weighting of residuals is not used. It should be set
to a value approximately equal to the overall timing
accuracy of P-arrivals in seconds.
For each iteration, if the epicentral adjustment is
greater than TEST(02), this step is recalculated
without focal depth adjustment. TEST(02) should
be set to a value approximately equal to the station
spacing in km.
Critical F-value for the stepwise multiple
regression. TEST(03) A value between 0.5 and 2 is
recommended.
If the hypocentral adjustment is less than TEST(04)
then Geiger’s iteration is terminated.
If the focal depth adjustment, DZ, is greater than
TEST(05), DZ is reset to DZ/(K+1), where K=
DZ/TEST(05). TEST(05) should be set to a value
approximately half the range of focal depth
expected.
If no significant variation is found in the stepwise
multiple regression, the critical F-value, TEST(03)
is reduced to TEST(03)/TEST(06) and the
regression is repeated.
Coda magnitude constant a, where
Mc = a + b log10(T) + c ∆ + ∆ M
Coda magnitude constant b.
Coda magnitude constant c.
If the latitude or longitude adjustment (DX or DY) is
greater than TEST(10) then DX is reset to
DX/(J+1), and DY is reset to DY/(J+1), where
J=D/TEST(10), D being the larger of DX or DY.
Maximum number of iterations in the hypocentral
adjustment.
If the focal depth adjustment (DZ) would place the
hypocenter in the air, the DZ is reset to DZ= -Z *
TEST(12), where Z is the focal depth.
Parameter for auxiliary RMS values
6.1.3 The Hypoinverse program, HYPINV (SUN and PC)
The Hypoinverse program has been implemented in a simple way and is mostly intended to be
operated interactively from EEV in order to compare locations. The main program has seen very few
72
changes and can be run according to the original manual (Klein, 1984) and will not be described here.
The program does not work well at large distances( > 1000km) so use it only for local earthquakes. If
original data, station and control files are available, it is just typing HYPINV and it will run according to
the manual. If none of these files are available, they can be made with the conversion programs. The
steps to run HYPINV without EEV are as follows:
1: Convert a CAT file to Hypoinverse file by typing norhin input file. The input file in Nordic format will
now be converted to a file norhin.out in Hypoinverse format.
2. Make the control files by typing makehin. This creates the instruction file hypinst, station file
hypinv.sta and model file hypinv.mod. These files are standard Hypoinverse files. The information is
taken from the STATION0.HYP file in either the working directory or DAT. Makehin cannot work with
an alternative STATIONx.HYP file.
3. Type hypinv and the program runs. There is a one-line output per event on the screen and the full
output is in a file called print.out.
Running HYPINV from EEV, the above 3 steps are done automatically when using the command H
and in addition, the print.out file is printed out on the screen.
6.1.3 HYP_ISC
Program implemented by Richard Luckett
ISC has for many years used a standard procedure to locate earthquakes and the ISC locations have
often been used as a reference. The earth model used is the Bullen tables. ISC has recently rewritten
the old location program and it was therefore possible to also port it to SEISAN. The purpose is that it
should be possible to compare standard ISC locations with location using other programs and
models. The implementation in SEISAN was done using the standard hyp program where only the
location routines have been changed. The program then behaves almost identical to HYP and uses
the same format input and output files.
Parameter files: STASTION0.HYP is used for station coordinates, magnitude scales and agency
code. The crustal model information is not used and only the RESET TEST parameters related to
magnitude are used. In addition, there is a new parameter file (in DAT) iscloc.def with parameters
specific for the ISC location routines, see file for explanation of parameters.
Input data files: Just like for HYP
Output files: Hyp.out is like before, print.out is different.
Not all crustal phases used with HYP may be available. The weights used in SEISAN do not apply
since the program uses residual weighing only, see parameter file.
Magnitudes are calculated exactly like in SEISAN.
In eev, the command to locate with HYP_ISC is ‘il’.
For more information about the ISC location program, see http://www.isc.ac.uk/Documents/Location/
6.2 Trace plotting, phase picking and spectral analysis, MULPLT
This program is the general plotting and signal analysis program. The program is capable of doing
general phase picking, correct for instrument response, and produce Wood-Anderson seismograms
for determining Ml, synthetic traces for Mb and Ms, determine azimuth of arrival for 3 component
73
stations and do spectral analysis. The program can also read in theoretical arrival times for global
phases for help in identifying phases. If a quick location is needed based on a waveform file only,
mulplt can both pick the phases and locate the event.
MULPLT operates either as a database independent program (started with command MULPLT) or in
connection with the database (started from EEV with command P or PO). If the program works
independently of EEV, it will create an output file mulplt.out in Nordic format with the readings and
results of spectral analysis. This file can directly be used with e.g. HYP.
Starting MULPLT from prompt line
Giving command ‘mulplt’, the question is
Filename, number, filenr.lis (all), cont for cont base, conts for large SEED
Filename, number, filenr.lis(all): The program asks for a file name or file number of a waveform file.
To use the number, it is assumed that a list of files has first been created and numbered in a file
filenr.lis using command DIRF, see section 6.8. By giving the number, the file corresponding to the
selected number is used. By giving a ?, the list with numbers is displayed and a new number can be
given. If many files are to be plotted with one command (hard copy only), give filenr.lis for file name
and all events in FILENR.LIS will be plotted. There will only be one question about filter and then all
events are plotted with all channels and the chosen filter.
Cont: Plot from a continuous data base. The program will use all data bases defined in SEISAN.DEF.
A question will be given for absolute start time and window length. See also 6.2.4.
Conts: Plot from a large SEED volume. A SEED file too large to be read in can be plotted in parts. A
question will be given for file name, absolute start time and window length. See also 6.2.4.
6.2.1 MULPLT main functions
The program has 7 main functions irrespective of type of input as illustrated below with the questions
given by the program:
Plot options:
Interactive picking
Return
Multi trace plot on screen, def (0)
Multi trace plot on screen
(1)
Multi trace plot on screen+laser(2)
Multi trace plot on laser
(3)
Continuous on screen
(4)
Continuous on screen + laser
(5)
Continuous on laser
(6)
Stop
(9)
Return: Picking phases, spectral analysis and 3 component analysis
0:
Initial plotting of new waveform data and registration in database using predefined
defaults. Phase picking.
1:
Initial plotting of new waveform data and registration in database, or general plotting
of multi-trace data. Phase picking. In this mode, the user is asked to select which channels to
work with and a graphical window will be shown for selection. If more than 250 channels are
available, selection will have to be made on several screens. A maximum of 1000 channels
can be used. Optimally, channels can be show in alphabetical order (see MULPLT.DEF).
2:
Same as 1, only hardcopies may be made at the same time.
3:
Making hardcopies of many waveform files with one command. No screen output.
4,5,6: Plotting one channel continuously like on a seismogram with several traces from left
to right on top of each other. One channel can be selected.
For continuous data, see 2.2.2 and 6.2.4.
74
Commands f and q: With a plot on the screen, q will always quit mulplt. F will, in single trace mode,
bring up the next channel, in multi trace mode bring up the next event.
Option 0 is particular useful for checking many new events, since the program does not ask question
about station choice (uses definition in MULPLT.DEF file in DAT, or working directory) and typing f
(when the plot is on the screen) automatically goes to the next file in filenr.lis.
If option 1,2 or 3 is used, a display will show the available channels and the user can click to select. If
MULPLT is operated from EEV, a * in a channel box will indicate that readings are available. When
the channel selection is shown, it is possible to quit the program with q or continue with f.
If option 4,5 or 6 is selected, continuous data is plotted, see below.
Running MULPLT using command MULPLT, the program asks for a file name or file number of a
waveform file. To use the number, it is assumed that a list of interesting files has first been created
and numbered in a file filenr.lis using command DIRF, see section 6.8. By giving the number, the file
corresponding to the number is used. By giving a ?, the list with numbers is displayed and a new
number can be given. If many files are to be plotted with one command (hard copy only), give filenr.lis
for file name and all events in FILENR.LIS will be plotted. There will only be one question about filter
and then all events are plotted with all channels and the chosen filter.
Hardcopies assume a PostScript printer. For each event plotted, a plot-file called mulplt.eps is
generated. The plot files are sent directly to the printer from within the program with the seisan print
command as soon as the plot is finished for one event but before the program is finished. In Unix, this
is lpr or lp while on PC, the command is given in a .bat file in COM (see installation section). This
means that the same plot-file is overwritten for each event plot. For setting up the printer, see
installation, section 3.
In multitrace mode, many traces (number limited by the SEISAN system definitions, see section 3)
can be plotted. If the plot is made via EEV, all picks are also displayed.
Location of waveform file:
Mulplt will search in current directory first. If not found there, the WAV directory is searched. If the
SEISAN.DEF file has been set up, MULPLT will thereafter try to locate the waveform file in one of the
databases or specific directory given in the SEISAN.DEF file (located in DAT or working directory).
Format of waveform files:
MULPLT can, on all platforms, use SEISAN, GSE, SAC ASCII and SEED and on Solaris and Linux, in
addition SAC binary. If started from EEV, files in different formats can be used at the same time.
Time gaps in waveform files:
Only SEED format has running time headers and therefore the only format where a possible time gap
can exist. SEISAN will replace missing data intervals with zeros.
6.2.2 Use of MULPLT from EEV
In order to process events more easily using SEISNET, EEV and MULPLT have been tightly
integrated. When MULPLT is called from EEV, command f will plot the next event in the database, to
go back to the same event in EEV, use quit. The event can be plotted with default parameters from
EEV using PO. If PO has been selected, the f command in MULPLT multi trace mode will show the
plot of the next event with default options. This is a fast way of plotting waveform files going through
S-files in a database.
If several waveform files are available, the user will graphically be shown the files and can select one
75
or several Files can optionally be displayed in alphabetical order (see MULPLT.DEF). If more than 75
files, several selection screens will be shown. A maximum of 1000 files can be used If the PO option
has been used, all default channels will be used.
6.2.3 Continuous plotting
This option is used to plot one channel in a multi line display and can therefore simulate a helicorder
plot. This is a very different option from the plotting from a continuous data base (see 6.2.4).
Interactive processing is not possible in this mode.
Using this option the program asks for the following input:
Low and high cut for filter: Give values or return for no filter. Sometimes low cut filters are unstable (a
different routine used for continuous plotting) so if e.g. a LP record is to be simulated, use filter limit 0
to 0.1 Hz. The zero means it is a low pass filter, not bandpass. A filter 10 to 0 would mean a high
pass filter. High and low pass filters are only available with continuous plotting, unless the FILTER
TYPE (MULPLT.DEF) is set to 1.0=RECFIL.
Seconds pr line: Number of seconds on each line.
End time: This question only appear if plotting from EEV. The list of files is then the list of events
belonging to the data base used. Give end time as e.g. 2000050203
Max count: The absolute maximum count to be used for full scale. Since many lines and possibly
many pagers are plotted, it is not possible to use autoscaling, and like on a seismogram, a fixed value
must be set.
Lines pr page: Number of lines per page.
Station code: Station code (max 5 characters)
Componet code: The component code, max 4 characters.
MULPLT will plot from the first file given from the filenr.lis file and then continue to plot as long as
more file names are given in filenr.lis. Alternatively if plotting from EEV, it will start with the current
event and continue until the end time. So if a month of data files are given, a month of seismograms
will be displayed. There is no requirement that the input files follow each other in time (no time gaps)
since each file is plotted on the page where it belongs in time. However, the files must be time
ordered. The continuous option can therefore be used to check availability and timing of continuous
data. Discrete events can also be plotted in this mode if one want to get a display of when the events
occurred. However, if filtering, it is assumed that the files follow each other in time since a few points
are carried over from one file to the next to make the filtering continuous. Figure 18 shows an
example.
6.2.4 Commands in MULPLT, overview
When the trace(s) are on the screen and the cursor is displayed, then several options are available.
Most options can be displayed by pressing the MENU button in the upper right hand corner. Pressing
MENU again removes the option boxes. Commands can be given by either pressing a letter or
clicking on a box in the menu (Figure 19). By pressing ? or clicking on the Help button, the following
help menu will be displayed:
Help on MULPLT
MULPLT has 2 modes:
multi trace mode:
Displaying many traces, some picking options
76
single trace mode: One trace, all picking options
Most commands are given by pressing one key, however several commands can
also be given by clicking in the appropriate menu box on top of the screen.
If the full menu is not displayed, select the menu on top right part of the
screen. The following commands are available, first given by the single
letter command and if also available as a mouse click, the letters in the
menu box is shown. COMMANDS ARE CASE SENSITIVE.
Channel selection: In multitrace mode, one or several channels can be
selected by clicking on the station code. If only one channel is selected,
and toggl (t) to single trace mode is done, a new toggl in single trace mode
will display all previous channels again.
Zooming:
Select a window with the mouse and a zoomed window
will appear below in single mode and replace the plot
in multi mode. In Single mode, it is also possible replot
the zoomed window in place of the original by placing the
cursor above the trace when selecting the zoom window. This
makes it possible to zoom in the zoomed window. In order to
go back to the original window in multi mode, do an opposite
zoom, meaning picking the last point first.
a:
Read amplitude. Position cursor at the top of
a wave and press a. Position cursor at the
bottom of the wave and press a. Amplitude(0-p) and
period are now stored. These values will be stored
with the NEXT phase pick. Amplitude and period are
displayed.
A:
Automatic amplitude reading, phase become AMP or
read phase AMP but only after using a twice for reading amplitude.
b:
Filter 5-10 hz, see below
B: Back
Go back one trace in single trace mode
From eev, multi: Go back one event
Continous data base: Go back one window
c:
Read coda.
C:
Read end of coda automatically
d:
Delete phase. Position cursor near phase and press d.
d: Del W: Delete waveform file, cursor outside area of trace plot, file must
be in working directory, else no files displayed
D: Del S: Delete S-file if operated from EEV, multi mode
e:
Phase E
f: Next:
Single: Go to next channel
Multi, One event: Go to next event
multi, continous data base: Go to next window
F: FK:
FK analysis of array data
g: Groun: Make a ground motion seismogram(s).
h: Azim:
Make 3 component analysis (single mode ONLY) to
determine azimuth of arrival. Select a window around
the P-phase on the Z-component. Azimuth and apparent
velocity will enter the S-file with the next phase
reading.
i:
Phase I
I: Iasp
Calculate IASPEI synthetic arrival times, which are then
Displayed, multitrace only.
77
j: Mb:
Generate a synthetic SP seismogram for reading amplitudes
for determining Mb.
k: Ms:
Generate a synthetic LP seismogram for reading amplitudes
for determining Ms.
l: Locat: Locate event, only multi trace mode
m:
Filter 15-24 hz, see below
M: Merge: Merge waveform file, mulplt called from EEV, only if files are
In working directory.
n:
Filter 10- 15 hz, see below
o: Oth C: Select other channels
O: Out:
Makes an output waveform file (waveform.out) of current
data on screen. Only multi mode. Response info not saved.
p: Regis: PUT (Register) event in database, you will be
prompted for event type and waveform file will be copied
to WAV.
q: Quit:
Quit program
r: Plot:
Replot same event, useful when screen is cluttered
up with e.g. many picks. Also used when a replot is
wanted with new parameters e.g. filter.
s: Spec
Make a spectrum, single mode ONLY. Press s, select
window with cursor, answer questions and the spectrum
will appear below with noise spectrum.
S:
Same as s without the noise spectrum.
t: Toggl: Toggle between multi and single mode
u: Rotat: Rotate components. Backazimuth is calculated from location
in header or as a second option, taken from observed + azimuth
residual for respective station. R is plotted instead of N and
T instead of E. Combining 3 component option (h) with u, the
user is asked for the backazimuth angle.
v:
Filter 1-5 Hz, see below.
y:
Pick a theoretical phase if displayed. Place cursor
where the phase might be. Press y and the program will
select nearest synthetic phase.
w: WA:
Remove system response and display synthetic
Wood-Anderson ground motion in nanometers (nm)
on next plot (using R or zoom).
W: Oth W: Select other waveform file, same event
z:
x:
Z: <W>
X: >W<
Filter 0.001 to 0.1, see below
Filter 0.1 to 1.0 Hz, see below
Increase window length in plotting from a continous data base
Decrease window lenght when plotting from a continous data base
>: Print: Will make a hardcopy of all channels of current event with
the last selected filter, only in multitrace mode.
<:
Same as D
*: Scale: Fixed scaling of trace amplitudes.
_: Dist:
Select plotting channels in distance order.
78
:
Resp:
Plot response file, single trace mode only.
TAB: NextW: Next window if multiple windows
Filter options: The fixed filters (4/8 pole Butterworth) are placed
on keys z,x,v,b,n,m with the following frequencies:
z:
x:
v:
;
b:
n:
m:
01-.1:
.1-1.:
1 - 5:
2 - 4:
5-10 :
10-15:
15-23:
0.01
0.10
1.0
2.0
5.0
10.0
15.0
- 0.1 Hz
- 1.0 Hz
- 5.0 Hz
- 4.0 Hz
- 10.0 Hz
- 15.0 Hz
- 23.0 Hz
Pressing key once gives a 4 pole filter one way
Pressing the key twice, and the filter also
go the other way and it is now an 8 pole filter.
From menu, only 4 pole filters are used.
.: Filt: Variable filter, question of filter limits is given in text
window.
,: FixF Fix filter. If pressed aftrer selecting as filter, the filter
remains fixed until pressing ',' again.
Phase picking: This is possible in both modes. In Single mode, the phases
defined are shown on top with some of the options, while
they are not shown in Multi mode, but have the same
definitions.
Combining options: Note that you can select several options together.
E.g. V and S will first filter the signal and then
make the spectrum.
Saving observations: When you go to the next trace or another event (F),
the readings are saved in the S-file. They are also
saved when you pick the next trace in Multi mode.
This is also true for spectral parameters and azimuth
determination.
79
Filters in MULPLT
MULPLT uses a 4 pole Butterworth filter that can be used forward and backwards. Normally when a
filter character or filter menu press is selected, the filter is only run one way and the number of poles
is then 4. This will make a small phase delay where the first onset might appear a bit later, so if
possible, read on unfiltered traces. If an 8-pole filter is desired, press the filter key twice and the filter
will also run backwards. This gives theoretically a zero phase shift filter, however in practice, some of
the onset energy is seen well before the first arrival, so it seems to distort the arrival times much more
than using the 4 pole filter. When the program asks for a non fixed filter like when using the ‘.’(Filt)
command, the filter is always 4 poles. When doing spectral analysis and specifying a filter before the
spectral analysis, the filtering is done in frequency domain and the filter is 8 pole Butterworth. When
reading polarities, DO NOT USE FILTER, if possible. If a filter is chosen from the menu, it is always 4
pole.
The filter pass-band limits can be changed in MULPLT.DEF.. The user can also chose between two
filter routines: bndpas (default) and recfil.
The filter used in continuous mode can be either bandpass, low pass or high pass. Specifying a filter
limit of zero, means that the filter is low pass or high pass. Limits of 0 10 Hz means a 10 Hz low pass
filter.
Displaying uncertain time
In each trace header in the SEISAN -waveform file, there is a flag to indicate if the time might be
uncertain (see Appendix 2). If that flag has been set, the message ‘UNCERTAIN TIME’ will be
displayed on top of the trace. Currently this flag is only put into the waveform files if the data comes
from a SEISLOG system that has detected a timing error.
80
Figure 15. An example of using MULPLT in multitrace plot mode. Notice that start and stop
times are different for different channels. The horizontal line at the start of the plot is the DC
level. The small number above each trace to the right is the max absolute count with the DClevel subtracted and the small number to the left above the trace is the DC level. If plotting
from EEV, the phase picks available are shown.
81
Figure 16. Examples of MULPLT with theoretical arrival times of some global phases. Short
period seismograms are shown. The theoretical phases are marked with onset y below the
trace and the read phases are marked normally above the trace.
82
Figure 17. Example of MULPLT with theoretical arrival times showing global phases on a long
period seismogram. The filter used from 0.01 to 0.1 Hz. Without filtering, almost nothing would
have been seen on this broadband station.
83
Figure 18. MULPLT in continuous mode.
The plot shows 6 hours of long period data. The scale is 3000 counts between the traces and
the filter used is from 0.01 to 0.1 Hz. The trace start time in hours and minutes is given on top
of each trace. On the header line, P1 means the first page and DC is the DC level subtracted.
Note that the numbers on the time scale at the bottom only are valid for the first trace unless
all traces are 60 sec or 60 min long.
84
Figure 19. Example of the menu, which can be displayed on top of the plot
Below is some more detailed description of some of the options. The one letter command is given
with the menu command in parenthesis:
To apply filters, first make a selection of options (filter, window, channel selection) and then execute
by pressing R(Plot) (or selecting a zoom window). Figure 15 shows an example.
Single trace mode:
In this mode, one trace is initially displayed on top of the screen, see example on Figure 20.
The traces used are the ones earlier selected and will be displayed one by one. Several options are
now possible as can be seen on the menu. Normally no hardcopies are made in single trace mode
since it is intended for fast routine work. However, by starting MULPLT in multitrace mode (option 2)
and then go to single trace mode (command T(Toggl)), hard copy files are made.
Multitrace mode:
In this mode hard copies can be made. If option 2 is used, both screen plot and hard copy files are
made. If replot is made, only the last plot is available in the hard copy file. If option 3 is used, which is
only hardcopy, there will be additional questions about, window length, start time, scaling and filters. If
the scaling is set so that the plot occupies more than one page, several pages will be printed. If in this
mode, filenr.lis is given as file name, the program assumes that all the files should be plotted and the
only questions will be about the scaling and filters. All channels in each file will be plotted. This option
is useful for plotting a large number of events with a single command.
Multiple screens in multitrace mode
If many channels are available (like more than 30), it might be difficult to distinguish all and the
channels can be displayed in multiple screens. The number of channels per screen is set in
MULPLT.DEF. The number of windows or screens for a particular data set is given in top left hand
corner as e.g. ‘Win 2 of 7’ meaning current window is number 2 of 7 windows. To move to the next
window, use TAB or NextW in menu. In each window, normal operation can be done. Channels
selected will be kept. Using a large data set, the user can then view each window separately, select
the channels of interest and when all channels have been viewed, only the selected channels will
remain for display.
Channel order in multitrace mode:
Normally channels are plotted in alphabetical order. They can also be plotted in the order they are
stored in the waveform file(s) (option set in MULPLT.DEF). By setting the channel order parameter in
the MULPLT.DEF file, it is also possible to plot the channels in time order. If MULPLT is started from
EEV (and distance ordering is set), the channels will be plotted in the order in which the stations
appear in the S-file, which normally is in distance order if the S file has been updated. However, if a
particular order is required, the stations can be ordered like that in the S-file. Since there is no
consideration for channels for the same station, the channels for one station, will be plotted in the
same order as given in the waveform file. If a station is not found in the S-file, it will be plotted last. If
plotting is done with MULPLT directly with a waveform file, the plotting order will be the start times as
given in the waveform file header. Channel ordering can be turned on or off with the key "-"(Dist). If
set in the MULPLT.DEF file, it is set when MULPLT starts up.
85
Plotting from continuous data base
If a continuous data base is set up (see section 2.2.3), it is then possible to plot all traces from the
continuous data base with MULPLT. When MULPLT starts up, use option cont and the user is prompted
for a start time and interval. MULPLT will now check all continuous data bases for available data in
required interval and display the available data. The forward (next) or back option will display previous or
next window respectively. There is an 25 % overlap between windows. If no data is available for the whole
window, no trace is shown. If the beginning and the end is available, a line will join the two segments. If
only end or beginning is available, only the available data is shown. All normal operation can be done on
the window plotted so it is possible to e.g. extract data. If the register option is used, the whole window is
extracted from the continuous data base as one file, copied to WAV and the S-file created.
6.2.5 Registering new events into SEISAN
Mulplt is the main tool for checking and putting new events into SEISAN. New events with waveform
data can appear in two ways in SEISAN:
(1) Unprocessed waveform files are available in a work directory and have to be inspected and
possibly put into the database. No S-files have been made.
(2) Raw data has already been put into a SEISAN database with S-files and corresponding waveform
files in some work directory, the data has not been checked. This process has most likely been done
with the automatic data collection software SEISNET (Ottemöller, 1999), however. events can also
have been auto registered with program AUTOREG.
In both cases above, the aim is to inspect an event and decide if the event is real and should be put
into the database using option ‘p’. All work must be done from the directory where the raw waveform
files are located. The process of putting an event into the database results in creating the S-file
(option1), giving the event identifiers and copying the waveform files of registered events to the
waveform directory.
By pushing p(Regis), the user will be prompted for distance indicator, which has to be L, R or D for
local, regional or distant event. It is possible here to enter 2 characters like LE or LV for local
explosion or local volcanic event. The event type or event ID can be any character. Three characters
are predefined and should only be used if the following definition correspond: P(probable explosion),
E(explosion) or V(Volcanic event). The volcanic events have a sub classification which can be
entered when registering an event as volcanic, see section 6.28. The process of registering the event
into the database implies that a new S-file is created or registered and in the S-file. An operator ID will
be asked and the operator ID will be put on the ID-line. The question about operator will only be
asked for the first event since it is assumed that all subsequent events are put into the same
database by the same operator.
The event ID, can later be used with the SELECT program to select out particular event types.
When first putting an event into the database, the user is also prompted for database.
Option (1)
Data is available as waveform files only and a list of files must be made first with DIRF. Main option 0,
1 or 2 can be used for plotting. The ‘p’ option creates the S-file and copies the waveform file to the
WAV directory. The waveform file remains in the working directory. Unwanted waveform files can also
be deleted so that when all events have been put in, only waveform files of ‘real’ events remain in
working directory. These can then be plotted with one run of MULPLT, see section 6.2.1.
Option (2)
Data is available already in a database, however since the data has not been inspected, the
waveform files are still in a work directory. In EEV, the first unprocessed event in the month is found
with command ‘ss’ and MULPLT is started with command ‘po’ to invoke all defaults. If the event is to
remain in the database, it must be registered with option ‘p’. The process and the questions are the
same as in option (1) except that the S-file is not created since it is already there. The S-file is
cleaned for all processing information from SEISNET if present. This normally also includes automatic
phase picks. However, they can be kept if parameter REG_KEEP_AUTO is set in the SEISAN.DEF
file. The status of the files also changes to being newly registered as under option (1) (see definition
86
of processing codes in Appendix 1) and waveform file(s) copied to WAV. Before registering, it might
be an advantage to merge waveform files and delete unwanted files (could be false triggers), see
section 6.2.4. Files can only be merged and deleted in working directory with commands Delw and
Merge (Menu).
In this process of putting new events into the database, it is also an advantage to delete unwanted
events. This is done with option ‘S’(Del S)’. The S-file is deleted, but the waveform files remain in the
working directory.
Preprocessing of data while registering new events, option (1)
Normally a series of events are registered first and MULPLT terminated. Then EEV is started up for
interactive picking and location. However, if preliminary processing is desired while registering the
event, this is also possible.
Phase picks:
If phases are picked before the event is registered, these readings are saved in the database at the
time of registration. After the event has been registered, MULPLT automatically goes to the next
event in FILENR.LIS and no more phase picking can be done.
Processing with a given program:
Optionally MULPLT can, after registration, start any program processing the newly registered event.
E.g. the AUTOPIC program can be started or a program reading amplitudes etc. The program name
is defined in MULPLT.DEF.
Locating the event:
As the final step after registration, the event can optionally be located and the location optionally
placed in the database.
The above options have been put in on the suggestion of Brian Baptie, who is using it for rapid
processing of volcanic events, where in most cases the operator only wants to look at the event once.
6.2.6 Phase picking, amplitude, weight and polarity
Picking phases:
The plot will display any pick present in the database (current S-file). To pick new phases, position
cursor at phase, and press the key as indicated on top of the screen (if in Single mode). E.g. pressing
1 will read IP. Pressing the same key again with the cursor at a different place will delete the old one
(indicated with a D) and display the new one. Additional default phases, which can be picked, are i for
I, e for E and A for AMP (note upper or lower case). Keys for phases have default definitions, but can
be redefined using the file MULPLT.DEF, see below. The end of the coda is picked as a phase (C)
and the program calculates coda length IF AND ONLY IF A P-READING IS PRESENT.
Picking amplitudes:
Position the cursor at the bottom or top of a wave and press a, then at the other extreme (bottom or
top) and press a (do not use upper case, see below). There is no requirement for going left to right or
top to bottom, it can be done in any order as long as the two extremes are marked. At each press, a
cross is marking where the pick was made. In case a filter, like WA, MS or Mb is applied, the program
will associate the amplitude with the respective amplitude reading (AML, AMS or Amb). Amplitude
and period are calculated and stored with the phase. Otherwise, if none of these filters are applied, a
menu pops up and the user needs to select a phase name to which the amplitude and period
readings are associated. It is often a good idea to store amplitudes with the nondescript phase E, I or
AMP since it then will remain even if the phase is deleted or changed. If an attempt is made to pick
amplitude on a trace which is not in nm, the reading must be confirmed since SEISAN assumes all
amplitudes to be in nm (see section on instrument correction). If no phase is picked, no amplitude is
stored. The amplitudes are always assumed to come in pairs so if e.g. 3 amplitude values have been
picked, and the user tries to pick a phase or quit the program, it will appear frozen since the program
87
is still waiting for the next amplitude measurement. It is always the last pair of amplitude
measurements, which are used. Amplitudes can be picked on both corrected and uncorrected traces.
If A is pressed instead of a, the amplitude is read and marked automatically. Woks in most cases, but
sometimes two subsequent peaks are not correctly chosen and the amplitude reading has to be done
manually.
Component names when picking phases:
In the S-file, the component only has 2 letters while in the waveform file it has 4 letters. There must
therefore be a unique translation between the two. This definition is given in the subroutine
componen.for in LIB. Most common combinations are now defined, however if a new one is defined in
the waveform file which does not exist in componen.for, the first and last letter of the input component
will be used. If e.g. an input component is called SS Z, then the code in the S-file will be SZ. This
means that picks for stations with components, which do not differ in first and last character, cannot
be separated in the S-file. Component names for rotated channels will be e.g. SR and ST for short
period radial and transverse components respectively.
Reading polarity:
If the cursor is above or below the trace at a distance marked by horizontal tics on the sides of the
plot, the first motion is also picked and displayed. Do not use a filter if possible.
Assigning weight:
A phase can be assigned a weight. Move the cursor close to a pick and press one of the keys 1-9 in
UPPER case thus using e.g. !"# (default, can also be changed), and a HYPO style weight is assigned
and displayed. Although weights 0 to 9 can be put in, HYP only uses 0-4 and 9 (see section 6.1.1).
Phases with associated amplitude, period, azimuth or apparent velocity are displayed with a hat
below on the phase indicator line. The default keys for the weights might not be correct on all
keyboards, if not, set keys in MULPLT.DEF.
Automatic determination of coda length (C):
The coda length can be quite variable among different operators and a function has been made to
automatically determine the coda length. The signal is bandpass filtered and the end of the coda is
determined by a standard STA/LTA procedure. The parameters are set in the MULPLT.DEF file.
Press C to find coda length. The coda length can only be determined if a P-phase is present.
6.2.7 Theoretical arrival times for global and local phases and location
In order to assist in identifying seismic phases, there is an option for displaying the theoretical arrival
times of several global and regional phases while picking phases. The steps to do so are the
following:
1. Before entering MULPLT from EEV, the theoretical travel times have to be calculated for the
current event. This assumes that the origin time and hypocenter is given in the header line or a
subsequent type one line. If not, enter manually (from e.g. PDE) or use the EEV command INPUTEPI
or INPUTONE. Then proceed to calculate the theoretical arrival times using EEV command IASP with
the IASPEI91 traveltime tables (for more details, see section 6.20.3). The same command is also
available inside MULPLT in multitrace mode. All arrival times (or a subset, see 6.20.3) for all stations
in current S-file will now be calculated with program IASP and stored in file iasp.out (no importance
for the use, just for information). See Figure 15 for an example in multitrace mode. Note that very
many theoretical phases can be generated if the S-file has many stations. MULPLT will stop if more
88
phases are used than the dimensions are set up for (include file seidim.inc), and you must use fewer
phases (a warning is given when 500 phases are generated) or set up SEISAN with larger
dimensions, see section 3. Theoretical local crustal phases for the current model can be calculated
with program WKBJ and displayed, see section 6.19. Theoretical phases can also be calculated when
using the location option, see next section.
2. Pick phases: When a trace is displayed on the screen, all theoretical phases inside the time
window will also be shown. To distinguish the theoretical phases, they are prefixed with a y and
displayed below the trace (normal phases have I, E or blank and are displayed above the trace).
Position cursor where you see a phase which you think corresponds to a theoretical phase and press
y. The nearest theoretical phase will now be placed at that position with a prefix E. Only theoretical
phases selected in this way will be written in the S-file. Note that the phase names can be up to 8
characters long, see Appendix 1 for the definition of long phase names.
If the phases fit badly, start looking at the P-phase. If that does not fit the theoretical P-phase, change
the origin time in the S-file so that the P-arrival fits, and recalculate the theoretical phases.
PROBLEM: In multitrace mode, only one theoretical phase can be picked. Replot must be made
before picking the next.
Locate earthquake
If several phases have been read and saved in the S-file, the event can, in multitrace mode, be
located with command l (Locat), just as in EEV. The screen is cleared and the usual location rolls
over the screen. When the location is finished, the plot will reappear and the calculated travel times
will be displayed as synthetic phases (see previous section). In this way it is possible to immediately
visualize the differences between the read and calculated phases. The output files are hyp.out and
print.out as usual.
6.2.8 Instrument correction and magnitudes Ml, mb and Ms
The correction for instrument response is done by taking the spectrum of the selected window of the
trace, dividing with the response function and converting back to the time domain. Any filtering
specified is done in the frequency domain. Filtering is needed in most cases.
Ground motion
Option g(Groun) removes the effect of the instrument and displays a ground motion seismogram.
After selecting g and the zoom window, there is a question of which type of seismogram to calculate:
Displacement (d), Velocity (v) or Acceleration (a). The corrected trace is shown below in
nanometers(nm), nm/sec or nm/(sec*sec) (if response information is available). Note that this might
produce strange seismograms, since e.g. a SP seismograph has very low gain at low frequencies so
noise might be amplified very strongly. It is therefore recommended to also do some filtering when
using the g option.
Amplitude for determining Ml
The w(WA)-option (Wood Anderson) is similar to the g-option except that a fixed 8 pole bandpass
filter is used (1.25 hz - 20 hz). Filtering is done in the frequency domain. For noisy traces it might also
be required to put a filter at the high end. This can be specified in the MULPLT.DEF file. The default is
a the 1.25 - 20Hz filter band. The filters can be set to any other values and number of poles, see
MULPLT.DEF. Ideally, the low cut filter should be set with 2 poles to simulate the seismometer
correctly. However, that will often result in the seismogram blowing up at low frequencies. It probably
only makes a difference for very large events (Ml > 5). NOTE: In SEISAN version 7.1.1 and earlier,
the low cut filter was set by mistake to 0.8 Hz. Repicking amplitudes with the correct filter might
89
change magnitudes of larger events slightly.
Displaying response information
The response function for the current channel can be shown with option ‘:’ (Resp), see Figure 24. If
no response function is given, a message is shown. If the response function is taken from the
waveform file header instead of from the CAL directory, a message is given.
Amplitude for determining Mb:
The attenuation function for determining Mb assumes that the amplitudes are measured on classical
1 Hz WWSSN instruments having a peak gain around 1Hz. To pick ground amplitudes for
determining Mb on instruments with a broader frequency band, like most high frequency SP
instruments, some filtering must first be done. The j(Mb)-option by default filters the seismogram in
the band 0.5 to 5.0 Hz (8 pole filter) and displays the ground displacement. Experience so far
indicates that the filter is adequate. Other filters can be set in MULPLT.DEF.
Amplitude for determining Ms:
The attenuation function for determining Ms assumes that the amplitudes are measured near 20 sec
period. To pick ground amplitudes for determining Ms, some filtering must first be done. The k(Ms)option by default filters the seismogram in the band 0.042 to 0.063 Hz (16-24 sec, 8 pole filter) and
displays the ground displacement. Other filters can be set in MULPLT.DEF.
Problem: If a long trace (large number of samples) is used, the instrument correction might fail (funny
result seen) due to numerical overflow in the spectral conversion. Choose a shorter window.
6.2.9 Determine azimuth of arrival (3 comp or array) and component rotation
Azimuth of arrival from 3-component stations, h(Azim)
If a 3 component station is available, the azimuth of arrival can be determined using the method
developed by Roberts et al, (1989). Display any of the 3 components and press h (Azim). Then select
a zoom window around the P-arrival of a few secs duration for the analysis. The 3 components will
now be displayed below in order Z, N and E and the calculated azimuth, apparent velocity and
correlation will be displayed at the bottom line. In order to check the stability of the estimate, try
different windows and filters. Often, a filter must be used to get reliable results. The displayed azimuth
and apparent velocity is only saved in the S-file when an associated phase is picked. THAT PHASE
MUST BE PICKED ON THE SINGLE UPPER TRACE SEEN ON THE SAME SCREEN. If there is
none, use I or E. The velocity estimate is not very reliable and is dependent on the local velocities. In
order to calculate the apparent velocity, the P-velocity of the top layer must be given. The default
value is 5.0 km/sec, but another value can be set in the MULPLT.DEF file. To get a good estimate,
the correlation coefficient should be as high as possible and positive. The quality of the obtained
azimuth can be tested by locating the event with the calculated azimuth weighted out and observe the
azimuth residual. Figure 23 shows an example.
Azimuth and apparent velocity from array data, FK analysis f(FK)
Using this command, the traces seen on the screen will be put into the FK program and an FK plot
will be displayed. The azimuth and apparent velocity with the highest correlation is selected by the
program, however any other value can be manually selected. The values will ONLY enter the S-file if
associated with a phase in the same way as amplitudes are picked. For more details, see section
6.29.
90
Rotated seismograms
Option u(Rotat) will rotate the horizontal components for the next plot if the two horizontal
components are available. The rotation will display the radial component instead of the N-component
and the transverse component instead of the E-component. The back-azimuth used is displayed
above the trace. All channels will be displayed rotated until u(Rotat) is pressed again This means that
phases can be picked and spectra made with the rotated channel. When picking phases on rotated
signals, these will appear in the S-file with components R or T instead of N and E respectively. This
also means that only if the rotated signals are shown, will the phases read on rotated channels
appear on the plot. The station back-azimuth is obtained in the following way: If a hypocenter is given
in the header line, the angles are calculated using the current STATIONx.HYP file. If no hypocenter is
available, the angle will be read from the S-file under column observed azimuth (47-51) (if not blank)
and the azimuth residual will be added. This option permits the user to first determine the azimuth
with the 3-component option and then rotate the signals with the determined azimuth. Finally, if no
observed azimuth is available, the event to station azimuth + 180 deg. will be used if available
(column 77-79). If no back-azimuth can be found, no rotation is done and an angle of 999 deg. is
displayed. If in single trace mode and choosing the 3-component option AND the rotate option, the
user will be prompted for a rotation angle and the rotated channels will be shown in the usual 3component plot, however, the azimuth determined is done with the unrotated channels.
PROBLEM: In general, the R-channel will use the response of the N-channel and the T-channel will
use the response of the E-channel so for instrument response removal to be correct, the 2 channels
must have the same response curve.
6.2.10 Data manipulation commands
Select other channels: o(Oth)
The channel selection menu comes up again.
Go back one channel in single trace mode, go back one event in multitrace mode if MULPLT is
started from EEV: B(Back)
Select other waveform files from S-file: W(OthW)
If more than one waveform file available for the event, one or several others can be selected.
Delete waveform files:
This can only be done in multitrace mode: The command is d(DelW) and the cursor must be above
the top frame of the plot. There are two possibilities:
(1) Input is from filenr.lis: The current file is deleted and if in default mode, the plot moves on to the
next event.
(2) MULPLT is started from EEV: If only one waveform file is available, the program proceeds as
under (1). The waveform file is deleted and the waveform file entry in the S-file remains. However, if
more than one waveform file is available, the user can use a menu to select which files to delete. Only
the waveform file entries in the S-file are deleted, the waveform files remain. This option is mostly
used with SEISNET.
Delete S-files D(Del S)
This command deletes the current S-file. It can only be used if MULPLT is called from EEV. No
waveform files are deleted.
Merge waveform files given in S-file M(Merge)
91
The files will be merged to one waveform file and the old individual file names removed from the S-file
and replaced by the new file name of the merged file. The original waveform files remain. Files to be
merged will be shown on a menu. Mostly used with SEISNET. The user MUST have files in working
directory. If files are in the data base, they will not be shown on the merge menu.
Output of binary waveform file, O(Out)
It is often useful to be able to select part of a waveform file and save it. The Out option makes an
output file of the traces AS DISPLAYED ON THE SCREEN with exactly the same channels, and time
window in a file with a standard SEISAN waveform name. The output format is always SEISAN, even
if some input files have a different format. The network code in the file name will ALWAYS be the
station code if all channels are from the same station Otherwise the network code has the default
name MERGE. Alternatively the parameter MERGE_WAVEFORM can be set in SEISAN.DEF. The
data is output exactly as displayed on the last screen, so if filtering or instrument response has been
made, the output file will also be filtered or instrument corrected. If any filtering or instrument
response correction has been done, a note will be inserted in the SEISAN waveform header so the
user can see that this is no longer the original data. The note could be e.g. ‘ Displacement 1.250- 20.0
Hz’ indicating that output has been filtered and converted to displacement (nm). Note that numbers
have been scaled so only if the SEISAN file is read with standard SEISAN routines, will the numbers
be correct. If the output file is converted to ASCII by SEIASC, the number shown must be multiplied
with a given scaling factor, see SEISAN binary format description (Appendix 2). There is no response
information in the header other than the short text. Since the station code is still the same, it is
technically possible to correct for the response again using the response information in the CAL
directory, however, be aware that this will give wrong results.
Output of ASCII waveform file
This option only works if parameter SPECTRAL_OUTPUT has been set in MULPLT.DEF. The output
file signal.out contains the last data displayed in the single trace zoom window (in ASCII and real
numbers). This option is a another way (see option O(Out) above) of getting an output file that has
been filtered or instrument response corrected. The main difference is that this file is only for one trace
written in ASCII.
Fixed scaling
Normally all traces are plotted with autoscaling. However, it is sometimes useful to be able to scale
the traces with a fixed scale in order to e.g. compare traces or override the autoscale in case a spike
distorts the autoscaling. Option *(Scale) will prompt the user for a maximum count to use for the
scaling of all traces.
92
Figure 20. Using MULPLT for picking phases. The top shows the original trace and the bottom
the zoomed part. Note that the amplitude has been associated with the phase E and not the
ESg. This means that if the S-phase is deleted, the amplitude will remain.
93
Example of using MULPLT on SUN:
Comments are given with ! in front
This example shows how running MULPLT from EEV would look.
/top/seismo/REA/BER__/1991/01/01-0557-12L.S199101 ! S-file name
Read header from file /top/seismo/WAV/9101-01-0557-12.WNN_13
Plot options:
Interactive picking
Return ! first choice
Multi trace plot on screen, def (0)
Multi trace plot on screen
(1)
Multi trace plot on screen+laser(2)
Multi trace plot on laser
(3)
Multi trace plot on laser
(3)
Continuous on screen
(4)
Continuous on screen + laser
(5)
Continuous on laser
(6)
Stop
(9)
! now comes a menu for selection and then
! the plot appear in Single mode since a
! return was made
The next example shows how to plot many events in one go, first make a list with
DIRF.
dirf 9101-10*
# 1 9101-10-0915-15S.KMY_03
# 2 9101-10-1510-55S.NSS_12
# 3 9101-10-2333-44S.NNN_11
! events from January 10, 1991
mulplt
file name, number, filenr.lis for all
filenr.lis
! plot all events in filenr.lis
Resolution in cm/sec, 0: plot all on one page (default)
0
! scale will be different for each plot!!!
Read header from file:9101-10-0915-15S.KMY_03
Page
1
Channel:
1
Plotfile sent
Read header from file:9101-10-1510-55S.NSS_12
Page
1
Channel:
1
Channel:
2
Channel:
3
Channel:
4
Channel:
5
Channel:
6
Channel:
7
Channel:
8
Channel:
9
Channel:
10
Channel:
11
Channel:
12
! next event in list
Read header from file:9101-10-2333-44S.NNN_11
Page
1
Channel:
1
! etc.
Plotfile sent
Read header from file:9101-10-1510-55S.NSS_12
Page
1
Channel:
1
Channel:
2
Channel:
3
Channel:
4
Channel:
5
Channel:
6
Channel:
7
Channel:
8
Channel:
9
! next event in list
94
Channel:
Channel:
Channel:
10
11
12
Read header from file:9101-10-2333-44S.NNN_11
Page
1
Channel:
1
! etc.
6.2.11 Spectral analysis, s(Spec)
The spectral analysis option is selected in single trace mode. The spectral analysis is based on the
Brune (1970) model and various assumptions about the geometrical spreading and anelastic
attenuation.
The theoretical displacement spectrum d(f) (Brune, 1970) is:
d(f) = G(r,h) * D(f) * Moment*KK /(1+f**2/f0**2)* (4 * pi * DE * V**3))
where G(r,h) is geometrical spreading, r is epicentral distance, h is hypocentral depth, D(f) the
diminution function due to anelastic attenuation, f is the frequency, DE the density, V the velocity at
the source, f0 the corner frequency and KK a factor of 2.0*0.6 to correct for the free surface effect and
radiation pattern (see also application note qspec.pdf in INF)
The diminution function D(f) is written as
D(f) = P(f) * exp (-pi*f*trtime/(q0*f**qalpha)) where
trtime is the travel time from the origin time to the start of the spectral window and
P(f) = exp (-pi*kappa*f)
is meant to account for near surface losses (Singh et al, 1982) with the constant kappa having a value
of the order 0.02 sec. Anelastic attenuation Q is assumed to be frequency dependent following the
relation Q = q0* f**qalpha.
The geometrical spreading has been defined to be dependent on the wave type with several
possibilities, all made equivalent to a distance called geo_distance (GD) such that geometrical
spreading is expressed as 1/GD. There are several possibilities for GD:
P-waves:
GD is the hypocentral distance (HD) = sqrt (r*r +h*h) so body wave spreading is assumed.
S-waves:
The geometrical spreading has been made dependent on distance and depth. At short distances, the
geometrical spreading is assumed to be body wave spreading. For distances beyond the HerrmannKijko distance (default of 100 km) and a shallow focus, the following relation is used:
G(r,h) = 1/r =1/GD
G(r,h) = 1/sqrt(100*r) =1/GD
for r < 100 km
for r > 100 km
which is commonly used (Herrmann, 1985, Herrmann and Kijko, 1983). This relation assumes
surface wave dispersion for epicentral distances larger than 100 km. In SEISAN 100 km is the default,
however it can also be set to any other value by the parameter HERKIJ_DISTANCE (see later).
The above relation breaks down if the depth is large or comparable to the epicentral distance and in
95
that case body wave spreading is again assumed. In order to get a smooth transition from surface
wave to body wave spreading, it is assumed that the relation changes nearly linearly from surface
wave spreading to body wave spreading between the depths GEO_DEPTH1 to GEO_DEPTH2. For
depth less than GEO_DEPTH1(default 50 km), Herrmann-Kijko spreading is assumed, for depths
larger than GEO_DEPTH2 (default 100 km), body wave spreading is assumed with the transition in
between. In each case the geometrical spreading term is given as the equivalent GD, which is also
recorded in the database. These 3 parameters can be used to change geometrical spreading. If e.g.
HERKIJ_DISTANCE is 10 000 km, body wave spreading is always used. For more info, see
qspec.pdf in INF.
From the spectral parameters, source radius and stress drop can be calculated as follows:
Source radius = 0.35 * VS /f0
where f0 is the corner frequency and VS the S-velocity at the source.
Stress drop = 0.44 * Moment /(source radius)**3
The spectral analysis is used in two ways. The first and most common is to make the attenuation and
instrument corrected displacement spectrum and determine the flat spectral level OM0, and corner
frequency f0 from which the seismic moment, source radius and stress drop can be calculated. The
second option is to display the instrument corrected spectrum (displacement, velocity or acceleration)
and model the spectrum for corner frequency and attenuation parameters. In this case no correction
for attenuation should be made.
Spectral analysis to determine moment, source radius and stress drop:
Select the spectral option, s(Spec). Before the spectrum comes up, you will get a question of the type
of spectrum wanted. The possibilities are displacement (d), velocity (v), acceleration (a) or raw
spectrum (r). For determination of Moment etc, the displacement spectrum MUST be selected. Unless
raw spectrum is selected, the spectrum will be instrument corrected. If no response file is available in
CAL, a message will be displayed on the screen and the raw spectrum calculated. At this stage it is
also possible to change the velocity from the MULPLT.DEF value or the moment given in the S-file
(see spectral fitting below). The spectrum shown will normally show both the spectrum from the
selected time window as well as a noise spectrum from an identical length time window at the start of
the trace. IF NO NOISE SPECTRUM is desired, select spectrum with capital S instead of s.
The spectral analysis produces two output files:
com_spec.out: The complex spectrum with some additional information needed for surface wave
analysis, must be displacement spectrum.
amp_spec.out : The real spectrum given as frequencies and amplitudes. The files are only generated
if parameter SPECTRAL_OUTPUT is set in MULPLT.DEF. Setting this parameter will also generate
an ASCII waveform file with the input signal used.
Power spectra: The above spectra can also be displayed as power spectra if capital letters are used.
Using e.g. 'V' instead of 'v' will show the power velocity spectrum.
When the spectrum comes up (see example in Figure 22, the axis units are log amplitude in
nanometers-sec (displacement) versus log frequency (Hz). The cursor can be used to select the
level, corner frequency and slope by defining the spectrum with a 3 point selection. This 3-point
selection is finished with f, q or r with the same meaning as in picking mode. The spectral values are
displayed on the screen once q, f or r is pressed. The abbreviations are
General parameters
96
Vel:
Dens:
Dist:
q0:
qalpha:
k:
Velocity used (km/sec) (Vp or Vs)
Density (g/(cm**3)
Hypocentral distance (km)
q0 for spectral amplitude correction
qalpha for spectral amplitude correction
kappa
On top of the general parameter is indicated which kind of spectrum is assumed, P or S. In order for
the program to automatically determine which kind of spectrum to assume, there must be a P or S
reading displayed on the screen near the time window analyzed. The reading must be within 10 sec
of the start of the window. If both a P and S-reading is within 10 secs, the nearest phase is chosen. If
it cannot be determined which kind of phase is analyzed, the user will get a question to select type of
phase (can also be changed later when spectral choices come up) The determination of which phase
influences the further calculation of geometrical spreading and moment (uses P or S-velocity).
If f is selected, the spectral values together with calculated moment etc are stored in the S-file at the
next key press (see parameters below). Spectral values in S-files accumulate, since no old values are
deleted !!!. This is because the spectrum might be made under different conditions (start time, time
window etc). The input parameters for the spectral analysis is given in file MULPLT.DEF, which can
be in either DAT or the working directory, see below. Additional parameters for geometrical spreading
are given in SEISAN.DEF in DAT.
The spectral parameters are calculated using the relations
Moment = 4 * pi * DE * V**3 * 10**OM /( G(r,h) * KK)
where V is the seismic wave velocity at the source (P or S if P or S-spectrum respectively) and OM
the spectral flat level on the attenuation corrected displacement spectrum.
Moment magnitude = 2/3 * log10(moment) - 6.06 which is equivalent to the relation
Moment magnitude = 2/3 * log10(moment) -10.73 if moment is in dynes-cm (Kanamori, 1977).
The moment is calculated in Nm, the source radius in km and the stress drop in bars. All results are
written to the S-file. Below is an example:
SPECAITK S
SPEC ITK S
Z MO 13.0 ST
Z K 0.002 T
4.2 OM
7 GD
1.5 f0 9.45 R
.22 AL 2.50 WI 4.0 MW
52 VP 6.00 DE 3.00 Q0
.0 QA 1.00 VS
2.6 3
3.5 3
Note that no special line has been created in the Nordic format. Comment lines are used with SPEC
at the start of the line followed by station and component. Only the first 4 characters of the 5 character
station name is used. An A after SPEC, means automatic determination. The information is:
MO:
ST:
OM:
F0:
R:
AL:
WI:
MW:
T:
K:
GD:
VP or VS:
log of moment, unit Newton*m
Stress drop in bars
log spectral level (nm-sec)
Corner frequency (Hz)
Source radius (km)
Decay of log spectrum
Spectral window used (secs)
Moment magnitude
Start time of window for spectrum in hr, min, sec
Kappa
GEO_DISTANCE in km
Velocity in km/sec at source for P and S-spectra respectively. The P or S in this line
indicated if the spectrum is a P-spectrum or an S-spectrum. It MUST be P or S to be
used for magnitude determination. A ‘?’ is put in if MULPLT does not know which kind
of spectrum is (no P or S reading near start of spectral window). This can be changed
97
DE:
Q0:
QA:
by editing the S-file afterwards.
Density in g/cm**3
q0 in relation Q = q0 * f ** qalpha
qalpha
Note: In earlier versions (before version 7.0), the field for kappa was used for the travel time to start of
window. This can be calculated from origin time and the start time of the window.
NOTE: MOMENT IS NOT CALCULATED IF THE SPECTRUM IS NOT IN DISPLACEMENT.
When doing an UPDATE of the database or just a location with HYP, all distance dependent spectral
values are recalculated and average values written into the output file. Mw will be calculated from the
average value and written in the header line. However, the original distance dependent Q and
kappa correction is not changed, since this correction was used to modify the spectrum used for
reading parameters. Normally a small distance change has insignificant influence on the spectral level
or the corner frequency so the Q-correction should be no problem. Spectra of the same type (P, S or
?) and from the same channel are overwritten. Only in case of UPDATE are the values written back
into the database.
Display of spectral parameters: Program MAG can read and plot relations between spectral and
source parameters. Program REPORT can read spectral parameters and combine in a table.
Potential problem with Q-correction: If the origin time in header is wrong, the Q-correction can be very
wrong.
There must be a phase line in the S-file with component and distance corresponding to the spectra
made in order for the spectral values to be calculated.
Spectral fitting
Once the spectrum has been shown (displacement, velocity or acceleration), a theoretical spectrum
can be calculated and superimposed on the observed spectrum in order to forward model either
source parameters or attenuation.
Entering constants and modeling: The modeling can only take place when the spectrum is seen on
the screen.
Press s or S and a question will appear to enter the constants f0, k, Q0 and qa which are as defined
above except qa is Qalpha. Once these parameters have been entered (terminate with return), the
theoretical spectrum (displacement, velocity or acceleration depending on what is used for the
spectrum) is calculated and superimposed on the observed spectrum. The parameters used or
calculated are displayed. The level of the theoretical spectrum is adjusted so it approximately passes
through the observed spectrum and the level difference is printed out on the screen (see below). S or
s can now be pressed again and a new theoretical spectrum calculated and plotted. To get out of the
spectral fitting loop, type r or q as usual.
Which constants and parameters are used: The moment is taken from the S-file if an average
moment has been calculated (see UPDATE command). If no moment is available, it can also be
entered the first time the spectrum is shown. If no moment is given, a log moment of 1.0 is used. The
distance and depth is likewise taken from the S-file. If no distance is available, a distance of 1 km is
used. If all 4 parameters f0, k, Q, qa are entered, stress drop is calculated with the relation given
above. If the corner frequency is given as zero, the user will be asked to enter the stress drop and the
corner frequency is calculated from the stress drop. If Q is zero, no Q-correction is made.
IMPORTANT: The Q and qa used here are distinct from the Q0 and Qalpha used for making the
amplitude spectrum and both should not be used when modeling since this would imply a Qcorrection two times. The best way is to use Q0=0 and kappa=0, so that Q is only corrected for
when modeling. The distance used is everywhere is GEO_DISTANCE.
98
The spectral parameters shown are:
Obs - calculated level: The difference in log absolute level of the observed and calculated
spectra. If a correct moment is used it should be small,
in the order of 1.
Moment:
Moment used
Geo dist:
Geo distance used
Stress drop:
Stress drop in bars
f0:
Corner frequency
k:
Constant used in diminution function
q:
q0 used in spectral fitting
qa:
qalpha used in spectral fitting
Power spectrum and noise spectrum
The 3 types of spectra (displacement, velocity and acceleration) can optionally be made as power
spectra. Instead of selecting the type of spectrum by pressing d, v or a, just press the same
characters in upper case and the power spectrum will be shown.
In seismic noise studies, the seismic background noise is often displayed as acceleration power
spectral density in dB relative to ((1m/s**2)**2)/Hz. Instead of selecting d, v or a, press n instead. The
plot shows the Peterson (1993) new global high and low noise models superimposed on the observed
spectrum (Figure 21). When doing noise spectra, no attenuation correction is done. The normalization
of the spectrum is as follows
2
P = F DFT ×
Δt 2
×2
T
where P is the Peterson Power spectrum, F
is the discrete Fourier transform, Δt is the sample
interval and T is the length of the time window. The factor 2 comes from the fact that only the positive
frequencies are used so only half the energy is accounted for. The total power is proportional to the
length of the time window since the noise is considered stationary, so by normalizing by T, the length
of the time window should not influence the results. This noise option is a handy method of checking
the noise characteristics of a given seismic station and compare it to global standards. This kind of
analysis can also be done with the SPEC program (section 6.23). For more information, see
instrument.pdf in INF.
DFT
Figure 21. Example of a noise spectrum.
Problems: There is currently no check if a displacement seismogram has been calculated when
calculating the spectral parameters. If spectral analysis is done outside EEV (output in MULPLT.OUT)
99
or with EEV when there is no origin time and/or epicentral distance, the output results are wrong for
moment etc. Before calculating moment etc, the S-file MUST HAVE BEEN UPDATED SINCE BOTH
THE DISTANCE AND ORIGIN TIMES ARE USED. If the spectra get very high amplitude levels when
correcting for instrument, this might be caused by correcting for Q. With a Q of 100 and a distance of
10 000 km, this gives a very large correction. The Q-correction can be disabled in the MULPLT.DEF
file.
If picks are made, but no readings appear in the S-file or readings appear with wrong component, the
waveform file component might not have been defined in subroutine componen.for.
If poles and zeros are used to remove the response, rotation cannot be used at the same time.
100
Figure 22. Spectral analysis
On top the original trace is seen and on the bottom the displacement spectrum (log -log, unit
nm-sec and Hz). The level and slope has been indicated interactively. Note the noise spectrum
at the bottom of the figure.
101
Figure 23. Three component analysis
On top the Z-channel is shown together with the window used for the 3 channels Z, N and E
shown below. The signals below has been filtered between 1 and 5 Hz and the resulting
azimuth of arrival is 160 degrees and a correlation coefficient of 0.2. The apparent velocity is
9.8 km/sec.
102
Figure 24. Plotting response curves
The figure shows the amplitude and phase response for station SUE, component S Z. The
response is the one which will be used in analysis irrespective of whether it is taken from the
file header or the CAL directory.
103
6.2.12 The MULPLT.DEF and SEISAN.DEF files
In the MULPLT.DEF and the SEISAN.DEF files, it is possible to set the various parameters for
MULPLT. Nearly all parameters are set in the MULPLT.DEF except geometrical distance parameters,
which are set in SEISAN.DEF since these parameters also are used by HYP. MULPLT will operate
without DEF-files using hardwired defaults. The MULPLT.DEF can be located in the working directory
and or in DAT. The if a DEF-file is present in the working directory, it overrides the file in DAT.
In MULPLT.DEF, several groups of parameters can be set: The keyboard, default channels to use
and analysis parameters (e.g. for spectral analysis). The parameters are identified by keywords, see
example file below for explanation.
Example file:
This file is for defaults for MULPLT and called MULPLT.DEF. The name must be in upper case on
SUN. The following shows the parameters, which can be set. The file can contain any number of lines
in any order, only the lines with recognized keywords and a non blank field under Par 1 will be read.
The comments have no importance.
KEYWORD............Comments.............Par 1.....Par 2
X_SCREEN_SIZE
Size in pixels
90.0
PHASE NAME KEY
Phase key and phase
PHASE WEIGHT KEY
Weight key and weight
PHASE MOUSE KEY
Mouse key character
SPECTRAL Q0
Q0
440.0
SPECTRAL QALPHA
Q = Q0**Qalpha
0.70
SPECTRAL KAPPA
SPECTRAL P-VELOCITY P velocity
6.2
SPECTRAL S-VELOCITY S velocity
3.6
SPECTRAL DENSITY
Density
3COMP VELOCITY
velocity for 3 comp
RESOLUTIONX
# points pl. screen 1500.0
RESOLUTIONHC
# points pl. hc
3000.0
NSORT_DISTANCE
0: no sort, min ph.
NCHAN PER SCREEN
max # chan screen
CHANNEL SORTING
1.0
SPECTRAL F-BAND
0.01
20.0
AUTO_LOCATE
0,1,2
0,1,2
0.0
2.0
AUTO_PROCESS
0,1,2
name
0.0
ls
SPECTRAL OUTPUT
1.0
WOOD ANDERSON HIGH CUT
20.0
### Set Filter type 0 for bndpas 1 for recfil routine
FILTER TYPE
0.0
### Setup user-defined filters
FILTER 1
z-key
0.01
0.1.
FILTER 2
x-key
0.1
1.0
FILTER 3
v-key
1.0
5.0
FILTER 4
b-key
5.0
10.0
FILTER 5
n-key
10.0
15.0
FILTER 6
m-key
15.0
25.0
FILTER 7
;-key
2.0
4.0
ML LOW CUT AND POLES
ML HIGH CUT AND POLES
MS LOW CUT AND POLES
0.05
2.0
MS HIGH CUT AND POLES
0.1
2.0
MB LOW CUT AND POLES
0.5
2.0
MB HIGH CUT AND POLES
4.0
2.0
BANDPASS FILTER
5.0
10.0
CODA AUTO
1.0
AUTOCODA FILTER
0.1
10.0
AUTOCODA STA
3.0
AUTOCODA RATIO
1.2
DEFAULT CHANNEL
KONO
LU Z
DEFAULT CHANNEL
Station and channel ITK
S Z
-----------------------------------------------------------------------------In SEISAN.DEF, the following parameters can be set, format as in SEISAN.DEF
SPECTRAL GEO_DEPTHS
10.0
14.0
HERKIJ_DISTANCE
100.0
104
All parameters are within column 41 and 60 and each occupying up to 10 characters.
NOTE: If any of the phase or weight keys are redefined, all previous defaults disappear.
DEFAULT CHANNEL: All channels are default if not given. For routine display, it is useful to only
select some channels.
PHASE NAME KEY: The keys associated with given phases. Remember that I, E or a blank MUST
be part of the name so it is not possible to chose a name like "P", it must then be " P" (note the blank
in front of P). About 10 phase combinations are currently default as seen on the pick display. If a new
phase key is selected, you must define all the keys you want to use for phases including all the
predefined phases. The combined onset/phase key can be up to 9 characters.
PHASE WEIGHT KEY: The defaults are upper case 1,2… to 0 for weights 1,2... to 0 . Again,
choosing just one other key, and all must be redefined.
PHASE MOUSE KEY: The default is blank. Normally no redefinition is needed since the mouse
character is defined in SEISAN. The key can be defined as a character or the ASCII code written as a
real number.
SPECTRAL P-VELOCITY: P-velocity in km/sec, default 6 km/sec
SPECTRAL S-VELOCITY: S-velocity in km/sec, default 3.5 km/sec
Both above parameters must be set separately, the Vp/Vs in STATION0.HYP is not used to calculate
one from the other. The values go into the S-file the first time spectra are calculated. if values are
changed later in the MULPLT.DEF file, no change will be made in the S-file, old values remain.
SPECTRAL Q0: Q is defined as q0 * f**qalpha, default 0 meaning no Q-correction
SPECTRAL QALPHA : See above, default 1.0, NOTE: Q is only used when doing spectral analysis
and has no effect on the displacement seismograms.
SPECTRAL DENSITY: Density for spectral analysis (g/cm**3), default 3.5 g/cm**3
SPECTRAL KAPPA: Near surface attenuation, default 0.0 meaning no attenuation
SPECTRAL GEO_DEPTHS: Depth range where geometrical spreading changes from surface wave
to body wave spreading, S-waves only. Default 50 and 100 km. This is only used if distance is larger
than HERKIJ_DISTANCE. THIS PARAMETER IS NOT SET IN MULPLT.DEF, BUT IN SEISAN.DEF,
MENTIONED HERE SINCE IT IS IMPORTANT FOR SPECTRA.
HERKIJ_DISTANCE: Epicentral distance at which geometrical spreading changes from body wave
spreading to surface wave spreading, S-waves only. Default 100 km. THIS PARAMETER IS NOT
SET IN MULPLT.DEF, BUT IN SEISAN.DEF, MENTIONED HERE SINCE IMPORTANT FOR
SPECTRA
3COMPVELOCITY: Velocity used (km/sec) in 3 component azimuth analysis. Default is 5 km/sec.
CHANNEL SORTING: If set to 1.0, channels and filenames are sorted alphabetically, if 0.0, no
sorting. Default = blank is sorting.
NCHAN PER SCREEN: The number of channels to be displayed per screen. Default = blank is 99
channels.
NSORT_DISTANCE: If not there or zero, channels are plotted in the order as they appear in the
waveform file. If set to a positive number, that number is the minimum number of phases to be
present in the S-file in order for the S-file ordering to be used for plotting order, if plotted from EEV. If
not plotted from EEV, any positive number will indicate sorting in waveform file header time order.
105
Default 0.
X_SCREEN_SIZE: Size of initial X-window in % of total screen. Default 90 %.
RESOLUTIONX and RESOLUTIONHC is the number of points plotted on the screen or laser printer
respectively. If e.g. 1000 points are plotted, this means that the remaining points are skipped although
some primitive smoothing is done. Choosing too few points can lead to funny looking seismograms
with aliasing effects and using all points will slow down the plotting. Resolutionx is for the screen and
resolutionhc for the hardcopy. NOTE: If using MULPLT mode where both screen and hardcopy is
used, it is the hardcopy resolution, which is used for both. Default 1000 and 3000 respectively.
SPECTRAL F-BAND: Spectral range (Hz) used for spectral plots. Default values are 0.05 to 20.0 Hz.
AUTO_PROCESS: Immediately following registration, MULPLT can run any program specified here.
Since the event name has been put into memory, the program can operate on the newly registered Sfile. Parameter one has the options: 0: Do not auto process, 2: Ask the user if autoprocess, 3:
Autoprocess without asking the user. Parameter 2 gives the name of the process to run. The name is
limited to 10 characters. Default, no auto processing.
AUTO_LOCATE: Immediately following registration, MULPLT can locate the newly registered event
and put the location into the database. Parameter one has the options: 0: Do not locate, 1: Ask the
user if locate, 2: Locate without asking the user. Parameter 2: 0: Do not save in database, 1: Ask if
saving in database, 2: Automatically save in database. Default, no auto locate.
SPECTRAL OUTPUT: If parameter set to 1, two output files are created for each signal spectrum.
com_spec.out is the complex spectrum and amp_spec.out is the real spectrum. Default 0.0. In
addition, the single trace zoom window is saved in signal.out.
FILTER: Change definition of filters 1 to 7. The settings affect both the shortcut keys and the menu
boxes.
FILTER TYPE: Choice between two different filter routines that basically do the same. Use 0.0 for
routine ’bndpas’ (default) and 1.0 for ’recfil’. Bndpas is a recursive Butterworth bandpass filter routine.
Recfil is an IIR filter routine (various filter types supported, but in SEISAN only Butterworth is used),
which works as low-, high- or band-pass, however, gives problems for bandpass at low frequencies.
Low- and high-pass when using recfil are given by setting either the upper or lower frequency limit to
0.0, respectively.
WOOD ANDERSON HIGH CUT: A high cut filter can be set when using Wood Anderson simulation.
Default 20.0 Hz.
ML LOW CUT AND POLES and ML HIGH CUT AND POLES: Filter band for Wood Anderson filter.
Ml high cut is the same as WOOD ANDERSON HIGH CUT with respect to frequency. Default values
are 1.25 Hz to 20 Hz and 8 poles.
MS LOW CUT AND POLES and MS HIGH CUT AND POLES: Filter band for Ms filter. Default values
are 0.042 Hz to 0.063 Hz and 8 poles.
MB LOW CUT AND POLES and MB HIGH CUT AND POLES: Filter band for Mb filter. Default
values are 0.5 to 8.0 Hz
BANDPASS FILTER: When using all defaults from EEV (option PO), a bandpass filter can be set.
Default is no filter. The parameters are lowcut and highcut for parameter one and two respectively.
CODA AUTO: Enable automatic coda determination (YES or NO). Default is NO.
AUTOCODA FILTER : Filter band for automatic coda: Default 5 to 10 Hz.
AUTOCODA STA : Auto coda short term average: Default 5.0 secs.
106
AUTOCODA RATIO : Autocoda ratio. Default 1.5.
6.2.13 Distance trace plot with GMT, TRACE_PLOT (Unix only)
TRACE_PLOT is a simple program to create a distance trace plot using GMT programs (Generic
Mapping Tools, http://gmt.soest.hawaii.edu/). The axes of the plot are time and distance, and the
traces are centered on the respective epicentral distance. The input to the program is a single event
in Nordic format (S-file). From the S-file, the program reads the origin time, epicenter location and the
names of the associated waveform files. TRACE_PLOT reads the waveform data and writes the x-y
coordinates of the lines in the plot to a file that is then used as input to the GMT program psxy. The
TRACE_PLOT program removes the DC from the data and as an option can apply a band-pass filter.
The output of the program is a Postscript file (trace_plot.ps) and a batch file that can be modified and
used to rerun the GMT programs (trace_plot.bat). The parameters are set in the trace_plot.par file,
which can be located either in the DAT or in the working directory. An example is seen in Figure 25.
The parameters in trace_plot.par are:
FILTER: The pass-band filter limits can be specified through the FILTER parameter.
DISTANCE: The distance range (y-axis) for the plot.
TIME: The time range in seconds (x-axis).
AMPLITUDE_SCALE: The amplitudes are scaled for every trace individually, by [amplitude/(max
amplitude) * AMPLITUDE_SCALE].
STATION_SFILE_ONLY: This variable can be set to 1.0 to only plot traces that are listed in the S-file,
the default is 0., which plots all traces without checking if they are present in the S-file.
TIME_ORIGIN: In the current version, the origin of the time axis corresponds to the origin time of the
event.
COMPONENT: This can be used to select components for plotting, in case no component is defined,
TRACE_PLOT will show all vertical component traces.
Example of trace_plot.par:
KEYWORD............Comments.............Par 1.....Par 2
FILTER
DISTANCE
TIME
AMPLITUDE_SCALE
STATION_SFILE_ONLY
TIME_ORIGIN
COMPONENT
COMPONENT
filter range
dist range
time window
amplitude/max*scale
0. if any station
1. if station has
to be in s-file
1.=origin time
0.=file start time
0.1
0.
0.
15.
0.
1.
BH Z
HH Z
5.0
440.
250.
107
Figure 25. Example of TRACE_PLOT output.
108
6.3 Plotting epicenters, EPIMAP
The command for plotting epicenters is EPIMAP <file>, where the optional file is a file with EPIMAP
commands. If file is not given, the user will be prompted for the input. The program can plot land contours,
epicenters, macroseismic intensities, stations and level contours as well as depth profiles. It is possible to
zoom in on selected areas (option by Mario Villagran). The program has been much revised by Jim
Bolton.
Input files: Land contours and other contours
The program will look for all files ending with .MAP located in the DAT directory. The user can then
choose any one or a combination of files. The users own contour files (e.g. faults) can be added to the
DAT directory. A very detailed world map is available on the SEISAN CD and on the SEISAN web site.
Areas can be selected out of these files with program SELMAP.
Stations
Epimap will look in STATION0.HYP for station coordinates. It will search first in the working directory, then
in DAT.
Epicenters
The user will be prompted for epicenter input files. The format can be Nordic or Nordic compact.
Magnitudes are plotted proportional with symbol size unless the ellipticity option is selected in which case
the error ellipses are plotted (if smaller than 100 km). Name of intensity files (SEISAN standard format,
see Appendix 1) are also entered here. The file name must have the 3 letters ‘mac’ after the ‘.’ See also
section 6.32.
Input files for EPIMAP can be made e.g. with the COLLECT command which collects S-files into one file
or with the SELECT command selecting data from the database using several criteria. HYP also
generates a CAT-file (hyp.out) which may be used as input to EPIMAP.
Magnitudes
The program will read all 3 magnitudes (magnitude1, magnitude2 and magnitude3) in the header line. It
will use the first non-zero magnitude in the order magnitude1, magnitude3 and magnitude2. Epimap will
search the first header line only. If it is desired to use a particular magnitude from any header line for
plotting, use MAG program first to select particular magnitude type which is then placed in first header line
magnitude position one. Program NORHEAD can move magnitudes from following header lines to the
first line. Program REPORT can move magnitudes around on the header line.
A typical run is as follows, comment after !:
Projection menu
===============
1:
2:
3:
4:
5:
6:
7:
8:
POLAR STEREOGRAPHIC
ORTHOGRAPHIC
MERCATOR
LAMBERT EQUAL AREA
GNOMONIC
AZIMUTHAL EQUIDISTANT
STEREOGRAPHIC
EQUIDISTANT CYLINDRICAL
conformal, azimuthal
view from infinity, azimuthal
cylindrical, conformal
azimuthal
Great Circles are straight lines
distance from origin is to SCALE$
conformal, azimuthal
109
9: OBLIQUE MERCATOR
10: MOLLWEIDE ELLIPTICAL
11: SANSON'S SINUSOIDAL
cylindrical, conformal
pseudocylindrical, equal area
pseudocylindrical, equal Area
Please enter projection number
:
Enter latitude range of the map :
Enter longitude range of the map :
3
60 70
0 30
! N
! E
positive
positive
Center of geographical map space is ( 65.0, 15.0) degrees.
! can be used e.g. to make an accurate MERCATOR at high latitude
Press <return> to accept these as the reference latitudelongitude for the projection or <N>o to enter your own
co-ordinates :
Enter latitude of any grid line and also the grid spacing
: 60 2
! possible to have grid spacing at any value
Enter longitude of any grid line and also the grid spacing
: 0 4
DO YOU WANT THE EVENTS NUMERATED? (Y/N=RETURN)
! a sequential number will be plotted besides each hypocenter, the
corresponding hypocenters are found in output file epimap.out.
Plot
Plot
File
Plot
Plot
!
title (max 60 chars), or press <return> for none:
error ellipses (y/n=return)?:
! from HYP output
name for contour levels, or press <return> for none: ! format below
place names (P) or
! file format below
all (a) or some (s) stations with a label, ! if s, question given about
! which station, give in UPPER CASE all stations
without a label (X), or none <return>
...Enter in uppercase if you wish the symbols to be filled
and this facility is available...
! only filled out in Postscript
:
Available colour index values are:
1) Blue
2) Green
3) Red
4) Yellow
5) White
6) Black (default)
Enter epicenter filename and colour index,
separated by a blank, otherwise press <return>
collect.out 2
! plot first file green
Input file is Nordic
! full Nordic format of input file
Enter epicenter filename and colour index,
separated by a blank, otherwise press <return>
select.out 3
! plot second file red
Input file is Compact
! compact file format of input file
Enter epicenter filename and colour index,
separated by a blank, otherwise press <return>
Enter the following in uppercase if you wish the symbols
to be filled and this facility is available...
By default, symbols will be plotted according to
magnitude, do you wish them to be plotted according
to magnitude range ?
==== Loading Epicenters...
! now comes the plot, see below for options
Earthquake locations inside the window are in epimap.out
Coordinates of the surrounding area are in epimap.cor
Area-selected epicenters are in epimap.are
Plot file is called epimap.eps
Interactive options:
110
When the plot is shown, there appears in the lower left-hand corner a menu of several options:
Q: Quit
P: Profile
A: Area
Z: Zoom
Press one of the letters to continue.
P: Profile
One or several depth section windows can now be selected with the cursor. First move the cursor to
where the section shall start (from where distances are calculated), press any character to select point,
move cursor to end of profile, press any character to select. A line between the two points is now plotted.
Move the cursor to a point on the side of the line and press any character. A rectangle defined by the
three points is now drawn, which defines the area used for the section. If more than one section is wanted
(up to 9), press the number of sections instead. The selected number of profile boxes will now be plotted,
all the same size. Pressing any character will draw the depth sections auto scaled, while PRESSING
THE CHARACTER F, THE X AND Y SCALES ARE EQUAL and determined by the horizontal extension.
When the first section appears, you can either press q to quit or any other character to plot next profile or,
if the last profile, replot epicenter map and select new sections. IF YOU WANT ALL SECTIONS TO
REMAIN IN PLOT FILE, QUIT AFTER PLOTTING THE LAST PROFILE. The plot file always stores what
has been plotted so far, and is overwritten when a replot is made.
It is also possible to plot a previously defined profile by entering O. The parameters are then taken from
file profile.out. This file stores the last parameters selected by EPIMAP, but can also be edited by the user.
A: Area
Select, by clicking with the cursor, at least 3 points defining a polygon within which epicenters are
selected. A new plot is made enclosing only the polygon and showing the epicenters within the
polygon. The corresponding epicenters (S-files) are in file epimap.are.
Known bug: Sometimes epicenters are still left outside, SELECT can be used instead.
Z: Zoom
Similar to Area, however a rectangle is selected by defining just the 2 diagonal corners.
Plotting place names
If option P is used when the program asks for place names or station codes, the user will be prompted for
one or several files with place names. The place name file format is:
name
latitude_degrees longitude degrees
eg:
Edinburgh
Edinburgh
55.94422 -3.20096
55.94422 -3.20096
or
etc.
The only requirement is that at least 2 blanks separate the place name and the geographical co-ordinates.
Note that the place name can contain one or more blanks, however each blank must occur singly.
111
Epimap contour file
EPIMAP has a simple contouring routine accepting a regular spaced grid. Below is an example (output
from EQRSEI). The top part of the file is just comments, the data starts at "Fields to use". The data must
come in longitude, latitude pairs (+ value of contour) in order as shown below. The contour value is plotted
exactly as shown below. E. g. the value 117 is plotted as __117_____ where "_" is blank. By specifying
____117.0_, the value would be plotted as 117.0 and moved one space to the left on the plot. Currently
only programs EQRSEI (version 7.0) and CRISEI from SEISAN version 6.0 make contour files. In the
DAT directory, there is an example of an EQRSEI.OUT file
NB: In the input file shown below, the FIRST COLUMN MUST BE blank.
Test Case 1.
NSTEP =
10 JCALC =
0 JPRNT =
1 IATTE =
0
LIST OF EXAMINED INTENSITIES
3.91
4.61
5.01
5.30
5.52
5.70
RISKS DESIRED
.0100
.0050
.0000
.0000
.0000
.0000
.1000
ATTENUATION DATA=
.0200
C1
6.16
C2
.64
C3
-1.30
NO. OF GROSS SOURCES
3
NO. OF SUBSOURCES IN GROSS SOURCES
3
SIGMA
.50
2
RZERO
25.00
RONE
AAA
5.0010000.00
1
This file is input for epimap contour plotting. The input of
parameters must appear as listed below with the actual data
starting after the line Fields to use. There can be any
number of header lines. The contour values plotted are
plotted with the number of decimals given below. The last
part of the file is the actual longitudes, latitudes and
levels. Color use is optional and the field can be left
blank, the codes are as follows:
1: blue
2: green
Min and max level
3: red
115.1800
4: yellow
5: white
6: black
129.6700
Fields to use
Latitude range and number of values
Longitude range and number of values
Contour level to plot and color
Contour level to plot and color
Contour level to plot and color
Contour level to plot and color
Contour level to plot and color
Contour level to plot and color
Contour level to plot and color
Contour level to plot and color
Contour level to plot and color
Contour level to plot and color
5.00000
40.0000
117.620
6.00000
40.0000
118.490
7.00000
40.0000
119.080
8.00000
40.0000
119.390
9.00000
40.0000
119.390
10.00000
40.0000
119.080
11.0000
40.0000
118.490
12.0000
40.0000
117.630
13.0000
40.0000
116.510
14.0000
40.0000
115.180
5.00000
41.0000
119.680
6.00000
41.0000
120.620
7.00000
41.0000
121.260
8.00000
41.0000
121.600
etc
I.........I.........I.........I
40.00
49.00
10
5.00
14.00
10
117
2.0
119
3.0
121
123
125
127
129
131
133
135
BBB
.00
112
EPIMAP output files:
epimap.out: Gives a numbered list of all events within main window. This can be used in connection with
the number option.
epimap.cor and epimap.are: If option A (selecting area) has been used, the coordinates of the corners will
be given in epimap.cor and the complete events (S-files) within the selected area, in epimap.are.
epimap.num: A compact file of epimap.out with the numbers plotted.
epimap.eps: Postscript plot file of epicenters and possible profiles. If only one profile has been selected,
all is on one page. If several profiles are selected, there will be two profiles per page up to a maximum of 6
pages (one with map and 5 with profiles).
epimap.inp: This file is storing all input parameters of the run and can be used to run epimap again without
entering any parameters. The file can be edited if a run has to be repeated with e.g. a new epicenter file.
The file can have any name so several predefined plot definitions can be stored and thereby automate
map production.
profile.out: The file stores the parameters used with the profiles. The file is overwritten for each new profile
parameter selection. An example is:
60.93583
27.8
3
7.21519
63.29655
1.36709
63.39875
5.01266
The first line gives latitude and longitude of the 3 points used for selecting profile (see explanation for
interactive section), next line the azimuth calculated for the profile and the last line gives the number of
profiles. The file can be used to repeat the same profile as in an earlier run or to predefine a more exact
profile than can be selected with the cursor.
profile.num: Output of distance and depth of the profile in km. Distance is only correct in unzoomed plots.
Problems: Known bug: When selecting events with polygon, sometimes some events remain outside
Figure 26 and Figure 27 shows examples of plots made with EPIMAP.
113
Figure 26. An example of using EPIMAP. The top shows epicenters plotted and the bottom the
first of a series of profiles. The frames on the top plot show the location of the profiles.
114
Figure 27. An example of using EPIMAP with area selection. The top plot shows where the
area is selected, while the bottom plot shows the selected area.
115
6.3.1 W_EMAP, Version 4.3 Windows based map program
Program and documentation by Fernando Carrilho, [email protected]
Program must be installed in addition to SEISAN
This program was developed to be used on seismic routine processing. Its main features are the
capability of allowing visualization of epicentre locations, seismic stations, error ellipses, coastlines,
macroseismic data, focal mechanisms and simplified tectonics. From the previous public version
(4.1), some bugs were corrected and new features added. In particular: cartographic deformation is
taken into account in error ellipses and station-epicentral path draws; some bugs on printing were
corrected; compacted seisan files can be used within ‘additional events’ representation; travel time
curves can now be displayed; simplified relief can be displayed (if available as a MAP file); more than
one file can be used for each category layer (coastlines, tectonic, relief and places names).
The program can be integrated within the SEISAN (Havskov et al, 2000) environment, since it uses
SEISAN parameter files, macrosseismic files, MAP files and station/model files.
Figure 28. Example for pplot using w_emap
If the program is called from EEV or from the command line as W_EMAP, then it displays information
contained in hyp.out file, generated by the HYPOCENT (Lienert, 1994) location program, included in
SEISAN (Havskov et al, 2000), in the settable working directory.
During the first run, user is driven to edit the configuration file w_emap.def that is created in the users
personal directory (SEISAN_TOP/DAT/users/<username>), where most of the program parameters
can be changed.
The program can automatically detect changes in the hyp.out file so the user doesn’t need to restart
the program each time the epicentre changes.
The program can also display epicentres contained in any SEISAN parameter file, where the user
may choose between one single epicentre and all epicentres at the same time. Double clicking the
right mouse button will change the active epicentre to the one picked.
Multi-user individual configurations (color schemes, additional event files, tectonic files, coastline files,
relief files, cartographic projections, etc.) are supported.
116
Installation
All the files are included in the distribution file w_emap.zip. To install it, you need to use PKUNZIP™
(version 4.00 or higher) or the WINZIP™ program.
Using PKUNZIP:
• Position in your SEISAN_TOP directory (e.g. c:\seismo)
• Execute pkunzip –d w_emap
Using WINZIP:
• Extract all files to your SEISAN_TOP directory, with the flag ‘use folder names’ on.
Manual will only be found in INF after installation.
6.4 Searching in the database, SELECT
Whenever selective search and extraction is wanted SELECT is used. The program can run on the
CAT database, single CAT files (Nordic or Nordic compact) or the S-file data base. The output file,
select.out, will also be in Nordic format. Since the input CAT database can contain both normal and
compact files, the output will always be a normal file with blank lines between events. If however the
input is one compact file, the output will also be a compact file.
Note: If SELECT is used on the CAT database (normal operation), you need to UPDATE your S-file
database in order to transfer changes from the S-files to the CAT database.
Select can work with input in 3 different ways:
1: The user is asked for selections
2: The selection parameters are in a file
3: Parameters are given on the prompt line
The program is started by typing SELECT (parameters from screen), SELECT ‘input file’ (parameters
from input file) or SELECT -options. A typical user interactive run is shown below. Comments
following !
POSSIBLE INPUT IS:
STANDARD CAT DATABASE: RETURN
ALTERNATIVE DATABASE, GIVE 1-5 LETTER CODE:
FILENAME FOR ONE FILE, MUST BE 6 OR MORE CHARACTERS:
Type of base: CAT (Return) or Sfiles (s):
Updating database TEST_
! standard base
The database TEST_ has 10 files
The first file starts: 199309
The last file starts: 199909
Start time (at least year), yyyymmddhhmmss: 199309
End time, return for end of month
: 199607
PARAMETERS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
-
Fault Plane Solution
Earthquake Felt
Magnitude Type(s)
Distance ID(s)
Event ID(s)
Magnitude Limits
Latitude Limits
Longitude Limits
Depth Limits
RMS Limits
Number of Stations Limits
Hypocenter Errors Latitude Limits
Hypocenter Errors Longitude Limits
Hypocenter Errors Depth Limits
Minimum Number of Polarities
Hypocenter Agencies
! time range
117
17
18
19
20
21
22
23
24
-
Magnitude Agencies
Station Codes, components and distance range
Polygon
Use all header lines
Look for wave form file names
Gap range
Phases
Volcanic subclasses
SELECT NUMBER TO CHANGE PARAMETER, RETURN TO SEARCH: 6
Minimum Magnitude, return for default:
5
Maximum Magnitude, return for default:
7
PARAMETERS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
-
Fault Plane Solution
Earthquake Felt
Magnitude Type(s)
Distance ID(s)
Event ID(s)
Magnitude Limits
5.0
Latitude Limits
Longitude Limits
Depth Limits
RMS Limits
Number of Stations Limits
Hypocenter Errors Latitude Limits
Hypocenter Errors Longitude Limits
Hypocenter Errors Depth Limits
Minimum Number of Polarities
Hypocenter Agencies
Magnitude Agencies
Station Codes, components and distance range
Polygon
Use all header lines
Look for wave form file names
Gap range
Phases
Volcanic subclasses
Ok. Input realized successfully.
SELECT NUMBER TO CHANGE PARAMETER, RETURN TO SEARCH:
199309.CAT No of events:
1 Selected:
1 Selected
199406.CAT No of events:
1 Selected:
1 Selected
199410.CAT No of events:
1 Selected:
0 Selected
199508.CAT No of events:
1 Selected:
0 Selected
199511.CAT No of events:
1 Selected:
0 Selected
199606.CAT No of events:
6 Selected:
3 Selected
199607.CAT No of events:
5 Selected:
1 Selected
7.0
total:
total:
total:
total:
total:
total:
total:
1
2
2
2
2
5
6
TOTAL NUMBER OF EVENTS IN TIME INTERVAL
16
NUMBER OF DISTANT EVENTS
- - - - 7
NUMBER OF REGIONAL EVENTS - - - - 0
NUMBER OF LOCAL EVENTS
- - - - 9
--------------------------------------------------NUMBER OF EVENTS SELECTED *************
6
NUMBER OF WAVEFORM FILES SELECTED
9
NUMBER OF INDEXES SELECTED
6
SELECTED EARTHQUAKES ARE IN FILE: select.out
LOCAL INDEX FILE IN:
index.out
NAMES FOR WAVEFORM FILES IN FILE: waveform_names.out
SELECT COMMANDS IN FILE:
select.inp
Note above, that the second time the menu is shown, the choice of magnitude limits is shown.
For each CAT file in the catalog, the number of events in file, number of events selected from that file
and the accumulated number are listed. The last file might not show the correct number of events in
file since SELECT might stop before reading the whole file if the end time is in the middle of the file.
Input parameters:
In the input database (or file) a time window must always be given. If no more selection is done, all
118
data in time window is selected. Further selection can be done by choosing a number and giving
parameters. The chosen parameters are then shown on the next parameter selection menu as shown
above for magnitude. Parameters can be reentered. Parameters not entered will have no influence in
the selection. If several parameters (numbered selections below) are entered, conditions for all must
be true for the event to be selected. Within each numbered selection, usually only one of the entered
conditions must be fulfilled for the event to be selected. If e.g. Ml and Mb are selected, events, which
have either magnitude, will be selected. When no more parameters are desired, press enter.
1 - Fault Plane Solution
Selects events with a fault plane solution (F- line in S-file)
2 - Earthquake Felt
Events felt indicated by a type 2 line
3 - Magnitude Type(s)
Normally, all magnitudes for one event are searched to see if any magnitude fits the selection criteria.
With option 3 it is possible to use one or a combination of magnitude types e.g. L and B. If
magnitudes without type are to be selected, use underscore "_" for magnitude type. If there is no
magnitude in the first magnitude position, chose "N" for one of the magnitude types to be able to
select the other 2 magnitudes on the line. Magnitude types are: C: Coda magnitude, L: Local
magnitude, B: Mb, S: Ms and W: Moment magnitude.
4 - Distance ID(s)
Restricting the search to be for one or a combination of the distance id's L, R and D.
5 - Event ID(s)
Restricting the search to one or a combination of event id's, e.g. E and V for explosion and volcanic
events. The letters used for selection are not limited to the examples shown above, they are however
the ones used currently. It is thus e.g. possibly to label events as X for unknown type (column 23 in
header line) and then later on select out all those events by specifying X for event ID. For the 3
questions about types, up to 5 letters can be used. The currently used codes are: E: Explosion, P:
Probable explosion, V: Volcanic, S: Sonic boom, Q: Earthquakes which is equivalent to blank for type.
However, if blank is selected, all event types are selected, while if Q is used as input, only events with
no ID or Q ID are selected. So if all earthquakes and volcanic event are to be selected, use QV.
Without the Q, only volcanic events are selected.
6 - Magnitude Limits
Range of magnitudes to select. Note that if no magnitude type is given, the extreme of all magnitude
types reported is used.
7 - Latitude Limits
Range of latitude. NOTE: If no latitude or longitude values are chosen, SELECT will include an event
even when it is not located if the remaining criteria are OK. If it is required that only located events are
searched for, enter at least one value like an upper latitude limit of 95.
8 - Longitude Limits
Range of longitude.
9 - Depth Limits
Range of depths.
10 - RMS Limits
Range of rms travel time residuals.
11 - Number of Stations Limits
Range of number of stations.
12 - Hypocenter Errors Latitude Limits
Range of hypocenter latitude errors. Works only if error line (E-type) is present in S-file. Currently
119
error lines are generated by HYP and the ISC conversion program ISCNOR. There should only be
one error line in file associated with the prime solution in first header line. However, if more than one
error line is present, all are checked and if one fulfills the selection criteria, the event can be selected.
13 - Hypocenter Errors Longitude Limits, See 12.
14 - Hypocenter Errors Depth Limits, See 12.
15 - Minimum Number of Polarities, only P-phases are used
Counts all polarities, useful to find potential events for fault plane solutions.
16 - Hypocenter Agencies
Selects events only with given hypocenter agencies as indicated on header line.
17 - Magnitude Agencies
Select only events with given magnitude agencies as indicated on header line.
18 - Station Codes, components, distance range and phase
Selects only events with given stations, component, distance range and phase. A formatted help line
comes up for selecting items. Any one or a combination can be selected. The distance can be
hypocentral or epicentral.
19 - Polygon
Selects events within a given polygon of at least 3 latitude-longitude pairs.
20 - Use all header lines
All header lines are searched for relevant information
21 - Look for waveform file names
Search the database for particular waveform files, input can use a fraction of file name or * for any
name. No wildcards can be used in the string so e.g. ASK* will select all due to the *. Use just ASK in
this case to select all filenames with the string ASK.
22 – Gap range
The range of gap as given on the E-line (normally 2. header line). Only hypocenters calculated with
SEISAN version 7.0 have gap.
23 – Phase
Look for events with particular phases. Up to 6, 4 character phase names can be selected. The event
is selected if at least one of the phases is present for the event. For a more selective selection based
on phase, see option 18.
24 – Volcanic subclasses
Search for events of given subclasses given by up to 10 codes. Any code can be given, however,
normally they will be as defined in VOLCANO.DEF. The program searches for lines starting with
‘VOLC MAIN’.
Historical data: When working with historical data, it can be useful to work with catalogs of several
centuries. The century is available in the Nordic Format, so catalogs can go back to year 0.
Output:
Select.out: A CAT-file or compact file (depending on input) of selected events.
Index.out: A list of event id's of selected events can be used with EEV or other programs accepting
index files. This could be used e.g. to work on only distant events in the database by first selecting all
distant events and then working with these directly in the database using command EEV index.out.
Index files can have any name (must contain a ‘.’) so different subsets can be available with different
120
index files.
Waveform_names.out: A list of corresponding waveform files. It is mainly intended for copying to or
from tape specific waveform files. It has the format of the filenr.lis files and can be used directly with
e.g. MULPLT. See also program get_wav for selecting waveform files from the database.
Select.inp: A file with all the parameters used for the run. The file can be renamed, edited and used
as input for select. This is particularly an advantage if a complex set of selection parameters are used
and the selection is wanted again with just a small change. An example file is shown below
Base or file name
:
Start time
:
End time
:
Minimum number of stations
:
Maximum number of stations
:
Minimum latitude
:
Maximum latitude
:
Minimum longitude
:
Maximum longitude
:
Minimum magnitude
:
Maximum magnitude
:
Magnitude agencies
:
Hypocenter agencies
:
Minimum rms
:
Maximum rms
:
Minimum depth
:
Maximum depth
:
Minimum error in latitude
:
Maximum error in latitude
:
Minimum error in longitude
:
Maximum error in longitude
:
Minimum error in depth
:
Maximum error in depth
:
Magnitude types (L,C,B,S,W) :
Distance (ID) types (L,R,D) :
Event types (e.g. E,V,P)
:
Minimum number of polarities :
Felt earthquakes
:
Fault plane solution
:
Check all header lines
:
Waveform files to check
:
Minimum gap
:
Maximum gap
:
Phases
:
Volcanic subclasses
:
Stat., comp. dist range, phase
STAT CO Mindis Maxdis Phas All
BER SZ
1
999 P
TEST_
19930300000000
19961231235959
0
999
-90.000
90.000
-360.000
360.000
5.000
7.000
0.000
999.000
-99.000
99999.000
0.000
99999.000
0.000
99999.000
-99.000
99999.000
0
F
F
F
P
0.000
360.000
SSS PP
(1x,a5,a2,2i7,1x,a4) one pr line, end blank line:
stat hdist->TT
Polygon points (lat,lon), one pair pr line, end with blank line :
Note: The TT at STAT line indicates that all stations must be present (True) and hypocentral distance
is used (True)
Select with input from the prompt line
This option is particular useful when using select with automated operations and has been made
specifically to deal with extracting data out of the data bases using WEB based software. This option
do not have all of the above options. The following are implemented:
-base :
5 letter data base
-seisweb:
if set, WEB output parameters
-time :
time interval (2 variables)
-web_out:
complete path to where data is placed, only
active if seisweb set. 3 files are made:
web_out.id : id's, like index.out without #
web_out.all : like select.out
web_out.head: header lines
-area :
lat-lon grid, minlat,maxlat,minlon,maxlon
121
-depth :
-mag :
-nstat :
-gap :
-rms :
-magtypes :
-disttype :
-eventtype:
depth range, mindepth,maxdepth
magnitude range, minmag,maxmag
range of number of stations, min,max
range of gap, min,max
range of rms, min,max
up to 5 mag types, one string, e.g L
distance type, e.g D
event type, e.g. E
Problems: An event might be found and listed in index.out, but when looking for it with EEV, it is not
there. This can happen if an event has been deleted with EEV and no UPDATE has been made, so
that the event is still present in the CAT part of the database.
6.5 Extracting events from the database, COLLECT
The command COLLECT is used for collecting many event files from the database S-files into a
single file. This may be split into individual event files later using SPLIT. The file can be used for
exchanging data with other agencies or be used with the epicenter plotting program. The questions
are:
Base name, ,, for local directory, name of index file
or return for default base:
Start time
:
End time, return for end of month:
Compact output file (Y/N=default)
At the end, the program will give statistics of collected data, and file name. For getting data out of the
database represented by the monthly CAT files, use SELECT. If an update has been made, SELECT
will always be the fastest program to use. COLLECT and SELECT are the only programs that can
make a CAT file from the individual S-files.
Program input can also be on the prompt line, below is an example:
collect -start_time 19910912 -end_time 19911015 -base_name BER -compact
This means that a CAT-file (default) is collected from BER and is written in compact format (-compact
has no arguments). The time interval is between 19910912 and 19911015. Only start_time is
required, the other arguments are optional. The syntax is: -"keyword" value -"keyword" value etc.
6.6 Inserting events into the database, SPLIT
The program splits up a multiple event S-file in Nordic format (usually made by COLLECT or
NEWEVE) or compact file to single files in the database or in the users own directory. Type SPLIT to
start program and questions are:
122
INPUT FILE NAME
! Give file name
COLLECT.OUT
! File name
INPUT BASE NAME:
! Indicate which type
BER FOR STANDARD DATABASE:
! - database
THREE LETTER CODE FOR OTHER BASE
RETURN FOR SPLIT UP IN LOCAL DIRECTORY
BER
! Choice was standard base
OPERATOR ID, MAX 4 CHARS
! Operator id logged in file
jh
1988 2 5 13:51 35.0 L
RECORDS: 4 ! Listing of events split up
File already exists, options are:
! try to make a file with same
! id
Ignore (leave old event):
Return
Ignore all
I
Overwrite duplicate:
O
Overwrite all duplicates
A
Create a new event, different ID:
N
Create new events for ALL duplicates: * O
1988
1988
2
2
5
5
1992 11
1
NUMBER
NUMBER
NUMBER
NUMBER
NUMBER
OF
OF
OF
OF
OF
14:15 25.2
19: 4 10.0
1:32
1.0
D
D
RECORDS:
RECORDS:
D
RECORDS: 55
LOCAL EVENTS:
REGIONAL EVENTS:
DISTANT EVENTS:
EXPLOSIONS:
PROB. EXPLOSIONS
1
0
2
0
0
TOTAL NUMBER OF EVENTS:
3
TOTAL NUMBER OF RECORDS:
14
4
3
! Statistics of events
! - split up
FORTRAN STOP
In the above example, there was already an event in the database with the same file name and
therefore the same id. It is up to the user to decide if this is the same event in which case it should be
ignored or if it is a new event which happens to have the same id (start time or origin time to the same
second and same event type). In case of a new event, a new id with one second different will be tried.
Sometimes it can be desirable to overwrite the whole database event by event. If e.g. a station code
is wrong in all events, this can be corrected by making a collect to extract all events, edit the
collect.out file using a global substitute, and finally use split to put the events back in. In that case the
option of overwriting all should be chosen.
Compact files can also be split up. Since this is unusual to do, the user will be prompted 2 times to
confirm the split up. Since there is no ID line in a compact file, the database name will be generated
from the header time. This option to be able to split up compact files has been made to facilitate work
with seismic catalogs where it is often desirable to be able to access individual events even when no
readings are available.
6.7 Updating final locations in database, UPDATE and UPD
UPDATE
Both the monthly epicenter files in \SEISMO\REA\BER__\CAT and the updated S-files are generated
with program UPDATE which is a special version of HYP. Type UPDATE to start the program and
there will be questions about time period and database. The program will also ask for operator ID (4
chars), which is stored in the updated log file and the S-file, see below.
By updating, both the S-files and the CAT-files in the CAT-directory are updated. The reason for
updating both at the same time is to ensure that there is a correspondence between the two.
The program will go through as many months as specified by the user. When the program is running,
123
one line will be printed out for each event. The S-files will be overwritten with the updated location,
residuals etc. At the same time, a monthly CAT file is created in the CAT directory containing all
events, also events not located. If a monthly file is already present, it is overwritten.
Update can also work on a local database. The S-files are updated as described above. Since there
is no CAT database, the Update program makes a CAT file in the local directory called hyp.cat with
events in chronological order.
At this time, an S-file might contain several old ID-lines which in an append process have been
converted to comment lines. These are deleted when doing an update. The remaining ID-line is
updated with the action UPD, the operator ID and the time. At the same time, all the error lines are
deleted and only the one belonging to the prime location is kept.
The update process can also change all S-file names according to the origin time and the ID's are
changed correspondingly. This is done in order for the database to be in chronological order
according to origin time and not the more random times used when the events were first registered
into the database. Even if the event is marked not to be located with a * in header line column 45, the
ID will still be updated (same for program UPD). Like with the SPLIT program, if two events of the
same type (L, R or D) have the same origin time to the second, one second is added to the file name
part indicating seconds (see also section 6.6). The event will also be in chronological order in the CAT
database.
*****VERY IMPORTANT ******
The first time an update is done, the S-files get a new name according to the origin time now
calculated and the internal ID is changed accordingly. The ID is then locked indicated by an L in
column 76 of the ID line. For all future updates, by default, the ID will remain the same, the S-file
name will also be the same irrespective if the origin time changes. This is VERY important in case
data is taken out of the database for some special analysis and then put back in to overwrite the
original data. If the ID is the same, the correct event will be replaced. Optionally, Update can make a
new ID each time the program runs (not recommended). It might be necessary sometimes to allow
this in case the events are no longer in chronological order according to origin time (e.g. a teleseismic
event is put in with the ID corresponding to the recording time, when located, the origin time is many
minutes before and it will appear too late in the database). However, this is rarely a problem after the
first location is done and it is recommended to use the default option of locking the ID.
NOTE: When an update takes place, the old location, magnitudes (except 3. if a different agency from
the default agency), residuals etc are removed. If an event cannot be located, the old location etc is
lost. This is intentional since the updated database should represent the data available. If a location
should be retained, special flags must be set, see section 6.1, "Fixing location" (a '*' in column 45 in
header line).
In order to keep track of how and when the database has been updated, every run of UPDATE
creates a log file of the update process. This file is located in a subdirectory of the database directory
(default BER__). If e.g. updating REA, the logfiles will be in ../REA/BER__/LOG/ (unix). Filenames are
similar to S-files. Below is an example of a logfile with name 01-0000-00L.S199606
1996 06 kk
1996 06 jh
99-09-08 14:30 03-1955-35D. 25-0337-29L.
98-09-08 14:29 03-1955-40D. 25-0337-31L.
6
5
The content is as follows: date and time of file updated, operator ID, time of update, event id of first
and last event of the month, number of events for month. The example above shows that June 96 has
been updated 2 times, the last time on September 10, 1999. For each update, one line is added to the
top of the file, so the update history is saved.
Note: If the command UPDATE is used from EEV, only one S-file is updated (name stays the same),
and a general update should be made.
UPDATE recalculate moments if distances (or depths) change, however it does not change the Vp or
Vs velocities used if a change is made in MULPLT.DEF.
124
Problem: If UPDATE crash, there will not be a correspondence between S-files and the CAT data
base: Redo UPDATE.
UPD
The command UPD is very similar to the UPDATE command, however there is no modification of the
S-file except the ID line. The program is used to simply move single S-files into the monthly CAT-files
without relocating. It is mainly used to manipulate database events already processed. E. g. if ISC
data a available and it is desirable to have it in individual files to be able to use EEV, the same data
can then be copied into the CAT part of the database using UPD without modifying the original
solutions. The data must be in the CAT part of the database in order for SELECT to work fast.
KNOWN BUG: On Sun OS, it seems that UPD can only operate on up to a 4 year time period.
6.8 Using ‘filenr.lis’, DIRF and DELF
DIRF
The DIRF command is a useful program for making a file with a numbered list of files from a DIR (ls
on SUN) command. The command makes a file with file name filenr.lis e.g. when working with many
waveform files with long names, a DIRF is first made, and subsequent programs then get file names
from filenr.lis, either by using the whole list, or just a given number. This is handled with routine
filename (in LIB). Below are some examples of using DIRF with SEISAN data files.
Dirf 9101-10*
#
#
#
1
2
3
9101-10-0915-15S.KMY_01
9101-10-1510-55S.N2F_08
9101-10-2333-44S.N3F_06
dirf 9101-10-0915-15S.KMY_01 9101-10-2333-44S.N3F_06
#
#
1
2
(Unix only)
9101-10-0915-15S.KMY_01
9101-10-2333-44S.N3F_06
The wildcard * above indicates that all files from the 10’th is wanted. Many programs use the same
subroutine to get the file name from filenr.lis. This means that most programs using filenr.lis assume
that if a name given is less than or equal to 4 characters, it is a number so file names less than 5
characters cannot be used when the program asks for "Filename or number".
DELF
DELF is a simple program that allows the user to delete a file that is listed in a ‘filenr.lis’ file or another
index file. First run DIRF to list the files that you want to delete. Then start DELF and choose the
number of the file to delete, ‘?’ shows the contents of ‘filenr.lis’. In addition, DELF also has an option
to delete all the files in the filenr.lis or index file. This is a useful option if selected files in a data has to
be deleted. If e.g. all S-files from a particular agency has to be removed, run SELECT first and then
DELF.
6.9 Making a bulletin, BUL
The bulletin program BUL is intended for writing seismic bulletins in a nice format. The output is
written to a PostScript file.
Input files:
125
1. A monthly data file: This file can be made by COLLECT or SELECT
2. "BUL.INP" : This file must be in DAT or in the local directory. In this file the layout of the front
pages are decided, as well as the font selection for the main bulletin. There are ample
comments in the file on how the commands are written.
Some special format features:
Type 3 line: If the first 5 columns in a type 3 line are:" Bul:", then the rest of the line is interpreted as
text line that is written in the bulletin. In this way comments to certain earthquakes can be written into
the bulletin. Type 2 line: Maximum intensity and casualty/damage reports are included in the bulletin if
found in the S-file.
How to run the program:
bul -h ==> This gives you a list of the different options like this:
Options:
-frontpage:
-nofrontpage:
-onlyhypo:
-minmag x.x :
Only frontages are printed.
No frontages are printed.
Only hypocenter solutions are printed.
Only hypocenter solutions with magnitude than the requested are printed.
The last option may be used in cases where the number of earthquakes is very high, so that it is
preferable to report phases only for events above a given magnitude.
You can also run the program without any options, in which case the default values used are:
i) All phases are reported.
ii) Front pages are printed.
You will always be asked for the name of the S-file.
Output file:
The output file is called "bul.ps" and is a PostScript file that you can print.
Optionally, a limited number of pages can be selected from the bul.ps file for printing. The header
page is still included and the page numbers correspond to the original page numbers.
6.10 Reports and statistics
SEISAN has several programs for extracting and writing out data for plotting or printing statistics,
most of which will be listed in this section.
Report
The program extracts parameter data from all header lines in a CAT file and rearranges the data in a
table. In additions, there is an option to rearrange order and location of magnitudes on the header
line. Below is an example of a run where the input CAT file is called collect.out
report collect.out
Below is shown parameters which can be chosen for output.
A return will chose all, placing any character under a field
will chose that parameter in the output. Each field starts
with a capital letter and ends within the following blank.
The order of the output can be changed by placing a number
under the field and fields will be written out in the order
126
of the numbers. E after time, lat, lon and dep are errors,
L E is distance and event id s, F is both fix flags and A is
agency for magnitude.
The following example shows that Mc, Depth(Dep) and Time with
error are selected and written out in given order.
Date TimeE L E LatE LonE Dep E F Aga Nsta Rms Gap McA MlA MbA MsA MwA Fp Spec
30 45
20
10
Date TimeE L E LatE LonE Dep E F Aga Nsta Rms Gap McA MlA MbA MsA MwA Fp Spec
x
x
x
x
x
x
x
Number of output fields
8
Number
Number
Number
Number
Number
Number
Number
Number
Number
of
of
of
of
of
of
of
of
of
events
events
events
events
events
events
events
events
events
with
with
with
with
with
with
with
with
12
1
3
16
10
10
8
5
12
spectra:
fault plane solution:
error estimates:
mc
:
ml
:
mb
:
ms
:
mw
:
Output report file is report.out
Output nordic file is report_n.out
Output of choices used in report.inp
The report.inp is a file with the choices used. Report can use that file (or a file with the same format
and a different name) as second argument:
report collect.out report.inp
in order to use a fixed set of choises.
Content of report.out
Year Date Latitud Longitud Depth NST GAP
1996 6 3 47.776 153.222
0.1 12 348
1996 6 6 62.652
4.940 15.0 13 270
1996 6 6 62.634
5.068 15.0 13
1996 6 6 62.652
4.940 15.0 13 270
1996 6 7 59.841
5.127 12.0 12
1996 610 -13.481 167.130 200.1 301
1996 625 61.656
3.363 14.9 35
1996 7 5 61.282
4.825
7.1 10
1996 713 61.416
3.870 12.1
9
1996 718 60.156
2.070 15.0
9
1996 718 51.438 157.737 29.9 18
1996 726 61.755
2.293 22.1
9
Ml STRIK
2.9
2.9
2.9
1.9
DIP RAKE
28
61
-41
28
8
61
41
-41
75
3.2
2.0
1.5
1.8
1.8
The file report_n.out contains the input data with the only difference that the magnitudes have been
moved around on the header line. This can be practical for later plotting with EPIMAP. If no
magnitude selection has been made, the magnitudes will come in the order Mc, Ml and Mb. If no
magnitude of that type is available, the output field is blank. The magnitude selected is the first to
occur of the corresponding type. If other magnitudes are to be selected, numbers can be used to
select any 3 magnitudes in any order. If it is important to select magnitudes by agency also, use
program MAG.
NORHEAD, making a compact Nordic file from a Nordic file
You must give arguments: First is input file, optional second is output file, if an optional second or
third argument is -mag, magnitudes from following header lines are moved up to empty magnitude
spaces on first line. The program was earlier called COMPACT (version 7.2 and earlier).
STATIS, statistics of databases
127
This is a simple program for making statistics of stations used in the database or in a file. The
program will ask the following questions:
1. Information about which stations should be searched for in the database. There are several options
for entry:
a: Give a filename with the stations listed one per line. The format is a5. The file name MUST have a '.' not
to be confused with option (b) below.
b: Give stations directly at the keyboard, one pr line, terminate with blank line, format a5.
c: Just make a return and the stations hardwired in the program will be used. These are Norwegian
stations, so the program will have to be recompiled with new stations if the defaults are to be changed.
2. Standard questions about base or filename and time interval
3. Question about counting all phases. This means counting the occurrence of a station for each phase for
that particular station. This can give the total number of phases read at a particular station in a given time
interval which can be more than the number of events. If not counting all phases, the program gives the
number of events recorded at the station.
The output from the program could be as follows:
Station
KONO
KMY
ODD
EGD
ASK
HYA
SUE
FOO
NRA0
MOL
NSS
MOR
LOF
TRO
BJO
KBS
JMI
KTK1
ARA0
NET
NSN
JMI
KNN
W_L
W_E
W_S
Local Ev.
0
24
0
28
29
16
16
18
86
38
9
0
25
12
0
3
16
22
66
Local S.
10
Distant E
21
6
Distant S
8
0
0
0
1
4
1
9
17
6
3
3
1
0
0
0
0
2
3
0
14
7
13
2
12
13
0
1
2
0
1
6
8
6
2
0
0
1
0
NWAW
147
21
10
2
1
2
Number of events selected with given stations
Number of events selected with more than --Number of events with no phases
Number of events with waveforms
Number of events with only waveforms
Number of events with 2 or more waveforms
Total number of waveform files
Total number of local events
Total number of regional events
Total number of distant events
Total number of events
Total number of records
222
98
0
168
0
9
183
200
0
34
234
2830
Output files are: statis.out
statab.out (station statistics only)
The top part shows the event statistics by station. Local Ev is number of local events (readings if so
128
specified above) (type L and R) at the station, Local S means number of local events ONLY recorded
at that station, Distant E and distant S is the same for distant events (type D).
The middle parts shows the number of waveform files NWAV from different networks NET as
indicated by the first 3 letters of the waveform file name after the "."
At the bottom is a summary statistics most of which should be self-explanatory. The information about "..
more than given stations" means that in addition to the stations searched for, the event had additional
stations not used in the statistics.
CATSTAT
This program calculates the yearly, monthly and daily number of events from a given earthquake
catalogue and plots the results (written by Mario Villagrán). The input is a standard Nordic file
containing only the header lines (compact file). The output is given in three different files with
following default file names:
catyear.out: Output catalogue of the yearly number of events. This file contains two columns
of data corresponding to year and the number of events.
catmonth.out: Output catalogue of the monthly number of events. This file contains three
columns of data, corresponding to the year, month and the number of events,
respectively.
catday.out: Output catalogue of the daily number of events. This file contains four columns of
data corresponding to the year, month, day and the number of events,
respectively.
cathour.out: Hourly distribution of events within a day interval.
In addition, a series of files with gmt in name give similar output for use with gmtxy (only Unix).
The output files can then be used for plotting the histograms for the desired time interval at yearly,
monthly or daily intervals. If desired, the corresponding histograms can be plotted interactively on the
screen or can be printed. Several other routine programs such as grapher, xyplot, gnuplot or GMT,
etc., can also be used for this purpose. The general purpose of this program is to evaluate the
catalogue completeness. When run for different magnitude intervals, one can detect the magnitude
thresholds above which the catalogue can be considered complete.
SWARM, finding earthquake swarms
The program is used to identify seismic swarms in a catalog. Input to the program is a CAT file with many
events and some manually entered parameters. Output is identified swarms. The output file swarm.out
contains all swarms organized as 'events'. In the header line is given the center for area identified and the
'magnitude' is the number of events in the area divided by 10. The rest of the line is information from first
event in swarm.
Principle of selection:
The area is divided into a lat-lon grid. Around each grid point, there is a cell with radius small_r. The
program first checks the number of events in each cell for the whole catalog. It then checks each cell to
find which has more than the minimum number of events to constitute a swarm under the condition that
enough events are within the required time window. For each time window, with enough events, a swarm
is declared so a swarm lasting e.g. twice the time window will be declared as two swarms. An additional
condition is that the number of events is larger than the normalized background activity. The normalized
activity is calculated as the activity in the large cell normalized for area to the small cell, and normalized in
time to the window for the swarm.
STATSTAT, number of events per seismic station in catalog
The program reads Nordic file input data and writes out text files giving the number of events per
station.
129
LSQ
A simple program to make and plot a least squares relation between two parameters. Input is from a
file with two columns x and y. The program also makes an output used with GMT in order to make
nice plots. The PostScript output file is lsq.eps and the GMT file is lsq_gmt.out. In order to produce
the GMT plot (only Unix), use command gmtxy lsq_gmt.out.
6.11 Waveform file management tools
This section describes the programs used for modifying and checking waveform files. The most
important features are to add or subtract channels and modify headers. A special program in this
group is GET_WAV which checks data bases for availability of waveform files. New from version 7.1
is that SEISAN also can handle other waveform formats, however not all programs can work with all
formats. This will be indicated with each program. The following programs are available:
APPEND:
Append two or more waveform files following each other in time
AUTOREG:
Automatically register events
GET_WAV:
Check for available waveform files
RDSEED_MANY
Simple way to chop up a seed volume
RESAMP:
Resample waveform files
SEIASC:
Convert SEISAN waveform files between ASCII and binary form
SEICUT:
Extract an interval of a waveform file
SEIDEL:
Splitting up a SEISAN waveform file in 2
SEISEI:
Split and merge SEISAN waveform files
SELSEI:
Find waveform files with given stations
WAVETOOL: Extract waveform data
WAVFIX:
Fix waveform file header time correction, make standard file names,
change headers etc.
WAVFULLNAME
Print full file name including path for waveform file.
APPEND, Append two or more waveform files
The program uses a filenr.lis input file. All files are read, and then written out as one new file. The
maximum number of channels is max_chan_out which is set as a parameter (currently 7). Only the
first max_chan_out channels are used or less if fewer channels in file. A blank line followed by a new
group of files will make a new output file. The output file cannot have more than standard SEISAN
dimension number of samples ( more than 1 200 000, see file../INC/seidim.inc for exact number) per
channel.
It is assumed that all channels have the same sample rate.
AUTOREG, automatic registering of events
When a large number of waveform files are available and it is known that they are real events, it might
be an advantage to automatically register them into a database. Remember, the database can be
made with MAKEREA. It is possible to register events both to the standard BER, any other database
or the local directory. To run the program, make a filenr.lis of the waveform files and run AUTOREG.
It is possible to put blank lines into the filenr.lis to separate into events, in case there is more than one
waveform file from the same event. All waveform files before a blank line are put together into one Sfile.
You get the questions:
Event type for all events: Local:
L (default)
130
Regional: R
Distant: D
1-5 letter base name, return for standard base, ,, for local base
BER
Operator, max 4 chars
jh
92011022.STE
.....
Now comes a listing of waveform file names and S-file names. The program will check if the event is
already registered and the same options are available as in program SPLIT (section 6.6). Since
AUTOREG automatically creates S-files for all events in filenr.lis, they will all be given an event type.
WAVETOOL, extract and convert waveform data
The program was called extract in SEISAN 7.2 and earlier. The program extracts all or selected time
sections of waveform data, optionally applies some signal processing and then creates the output
file(s). The input formats supported are SEISAN, SEED, MINISEED, GSE and SAC (Unix only) and
output formats are SEISAN, MINISEED, GSE and SAC (Unix only). The program can be used as a
conversion program between these formats, instead of using e.g. SACSEI. It would also be possible
to convert for example SAC and GSE files to SEISAN in one go. When creating GSE, MINISEED or
SAC files, the respective format code is added to the filenames. The program can also extract data
from a SEISAN continuous data base.
Note that for MINISEED writing, only Steim1 compression is used. Integer format is also possible, but
requires a parameter change in WAVETOOL and recompilation.
There are two input options: (1) a single S-file or a list of S-files created with DIRF (is also an index
file), which points to the waveform data or a filenr.lis type of file that gives the waveform file names;
(2) a waveform file or a list of waveform files. The program can be started either interactively, without
arguments, or non-interactive by specifying the commands as arguments.
The arguments are: -sfile <sfile-name>: The S-file name of the event you want to extract waveform
data from.
-wav_files <file-name>: Extract from a list of waveform files in filenr.lis format. Input from S-file will be
ignored. This option merges all files from list (if within a 'reasonable time window') and program is
then partly doing what SEISEI is doing for only SEISAN format files.
-wav_out_file <file-name>: Name of waveform output file, not used if output format is SAC.
-start <time>: Start time can be used to set start time the same for all channels instead of using
chansel file, time can be absolute or relative to beginning of the first trace.
-duration <time>: Select duration of time window if –start is used.
-cwav: Input is from the SEISAN continuous data base, useful for extracting intervals
-cseed: Input from large SEED file, similar to -cwav
-wav_in_file <file-name>: Input of one waveform file
-sfile <file-name>: Input of S-file name
-format <output format>: The output formats supported are SEISAN, SAC, GSE (GSECM6) and
GSEINT. In case of SEISAN or GSE, multi-trace files including all selected traces are created, while
for SAC, single trace files are generated.
131
-chansel <file-name>: Input file to select channels and time windows. The first line contains number of
channels. The following lines give station code, start time (both absolute and relative to earliest trace
allowed) and duration. If start time and duration are set to 0, complete traces are selected.
Example
Select the first three complete traces
3
1 0 0.0
2 0 0.0
3 0 0.0
Use absolute start time, and duration of 180 seconds
3
1 19991001124500.000 180
2 19991001124500.000 180
3 19991001124500.000 180
Use relative start time of 60 seconds from beginning of earliest trace, and 300 seconds duration
3
1 60 300
2 60 300
3 60 300
The program assumes that a large number is absolute time.
-chan_out_file <file-name>: Name of text file containing a list of available channels from a list of
waveform files. If wav_out_file is not specified, program terminates after creating the list.
Example
1
2
3
4
5
KBS
LOF
MOL
FOO
HYA
BV
S
S
S
S
Z
Z
Z
Z
Z
1996
1996
1996
1996
1996
6
6
6
6
6
3
3
3
3
3
20
20
20
20
20
2
5
5
5
5
18.991 6000
5.531 5800
24.984 10000
34.156 9650
36.078 9900
20.000
50.000
50.000
50.000
50.000
2666.400
2832.940
2852.393
2861.565
2863.487
299.950
115.980
199.980
192.980
197.980
-filter <flow> <fhigh>: bandpass filter limits
If only a filter is used, a 4 pole Butterworth filter is used one way. If both a filter and response removal
is used, an 8 pole Butterworth filter is used in the frequency domain
-npole <n>: number of poles used for filter
-ground <1,2,3>: compute displacement, velocity or acceleration (1,2,3)
-ichan <id>: select one channel only
-interactive: Flag to specify non-interactive use, in which case the program does not ask any
questions (for example given by SEISAN-autodrm interface), default is interactive.
-seisweb: Flag to indicate that the program is started by SEISWEB.
-maxpoints <number>: Number specifies the total number of points desired for the total time window
covering all selected traces. This option is meant to reduce the number of points to what is needed to
visual correctly plot the traces. When plotting the trace using a number of pixels, which is smaller than
132
the number of points on the trace, samples are plotted on top of each other for one time sample. This
results in the maximum and minimum being plotted at the same place on the x-axis. The idea now is
to reduce the trace to these maxima and minima only. Then using twice the number of samples than
pixels will allow to visually correctly show the trace. Note that this is not a resample routine. Option
mainly used with SEISWEB.
-stat_out: write out station location file, simple xy output file (station_list.out)
-rsam: convert data to RSAM (1-minute absolute average), which is commonly used in volcano
seismology
-resp_out: write out list of all response files for channels given in waveform files (respfile_list.out)
Another way of extracting waveform data is using MULPLT where many traces can be extracted as a
binary SEISAN file (using WAVETOOL in the background) or a single trace as an ASCII file.
Accuracy of extracted data
If the data is filtered or corrected for instrument response, the number out can be less than one and
an output file of zeros can be made. If the output format is SEISAN, the values will be scaled to avoid
this and the appropriate scaling factor is included in the waveform file. Subsequently reading of these
files in SEISAN will produce the correct values. For this reason, it is advised to use SEISAN as output
format when filtering or correcting for instrument response.
Note: Parameter MERGE_WAVFORM in SEISAN.DEF sets the network extension of extracted files.
GET_WAV, get listing of available waveform files
The program uses a CAT file as input and checks for availability of all waveform files listed. A typical
run is shown below:
get_wav
INPUT FILE
select.out
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
path
path
path
path
path
path
path
path
path
path
path
path
NAME
name
name
name
name
name
name
name
name
name
name
name
name
:
:
:
:
:
:
:
:
:
:
:
:
\SEISMO\WAV\1996-06-03-2002-18S.TEST__012
\SEISMO\WAV\1996-06-03-1917-52S.TEST__002
\SEISMO\WAV\1996-06-06-0647-46S.TEST__011
\SEISMO\WAV\1996-06-07-1324-51S.TEST__009
\SEISMO\WAV\1996-06-23-0126-27S.TEST__013
\SEISMO\WAV\1996-06-23-0059-47S.TEST__001
\SEISMO\WAV\1996-06-25-0336-34S.TEST__032
\SEISMO\WAV\1996-07-05-0219-49S.TEST__010
\SEISMO\WAV\1996-07-13-0555-53S.TEST__007
\SEISMO\WAV\1996-07-18-0946-21S.TEST__010
\SEISMO\WAV\1996-07-18-2303-34S.TEST__015
\SEISMO\WAV\1996-07-26-0740-58S.TEST__010
Total number of events
11
Number of events without waveform files
1
Number of waveform files
13
Number of waveform files present
12
Output file with events is get_wav.out
Output file with waveform file names is copy_wav.out
In the above example, a select.out was used. For each file, it is checked if the waveform files are
available in the system. All waveform data bases and directories specified in SEISAN.DEF are
searched. In order to extract the waveform files corresponding to the input CAT file, the output file
copy_wav.out can be used to copy the files out of the data base to working directory. On Unix, just
source the copy_wav.out file, on Windows, change the file to a .bat file (e.g. ‘copy_get_wav.out
wav.bat’) and run it.
133
RDSEED_MANY, chop up seed file
The program reads a large SEED volume and divides it up into several files of the same size. It calls
rdseed, so rdseed must be installed. Rdseed can do the same, but RDSEED_MANY is simpler to
use.
Example:
rdseed_many
Seed file name
test.seed
start time YYYY,MM,DD,HH,MM
2005 01 01 01 01
Interval in minutes
20
Number of intervals to extract
200
The output format is SAC, other format require a change of the program.
RESAMP, resampling waveform files
RESAMP is a simple resampling program, which can resample one or several waveform files. It only
works with SEISAN format. All files are read, filtered and resampled. Then written out as one new file
with the data from one or several input files. The maximum number of channels is max_chan_out,
which is set as a parameter in the program, currently it is set to 7. Only the first max_chan_out
channels are used or less if fewer channels in input file. It is assumed that all channels have the same
sample rate and will be resampled to the same lower sample rate, which is an integer fraction of the
original sample rate. If e.g. the original sample rate is 50, new rates of 25,10,5,2 etc can be obtained.
The anti-alias filter is a single pass Butterworth with 6 poles. The user specifies manually both the
decimation rate (2,5,10,25 in the above example) and the filter frequency. The new file(s) can have a
new component specification, which is asked for interactively. Finally the user is asked for a new
network code.
The input files(s) come from a filenr.lis file generated with DIRF. If more than one file is given in the
filenr.lis, these will be put together in one file and some samples are saved from one file to the next in
order to assure that there are no overlap problems when using the filter. IT IS ASSUMED THAT ALL
FILES HAVE THE SAME LENGTH OF TIME. The program will check if a following file has the correct
header time based on the length of the previous file. If the following file starts before the end of the
previous file (err_samp samples, default 70), it is assumed that the timing is wrong and that the files
should follow each other. A warning is given and the program continues. If the following file has a
header time that is more that a given err_samp samples after where it should be, it is assumed that
the next file is missing and zeros are inserted in the channel data. The number of sample errors,
err_samp, is hardwired in the program, currently 70. The program will continue to put data together in
one file until there are no more file names in the filenr.lis file or a blank line is encountered. After a
blank line in the input file, a new output file will be created. This can be used to make daily files of e.g.
2 weeks on continuous data by manually placing a blank line in the filenr.lis file for every 24 hours.
The program recalculates the sample rate based on time in first and last file.
The output file name will be given the standard waveform file name with type R for resampled like
1999-07-02-1112-22R.BERGE_005.
Works ONLY with SEISAN format
SEIASC, converting SEISAN waveform files to or from ASCII
A simple program to make an ASCII equivalent of a binary SEISAN file, or vice versa. It is the same
call to use the program both ways. By using a filenr.lis file as input, many files are converted and the
original filenames are kept with the addition of an A for ASCII or B for binary. If the files are converted
134
back, the A or B is removed.
The program is useful for manually editing a waveform file or checking the content in case of
problems. The program is also useful for moving binary files between different types of computer
platforms (moved as ASCII files, not needed between platforms running SEISAN). Between PC, Sun,
Linux and MacOSX, SEISAN programs will automatically adjust for differences in binary structure.
The header format is exactly like the binary SEISAN files and the sample values are written in
multicolumn format.
Works ONLY with SEISAN format
SEICUT, extract part of a waveform file
A simple program to extract out a section of a waveform file (any seisan primary format). A similar job
can be done with wavetool. Syntax is:
seicut filename yyyymmddhhmmss.s interval
The first sample to use is the first sample found before the start time, the output time interval (in
seconds) will be the time from first to last sample, so if e.g. one second of data is asked for at a
sample rate of 100 Hz, the time interval in header will be 0.99 sec and the number of samples output
will be 100. If the interval is not available in any of the channels, the program will stop. The output file
name will use a network code reflecting station code of first channel in input file and 'CUT' is added to
the end of the file name.
SEIDEL, splitting a SEISAN binary file into 2 files
The program splits up one waveform file into 2 files. The questions are:
Filename or number
2
No of channels to remove
3
Channels to remove
1 3 6
! standard question
The program will generate 2 new files, one with the channels removed and one with the remaining
channels. The original file is still present.
Works ONLY with SEISAN format
SEISEI, splitting and merging SEISAN binary files
The program can merge several SEISAN waveform files to one file or take one SEISAN file and split it
into single channel files. The program is intended for editing waveform files and merging files from
different networks to one file. In order to use SEISEI for merging files, a DIRF must be made to make
a filenr.lis file containing the files to be merged. The program will sequentially read filenr.lis and merge
files which have start times within the time interval specified (3 minutes default). Once a gap of more
than 3 minutes occur, a new output file is made. Merging to a new file can be forced by editing
filenr.lis so the groups of files to be merged are separated by a blank line, however, within the group,
the time difference can still only be the given time interval.
If two channels to be merged have the same station and channel codes and the same start time, the
second occurrence will be ignored. If the station and channel codes are the same, but start time
different, the user will be asked to confirm merging.
The program can also split up a multichannel file to files with only one channel. This can be used to
135
remove unwanted channels by deleting selected channels and merging again. When the file is split
up, the channel component is added to the file name. A filenr.lis file can also be used for splitting
many files in one go. If a file is only to be split into only 2 files, it is more convenient to use the
program SEIDEL, see above. SEISEI is also used in connection with MULPLT for merging files
automatically based on waveform file names in an S-file.
Works ONLY with SEISAN format
Note: The network extension of merged files will be set by default to the value of parameter
MERGE_WAVEFORM in SEISAN.DEF.
SELSEI, searching headers in waveform files
A simple program to search headers in waveform files for files containing a particular station.
Works ONLY with SEISAN format
WAVFIX, fixing time correction and channel names in SEISAN waveform file headers and make
standard file names
It can easily happen that a waveform file has a wrong time in the headers, or that individual channels
have wrong timing, for example introduced by different delays in the acquisition system that are not
accounted for. WAVFIX can change all header times with a given constant time delay, or correct
individual channels as specified in a parameter file (wavfix.tim). In addition, the file name will also be
changed to reflect the header time change. Waveform file names were shorter in SEISAN version 6.0
so when using older files, the user might want to use standard file names. In cases where only
channel names or timing of individual channels are changed, the filename can be kept the same. In
this case a temporary file is created, which is later renamed to the original name.
WAVFIX can also change polarity. This is done by setting the output channel and station codes to the
same as the input values in wavfix.def.
In case channel names are to be changed, this can also be done with WAVFIX. A definition file is
needed for changing station, component or both. The parameter file name is wavfix.def and an
example is given in DAT. For definition of the wavfix.def, see next section 6.12 on “Conversion
programs definition file”.
WAVFIX can change header times and/or file names for one or many files. Before running the
program, a list of file names must be made with DIRF. Below is an example where the header time is
changed by 120 secs. No wavfix.def file is present (current or DAT directory).
To correct the timing of individual channels, you need to create the file wavfix.tim in either the DAT or
working directory. WAVFIX checks if the file is present and applies the correction from the file as
default. The format of this file is as follows:
Column 1-5: station code
Column 7-10: component code
Column 12-25: start date and time for time correction (can be empty)
Column 27-39: end date and time for time correction (can be empty)
Column 41:60: time correction to be added
Example:
wavfix time correction applied to individual components
stat comp start time
end time
correction in
a5
a4
yyyymmddhhmmss yyyymmddhhmmss seconds f20.3
-----|----|--------------|--------------|--------------------
136
TEST
SH Z 19500101120000 19600101120000 -0.015
File names of waveform files can be given to WAVFIX directly, from a filenr.lis file or from a Nordic
format file. In case you choose the Nordic input, the waveform file names will be changed in the
Nordic file (output file nordic.fix). This option is useful if you are correcting file names, since the
entries in the S-files are otherwise not fixed.
Example of running WAVFIX
No wavfix.def file, will use internal information for channel codes
This program will change header times in all headers
with the same amount. The waveform file name will be
changed at the same time and adjusted to the standard
name.
If no time correction is given, only the waveform
names are adjusted.
In addition, channels names can be changed if a
wavfix.def file is available
Time correction in seconds, return for no correction
Input options: (1) filenr.lis or waveform file name
(2) Nordic file
Filename or number, filenr.lis for all
Input file name: 1994-06-16-1841-57S.TEST__019
Output file name 1994-06-16-1843-57S.______019
Input file name: 1994-10-04-1324-00S.TEST__016
Output file name 1994-10-04-1326-00S.______016
Input file name: 1994-10-04-1324-24S.TEST__016
Output file name 1994-10-04-1326-24S.______016
Note: WAVFIX ONLY works with SEISAN format
WAVFULLNAME
Prints full filename including path for a waveform file by searching directories and databases specified
in SEISAN.DEF. Filename is to be given on prompt line, e.g. ‘wavfullname 1996-06-13-124815S.NSN__003’.
6.12 File conversion and modification programs
There are mainly two types of files to convert, parameter files with readings and related parameters
and binary waveform files.
PARAMETER FILES
CAT_AGA:
GIINOR:
HYPNOR:
HINNOR:
HSUMNOR:
ISCNOR:
ISCSTA:
ISFNOR:
KINNOR:
NORGSE:
NORHIN:
NORIMS:
NORHYP:
PDENOR:
Records the S-file header lines according to agency
Converts from Geophysical Institute of Israel parameter format to Nordic
Converts from Hypo71 readings files to Nordic format files
Similar to HYPNOR for Hypoinverse files
converts from Hypo71 summary file format to SEISAN format
Converts from ISC 96 column format to Nordic format
Converts ISC station list to SEISAN station list selecting specific stations.
Converts between ISF1.0 and Nordic
Converts from Kinemetrics to NORDIC
Converts between Nordic format and GSE parametric format
Converts from Nordic format to Hypoinverse format
Converts from Nordic to and from IMS1.0
Converts from Nordic to HYPO71 format
Converts a PDE bulletin file to NORDIC format
137
RSANOR:
SEIGMT:
SELMAP:
USGSNOR:
Converts Andalucian Seismic Network data to NORDIC format
Converts from NORDIC file to input for GMT
Selects out a part of a MAP file
USGS/NEIC CDROM catalog conversion to NORDIC format
CAT_AGA, reordering of CAT file header lines
When plotting hypocenters or doing seismic hazard work, it is the first header line in an S-file or CATfile that is used since it is assumed that it is the prime estimate. When making compact files it is also
the first header line, which is used. However, there can be a need for resorting the many type 1
header lines for one or several events so that they are ordered according to agency. It could e.g. be
needed to put priority on all the ISC solutions, which then should be the first line in the file. CAT_AGA
will reorder the type 1 lines in a CAT file according to the order in which the agencies (3 character
codes) are given by the user. If there are many agencies, they can be given in an input file named
cat_aga.par, format is one agency per line in the first 3 columns. If the file is not present, the program
will ask the user to enter the agencies manually. The output file cat_aga.out will contain the sorted
events.
GIINOR, Geophysical Institute of Israel to SEISAN
The input files are the bulletin type files.
HYPNOR, converting HYPO71 files to Nordic files
Input is just filename of HYPO71 file. A similar program for HYPOINVERSE files is HINNOR.
HINNOR, converts from Hypoinverse to NORDIC format
This program works like HYPNOR.
HSUMNOR, HYPO71 summary file format to NORDIC format
Note that the program only converts to header lines.
ISCNOR, converting ISC bulletin file to Nordic format
This program works with the ISC fixed 96-column format as e.g. distributed on CDROM (files of type
FFB). The program can select out subsets of ISC data using a latitude-longitude window, depth and
prime magnitude. Any of the magnitudes Ms and mb are used. Before 1978, there was only mb on
the CD’s. More detailed selection can be done on the output file later with SELECT. Since the amount
of data is very large it is also possible to write out only the hypocenters. Note that ISC now writes in
ISF format also, which can be converted with ISFNOR.
Newer CD's have compressed data and cannot be used directly. files must be copied to disk first,
decompressed and then handled as single files.
The program will first check if a file with agency codes called agency.isc is present. If so the station
codes are read from this file (same format as files on CDROM). The program will also check the
beginning of the data input file for a possible list of agencies and station coordinates. If present, the
stations coordinates are read and converted to SEISAN format and additional codes read in. The
agency codes are needed in order to identify in plain text the various agencies used.
Principles in conversion:
Phases: The phases out can be either the phase ID's sent to ISC or the ISC reinterpreted phases
(given with a number code in the input file). If the user supplied phases are used,
parenthesizes are removed, and if P/PKP etc is given, it is replaced by P.
Times: If day is incremented relative to origin time day, it is carried into the hours, which can be
more than 24.
Agency: It is assumed that it is the same agency for hypocenter and first magnitude. Magnitude is
checked for agency, if blank, it is assumed also to be the same as for hypocenter. Only first
3 characters of code is used.
Stations: Only first 4 characters of code are used.
Depth: If no error on depth, a depth fix flag is set.
First motion:
Only C or D are used, ISC codes J and B are ignored.
138
Hypocenter orders: ISC put the best solution last, here the order is reversed, and the prime estimate
is first.
Duration magnitude: Change D to C for type.
Distance indicator: If station furthest away is less than 1000 km indicator is L, between 1000 and
3000 km indicator is R and if more than 3000 km indicator is D. If no stations are present
the type is set to D.
In order to relocate an event and compare to ISC location, the ISC reidentified phases must be used
(option 2, see below). This has the disadvantage that phases not used by ISC (mainly S-phases of
local earthquakes) are weighted out in the output file. If option 3 is used, the ISC identified phases are
selected if there and if no ISC identification is given, the local reported phase is used. The output file
for option 2 and 3 looks the same except that for option 2, the user-defined phases are weighted out.
The residuals given in the output file are always relative to the ISC identified phases.
Running ISCNOR:
Below is an example of a run where a latitude - longitude window has been used.
Phases selected can be:
User reported phases (default=return)
: 1
ISC identified phases only
: 2
ISC identified phases and user reported phases
when not identified by ISC
: 3
3
Output: All hypocenters and phases : Return
All hypocenters
: 1
Prime hypocenter
: 2
Latitude range, return for all
60.2 70.5
Longitude range, return for all
10, 20
Depth range, return for all
Magnitude range, return for all
Write selected events on screen (y/n=return)
No agency.isc file present
If ISC CDROM, give drive letter, else return
d
Give first year and month, e.g. 199501 198601
Give last year and month, e.g. 199602 198602
Now opening d:\1986\198601.FFB
Number of agencies in input file
244
etc, for each month
493 events converted
Output file name is iscnor.out
File with stations is isc.sta
The file input can be from a CDROM as in the example above. In that case, the whole CDROM can
be read or a smaller time interval can be given. The input can also be from a single file and the
program will then ask for the next file when the first has been converted. If many files are to be
converted, a list of file names can be made with DIRF and filenr.lis entered as an input file name. The
Nordic format output file is iscnor.out and the station list is in isc.sta which has the format used by
SEISAN. Optionally, output can also be in the original isc format, however that requires setting a flag
in the program and recompiling, see program source code.
ISCSTA, selecting stations in the complete ISC station file
The complete station list in the ISC list is very large and it is often an advantage to use a smaller
subset, although HYP can use the whole list. The program can select out subsets of stations in both
SEISAN and an older ISC format.
139
The program will read an S-file, find how many different stations there are and select those stations
out of a station file, which can either be in SEISAN (=HYPO71) format or ISC format (automatically
determined). The output is in SEISAN format. If no S-file is given the input station file is assumed to
be in ISC format and the whole file will be converted to SEISAN format.
KINNOR, Kinemetrics to NORDIC
Converts .PCK file output of EDPPICK to file in SEISAN format. Many events are converted from one
file. The program is based on program from Kinemetrics by Christopher S. Lim. For info on how
conversion is made, see program source code.
ISFNOR, ISF1.0 to and from Nordic
The ISF format is used by the ISC and is an extension to the IMS format. The program is based on
the routines provided by the ISC for reading and writing ISF, and the SEISAN standard routines for
reading and writing Nordic data. The program converts in both directions. All possible information is
converted. Information on the ISF format can be found on the ISC website (http://www.isc.ac.uk). It is
recommended to use ISF format for data exchange with ISC.
NORIMS, IMS1.0 to NORDIC format .
The IMS1.0 (International Monitoring System) is a new version of the GSE format and very similar.
The program can partly be used for the new ISF (IASPEI Seismic Format) which will include all of the
IMS format an additional information needed by ISC and NEIC. The program and the following
description is by Mario Villagran. The program works with the IMS1.0:SHORT format (phasereadings/origin files) and the program works both ways.
IMS1.0:SHORT -> Nordic
Nordic -> IMS1.0:SHORT
The IMS1.0:SHORT format is exactly the one used at the IDC International Data Center (Vienna,
Austria). In addition some features used by the ISC International Data Center and the Spanish NDC
National Data Center had been added. Magnitudes in IMS format use many characters, the Nordic
format allows only one; the following rule is followed:
IMS
Nordic
For mb --> 'b'
For MS --> 'S'
For ML --> 'L'
For MD --> 'D'
For Ml --> 'l'
For MN --> 'N'
For mblg--> 'G'
For ms --> 'S'
For MB --> 'B'
The maximum likelihood magnitudes mb1, mb1mx, ms1, ms1mx, etc are pending. IDC still does not
have documentation and they may be changed.
Single measurements of magnitude/station are parsed as comment lines (type 3) starting with symbol
"$". When importing data from IMS format, only the "Event IDC" number is parsed and included into a
comment line (type 3) of Nordic, together with the ellipse dimensions orientation and the mb standard
deviation.
All parameter values read that exceed field limits of Nordic (Amplitude, velocity, snr, etc) have been
set to the maximum or minimum possible, example: if snr > 999.9 snr=999. For conversion from
Nordic to IMS it is necessary to use both the hyp.out and print.out files; The reason is that IMS
includes many parameters that need to be searched in both files.
When converting to IMS format, the user can specify the start numbering for the first event and phase
in the file; ignoring will assume (1,1).
NORGSE, NORDIC from and to GSE parametric format
The program (written by Mario Villagrán) converts parametric data between Nordic and GSE2 format.
It can be used interactively or by giving the options as arguments. Type ‘norgse –help’ to see the
140
options.
NORHIN, From Nordic to Hypoinverse format
The program is started by typing 'norhin input-file'. The output file is norhin.out.
NORHYP, From Nordic to HYPO71 format (SUN and PC)
The program is written by F. Courboulex. The program asks for the input file name and the output file
name is norhyp.out.
PDENOR, converting PDE bulletin file to NORDIC format
PDE distributes bulletins on e-mail, both a monthly bulletin and a weekly bulletin (different formats).
The program converts one of these files to Nordic format and put the file into a standard SEISAN
database called PDE for the monthly files and PDEWE for the weekly files. This database must have
been created before running the program. Both CAT and S-files are made and SELECT and EEV can
be used afterwards
RSANOR
Program converts between format used by “Red Sismologica de Andalucia” and a few others in
Spain.
SEIGMT, Nordic to GMT input
The program SEIGMT reads information from Nordic or compact files and writes the parametric data
to files that can be used as input for GMT (Generic Mapping Tools, http://gmt.soest.hawaii.edu/). The
user can choose a scaling for the magnitudes and also select a magnitude type order. The scaling
option is useful if you wish to scale the symbol size of your epicenters with magnitude. The magnitude
type order defines, which magnitude should be taken in case several magnitudes have been
determined for one event. If you don’t give a magnitude order, the program chooses the largest
magnitude.
The files written by SEIGMT are:
gmtxy.out – event locations, to be plotted with psxy
gmtxyz.out – event locations and depths, to be plotted with psxy
gmtpath.out – travel path data, to be plotted with psxy
psmeca.out – fault plane solutions, to be plotted with psmeca (Aki and Richards convention)
SELMAP, selecting a subsection of a MAP file
The program can retrieve parts of a large MAP file written in SEISAN map format. On the SEISAN
web site or on the SEISAN CDROM, very detailed global mapfiles are available in SEISAN format.
The file originally comes from the USGS. SELMAP can select out part of a MAP file in a latitudelongitude grid. The MAP files consist of several small segments and a segment is selected if at least
one point is inside the specified grid.
USGSNOR, USGS catalog to NORDIC format
The program converts USGS CDROM hypocenters to NORDIC format. Most of the information is
used. If more than 3 magnitudes are available, only the 3 first are used. The number of stations is
included when available. The depth is indicated as fixed in all cases where the operator has been
used (A,N,G). Macroseismic information is included with max intensity. The residual standard
deviation is put into rms column. Event types are set to R. Magnitude types are converted as follows:
UK is made blank
b is replaced by B
s is replaced by S
D is replaced by C
w is replaced by W
WAVEFORM CONVERSION PROGRAMS
141
This group of programs are mostly converting waveform files from some format to SEISAN although a
few also convert from SEISAN to some other, mostly standard, formats. Most programs convert from
binary to binary formats.
Many instruments come with conversion programs to some standard format like PCSUDS or
MINISEED, and these have often been used to convert to SEISAN instead of writing programs
reading the original files directly. Many such conversion programs work on PC so the corresponding
SEISAN programs only work on PC. However, since the PC files can be read directly on Sun, this
should not present a problem. Many programs have VERY LITTLE documentation, look in source
codes for more information.
The number of programs are forever increasing with new recorders coming onto the market and new
formats coming in use and others going out of use and it is becoming increasingly difficult to keep
track of it all. For this release of SEISAN it has not been possible to test all programs on all platforms.
An attempt has been made to standardize the programs. A general problem is that many seismic
recorders and formats do not provide proper identification of the channels. In the worst cases, there
are no station codes, only channel numbers and in very many cases, there is no room for proper
component information. This is being taken care of by having a definition file, and only one format for
the definition file is used, see below. This is also used with program WAVFIX.
Most programs work in the standard way with a filenr.lis file as input (made with DIRF).
The response information is seldom in the original files and in most conversion programs, the
response information is taken from the CAL directory. If no response information is available, a
message will be given. For each program, a comment will be given as to the status of testing and on
which platforms they operate. If the channel definition file option is implemented, the array dimensions
will be SEISAN standard.
The program SEIPITSA might be an easy way to convert between 1-column ASCII data and SEISAN
(see below).
Conversion programs definition file
The conversion programs use a common format for the definition file for naming station and channels.
The definition file is named programname.def as e.g. sudsei.def. The definition file can be in the
working directory or the DAT directory. The conversion program will first look in the working directory
for the file and then in DAT. The conversion of codes can take place in 2 ways (see below for details):
(1) An input station and component code is converted to an output station code and component, (2)
an input channel number is assigned a station and component code. The advantage of (1) is that the
conversion is independent of the channel number or order, however, the user must then know the
default station and component names generated by the conversion program.
Default assignment of station code and component:
This is very much dependent on the conversion program used since some data files have complete
information and others very little, see description of individual programs in manual or at start of source
codes. In all cases, the conversion program will make both station and component codes based on
what is available of information in the input files. IT IS THESE CODES THAT are used for input code
as described below. In order to find out what they are, it is easiest to run the conversion program
once (without a def file) and see what codes the program assign. Alternatively, some of the programs
have documentation in the manual. Some of the station codes might be instrument serial numbers,
which are not always known. Therefore, running a test might be the best way to find out.
In addition to converting channel codes, the def file can also give SEISAN waveform file header
information and network code as it appears in the file name. If no network code is given, the network
code will be the station code of the first channel.
Principle of conversion in order of precedence:
(1) Both station and component given on input: Converted to what is given
for output station and component.
(2) If both are not present, the channel number is used.
Header line text (29 char)... NetCd
Header for REFTEK
NEWNT
(5 chars), Comment for next line
142
chan stati
1 BO11
BO12
BO13
comi
S Z
S N
S E
stato
BOM
BOM
BOM
como, In and output definitions, comment for next line
B Z
B N
B E
The first line is just a comment line, must be there in any format. Here it shows where the network
code is positioned as indicated by NetCd.
The second line gives the header information for the SEISAN main header, which are the first 29
characters. The file name network code is also given and is here NEWNT. Format a29,1x,a5.
The third line is just comment to indicate the position of the columns in the following lines (max 200).
A line must be there. The abbreviations are:
chan:
stati:
comi:
stato:
como:
Channel number, optional unless no input station and component given.
Input station code, 1-5 chars
Input component code, 4 characters
Output station code, 1-5 characters
Outut component code, 4 characters. First character MUST be S, L, B, A, or I,
last character MUST be Z, N or E, all upper case.
Format i5,1x,a5,2x,a4,1x,a5,2x,a4
The conversion programs are listed below
ARCSEI
Reftek archive to Seisan
ASCSEI
ASCII to SEISAN
BGISEI:
Beijing Geodevice Institue to SEISAN
CITSEI:
CityShark to SEISAN
CNVSSA:
Kinemetrics SSA to Kinemetrics Dataseis
CNVSSR:
Kinemetrics SSR to Kinemetrics Dataseis
DRSEI:
Sprengnether recorders to SEISAN
GIISEI:
Geophysical Institute of Israel to SEISAN
GSRSEI:
GeoSig to SEISAN
GSESEI:
See WAVETOOL
GSERESP:
Conversion between GSE and SEISAN response files
GURSEI:
Güralp to SEISAN
IRISEI:
From IRIS ASCII to SEISAN
ISMSEI:
ISMES to SEISAN
KACSEI:
Kinemetrics ASCII acceleration to SEISAN
KINSEI:
Kinemetrics Dataseis to SEISAN
K2SEI:
Kinemetrics K2 to SEISAN
LEESEI:
Willy Lee system to SEISAN
LENPCQ:
Converts from Lennartz to PCEQ to PCEQ format
LENSEI:
Lennarts ASCII to SEISAN
M88SEI:
Lennartz MARS88 to SEISAN
NANSEI:
NEISEI:
OS9SEI:
PITSA:
PCQSEI:
PDASEI:
PSNSEI:
QNXSEI:
RDSEED:
RSASEI:
RT_SEIS:
SACSEI:
SEIM88A
SEISAF:
Converts from Nanometrics to SEISAN
Converts from NEIC CDROM waveform data to SEISAN
Converts SEISLOG files to SEISAN waveform files
Conversion programs described with program PITSA
Converts from PCEQ to SEISAN
Geotech Instruments PDAS to SEISAN
Public Seismic Networks to SEISAN
SEISLOG QNX to SEISAN
IRIS program to read SEED volumes
Conversion from Andalucian Seismic network to SEISAN
Reftek Passcal format to SEISAN conversion
SAC to SEISAN
Conversion from SEISAN to MARS88 ASCII format
SEISAN to SESAME ASCII
143
SEIPITSA:
SGRSEI:
SISSEI:
SILSEI:
SUDSEI:
TERSEI:
WGSSEI:
SEISAN <-> PITSA ASCII
SeisGram to SEISAN
Sismalp format to SEISAN format
SIL network ASCII files to SEISAN
PCSUDS to SEISAN
Terra ASCII to SEISAN
WGSN format to SEISAN
For each program, a summary of capabilities is mentioned: The platforms available (all for all or
specific name), channel definition file available (chan. def. yes or no) and if the program will look for
response files in the CAL directory to insert in the headers (resp. yes or no).
If you do not find the conversion program here, look on the ORFEUS website for other programs that
might convert to one of the formats used above
(http://orfeus.knmi.nl/other.services/conversion.shtml).
ARCSEI, Reftek archive to SEISAN
ARCSEI is a program to automate the extraction of data from a RefTek data archive and the
conversion to SEISAN format. The program works interactively and with a simple text interface that
asks the user to enter the details for the data extract. Based on the user selected criteria the program
(1) extracts the data from the archive in Passcal format using ARCFETCH, (2) converts the Passcal
data files to SEISAN format using RT_SEIS, and (3) merges the SEISAN files if merging is activated
by the user, using SEISEI. The program is written in Fortran and works on Windows only.
The ARCSEI program can be used in various ways:
•
•
•
to extract a single time window from one or several stations
to extract several time windows from one or several stations
to extract sequential time windows from one or several stations
The ARCFETCH and RT_SEIS programs, both part of the RefTek software package, have to be
installed (see RefTek documentation) and the PATH variable set to include the directory where the
programs are stored. It is assumed that the RefTek data archive exists and that the user is familiar
with the content of the archive. The archive content can be shown with the command ARCINFO.
To test that the program is installed correctly, open the Windows command tool (from the menu, or by
selecting Start ⇒ Run ⇒ cmd) and type ARCSEI <RETURN>.
The definition file: arcsei.def
The purpose of the definition file is to set some parameters needed to run ARCSEI, however, the
program also works without. The arcsei.def file can either be stored in the seismo/DAT directory, or
the current working directory. The program first checks in the current directory. The arcsei.def file
should be adjusted to the user’s set-up, before ARCSEI is started.
The parameters are:
ARCHIVE: The path of the RefTek data archive, can also be entered manually at run time.
OUTPATH: The directory in which the SEISAN files are to be stored. The default is ‘.\’
(the current directory).
MERGE: Select if SEISAN files from several stations for the same time interval should
be merged (Y), or not (N).
NETWORK_CODE: Network code used in case SEISAN files are merged.
CHANNEL: Data channel in RefTek archive consisting of the unit, stream and channel
(unit,stream,channel). The * can be used as wildcard to select all streams or
channels, BUT not to select all units (since ARCFETCH is used in cooked mode,
which means that the time interval extracted matches the input start- and endtime.
Example of the arcsei.def file
144
KEYWORD............Comments.............Par1........................
ARCHIVE
OUTPATH
MERGE
NETWORK_CODE
CHANNEL
CHANNEL
CHANNEL
CHANNEL
CHANNEL
CHANNEL
CHANNEL
Path to archive
Path to put
converted SEISAN
files
Y or N for
merging Seisan files
used for merging
Name of channel
Name of channel
Name of channel
Name of channel
Name of channel
Name of channel
Name of channel
G:\CTBTO\ARCHIVE
C:\test
Y
ARCSE
8020,1,*
8021,1,*
8022,1,*
8023,1,*
8024,1,*
8025,1,*
8028,1,*
How to run the program:
ARCSEI is started from the Windows Command Tool (cmd) by typing
arcsei <RETURN>
ARCSEI - DATA EXTRACTION FROM REFTEK ARCHIVE AND CONVERSION TO SEISAN FORMAT
SEISAN FILES ARE MOVED TO DIRECTORY: C:\test
ENTER ARCHIVE, OR <RETURN> TO USE DEAULT (default: G:\CTBTO\ARCHIVE)
Return to accept default, which is set in the arcsei.def file, or give the archive path.
AVAILABLE CHANNELS ARE:
8020,1,*
8021,1,*
8022,1,*
8023,1,*
8024,1,*
8025,1,*
8028,1,*
ENTER CHANNEL SELECTION (UNIT,STREAM,CHANNEL)
OR TYPE ALL TO USE ALL AVAILABLE CHANNELS
Type channel and <RETURN>, if defined in arcsei.def channels are listed, otherwise an example is
shown. The channel is given as unit,stream,channel. Wildcards can be used for stream and channel,
but not for the unit.
NEXT CHANNEL OR RETURN TO CONTINUE
Additional channels can be entered, to continue press <RETURN>.
ENTER START-TIME (YEAR:DAY-OF-YEAR:HOUR:MINUTE:SECOND)
EXAMPLES: 2000:200:12
2000:200:12:15
2000:200:12:33:15
Type start time as year:day-of-year:hour:minute:second. Minute and second can be omitted.
NEXT START-TIME OR RETURN TO CONTINUE
Additional start times can be entered, to continue press <RETURN>.
ENTER END TIME USING ONE OF 3 OPTIONS:
- ABSOLUTE TIME AS YYYY:DDD:HH:MM:SS (LIKE START-TIME)
- +SECONDS FOR TIME INTERVAL (e.g. +300)
- ++SECONDS FOR MULTIPLE INTERVALS(CONTINUOUS EXTRACT, e.g. ++300)
Specify the end time, either in the same style as for the start time (only if one start time), or in some
cases more useful, the desired time window in seconds, by entering +seconds. If sequential time
windows are to be extracted, use ++ seconds. The user is then asked how many time windows
should be extracted. It is thus possible e.g. to extract 10 consecutive windows of 600 seconds.
Only if sequential extract windows specified:
145
ENTER NUMBER OF CONTINUOUS WINDOWS
After the program has finished, the data in SEISAN format can be found in either the current directory
(default) or in the OUTPATH directory if the variable is specified in arcsei.def. The temporarily created
files are deleted automatically.
How it works
ARCSEI reads the user input that specifies what should be extracted from the RefTek archive and
then calls the programs ARCFETCH, RT_SEIS and SEISEI. For temporary data storage ARCSEI
creates the directory arcsei_temp under the current directory. The arcsei_temp directory is
automatically deleted upon program completion.
(1) Create empty arcsei_temp directory
(2) Arcfetch
The arcfetch program performs the data extraction from the RefTek archive. A complete list of the
command line input of arcfetch can be obtained by starting the program without additional options.
ARCSEI starts arcfetch in the following way:
arcfetch archive channel,start-time,end-time –o OUTPATH -c
Where:
-o OUTPATH:
-c:
Specifies the output path for arcfetch, always arcsei_temp
Specifies cooked mode, which means that the time interval extracted
matches the input start- and end-time (this is not the case, when not running in
cooked mode)
Example:
arcfetch G:\ARCHIVE 8020,1,*,2000:200:12,+10 -oarcsei_temp –c
(3) rt_seis
RT_SEIS converts all files with the suffix ‘rt’ in arcsei_temp to SEISAN format. RT_SEIS reads the
RTU.INI file for station definition, if the environmental variable RTU is set to point to the RTU.INI file
(see RT_SEIS section below).
(4) SEISEI
SEISEI, if merge is selected, merges all SEISAN files in the arcsei_temp directory.
(5) move
Finally all files (single or merged) are moved to the OUTPATH directory or the current directory if
OUTPATH is not defined.
In case multiple stations are selected, ARCSEI runs steps (1) and (2) in a loop, before the data is
merged and moved.
In case several time windows are selected, ARCSEI runs steps (1) to (4) in a loop, and in addition a
second loop over multiple station (1) and (2).
If sequential time windows are specified, ARCSEI computes multiple start times and works as if these
time windows were user specified. All, def. File yes, resp yes
all, chan. def. yes, resp yes
ASCSEI, ASCII to SEISAN
Converts a single column file to a one channel SEISAN file. Date, time, sample rate, station and
component must be entered manually.
Linux, PCl, chan. def. yes, resp yes
BGISEI, Beijing GEODEVICE FORMAT (BGI) to SEISAN.
The program to convert waveform files from BGI to SEISAN format is called BGISEI... The instrument
146
response in the original files is not used. The program has only been tested with data recorded in
Cuba. The program is written by Bladimir Moreno.
all, chan. def. yes, resp yes
CITSEI, CityShark to SEISAN
Converts from CityShark ASCII to SEISAN. Components S Z, S N, S E are assumed for the 3
channels. Assume 3 channels files only, all channels same sample rate and number of samples.
CNVSSA and CNVSSR Kinemetrics accelerometers to Kinemetrics Dataseis PC
The programs are supplied by Kinemetrics to convert from SSA and SSR formats to Kinemetrics
Dataseis. To further convert to SEISAN, use program KINSEI. Only PC executable programs are
available. The data is 16 bit.
CSS
At the moment there is no direct conversion from CSS to SEISAN. It is possible to convert CSS data to
SAC or GSE using other tools like codeco, Geotool and SAC, and then convert to SEISAN format.
all, chan. def. yes, resp yes
DRSEI, Sprengnether data recorders to SEISAN
Converts Sprengnether DR3024 and DR3016 to SEISAN format. These two formats are slightly different,
but the program makes the adjustment. Only essential information is read in and only 4 lowest digits of
serial number are used. If station codes are set up, these are used, else the serial numbers are used for
station codes.
all, chan. def. yes, resp yes
GIISEI, Geophysical Institute of Israel to SEISAN
Converts Geophysical Institute of Israel imported DAQ files to SEISAN format. The initial station
codes are as defined in file, can be converted with the normal .def file. If 4.th character of station
name indicates component (N or E), that is blanked out and transferred to 4.th character of
component name BEFORE using the def file conversions.
all, chan. def. yes, resp yes
GURSEI, Güralp to SEISAN
Converts Güralp GCF files to SEISAN format, only works with one channel data. Maximum number of
samples as defined in seisan, at least 1 200 000, channels codes can be defined using the gursei.def
definition file. If no definition file, the station name is the first 4 letters from internal station name and
the component is B Z.
GSERESP, conversion between GSE and SEISAN response files all
The program provides conversion between SEISAN, GSE1 and GSE2 response files. The response can
be given in frequency, amplitude and phase (FAP) triplets or in poles and zeros (PAZ). Since the number
of values in the GSE format is unlimited the conversion from SEISAN to GSE only changes the format,
whereas converting from GSE to SEISAN, if the number of FAP triplets is more than 30 or the number of
poles and zeros larger than 37, the response in SEISAN format will be approximated by 30 FAP triplets.
The output files in SEISAN format will have the default SEISAN response filenames (see RESP program
and SEISAN response format). Output files in GSE format will include the station name, the component,
number 1 or 2 for GSE1 and GSE2 respectively and end on ‘.CAL’ (e.g. MOR_SHZ2.CAL (GSE2),
KONO_BZ_1.CAL (GSE1).
all, chan. def. yes, resp yes
GSRSEI, GeoSig to SEISAN
Converts from GBV recorders to SEISAN. GeoSig was earlier GeoSys. Before version 8.1, there was
a bug in program so start time was wrong by the amount of the prevent time.
all, chan. def. no, resp yes
IRISEI, IRIS ASCII to SEISAN
The input format is the variable ASCII download format used on the GSN Quanterra stations.
The format is used in connection with SEISNET. The program only works if input file has more than
1000 samples.
PC, chan. def. no, resp no
ISMSEI, ISMES to SEISAN
ISMES is an Italian seismic recorder. This is the first version of the program made by IIEES in Iran.
The program can convert one file with up to 3 channels.
KACSEI: Kinemetrics ASCII acceleration to SEISAN
all, chan. def. yes, resp yes
147
Kinemetrics ASCII film record acceleration files (type *.v1) to SEISAN. It is assumed that:
- channel 1 is N, 2 is Z and 3 is E
- there are always 3 channels in file
- input values are in 1/10 g, the output is in 1/1 000 000 g
- station code is taken from file name as given in first line of input file
- the 3 channels can have different number of samples, however it is assumed that they all start at the
same time
PC, chan. def. yes, resp yes
KINSEI, Kinemetrics DATASEIS to SEISAN
The program takes the station code from the input files. The component codes are also taken from
the input file as far as Z, N and is E is concerned, but the first letter is always set to S, like 'S Z'. The
program is also used if CNVSSR or CNVSSA have been used first.
PC,Linux, chan. def. yes, resp yes
K2SEI, Kinemetrics K2 to SEISAN
Program for K2 binary files. The program works by first converting the binary files to ASCII by
internally running the Kinemetrics program kw2asc (PC only).
If no definition file is present, channel 1-3 will be A Z, A N and A E. If more channels they will be
called A 04, A 05, etc.
PC, chan. def. no, resp no
LEESEI, Willy Lee binary files to SEISAN
The number of channels is fixed to 16 and the time information is not read, it must be entered when
converting the file.
LENSEI, Lennartz ASCII to SEISAN
all, chan. def. yes, resp yes
PC
LENPCQ, converting Lennartz to PCEQ format
Only executable code for this program and only PC (made by the Royal Belgian Observatory). The
format is used by an older version Lennartz tape recorder. The output files have the same names as
the input files and are placed in a directory c:\qcoda, WHICH MUST BE THERE.
M88SEI, Lennartz MARS88 to SEISAN
all, chan. def. yes, resp yes
PC, Sun, chan. def. yes, resp yes
NANSEI, Nanometrics to SEISAN
The program converts from the Y-file format to SEISAN. This is done by first making an ASCII file with
Nanometrics y5dump program (done internally in NANSEI). NOTE: The y5dump program requires
some special Nanometrics libraries (Solaris) or *.DLL files (PC), which are included and installed with
SEISAN (see installation section). The program converts single channel files only.
PC, chan. def. no, resp no
NEISEI, NEIC digital data to SEISAN
NEIC earthquake digital data comes on CDROM. The data is extracted with a program coming with
the data and then converted to SEISAN binary waveform data. The response information is given as
poles and zeros in the SEISAN waveform file header.
PC, SUN, chan. def. no, resp yes
OS9SEI, converting SEISLOG files to SEISAN
The program takes a SEISLOG ASCII (downloaded in CMP6 format) or binary file and converts to a
SEISAN file. The input can be several files from a filenr.lis or an ASCII downloaded file either
compressed or uncompressed. The program will look for the calibration file in the CAL directory and
add it to the SEISAN file, or give a message if it is not there. The program will work with SEISLOG
files recorded under operating system OS9 or QNX up to version 7.6. For QNX version 7.0, use
program QNXSEI.
PC, chan. def. yes, resp no
PCQSEI, converting PCEQ format to SEISAN
PCEQ format to SEISAN. Earlier used with IASPEI software libraries.
all, chan. def. yes, resp yes
PDASEI, converting PDAS files to SEISAN
The program converts a single channel PDAS file to a single channel file in SEISAN format. Several
of these files can then be merged with SEISEI. PDASEI in previous SEISAN versions (before version
6.0) only worked with PDAS in 16-bit format, so if 32 bit or gain ranged format was input, the output
148
would have been in error. The current version of PDASEI should be able to convert all 3 types of input
files. A description of the PDAS format is found in the PDASEI program.
all, chan. def. yes, resp yes
PSNSEI, Public Seismic Networks to SEISAN
The Public Seismic Network recording system makes one file pr channel. Since component is not well
defined, several files from the same recording system might get the same SEISAN file name. Do some
testing when setting up the recording system. The one component files can be assembled into
multichannel files with SEISEI. There might be a newer version of PSN format not supported.
all, chan. def. no, resp yes
QNXSEI, SEISLOG QNX version to SEISAN
This program works as OS9SEI except that it does not read the ASCII files. The program must be
used with Seislog 8.0. The program is currently the only program that put in the time synchronization
flag in SEISAN waveform files except for data logging programs under Seislog Windows. See format
description in Appendix 2. The program recalculates the sample rate based on the time in the first
blocks in the file and the last blocks in the file (each block is one second long). For very long files, this
might be of importance since the digitizer might not have exactly the nominal sample rate.
all, chan. def. yes, resp yes
RSASEI, Andalucian Seismic Network to SEISAN
Conversion of network and broad band files to SEISAN format. Covers several versions of the DTS
format also used by other institutions in Spain. Not tested on Linux.
PC, chan. def. no, resp no
RT_SEIS, Reftek Passcal to SEISAN
The RT_SEIS program converts Reftek Passcal format to SEISAN. This program is provided by
Refraction Technology Inc. The program does not use the filenr.lis as input file. To see the options of
RT_SEIS, start the program without any arguments.
In order to make use of the RTU.INI definition file, the environmental variable RTU needs to be set to
for example c:\seismo\dat (see RefTek documentation for more details). This file can be used to set
station names for respective unit IDs.
Example of
RTU.INI:
[8020]
Station=SB00
Network=CTBTO
CH1Band=
CH1Type=
CH1Axis=a
CH1Loc =
CH2Band=
CH2Type=
CH2Axis=b
CH2Loc=
CH3Band=
CH3Type=
CH3Axis=c
CH3Loc =
[8021]
Station=SB01
Network=CTBTO
CH1Band=
...
Linux, Sun, chan. def. yes, resp no
SACSEI, SAC to and from SEISAN
Note: From version 7.2, WAVETOOL should be used instead of SACSEI.
The SACSEI program converts between SEISAN and SAC ASCII/BINARY. SEISAN multi trace files are
split up into single trace files when converting to SAC.
PC, chan. def. yes, resp yes
SGRSEI
SeisGram binary to SEISAN. Only 3 component data has been tested. Channel order is assumed to be Z,
N, E. The input real values have been multiplied by 100 000 before being converted to integers. Program
little tested.
SEED
The Standard for Exchange of Earthquake Data (SEED) format is defined by the Federation of Digital
Seismographic Networks (FDSN). The rdseed program is distributed with SEISAN to extract data from
149
SEED volumes. RDSEED is an IRIS program to read SEED volumes. The program provides conversions
to SAC (ASCII and binary), AH, CSS and miniseed. It is described in the file ‘rdseed.txt’ in the INF
directory. Updated versions of rdseed will be available at ‘http://www.iris.washington.edu/pub/programs’. A
PC version (rdseed.exe) is distributed with SEISAN CD (also on home page). SEED volumes contain the
complete response information, details on how to convert the SEED response to GSE response format
can be found in Havskov and Alguacil (2004).
all, chan. def. no, resp no
SEIM88A, conversion from SEISAN to MARS88 ASCII format
The program converts SEISAN waveform files to Lennartz-ASCII MARS88 format. The program will
write one file per channel. Output files are either mars.xxx if a single file is converted or marsxxx.yyy if
the ‘filenr.lis’ file is used as input.
all, chan. def. yes, resp yes
SEIPITSA
The program converts from SEISAN to PITSA ASCII format and back. The ASCII format has one file
per channel. The user will be asked for a name of the output file-system. If a single file is converted,
the channel number will be added to the output file-system name (e.g. data.001). If the ‘filenr.lis’ file is
used the filenumber will be added to the file-system name (e.g. pitsa001.004, first file and fourth
channel). The program is no longer used for conversion when PITSA is started from EEV, but might
be useful, since it creates one column ASCII data and can easily be modified.
all, chan. def. no, resp no
SEISAF, SEISAN to SESAME ASCII
The 3 first channels in SEISAN file are read. There is no check if from same station. It is assumed
that the order in SEISAN file is Z,N,E, that all 3 channels have the same start time, number of
samples and sample rate. These values are taken from the first trace.
all, chan. def. no, resp no
SILSEI
Conversion from the Icelandic SIL system to SEISAN. Only conversion from ASCII files.
all, chan. def. yes, resp yes
SISSEI, Sismalp to SEISAN
The program converts from Sismalp to SEISAN. Sismalp is a French field recording system. The input
consists of 2 files pr event, a header file and a data file. It is assumed that the Sismalp ndx files have
the same file name as the header file except for the file extension. It is also assumed that the file
names are 12 characters long.
PC, chan. def. yes, resp yes
SUDSEI, PCSUDS to SEISAN
The program converts from PCSUDS to SEISAN. This is done by first running the program SUD2ASC
(included) and then converting to SEISAN. The SUD2ASC program and test data was supplied by
REFTEK through the distribution of PC-SUDS Utilities by Robert Banfill (1996).
all, chan. def. yes, resp yes
TERSEI, Terra ASCII to SEISAN
Program converts from Terra Technology ASCII files to SEISAN. Only tested with 1-3 channel files
all, chan. def. yes, resp yes
WGSSEI to SEISAN
Program converts from WGSN files to SEISAN. The format is used on IRIS stations as processing
format. Little tested.
6.13 Signal processing programs
The two processing systems PITSA and SAC are interfaced to SEISAN and can be started directly from
EEV. This is done since both systems support functions that SEISAN does not have. They only operate
on Unix. The Degtra program only operates in Windows. It is not interfaced directly to SEISAN but
operates on SEISAN format.
6.13.1 PITSA
PITSA (Programmable Interactive Toolbox for Seismological Analysis) is a program written by Frank
Scherbaum and James Johnson. The program is included in the SEISAN package, updated versions are
150
available at http://lbutler.geo.uni-potsdam.de/service.htm. From this version, PITSA is interfaced with the
SEISAN system through the program WAVETOOL, which converts waveform files from SEISAN to the
GSE2 format. PITSA since version 5.0 supports reading multi channel GSE2 files. PITSA can be started
from EEV by typing ‘pitsa’ on the prompt line. All waveform files listed in the S-file will be converted to
multi-channel GSE2 files. The multi-converted files are put into your local directory and are named ‘gse1’,
‘gse2’ etc. The response is converted to GSE format. When PITSA is started, the waveform files have to
be loaded using the GSE2 input format. The response file names will be given as described in the
GSERESP section.
6.13.2 SAC2000
SAC2000 (seismic analysis code) is currently developed by Lee Minner and Peter Goldstein
(Goldstein, 1999). SAC is not distributed with SEISAN, information on SAC can be obtained from the
SAC homepage (http://www-ep.es.llnl.gov/www-ep/esd/seismic/sac.html). The main features of SAC
include general arithmetic operations, Fourier transforms, three spectral estimation techniques, IIR
and FIR filtering, signal stacking, decimation, interpolation, correlation, and seismic phase picking.
SAC also contains an extensive graphics capability.
With SAC it is possible to write macros, which helps to process large amounts of data. The SAC
format is used in several research oriented programs. SAC can be started from EEV using the
command ‘sac’. EEV will start the WAVETOOL program to convert the data to SAC and then execute
the command sac. In case your sac executable is called sac2000, it is necessary to rename it (to sac)
or alternatively to create a link in either the SEISAN PRO directory or the SAC bin directory. This is
done for example by the command ‘ln –s /sac/bin/sac2000 /sac/bin/sac’. Since the SAC format is a
single trace format, the SEISAN multichannel files are split into single trace files. The station and
component names are included in the file name and the suffix ‘SAC’ is added to all SAC files.
For both systems, waveform data can be converted to the respective format outside EEV using
WAVETOOL, GSESEI or SACSEI, and the programs can be started without using EEV.
6.13.3 Degtra A4
Degtra A4 is a Windows-based computer system designed to process ground-motion time histories
for seismology, engineering seismology and earthquake engineering applications. It has been
developed at the Institute of Engineering, UNAM, Mexico, where it has been used as a professional
and teaching tool for more than 10 years. Figure 29 shows a typical view of Degtra A4.
151
Figure 29. Example of DEGTRA A4.
Degtra A4 accepts input ground-motion recordings written in several formats: simple ASCII or binary
files, SEISAN format and the standard format of the Mexican Strong-Motion Database.
The program has a nice and user-friendly interface, in which several time-histories can be
manipulated at a time. It includes zoom-in/zoom-out capabilities that can be used if only a portion of a
recording is of interest.
There are two types of operations that Degtra A4 can perform using strong-motion recordings:
operations involving a single recording and operations involving two recordings.
Among the first group, the main functions are the following:
•
•
•
•
•
•
•
•
•
•
•
•
Scaling. Multiplies the recording by a constant
Decimation. Decimates the recording by a user-given factor
Base-line correction: Includes several methods.
Differentiation: Numerical differentiation with respect to time.
Integration: Numerical integration with respect to time.
Computation of Arias’ intensity.
Filtering: Includes high-pass, low-pass, band-pass and band-stop Butterworth filters, as well
as Gaussian and Futterman filters.
Computation of elastic response spectra. Computes absolute acceleration, relative velocity,
relative displacement and pseudo acceleration response spectra.
Computation of inelastic response spectra. Computes required-strength response spectra for
fixed ductility demand of bilinear systems.
Computation of Fourier spectra. Computes and (optionally) smoothes FFT of a time signal.
Computation of response (acceleration, velocity, displacement) of elastic and inelastic
(bilinear) single-degree-of-freedom oscillators.
Computation of the S-wave response of a soil column of given properties assuming that the
152
time-history is the incident wave field at the base of the column.
The main operations involving two recordings are:
- Spectral ratio. Computes the (Fourier) spectral ratio between two recordings.
- Odogram. Displays an X-Y parametric representation (time is the parameter) of two recordings,
typically ground displacements, in order to observe particle trajectories.
- Addition/Difference. Computes the sum or difference of two recordings. This is useful, for instance,
to identify rocking or torsional motions in buildings.
- Rotation: Rotates two components of the ground motion to arbitrary angles.
- Cross-correlation: Computes cross-correlation between two signals.
- Coherence: Computes coherence between two signals.
Degtra A4 includes an on-line help file with details about the meaning of required parameters,
techniques used, and use of Degtra A4 itself. It is available in two versions: one for Windows 2000 or
lower and another for Windows XP. Installation is made with a typical Windows executable setup file.
The distribution is found in directory SUP.
Degtra A4 is currently in Spanish, English version in preparation.
For questions please contact:
Mario Ordaz
Institute of Engineering, UNAM
[email protected]
6.14 Calculating b-value, BVALUE
BVALUE is a program to make b-value plots using a NORDIC input file (also compact). A postscript
plot file is generated.
The questions are:
Input file name, select.out or collect.out are defaults
! Give filename or return
Which magnitude type, C,L,B,W or S, return for no type
! C: coda, L: Ml, B: mb and S: surface wave magnitude, W: Moment mag.
blank: no magnitude type
Output:
Number of events selected from file:
91
Duration of catalog in years: 0.502
! Output number of events selected and duration of catalog
New input:
Magnitude step 1.0, 0.5, 0.25, 0.2 or 0.1
! Magnitude step for summing number of events, MUST be one of the above.
Magnitude range m1,m2 for b-value and fixed b-value
! Range for calculating b-value, and the fixed b-value for which a-value is calculated.
The a value is calculated as the average of N(i) + bfix * mag(i), where N(i) is the
accumulated number of events at magnitude mag(i) and bfix is the fixed bvalue.
Output is now:
n
m1
m2 maxl a maxl b
sd
52
2.0 4.0 3.25
0.68 0.46
Normalized
3.55
Normalized m1
2.19
Norm. lin. m1
154.9
lsq a
3.77
4.07
2.21
162.2
lsq b
cor
0.93 -0.93
rms
0.14
bfix
1.0
afix
sd
4.02 0.16
4.32
2.32
208.9
! Normalized means normalized to one year (m=0), Normalized m1 (m=m1) and Norm. lin. m1 is
just antilog the previous. So in the above example, it is expected that 162.2 earthquakes
occur larger than or equal to 2 in one year (least squares relation).
mag nmag cmag
0.4
1
91
0.6
1
90
0.8
3
89
1.0
6
86
1.2
6
80
1.4
7
74
1.6
4
67
1.8
11
63
153
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
8
11
6
7
5
4
7
52
44
33
27
20
15
11
n: Number of values used
m1 and m2: Magnitude range used for b-value calculation
maxl a and b: Maximum likelihood a and b
sd: Standard deviation
lsq a and b: Least squares a and b ( the one plotted)
cor and rms: Correlation coefficient and rms of above
bfix: fixed b-value given at input
afix: a-value for above
sd: standard deviation for above
mag: magnitude
nmag: number of events within mag interval
cmag: cumulated number of events
Questions:
Plot(y/n)
! This will make a plot and a postscript plot file.
Note that only the last plot on screen is saved. The plot
must be sent manually to the printer.
Which b-value 1: Least squares (default)
2: Fixed bvalue
3: Maximum likelihood
! Choice of b-value to plot
Run again with other parameters (y/n)
! Another step length or range can be selected
The final output is:
Output file in bvalue.out
Last plot in file bvalue.eps
The output file bvalue.out contains the same information in the same format as shown in the example
above. The file can be used with other plotting programs to make 'nicer looking' b-value plots. An
example is shown in Figure 30.
154
Figure 30. An example of a b-value plot. The bars are number of events and crosses the
accumulated number of events.
155
6.15 Automatic phase picking, AUTO, AUTOPIC, AUTOSIG
AUTO and AUTOPIC
This program will make picks automatically of events registered into the database. The program will go
through a series of events in the usual way using start time and end time. If an event file (S-file) has any
readings, the program will not reread in order to not destroy old picks. The automatic readings in
the file are marked with an A after the weight column to indicate automatic pick. Each pick is evaluated by
using the signal to noise ratio and an indication of the quality is given with the weight. The program will run
on all waveform files given in an S-file. Each time the program runs, there is a file called autopic.out
containing information about the run.
If there are any 3-component stations, an azimuth will also be calculated, and the S-phase will be more
reliable.
The pick program can also be used from EEV by typing Z (will run program AUTOPIC). When it is used
from EEV, there is always an output in the S-file, which will be grouped at the bottom of the file, making it
possible to compare manual and automatic readings. THE S-FILE MUST THEN BE EDITED MANUALLY
IN ORDER TO REMOVE DOUBLE READINGS.
The program requires an input parameter file in the working directory or DAT with the name
AUTOPIC.INP. The program will first look in the working directory. The parameters in that file are
explained in file AUTOPIC.INF. Both files are shown below.
The program uses a 4-pole filter running one way. This might result in phases being picked a bit late.
However, it seems more accurate than the earlier version where the filter run both ways and picks were
often far too early.
The program is made mainly by Bent Ruud. For more information about how it works, see Ruud and
Husebye (1988) and Ruud et al., (1992).
Description of parameters
------------------------------------------------------------------%
% Input parameters common to all filters:
%
% LWIND : used to define step length (DELTA=WINDOW/LWIND)
% ISHIFT : defines time shift between STA and LTA window (ISHIFT*DELTA)
%
Delay for LTA window (15 * 0.1) ==> 1.5 sec. after STA window.
% ISIGMA : defines fall-off rate of LTA window (larger values - longer windows)
%
LTA(i) = (1 - 2^(isigma)) * LTA(i-1) + 2^(-isigma) * STA(j)
% COHMIN : Polarization threshold.
%
Minimum coherence (see thresh_1 and thresh_2)
% NDMIN : Mimimum number of consecutive triggered windows in a detection
% SVELO : S-wave velocity of the medium below the station (used for 3-comp)
% NFILT : number of filters
% CRAT
: Ratio for calculation of coda duration ( range 1 - 4)
% LWIN
: Window used in coda duration routines (range 20 - 50 seconds)
% THRES : Quality threshold (range 2 - 5). Used on the maximum to average
%
amplitude ratio in order to sort out the most noisy traces.
%
% Input parameters defined for each filter:
%
% WINDOW : length of the moving time window (sec)
% F1
: lower cutoff frequency (Hz) of band pass filter
% F2
: higher cutoff frequency (Hz) of band pass filter
% THRSH1 : STA/LTA threshold for polarized signals
% THRSH2 : STA/LTA threshold for unpolarized signals
%
If coherence > cohmin then detection is made on thresh_1
%
If coherence < cohmin then detection is made on thresh_2
%
% Output parameters:
%
% D
: day of year
% H
: hour
156
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
M
SEC
DUR
FRQ
SNR
STA
NT
NH
NV
NC
Q
PS
AZI
DA
VEL
DV
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
minute
second
duration, i.e. time in detection state (sec)
centre frequency of filter giving the best detection (Hz)
signal to noise ratio (SNR=STA/LTA)
short time average (root mean square of amplitude)
total number of triggered time windows in the detection
number of windows with best SNR on one of the horizontal comp.
number of windows with best SNR on the vertical comp.
number of windows with acceptable polarization
quality class, 1(best) - 4(worst)
P/S wave discriminator, 0(S) - 10(P)
backazimuth in degrees measured from North through East
variability in azimuth (deg)
apparent velocity (km/s)
variability in apparent velocity (km/s)
Note : azimuth and apparent velocity calculations are based on the
assumption of P-wave, so that these variables should be
neglected for S-waves.
Example of input file AUTOPIC.INP for AUTO
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
This is the parameter file needed by program: --- AUTO --The following rules apply:
1. All lines with % in the first column are comment lines
2. Lines with a blank in column 1 are read for fixed parameters.
3. All lines starting with "filter_x", where x is a number,
are read for filter variable parameters
4. All lines with * in the first column are read for stations to process
5. A brief explanation of all parameters is given in preprocess.inf
FIXED PARAMETERS THAT ARE USED THROUGHOUT THE PROGRAM
---------------------------------------------------------------------Lwind Ishift Isigma Cohmin Ndmin Svelo Nfilt
Crat
Lwin Thres
!
!
!
!
!
!
!
!
!
!
---------------------------------------------------------------------4.0
15.0
6.0
.4
.0
2.75
3.0
1.6
30.0
3.0
%
%
% PARAMETERS THAT ARE FILTER DEPENDANT
% ---------------------------------------------------------------------%Filter_nr
Window
F1
F2
Thrsh1
Thrsh2
%
!
!
!
!
!
!
% ---------------------------------------------------------------------filter_1
.0
1.0
4.0
1.70
2.55
filter_2
0.6
4.0
8.0
1.80
2.60
filter_3
0.4
8.0
16.0
1.90
2.75
%
%
% STATIONS TO USE IN THE PROCESSING
% ---------------------------------------------------------------------*SUE S Z
*BER S Z
*HYA S Z
*KMY S Z
*ODD S Z
*ODD1 S Z
*BLS S Z
*BLS1 S Z
*ESG S Z
*EGD S Z
*KTK1 S Z
*NSS S Z
*MOL S Z
*JNW S Z
*JNE S Z
*FRO S Z
*JMI S Z 3 component
*ASK1 S Z 3 component
*ASK S Z 3 component
*MOR7 S Z 3 component
*LOF S Z 3 component
*OSG S Z 3 component
*TRO S Z 3 component
*FOO S Z 3 component
157
AUTOSIG
AUTOSIG is a program to perform some automatic processing. The program includes routines for Pphase picking, determination of signal duration, amplitude determination, determination of spectral
parameters (Ottemöller, 2003) and determination of distance type (local, teleseismic, noise). The
program can still do with improvement. The input to the program can be either a parametric Nordic file
(with one or several events) or waveform files. In both these cases, the output is written to the
autosig.out file. Additional output files are autosig.trace and autosig.err, which will help to find
potential problems. Alternatively, the program can also be started from EEV; the output is then directly
written to the S-file. The input parameters are defined in the file autosig.par, which is located either in
the DAT or the working directory.
Following are descriptions of the automatic processing routines:
P-phase picking:
The phase picking is based on changes in the STA/LTA ratio. A band pass filter can be specified. The
routine gives options to enhance the changes in the signal before computing the STA/LTA ratio. It is
recommended to first remove the linear trend and then to compute the characteristic function which is
given by y**2+k*(dy/dt)**2, which enhances changes in both amplitude and frequency content. Then
the STA/LTA ratio is computed to detect changes in the signal. The routine can also compute the
squared STA/LTA. When a change is detected (STA/LTA ratio above trigger level), it is tested
whether the signal spectral amplitudes are significantly higher (factor of 2 in amplitude) than the presignal noise spectral amplitudes. This is done to avoid triggering on spikes.
Signal duration:
The signal duration is determined by comparing the signal amplitudes with the amplitudes of the presignal noise. The duration is determined by the point from which the ratio of these amplitudes is lower
than a given value. A filter is applied if specified in the parameter file.
Amplitude:
Routine finds maximum amplitude between two peaks.
Spectral parameters:
The routine computes the displacement amplitude spectrum for P or S waves (see section 6.2.11)
and, using either a converging grid search or a genetic algorithm determines the seismic moment and
the corner frequency by minimizing the difference between observed and synthetic source spectra.
The frequency band is determined by comparison with the pre-signal spectrum. The grid search is
generally more cost effective and produces better results. The method is described in Ottemöller and
Havskov (2002) (see also the file qspec.pdf in the INF directory). The displacement spectrum is
corrected for geometrical spreading and attenuation (both along the travel path and near surface).
Therefore, the hypocentral distance has to be known. The time domain window for extracting the data
from the trace can be given by either a group velocity (Vg=distance/travel time) window or a fixed
window in seconds around the phase pick.
Distance type:
The routine determines whether the signal is from a local or teleseismic event, or noise. If signal
spectral amplitudes are not significantly higher than pre-signal noise amplitudes, it is assumed that
the signal is noise. Otherwise the amplitudes at two selected frequencies given by ‘DIST FREQ
SELECT’ are compared, the rules are (f1<f2):
Spec signal amp(f1) - Spec noise amp(f1) > Spec signal amp(f2) - Spec noise amp(f2): teleseismic
Spec signal amp(f2) > Spec noise amp(f2): local
There are a few command line options that can be used to run autosig in non-interactive mode,
syntax is
autosig -infile <filename> [-spec on/off -phase on/off -clear on/off]
158
where
-spec on/off: determine spectral parameters if option given
-clear on/off: remove phases from input S-file before start if option given
-phase on/off: detect phases if option given
-infile <file>: give name of input file, either S-file or waveform file
-help: get help
Note: When running the program the first time and the hypocenter location is not known, the
determination of spectral parameters is not done. To run the determination of spectral parameters,
the hypocenter location has to be given in the S-file.
The meaning of most parameters in the parameter file is clear from the keyword. The spectral
parameters are as described in the MULPLT section. Other parameters that need explanation are:
AUTO PHASE, AUTO SPECTRUM and AUTO AMPLITUDE: Logical flag to activate phase picking,
spectral analysis and amplitude reading, respectively (1. for true)
GA POPULATION SIZE: Number of elements in the population, used only if SEARCH ALGORITHM
is 1.
GA GENERATIONS: Number of generations in one run, used only if SEARCH ALGORITHM is 1.
Note: Increasing GA POPULATION SIZE and GA GENERATIONS will increase the computation time.
GRID NLOOP: Number of loops in converging grid search for spectral parameters, used only if
SEARCH ALGORITHM is 2. Resolution increases with every loop.
NGRID FREQUENCY: Number of grid points in search for corner frequency, used only if SEARCH
ALGORITHM is 2.
NGRID SPECTRAL AMP: Number of grid points in search for spectral amplitude, used only if
SEARCH ALGORITHM is 2.
NORM: Norm for computation of residuals in spectral fitting can be set, however, tests show that 1 or
2 produce the same result, and generally default of 1 can be used.
SEARCH ALGORITHM: Defines whether genetic algorithm (1) or converging grid search (2) should
be used. Converging grid search is recommended.
SELECT PHASE: Defines, which phase to use for spectral analysis, choices are: 0 for P by
AUTOSIG, 1 for computed P arrival for given location, 2 for computed S arrival, 3 for P from s-file, 4
for S from –s-file or 5 for S or P from s-file.
SEPCTRUM F LOW: Lower limit of frequency band to be used.
SPECDURATION CHOICE: The time window for computation of the spectrum can be given either as
a time window starting from the phase onset (0.) or can be defined by a group velocity window (1.).
SPECTRUM P LENGTH: Duration in seconds of signal starting from P arrival.
SPECTRUM S LENGTH: Duration in seconds of signal starting from S arrival.
SPECTRUM PRE LENGTH: Duration in seconds of signal to be included prior to phase arrival.
GROUP VEL WINDOW P: Range of group velocities defining time window to be used for P spectrum.
Time window is given by (distance/group velocity)
GROUP VEL WINDOW S: Range of group velocities defining time window to be used for S spectrum.
159
Time window is given by (distance/group velocity)
STALTA NREC/REC: There are two STA/LTA algorithms, recursive (0.) and non-recursive (1.).
STATION LINE: One line with processing parameters for phase detection is given for each channel.
The parameters are (also see example below):
STAT – station name
COMP – component name
STA – duration of STA
LTA – duration of LTA
RATIO – trigger ratio
MINCOD – minimum coda required for trigger
DTRLE – de-trigger level
FILL – bandpass filter low cut
FILH - bandpass filter high cut
Example of the parameter file autosig.par:
#
KEYWORD............Comments.............Par 1.....Par 2
#
# spectral parameters
#
SPECTRAL S-Q0
Q0
SPECTRAL P-Q0
Q0
SPECTRAL S-QALPHA Q = Q0**Qalpha
SPECTRAL P-QALPHA Q = Q0**Qalpha
SPECTRAL KAPPA
SPECTRAL P-VELOCITY P velocity
SPECTRAL S-VELOCITY S velocity
SPECTRAL DENSITY
Density
#
# auto signal processing parameters
#
REMOVE MEAN
1. for true
REMOVE TREND
1. for true
CHAR FUNCT
1. for true
K IN CHAR FUNCT
K IN X=Y^2+K*Y'^2.
STALTA NREC/REC
rec 0./ non-rec 1.
STALTA SQUARE/ABS
square 1.
AUTOCODA SQUARE/ABS square 1.
AUTO PHASE
1. for true
only if no phase
AUTO SPECTRUM
1. for true
AUTO AMPLITUDE
1. for true
AUTO LOCATE
1. for true
NORM
SEARCH ALGORITHM
#
# window selection
#
SPECTRUM P LENGTH
SPECTRUM S LENGTH
SPECTRUM PRE LENGTH
GROUP VEL WINDOW P
GROUP VEL WINDOW S
SPECDURATION CHOICE
#
# select phase
#
SELECT PHASE
1.=GA 2.=GRID
in seconds
in seconds
in seconds
0:SPEC. P/S LEN.
1:GROUP VEL W. P/S
0:auto P
1:synth P
2:synth S
3:P from file only
4:S from file only
440.0
85.0
0.70
0.70
0.02
6.2
3.6
2.8
0.
1.
1.
3.
1.
1.
0.
0.
1.
1.
0.
1.
2.
5.
2.
.5
5.0
2.7
1.
2.
6.5
3.7
160
5:S or P from file
--- for 3-5, if no phase and AUTO PHASE is 1. use auto phase pick --DIST FREQ SELECT
1.
10.
#
# parameters used in the genetic algorithm searching for spectral parameters
#
GA POPULATION SIZE
50.
GA GENERATIONS
250.
#
# grid dimension in case of grid search
#
GRID NLOOP
5.
NGRID FREQUENCY
100.
NGRID SPECTRAL AMP
100.
#
# low filter limit to use for auto spectrum
SEPCTRUM F LOW
.05
#
# parameters controlling output
#
CREATE WAVEOUT
SPEC OVERWRITE
PHASE OVERWRITE
1.
0.
0.
#
# station parameters
#
#
STAT- COMP -sta-- -lta-- -ratio mincod -dtrle fill filh
STATION MOL
S Z
3.0
20.0
10.0
2.5
1.5 5.0 10.0
STATION MOL
B Z
3.0
20.0
10.0
2.5
1.5 5.0 10.0
STATION HYA
S Z
3.0
20.0
10.0
2.5
1.5 5.0 10.0
STATION LSA
L Z
3.0
20.0
10.0
10.0 9999.0 00.1 10.1
STATION CHTO L Z
3.0
20.0
10.0
10.0 9999.0 00.1 10.1
STATION XAN
L Z
3.0
20.0
10.0
10.0 9999.0 00.1 10.1
6.16 Fault plane solution, FOCMEC
The program can be used to determine double couple earthquake focal mechanisms using polarities
and/or amplitude ratios. The program also provides an interactive graphical display. The existing
solution can be plotted without any station data or location being available, however if existing
polarities should be plotted, the event must be locatable in order to calculate angles of incidence.
Several solutions can be plotted on the same figure in order to compare solutions.
The SEISAN program FOCMEC provides the interface between the database and the program that
determines focal mechanisms, which in SEISAN is the program FOCMEC_EXE. This program is written
by Arthur Snoke (Snoke et. al., 1984) and distributed as part of the FOCMEC package
(http://www.geol.vt.edu/outreach/vtso/focmec). FOCMEC_EXE is identical to FOCMEC in Snoke’s
package and can be easily upgraded (unless formats are changed). Generally the user will use FOCMEC
when working with SEISAN data, however, it is also possible to run the original version (see
documentation by Snoke: INF/focmec.pdf).
The program works with polarities and amplitude ratios. See the MULPLT section on how to read
polarities and amplitudes. Amplitude ratios are computed from amplitude readings given in the S-file.
While amplitude ratios can provide additional constraint on the solution, they should be used with caution.
Generally, the solution should be well constrained by polarities only, and then amplitude ratios can provide
confirmation of a solution or help to select one of several equally good solutions. Amplitude ratios must be
determined from the same wave type for example Pg and Sg. While in principle it should be possible to
use ratios determined from refracted waves, generally ratios determined only from direct waves are used.
The ratios that can be used are SV/P (either direct or refracted waves), SH/P and SV/SH (both direct
waves only), where P is the P wave read on a vertical component, SV is the S wave amplitude read on a
vertical or radial component, SH is the S wave read on the transverse component. FOCMEC applies the
free surface correction, but assumes a constant vp/vs and the same Q for P and S waves and thus the
ratios are not corrected for possible difference in attenuation or geometrical spreading. SV amplitude and
161
phase change rapidly around the critical angle and should not be used at those angles (see
INF/focmec.pdf for details).
The program makes a grid-search and finds how many polarities fit each possible solution. All solutions
with less than a given number of wrong polarities and/or amplitude ratios within given error limits, are then
written out and can be plotted. With a cursor, the user can then select the preferred solution, which can be
stored in the input file or the database. The program is intended to work from within EEV (option F),
however it can also work independently (see below). The program uses an input file called focmec.inp.
This is a Nordic format file. The angle of incidence is put in column 58:60, format I3 (although from version
8 onwards, this information is part of the S-file). Direct waves have angle > 90 and refracted arrivals angle
<90 degrees. If the angle is >90, the polarity is plotted at an azimuth+180. If the program is operated from
within EEV, this information is automatically put in and the focmec.inp file created. If the user wants to use
FOCMEC as a freestanding program, the angle of incidence information must be put in manually in a
standard CAT-file, which is then renamed focmec.inp. This can be done automatically by FOCMEC if a
hyp.out and corresponding print.out file is available.
FOCMEC can also be used to convert angles, like dip, strike and rake to T and P-axis, simply say ‘focmec
a’, where argument a stands for angles and you will be prompted for input.
When the program runs, the following menu is put up:
Stop
(0)
Plot saved solution
(1)
Plot new solutions
(2)
Plot selected solution
(3)
Find new solutions
(4)
-1, -2, -3 also plot station
1: This is the solution(s) already stored in the data base (S-file). The rule is that there should only be
one prime fault plane solution. The prime solution has ‘ F’ in the last two columns of the line. If any
other character is put into column 79, the solution is not considered prime, however it will be left in the
file when new solutions are generated. There might be a need to plot several solutions in order to
compare solutions. In that case the character in column 79 must be ‘O’.
2: Plotting new solution after having used option 4
3: Plotting the selected solution after using option 4
Using e.g. -1 instead of 1, also plots the stations to help identify them on the plot, see Figure 31
4: Starting a search for new solutions
Option 4 gives the following information and questions:
Number of polarity values: Number of polarities found for event with P-phases.
Any P-phase can be used like Pn and Pg. When few polarities are available, it is an advantage to use
both Pg and Pn since these phases have different angles of incidence. Polarities associated with
other phases are not used. There is no check if a P-phase has been duplicated.
Number of amplitude ratios: Total number of amplitude ratios derived from amplitude readings in Sfile.
The program now asks:
Maximum number of polarity errors: Depending on number of data values, 0-5 is a good answer. To let
the program find the minimum number of polarity errors, type ‘-1’, which is particular useful if there is a
significant minimum number of polarity errors.
Maximum number of amplitude ratio errors: Equivalent for ratios to ‘Max number of polarity errors’,
however, error is defined by amplitude ratio error.
Maximum amplitude ratio error: Give maximum allowed difference between observed and computed
amplitude ratio, default is 0.2.
162
Degree increment in search: Initially use e.g. 20 deg to make a fast search, later use e.g. 5 deg to make a
final solution.
The program will now start the searching and write out on the screen (and in a file) the solutions
which fit the requirement of number of misfits. The maximum number of solutions is limited to 100 as
a default, or to the value defined by ‘FOCMEC MAXSOL’ in SEISAN.DEF. At the end, the number of
acceptable solutions is written out as well as the minimum number of bad fits. This can then be used
for the next search. Now option 0 to 4 can be used again.
When plotting the solution with option 2, the cursor comes up. Also, the solutions will be printed in text
form to the screen, e.g.:
Strike
358.0200
358.1700
Dip
15.7900
28.9000
Rake Pol: P
-71.3200
1
-57.6200
1
SV
SH
0.0
0.0
0.0
0.0
Rat Err
0
0
RMS RErr
0.25
0.09
RErr (All)
0.25
0.09
The polarities and amplitude ratios can be plotted on the focal sphere using the same convention as
the original FOCMEC program, which is:
○ = compression
+ = emergent compression
∆ = dilatation
- = emergent dilatation
V = amplitude ratio SV/P
S = amplitude ratio SV/SH
H = amplitude ratio SH/P
The last three fields Rat Err, RMS Rerr and Rerr refer to the number of ratio errors, the RMS error for
the ratios used and the RMS error for all ratios, respectively. When saving a solution, the same text
field is included into the S-file.
The user can select a preferred solution by moving the cursor near one of the letters T or P (T and P
axis). By pressing T, the program will find the nearest T axis (same for P and nearest P-axis) and
corresponding fault plane solution, which can be stored in the database and/or plotted with option 3. If
no solution is to be selected, press q for quit. If a solution has been selected, the user will be asked if
it is to be saved or not after selecting option 0. The saved solution goes into the focmec.out and from
there into the S-file (type F-line) in the database if FOCMEC is operated from EEV. NOTE: The
previous fault plane solution will be overwritten unless a character is written in column 79 of the fault
plane solution line. If e.g. the last 2 characters are ‘OF’, this solution remains in the S-file.
When working from EEV, the event will always be located before the FOCMEC program starts up. In
the Nordic format the solution is stored simply as strike, dip, rake and number of bad polarities
(3f10.1,I5). Aki and Richards convention is used. In addition, the name FOCMEC will be written near
the end of the line to indicate that the fault plane solution was made by FOCMEC. The other program,
which can make a fault plane solution, is INVRAD (see EEV). The line type is F.
The following files are created:
focmec.dat: Input parameters to FOCMEC_EXE.
focmec.log: Log of the FOCMEC_EXE run.
focmec.lst: More details on solutions
focmec.out: Gives input parameters and solutions
focmec.eps: A Postscript plot file of LATEST plot
focmec.run: Run parameters for FOCMEC_EXE, you can re-run FOCMEC by ‘focmec_exe < focmec.run’
Running FOCMEC independently of EEV and composite fault plane solution:
This can be done in two ways:
1. Locate event(s) with HYP, then give command focmec. The program then combines the files print.out
163
and hyp.out to make the focmec.inp file and proceeds as usual. This is actually the way FOCMEC works
from within EEV. However, if more than one event is located, FOCMEC assumes that all events shall be
used in a composite solution, and focmec.inp will therefore contain the header from the first event and
phase lines from all subsequent events. This is the easiest way to make a composite solution.
2. Manually make the focmec.inp event or use option FC in EEV. Then run focmec with the argument c to
indicate that the focmec.inp file already exists and a composite solution is to be made.
Using EEV and option FC (see EEV) an input file focmec.inp is generated containing polarities etc for
several events.
NOTE, when running FOCMEC outside EEV, the fault plane solution is not put into the database (it does
not belong to any particular event !), however it is written out in file focmec.inp.
Computer limitations: Total number of polarities must be less than the dimension of array DATA
(parameter max_data) for Nordic data (see file seidim.inc in INC directory).
Figure 31 shows an example of a fault plane solution calculated with FOCMEC.
164
Figure 31. Top: An example of a fault plane solution plot. Symbols are explained in the text.
Bottom: A fault plane solution also showing the stations with corresponding polarities.
165
6.17 Calculation of coda q, CODAQ
The program will calculate coda Q (hereafter called Q) for a series of events and stations at given
frequencies. On completion, the average values are calculated and a Q vs f curve is fitted to the
calculated values. The program will also plot the individual events and filtered coda windows.
The principle for calculation is the standard coda Q method, whereby a coda window is bandpass
filtered, an envelope fitted and the coda Q at the corresponding frequency calculated. The envelope is
calculated RMS value of the filtered signal using a 5 cycle window. The program used here is the one
described in Havskov et al. (1989).
The program can only operate in connection with the SEISAN format S-files and waveform files and
will also take advantage of the SEISAN database structure.
Input:
The calculations are controlled by a parameter file called codaq.par and the actual event-station
combinations to use are given in codaq.inp. Example files are in DAT, and with the test data set and
the example files in DAT, a test run can be done. An example of a parameter file is shown below:
-----------------------------------------start in s-times
2.0
absolute start time (sec)
0
window length (sec)
20
spreading parameter
1.0
constant v in q = q0*f**v
1.0
minimum signal to noise ratio
2
Noise window in front of signal and length of RMS noise window
15,5
minimum correlation coefficient
0.50
maximum counts to use
64000
number of frequencies
3
frequencies and bands
8,3
12,4
16,6
default stations(1. line) and components (2. line), 30a5
HYA ASK SUE
S Z S Z S N
---------------------------------------------------------
Start in s-times: Normally the coda window starts at twice the S-travel time from the origin, this factor
can be varied and might be chosen differently in special cases. Note that the S-time is calculated from
the P-time so a P-time must be present. This also means that if a Pn is used, the coda window will
start at 2 times the Sn travel time, which might be substantially different from 2 times the Sg travel
time.
Absolute start time: If 0.0, above parameter is used. However if different from zero, an absolute start
time relative to the origin time is used for the start of the coda window. This might be useful since
different start times (meaning different lapse times) might produce different q-values. To use this
parameter, one must be certain to choose it long enough which can be checked with the plots. If the
absolute start time is smaller than (Start in s-times) multiplied by the s travel time, the station will be
skipped and a message given.
Window length: This is the coda window length in secs. Use at least 20 secs to get stable results.
166
Spreading parameter: The geometrical spreading parameter used in q-fit, normally 1.0 is used.
Constant v in q = q0*f**v : For all q(f) values, q0 is calculated using a fixed v, use e.g. 1.0. This
parameter has no influence on the individual q calculations.
Minimum signal to noise ratio: In order to accept a q value for the average, the signal to noise ratio
must be above this value. The signal to noise ratio is calculated using the last tRMS ( see next
parameters) secs of the filtered coda window and the first tRMS secs of the data file window. If the
data file starts with noise or in the P signal, the s/n ratio will be in error. A reasonable value is 5.0.
Maximum counts to use: If the count value in a coda window is above this value, the window is not
used. The intention is to avoid using clipped values. From SEISAN version 7.2, there is also an
automatic checking for clipped values in addition to ‘maximum counts’.
Noise window in front of signal and length of noise window, tnoise and tRMS: The first number is the
number of seconds of noise to plot in front of the signal. In previous versions, 15 secs was hardwired,
but sometimes there was not 15 secs of noise before the P. The second number is the length of the
noise window used for calculation of the signal to noise ratio. This was earlier hardwired to 5 secs.
Minimum correlation coefficient: In order to use the q value in the average, the correlation coefficient
of the coda q fit must be larger than or equal to this value. NOTE. Correlation values are in reality
negative, but are always referred to as positive in the following. An acceptable value depends on the
data, try to use a value higher than 0.5 (in reality -0.5)
Number of frequencies: Number of frequencies to use, maximum 10, 5 is a good number.
Frequencies and bands: The corresponding center frequencies and frequency bands. The frequency
band should increase with increasing frequency to avoid ringing. E.g. 8,3 means that the signal is
filtered between 6.5 and 9.5 Hz. It is advisable to use constant relative bandwidth filtering, to get an
equal amount of energy into each band. The relative bandwidth is defined as RBW = ( fu - fl )/ fo where
fu and fl are upper and lower frequency limit respectively. Such a filter would be e.g. 4±1, 8±2. 16±4. The
frequency representing the energy in a particular filter band, is the geometric center frequency calculated
as f c =
f u f l . Since the user probably wants to calculate coda Q at the given frequency, the normal
option (new in SEISAN7.2) is that fu and fl are calculated such that the given bandwidth (e.g. 4 Hz) is
used, but the actual fu and fl will give the specified central frequency. It is still possible to calculate as
before, where fu and fl will be exactly as specified (but the geometrical center frequency will not
correspond to specified center frequency) by giving the bandwidth as a negative number.
Default stations: The stations that will be used if not specified in the codaq.inp file. THE LINE MUST
CONTAIN AT LEAST SOME BLANK CHARACTERS, if not, stations will not be read from codaq.inp
file and the program will crash. Note also that the program assumes that you use SP Z channels if
there is no components given in the line following. After reading the parameter file, the program will
by default use the codaq.inp file to get the event station information. However, any other name can be
used if specified interactively, see below. The station codes can have up to 5 characters.
The codaq.inp file will consist of a series of lines each giving an event identifier (an INDEX file). An
easy way to generate the file is using the SELECT program. The file can also be generated with EEV
using the (C)opy option making a file called indexeev.out. An example is shown below:
1
3
7
/top/seismo/seismo/REA/BER__/1992/06/16-0343-38L.S199206
/top/seismo/seismo/REA/BER__/1992/06/16-1311-58L.S199206
/top/seismo/seismo/REA/BER__/1992/06/30-1504-30L.S199206
The above example only uses the default stations given in codaq.par. Below is an example where
particular stations and components have been selected with particular events, for this to work the
station line in codaq.par MUST be blank.
HYA
1 /top/seismo/seismo/REA/BER__/1992/06/16-0343-38L.S199206
KMY BER ASK TRO
167
S
Z S E B E S Z S Z
3 /top/seismo/seismo/REA/BER__/1992/06/16-1311-58L.S199206
HYA
HYA
S
7 /top/seismo/seismo/REA/BER__/1992/06/30-1504-30L.S199206
EGD
E S
Z
Note that the numbers to the left originate from the index file and do not have any importance. The
long name with the directory structure, is the name of the pick file (S-file) in the database, if the S-file
is in the local directory, it can have just the event id, in this example starting with 30-....The waveform
file name is in the S-file. Following the S-file name is, (like in the parameter file), first a line with station
codes followed by a line of component codes. Like in the parameter file, if a component is not given, it
will be assumed that the component is S Z. THE COMPONENT LINE MUST BE THERE, EVEN IF
BLANK.
Below is an example of a codaq.inp file, where it is assumed that the S-files are the current directory.
This file can also be generated with DIRF.
HYA
16-0343-38L.S199206
KMY BER ASK TRO
16-1311-58L.S199206
HYA
S E
30-1504-30L.S199206
HYA EGD
S N S E
Program operation:
The program first reads the parameter file, default ‘codaq.par’ which must be in your current directory.
It then reads the ‘codaq.inp’ file with the events to analyze (also in current directory). The S-file
names given here can, as shown in the examples above, be in the database or elsewhere, e.g. in
your local directory. In the S-file, the name of the waveform file is given. If more than one waveform
file is given, all files will be searched for the specified station and component. The program will first
look in the current directory, and then in WAV and thereafter in the WAV database and other
directories as given in the SEISAN.DEF file in DAT. The program can therefore work without moving
the data from the database, however you can also move both the S- files and waveform files to your
local directory. Remember that the S-files must be updated in order to have origin time, since the
program uses the origin time and P arrival times from the S-files.
Running the program:
Type codaq, the program asks about output:
0: Only q is calculated
1: Q is calculated and a plot on the tek-screen is shown
2: Q ---------------------------------------------------,
and at the same time hard copy plots are made.
3: Q is calculated and hard copy plots are made, but
no screen plot.
Parameter file, name codaq.par is default (return)
Just hit return if default file, otherwise give name.
File with event-stations, codaq.inp is default (return)
Just hit return if default file, otherwise give name.
The program will now start to run. If no plot is chosen, one line will appear on the screen for each
station used and one for each frequency. The program will start a new page for each new event. If
you are plotting on the screen, you will therefore have to hit return to get the next plot. The screen
might not have been filled out if there are few data.
All questions will appear in the text window. At the end, a summary is given, which is the same as
logged in the output file codaq.out.
168
The abbreviations are:
H:
M:
TP:
TC:
F:
Q:
S/N:
AV Q:
SD:
NT:
N:
q:
1/q:
f:1/q:
cq0:
v:
corr:
Focal depth
Magnitude
P travel time
Start time of coda window relative to origin time
Frequency
Corresponding coda q, if 0 value is > 10000 or negative
Signal to noise ratio
Average q
Standard deviation for average
Total number of q values at all frequencies
Number of q values at given frequency
Average of q values
q is calculated as 1/q averages, probably the best to use
Q values calculated using the relation derived from the 1/q averages
q = q0*f**v obtained with the average 1/q-values
Constant q0 obtained using the fixed user selected v
Constant v determined
Correlation coefficient in determining q vs f
If a station is not present or no P is read, a message will be given. The program will search for the
first P arrival time in the S-file. If several are present for the same station, it will use the first.
Output:
A file called codaq.out is generated. It contains a copy of the parameter file, one line for each event
station combination accepted by the program (correlation and s/n ratio) and the average q values.
The q values are averaged directly (indicated by q) and 1/q are averaged (indicated by 1/q). At the
end are the fits to the q = q0*f**v relation.
Output file codaq1.out contains the same output as codaq.out except there is no print out for each
event.
Example of codaq.out:
start in s-times
absolute start time (sec)
window length (sec)
spreading parameter
constant v in q = q0*f**v
minimum signal to noise ratio
noise window in front of signal and len
minimum correlation coefficient
maximum counts to use
ASK SUE KMY EGD HYA
S Z S Z S Z S Z S Z
2.00
0.00
15.00
1.00
1.00
5.00
15.00
0.50
500000
\SEISMO\WAV\1996-06-07-1324-51S.TEST__009
\SEISMO\WAV\1996-06-07-1324-51S.TEST__009
\SEISMO\WAV\1996-06-07-1324-51S.TEST__009
1996 6 7132458 KMY
tc 41.3 f 16.0 s/n 37.4
\SEISMO\WAV\1996-06-07-1324-51S.TEST__009
\SEISMO\WAV\1996-06-07-1324-51S.TEST__009
\SEISMO\WAV\1996-06-25-0336-34S.TEST__032
1996 625 33715 ASK
tc 87.2 f 4.0 s/n121.9
1996 625 33715 ASK
tc 87.2 f 8.0 s/n 72.5
\SEISMO\WAV\1996-06-25-0336-34S.TEST__032
1996 625 337 5 SUE
tc 57.3 f 4.0 s/n 96.9
\SEISMO\WAV\1996-06-25-0336-34S.TEST__032
1996 625 33730 KMY
tc 143.5 f 8.0 s/n 13.6
\SEISMO\WAV\1996-06-25-0336-34S.TEST__032
\SEISMO\WAV\1996-06-25-0336-34S.TEST__032
5.00
Q 1077
corr -0.55
rms 0.30
Q
Q
340
551
corr -0.60
corr -0.56
rms 0.19
rms 0.28
Q
193
corr -0.61
rms 0.34
Q
506
corr -0.61
rms 0.27
169
1996 625 33653 HYA
1996 625 33653 HYA
1996 625 33653 HYA
Freq
Band
2.00
1.00
AV Q
NT= 8
q
1/q
f:1/q
q
1/q
cq0= 84
cq0= 82
84.4
84.4
84.4
4.00
2.00
SD AV Q
N= 1
288
0
288
0
206 91
tc
tc
tc
8.00
4.00
SD AV Q
N= 3
320 118
287 123
333 147
sd= 37
sd= 38
f
f
f
2.0
4.0
8.0
s/n 51.5
s/n239.5
s/n112.1
288
427
504
corr -0.54
corr -0.56
corr -0.61
rms 0.12
rms 0.17
rms 0.27
16.00
8.00
SD AV Q
SD
N= 3
N= 1
520 27 1077
0
519 26 1077
0
537 237 867 382
q0=143
q0=128
Q
Q
Q
sd= 49
sd= 57
N=
v= 0.65
v= 0.69
sd= 0.16
sd= 0.20
cor= 0.94
cor= 0.93
Corr: 0.540.00 0.590.03 0.590.03 0.550.00
Average lapse time with sd
83.704498
29.501974
Above, the one line per q calculations is showing results from different stations. Only the traces
selected (fulfilling selection criteria) are shown. The time indicated, is the start time in the waveform
file for that particular station. In the SEISAN format that does not have to be the same for each station
as shown above. If some data is missing, it is also show in the codaq.out file. Corr is the average
correlation coefficient (with standard deviation) for the data selected for that frequency. The average
lapse time is the average of the tc - values.
In the DAT directory, there is an example codaq.par and codaq.inp set up to run on PC assuming that
SEISAN has been installed under \seismo. If installed differently, edit the codaq.inp file to reflect the
installation. For Unix testing, the codaq.inp MUST be edited to reflect the installation path or the file is
regenerated using EEV as described above.
General recommendations: Coda window should be 15-25 seconds, minimum correlation coefficient
larger than 0.5. For comparing coda values in different regions, ALL processing parameters must be
identical and average lapse times should be very similar.
Figure 32 gives an example of a codaq plot. There are no options for the codaq plots and the length
of the window is always the first 200 secs from the original trace. If origin time or coda window is
outside this 200-sec window and data is available, the program continues, but the coda window is not
plotted on the figure.
170
Figure 32. An example of a coda Q plot. On top is shown the original trace and below the
filtered coda windows. Note that 15 secs of noise are shown in front of the selected filtered
coda window. The first 5 secs of the noise shown is used for calculating the S/N ratio. On each
filtered plot is given F: Center frequency, Q: Q-value, zero means no Q-value could be
calculated, S/N: Signal to noise ratio.
171
6.18 Merge events near in time ASSOCI
The program will check if two events are close together in time and merge the events if requested.
This is partly an alternative to use append in EEV. The program asks for maximum time difference
between events to associate. The user will then be asked if events should be physically associated or
not. The program is useful when merging a large number of events. The program has two alternatives
for merging:
(1) : Merge events in same data base: One event is compared to the next event in the same data
base. If they are close enough in time, the two events are merged and the program moves on to the
next event. If 3 events are close in time, only the 2 first are merged. In order to also merge the third,
the program has to be run again.
(2): Merge events from a file into the data base: This option makes it possible to merge from another
data base (use SELECT or COLLECT to create a file) without first completely mixing the two. The
event from the file will be merged with as many files from the data base as fit the time difference
criteria. So e.g. 2 events from the data base can both get the same event from the file included. At the
end of the run, two files are output (associ_rest.out associ_merg.out) with events which were not
merged and merged respectively. These can then be put into another data base with split, if desired.
This function can also be used to separate the input file in two files.
6.19 Making synthetic seismograms
BOUCH and BOUSEI, HERRMANN and HERRSEI and WKBJ are all programs which is used for
generating synthetic seismograms.
The full wave modeling programs are written by Bouchon and Herrmann, and for WKBJ, Chapman
and Valerie Maupin. Valerie Maupin has integrated WKBJ for SEISAN and written the routines that
makes it possible to use specific phases. She has also made many improvements in the original
installation of BOUCH and HERRMANN and written a large part of this chapter.
Bouchon:
The Bouchon program is somewhat modified for SEISAN. The theory, which is quite straight forward,
is given in a series of papers (e.g. Bouchon, 1981). It is based on a discrete wave number
representation of the wave fields. Basically, the source is repeated periodically in space, so that
integration over the k-domain is replaced by a series. This implies that the periodicity of the source, L
(in km), should be large enough so that the information from fictitious sources does not arrive during
the time interval of interest. Roughly r < L/2, sqrt((L-r)**2+Z**2) > Vp*t where r is the epicentral
distance and Vp is the highest P-wave velocity of the model, t is the travel time and Z the
hypocentral depth. Only layered (horizontal, parallel) earth model is used. The earthquake source
cannot be in the bottom layer or at the surface.
There are 2 programs, BOUCH and BOUSEI. BOUCH computes the frequency response given the
model, the source depth, the focal mechanism, the receiver locations and the orientations of the two
horizontal components. BOUSEI takes the output file from BOUCH, multiplies it by the source
spectrum and uses an FFT to get the synthetic ground motion (displacement, velocity or
acceleration). The user must provide the source function (see below) and the original waveform files
must be available in WAV or working directory if a file containing both real and synthetic signal is to
be generated. Otherwise, only synthetic data will be seen in the output file.
Herrmann:
The Herrmann programs HERRMANN and HERSEI work the same way as BOUCHON and BOUSEI
respectively. The major difference is that once HERRMANN has been executed, HERSEI can be
executed with different fault plane solutions to obtain the time series, while for the Bouchon programs,
both programs must be run again. The Herrmann programs are thus faster for testing many different
fault plane solutions.
HERRMAN’S PROGRAMS ARE NOT RUNNING ON PC DUE TO COMPILER ROUNDOFF
172
PROBLEMS.
THE PROGRAMS ONLY USE SEISAN FORMAT
The description in the following is for the Bouchon programs, but the steps are the same for
HERRMANN.
WKBJ:
As opposed to the seismograms calculated with the Bouchon and Herrmann programs, the WKBJ
synthetic seismograms contain only the number of phases selected by the user. The execution time
for one run of the program is very short. In addition to making the synthetic seismograms, the
program calculates the arrival times of these phases, and write them both on the screen and in the
iasp.out file for later plotting (see MULPLT). This is intended to be a tool to help identify phases on
the data or on the Herrmann or Bouchon synthetic seismograms: it can by no means replace these
two programs, which are much better than WKBJ to model the frequency-dependent character of
crustal phases at regional distance.
WKBJ seismograms have been introduced in seismology by Chapman (1978). More details on the
method can be found in Dey-Sarkar and Chapman (1978) and in Chapman and Orcutt (1985). The
core of the present program is a code written by Chapman et al. (1988) and is part of the
seismological software distributed freely by IASPEI.
The synthetic seismograms are given in displacement. Although their spectra contain low
frequencies, one should bear in mind that they represent a high-frequency approximation of the wave
field. They include a number of non-physical phases due to truncation of the integrals in slowness p.
For the most interesting crustal phases, the epicentral distance is usually much larger than the source
depth, and these phases interfere with the physical phases and modify their amplitudes.
The head waves on an interface appear automatically as a by-product of the reflected phases, as
soon as the epicentral distance is larger than critical. That means for example that the Pn phase
appears automatically on the synthetic seismogram as a by-product of the PmP phase. In order to
synthesize or calculate the arrival time of a Pn or Sn phase, you must then specify 'PmP' or 'SmS'
(see below).
For a receiver at the free surface, the synthetic seismograms must include the free surface reflection
coefficient to yield correct amplitude and waveform for the different phases. For S phases, at
epicentral distances larger than critical, this includes automatically the SP phase (a P phase which
propagates horizontally along the free surface, and which originates from the critical conversion of S
to P at the free surface). The critical distance is of the order of the source depth for the Sg phase, and
its SP phase usually appears as a large arrival between the P and S wave. The SP phases are
physical, but the amplitude of their high frequency part is overestimated with WKBJ. If one wishes to
suppress them from the synthetic seismograms, one may optionally do so. With this option, the
surface reflection coefficient is omitted and the synthetic seismograms contain only the upgoing
wavefield, that is the wavefield one would get in a borehole, after filtering out the downgoing
wavefield. Let us note that this option may strongly modify the amplitudes and waveforms of the
different phases compared with those at the free surface.
In addition to the synthetic seismograms, the program calculates the arrival times of the phases you
have specified, and write them in the iasp.out file. These times are calculated by interpolation in
epicentral distance of the values tabulated in wkbj.tab. For sources close to an interface (in practice
for Pg and Sg phases and the source under an interface), there is a limited epicentral distance range
in which an arrival time can be calculated. For example, the maximum epicentral distance for Pg is
about 250km for a source 0.1 km under Moho in the default SEISAN model. In order to increase the
maximal epicentral distance, you may move the source away from the interface, or you may increase
the number of ray parameters used in program wkbj_or.for (parameter 'nnpp') called from wkbj.for.
All three programs are hardwired to use triangular sources.
Running the programs
173
The programs require input about distances, azimuths, depth, crustal model, fault plane solution, time
window, number of points and some modeling parameters. Almost all of these parameters are
available within SEISAN. The programs have therefore been modified to use an S-file (Nordic format)
as input file with additional information about time window, number of points to model and crustal
model. A special format has been used to keep the modeling information separate from other
information in the file (see below for an example). The steps to model a particular event are as
follows:
Problem Bouchon: Use fewer layers, ideally just a halfspace under the deepest ray. The programs
seems to become unstable if too many layer are used there.
Step 1
Edit the event in EEV and mark the stations wanted for modeling with a minuscule s in column 1,
ONLY mark the station once. Exit from editor and, within EEV, give the command "synt". This will
generate all the necessary default input parameters for modeling, which are stored as comment-lines
starting with SYNT in the S-file (see below). At the same time, the s's used as markers are removed.
Any old modeling information present will remain and override the defaults. However, in case
the F-flag is set for the DEPTH parameter, distances and azimuths will be reset according to the
current location.
Step 2
Edit event again and check if default parameters are ok (see explanation below).
Step 3
Run one of the programs BOUCH, HERRMANN or WKBJ. These are known commands in EEV.
BOUCH: The program will now run for a certain amount of time depending on number of points
required. At the standard output, the input parameters used will be printed out and for each
frequency, the number of terms in wave number integration is printed out. If the limit of the number of
terms is reached, something is wrong, try other parameters. The speed of this output (NPOINT/2+1
lines) gives a good indication of how long time it will take.
HERRMANN: Takes longer than BOUCH.
WKBJ: Very fast.
Step 4
Generate the seismograms.
BOUCH: Use program BOUSEI. The program is interactively asking the seismogram type
(displacement, velocity or acceleration). BOUSEI will generate a file bousei.out in SEISAN format
containing both original and synthetic traces. The number of traces is determined by the
specifications for each station, see below. Output file is bousei.out.
HERRMANN: Use program HERSEI, similar to BOUSEI. Output file is hersei.out.
WKBJ: The first command WKBJ also makes the seismograms. Output file is wkbjsei.out.
In all cases it is possible to shift the original trace relative to the synthetic trace and the program will
ask, for each channel, how much it should be shifted. A positive value shifts the real trace up in time
(to the left). The default is to shift the trace the amount of the P-travel time residual of the first P found
in the S-file for that station in order to line up the P - phases. NOTE: These phases MUST be the
same phase types in order to be lined up. If the first modeled phase is Pn and the first observed
phase given in the S-file is Pg, there will be a no alignment. The amplitudes for Bouchon are in nm,
nm/sec or nm/sec*sec (hopefully !!) assuming a seismic moment of 10 **22 dyne-cm. The output file
will normally contain both the original and synthetic traces. However, if no waveform file is available
(in local or WAV directory), the output file will contain an empty channel where the original data
should have been. The specifications in the hyp.out file determine which traces from the modeled
stations are included in the output file. If the specification after STATION is only component (e.g. S),
then all 3 channels are shown. If a particular channel is given (e.g. S N), then only that channel is
shown. Only one or 3 channels can be displayed.
All output traces are given in Z, N and E or Z, R and T depending on the parameter file (see below).
The channel names are SH, SB and SW for Herrmann, Bouchon and WKBJ respectively.
Step 5
174
Plot the traces with mulplt. This can be done within EEV using the command pw, ph or pb for WKBJ,
Herrmann or Bouchon respectively. Since there is no instrument correction, it is a good idea to plot
both the modeled and observed signals narrow band pass filtered. E.g. for regional events 0.1-1 Hz
and for small local events 2- 5 Hz (depending on sample rate).
shows an example of the modeling.
Note: The whole modeling process can be done entirely within EEV and it is intended to be done so.
Since the modeling requires updated distances, depths etc when changing model etc, it cannot take
its input from the location in the S-file, which only changes when doing an update (see UPDATE
program). So when running from within EEV, a location will always be done first to get an updated Sfile (in this case the hyp.out file) and this is the reason that the modelling programs use the hyp.out
file instead of the S-file for input. This also means that the modeling program can be run separately
from any hyp.out file, however it is then up to the user to keep it updated.
The modeling parameters
Below is shown an example of part of an S-file prepared for modeling. The file is one of the events in
the test data set and by using EEV to find the event, modeling can start immediately. All parameters
have been set automatically.
1996
1996
6 7 1325 29.1 L* 59.841
5.127 12.0F BER 12 1.1 2.2CBER 1.9LBER 2.0LNAO1
6 7 1325 29.1 L*
BER
2.0WBER
1
8.3
41.0
74.7
1
F
1996-06-07-1324-51S.TEST__009
6
535 SOUTHERN NORWAY
3
depth fixed to 12 km, rms lowest with near station (less than 110 km) location3
fault plane solution ok within 10 deg
3
SYNT: MODEL--:
THICK
VP
VS
DENS
QP
QS
3
SYNT: MODEL--:
12.000
6.200
3.563
2.600
0.000
0.000
3
SYNT: MODEL--:
11.000
6.600
3.793
2.800
0.000
0.000
3
SYNT: MODEL--:
8.000
7.100
4.080
3.000
0.000
0.000 B 3
SYNT: MODEL--:
19.000
8.050
4.626
3.200
0.000
0.000 N 3
SYNT: MODEL--:
30.000
8.250
4.741
3.400
0.000
0.000
3
SYNT: MODEL--:
50.000
8.500
4.885
3.600
0.000
0.000
3
SYNT: ST-D-RK:
8.3
41.0
74.7
3
SYNT: DEPTH--:
12.0
3
SYNT: NPOINTS:
256
3
SYNT: TIMES--:
TOTAL
60.000
INITIAL
0.000 SY-TRACE
60.000
3
SYNT: BOUPAR-:
300.0
600
0.001
3
SYNT: PHASES-:
Pg
Sg
PmP
SmS
SmP
3
SYNT: DT-Tsou:
0.050
.100
3
SYNT: REDVELO:
8.0000
3
SYNT: COMPON-:
RADIAL
3
SYNT: STAT-AT: no
3
SYNT: NSTAT--:
3------------------------------------------------------3
SYNT: NEW STAT:---------------------------------------------------------------3
SYNT: STATION: EGD S Z DISTANC:
49.0
3
SYNT: STATION: EGD
AZIMUTH:
9.0 BAZIMUTH:
189.2
3
SYNT: NEW STAT:---------------------------------------------------------------3
SYNT: STATION: KMY S Z DISTANC:
71.0
3
SYNT: STATION: KMY
AZIMUTH:
172.0 BAZIMUTH:
352.7
3
SYNT: NEW STAT:---------------------------------------------------------------3
SYNT: STATION: ASK S Z DISTANC:
72.0
3
SYNT: STATION: ASK
AZIMUTH:
5.0 BAZIMUTH:
185.1
3
STAT SP IPHASW D HRMM SECON CODA AMPLIT PERI AZIMU VELO SNR AR TRES W DIS CAZ7
EGD SZ IP
C 1325 35.95
93
-1.110
48
6
EGD SZ ES
1325 42.03
-1.010
48
6
BER SZ IP
C 1325 38.12
55
-1.010
62 11
BER SZ ES
1325 45.44
-1.110
62 11
BER SZ E
1325 46.71
31.7 0.2
62 11
ASK SZ EP
D 1325 39.59
68
-0.910
70
3
ASK SZ ES
1325 48.07
-0.810
70
3
ASK SZ E
1325 50.90
111.0 0.3
70
3
KMY SZ IP
C 1325 40.26
62
-0.410
71 175
KMY
KMY
SZ ES
SZ E
1325 48.74
1325 48.92
-0.410
83.6
0.2
71 175
71 175
MODEL: The model to be used. THICK is layer thickness, VP is Vp velocity, VS is Vs velocity, DENS
is density and QP and QS, are P and S q-values respectively. The model, velocities and Q-values are
taken from the STATION0.HYP file with first choice from current directory and second choice from
DAT directory (like the HYP program). The S-velocities are calculated using the Vp/Vs ratio given
there. Moho is indicated with N at the end of the line with the first mantle layer. A Q of zero means
infinite Q. The densities are approximate values and should be modified. See below for maximum
175
number of layers.
ST-D-RK: Strike dip and rake is taken from an existing fault plane solution for the given event (F-line)
if it exists, otherwise arbitrary values are supplied. The convention is Aki and Richards.
DEPTH: Focal depth is taken from the current solution. The second field can optionally have the letter
F (right justified). If this flag is set, the user can give the synt command to update all distances and
azimuths used for modeling which will correspond to the latest location determined as e.g. a result of
a changed fixed depth or a changed model. The intention with this flag is that the user should be able
to set a fixed depth in the S-file header line, give the synt command to update the parameters for
modeling corresponding to this depth and then model.
NPOINTS: Number of points to model, 256 is set as default, must be 2**N. Used by BOUCH and
HERRMAN only.
TIMES--: Three different times:
TOTAL: The total time window for generating data and synthetic seismograms for all
channels, see also REDVELO.
INITIAL: The initial time of the earliest trace in the output file, with reference to the source origin
time. The synthetics at the station with smallest epicentral distance automatically start also at this
initial time.
SY-TRACE: The duration of the synthetic seismogram for each channel, might have different start
times, see REDVELO.
DT-Tsou: Sampling interval (used for WKBJ seismograms only), and half-duration of the source used
for
all three programs. In all programs, the source is triangular, however BOUCH can optionally use
several sources, see below.
REDVELO: Reduction velocity to calculate the initial times at subsequent distances (put 0. for no
reduction velocity).
PHASES-: The names in format A4 (right justified) of the phases to be
synthesized with WKBJ. The phases may be given in any order, with a maximum of
6 phases per line, and there may be several "SYNT: PHASES-" lines.
Possible phases:
Pg (direct P from source to receiver)
Sg (direct S)
PmP (includes automatically Pn at distances larger than critical)
pPmP (includes automatically pPn at distances larger than critical)
sPmP (includes automatically sPn at distances larger than critical)
SmS, pSmS, sSmS (includes automatically Sn, pSn, sSn at distances larger than
critical)
SmP, PmS
P1P, P2P, S1S, etc: the same as PmP, SmS etc, but on interface number 1, 2, etc.
(The free surface gets interface number 0 in the convention taken here). Their
associated head waves are labeled Pn1, Pn2, Sn1, etc.
COMPON-: RADIAL for radial-transverse components, NORTH for North-South, East-West
components.
STAT-AT: Is "not free" or "NOT FREE" anywhere within column 16 to 25: Optional line. If this option is
chosen, the WKBJ synthetic seismograms are calculated omitting the reflection coefficient at the free
surface, at the receiver location.
BOUPAR: Modeling parameters L, Nt and e. L is length of periodicity (should be a few times the
hypocentral distance), Nt is maximum number of terms in wave number summation and e is the value
used in truncating the summation. Increasing e and decreasing Nt will speed up convergence, but the
results might be unreliable.
NEW STAT: Comment line
STATION: Station to be modeled with component(s) to be displayed. The S means that short period
instruments are used. If no component is given, all 3 components are assumed. The other option is to
176
indicate a component (e.g. Z) and only that component will be displayed (see also description of
BOUSEI). DISTANC is epicentral distance used, this distance is taken from the current location,
AZIMUTH is azimuth from the source to the station taken from current location, BAZIMUTH is the
back azimuth at the station, calculated by EEV, used to rotate if so specified. Each new station is
represented by the above 3 lines.
The source time function
The time duration of the triangular source time function for Bouchon is given as Tsou above, and is
also used in WKBJ and Herrmann.
Hints on modeling
Event 199606071325 in the test data set is set up with modeling parameters and can be tested
immediately.
The model
The standard model given in STATION0.HYP might be too detailed for most cases and should be
simplified to include 3-4 layers by just editing the S-file, this also speeds up modeling. However, if you
located the event with one model and model with another, the distances and residuals might not fit. A
solution could be to have a STATION0.HYP in the local directory with the simplified model.
Alignment of P and S
If the distance calculated by HYP is not correct as indicated by P and S residuals, the synthetic and
observed signals will not be aligned. The distance for that station can then be changed manually in
the S-file under DIST and/or delays can be applied when generating the seismograms.
Testing different parameters
There is no need to go back to EEV to test for the parameters that do not change the location. Thus
to test for different fault plane solutions, time windows, number of points, edit the hyp.out directly and
rerun. However, if depth or model is changed, relocation must be made. To test for different depths,
locate with fixed depths, see HYP.
NOTE: THE SOURCE AND RECEIVER CANNOT BE AT THE SAME DEPTH (BOUCH AND
HERRMANN) AND IN NO CASES CAN THE SOURCE BE AT DEPTH ZERO.
Running time
This depends mostly on the number of points and to some degree on number of layers. The number
of stations has an insignificant effect on running time. A 4-layer model and 256 points will take under
a minute with Bouchon.
Program limitations: HERRMANN and WKBJ is set up with max 20 layers and Bouchon with 10
layers. Maximum of 32 stations Change programs and recompile if more layers are needed.
Bouchon is compiled for 512 frequencies (1024) points.
Computer notes:
The original Bouchon program BOUCH is almost unchanged. The only modification is that it uses a
subroutine to generate its original input file bouch.inp from the hyp.out file. This file still remains after
running BOUCH for debugging purposes. The output from BOUCH is bouch.out, which in turn is input
to BOUSEI.
Herrmann:
The Herrmann waveform modeling is based on a concept where the synthetic seismograms are
177
computed through a sequence of four distinct processes (programs).
1. The program "hspec8" will calculate the medium response for 10 basic Green’s functions, where
the response is given in frequency - wavenumber domain F(f,k).
2. The program "rhwvinta" will integrate and take the medium response from F(f,k) --> F(f,r)
3. The program "rhfoc10" will convolve the response function with a source time function and with
inverse Fourier transform take F(f,r) --> F(t,r)
4. The program "mech" will construct a 3 component synthetic seismogram given a focal mechanism.
Herrmann's programs originally had several optional source time functions, however, a triangular
source has been hardwired (for all 3 programs) so it is easier to compare the results. The original
options can be reactivated by editing the program.
The programs HERRMANN and HERSEI run these 4 programs in an automated sequence.
All References, a detailed manual, source code and parameters as well as other related programs:
"Computer Programs in Seismology", Volumes I - VIII. By Robert B. Herrmann, Saint Louis University,
Saint Louis, Missouri.
WKBJ:
Input file hyp.out, read by "WKBJ.
Output file iasp.out, written by "WKBJ". Contains the arrival times of the different phases at the
stations, in SEISAN format.
Output file wkbjsei.out, written by WKBJ in the SYNTSEL.FOR subroutines. A waveform file (SEISAN
type) containing the data and the synthetics, which can be plotted using "mulplt". Note that there is a
code for each synthetic seismogram giving the modeling method (SH: Herrmann, SB: Bouchon, SW:
WKBJ), and the component (Z, R, T, N or E).
INTERMEDIATE FILES
wkbj.inp, created by WKBJ for input to WKBJ_OR. The same information as in hyp.out, in a
WKBJ_OR format.
wkbj.tab, output from WKBJ_OR, reprocessed by WKBJ. Contains tables as a function of ray
parameter.
wkbj.out, output from WKBJ_OR, reprocessed by WKBJ. Contains the Green functions.
178
Figure 33. An example of synthetic seismograms using Bouchon(2), Herrmann (3) and WKBJ
(4). The original seismogram is shown in channel 1. All synthetics are displacement. Also
shown are the theoretical travel times calculated by WKBJ.
179
6.20 Calculation of travel times
In SEISAN, travel times are generated from a flat crustal model or using the IASP91 global travel time
model. It can often be useful to generate travel times for given distances and two programs are supplied
to do these calculations. TTIM will calculate travel times for global phases at one given distance and depth
and TTLAYER, will calculate a travel time table (layered flat model) for a given depth and a distance
range. A special version of TTIM called IASP is used in connection with EEV and MULPLT.
6.20.1 IASPEI travel time software, program TTIM
This program can be used for calculating global travel times, see below for details on phases calculated.
The program assumes that you have the travel time tables in the working directory or in DAT, see
computer notes below on how to generate these file if not already there. The same files are also used by
HYPOCENTER.
After starting the program, the first two questions 'do you want xxxxx’ relate to range summaries, etc., that
are normally not required and can be answered with n(no) followed by ENTER. The program then asks
'Enter phases, one per line...' You can then enter a specific phase, or a keyword defined as follows:
All
P
P+
gives all phases
gives P-up, P, Pdiff, PKP, and Pkikp
gives P-up, P, Pdiff, PKP, Pkikp, PcP, Pp, Ppdiff, PPKP,
PPKIKP, Sp, Spdiff, SPKP, and SPKIKP
S+
gives S-up, S, Sdiff, SKS, Ss, Ssdiff, SSKS, Ps, Psdiff,
and PSKS
basic
gives P+ and S+ as well as ScP, SKP, PKKP, SKKP, PP, and
P'P'
Writing all individual phases, separate by ENTER, terminating the list with an additional ENTER. The
program will then enter a loop where phase times are calculated for new distances entered on request.
The program is terminated for a particular distance by entering -1, and a new depth can be used, or the
program can be terminated by entering -1 again.
A special version of this program used in connection with MULPLT is IASP.
In order to generate the earth model files IASP91.HED and IASP91.TBL, first run program REMODL, then
program SETBRN. The program REMODL has the earth model hardwired. Note: These binary files
CANNOT be moved between platforms. They are included with SEISAN for each respective distribution. If
lost, they must be regenerated on the same platform.
For more information about IASP91 programs, see HYPOCENTER manual by B. Lienert.
6.20.2 Calculation of travel times for layer and gradient model, TTLAYER
The TTLAYER program is written by Barry Lienert to calculate travel times for both layer and gradient
model. In this version the program only works for zero depth, and therefore might not be very useful.
The program reads a set of velocities and depths from an input file in ‘STATION0.HYP’ format and
calculates travel times for P and S velocities for a set of uniform-velocity layers, using the
HYPOCENTER dtdx2 routine and also for a set of uniform gradient layers, using dtdxg, a new routine
written to have the same input arguments as dtdx2.
The routine to calculate travel times for a gradient model uses an adapted version of Fred Klein’s
TTCAL routine, which he uses in his program TTGEN to generate a table of values from which to
interpolate travel times and their derivatives in HYPOINVERSE.
The program is easy to run and the output can be plotted with some standard xy plotting tool.
6.20.3 IASP, travel times for MULPLT
This program is a special version of IASP91 to be used in connection with EEV and MULPLT. Giving
180
command iasp from the EEV prompt (or from within MULPLT), the program will read the current
active S-file, and for each station, calculate possible IASP91 phases and arrival times relative to the
hypocenter and origin time given in S-file. The origin information can be obtained from two places in
the S-file: (1) The header lines are searched for hypocenter lines and the first found after the main
header will be used, (2) If no secondary header lines, the main header line is used. The intention of
this order is that it is possible to put in a PDE solution in a secondary header line (option INPUTONE
in EEV) so that theoretical travel times are calculated relative to a fixed solution and not the temporary
solution made by the local agency.
The IASP91 tables can be found in the local directory or DAT and have the same names as used in
HYP and TTIM. The program generates an output file iasp.out in Nordic format. This file is read by
MULPLT and the theoretical phases displayed on the screen. The number of phases calculated can
be very large making it hard to see which phase is which. IASP therefore has a definition file,
IASP.DEF, where phases to be written out are given. The file can be in the working directory or in
DAT. If no definition file is available, all phases will be written to the iasp.out file. Below is an example
of a IASP.DEF file.
This file contain the definitions of phases to be used when calculating
synthetic phases to be plotted with mulplt. There is one phase pr line
and each phase is preceded with the keyword IASP-PHASE. Only lines with this
keyword will be read. The defined phase then follows in column 13 to 20.
If no phases are defined, all ISPEI91 phases will be used.
Phase ID
Phase---------------------IASP-PHASE P
IASP-PHASE PP
IASP-PHASE PPP
IASP-PHASE PKP
IASP-PHASE pP
IASP-PHASE sP
IASP-PHASE PcP
IASP-PHASE S
IASP-PHASE SS
IASP-PHASE SSS
IASP-PHASE SKS
IASP-PHASE ScS
IASP-PHASE PS
IASP-PHASE SP
IASP-PHASE ScP
6.21 Inversion for QLg, QLG
The QLG program can be used to determine an average QLg or to perform a tomographic inversion.
The method is described in Ottemöller et al. (2002). The program can also produce the input for
distance trace plots. Note that using the program is no trivial task. The data set needs to be carefully
selected and the instrument calibration has to be known. The input to the program is a Nordic file,
which includes several events. The parameter file needs to be carefully set up.
The program can be used in the following way:
1) Determine average QLg
2) Perform checker-board test to chose damping parameters
3) Tomographic inversion
Note: In this version of SEISAN, a first version of the program is included. It is expected that the user
interface will be changed. It is suggested that the user contacts the editors in case he/she wants to
use the program. The main purpose of including the program is to give an example source code so
that the user can make use of it when implementing similar programs.
Limitations: The program assumes that the grid is defined over a flat area. The tomographic inversion
using QLG, therefore, should only be done for regions near the equator. The problem otherwise
would be that the grid cells would not have the same size, which may lead to artefacts.
181
Example of the parameter file qlg.par:
KEYWORD............Comments.............Par 1.....Par 2.....
FILTER
DISTANCES
GROUP VEL LG
GROUP VEL P
for distance plot
min and max
lg group vel window
p group vel window
0.01
200.
3.0
5.0
INVERSION TYPE
1.
ORIENTATION
PHASE ONLY
1. for tomography
0. for average
0.=vert, 1.=horiz.
1.=phase pick requ.
in s-file
FREQUENCY
FREQUENCY
FREQUENCY
FREQUENCY
FREQUENCY
FREQUENCY
FREQUENCY
FREQUENCY
frequency
frequency
frequency
frequency
frequency
frequency
frequency
frequency
1.
1.25
1.60
2.
2.5
3.15
4.
5.
STATION MIN
VELOCITY LG
DAMPING ALPHA
DAMPING SIGMA
DAMPING BETA
DAMPING LAMBDA
NSMOOTH
CHECKERBOARD
VERBOSE
min # of stations
#
#
#
X
Y
X
Y
X
Y
and
and
and
and
and
and
and
and
1/q
1/q
1/q
1/q
1/q
1/q
1/q
1/q
15.
3000.
3.7
8.0
0.
1.
damping parameters
---------------------------------------------------smooth spec # times
1. for cb-test
0. for quite mode
4.
3350.
500.
100.
500.
0.001
0.
0.
1.
x
y
x
y
x
y
-92.5
6.50
1.
1.
17.
13.
5.2392E-03
4.5246E-03
4.1239E-03
3.5312E-03
2.9081E-03
2.2568E-03
1.7029E-03
1.1228E-03
Grid
START
START
DELTA
DELTA
NPOINTS
NPOINTS
start of grid
start of grid
delta grid
delta grid
# points
# points
6.22 Wadati
This is a program to make Wadati diagrams and apparent velocity from a Nordic file with one or many
events. The apparent velocity is calculated from the arrival times and the calculated epicentral
distances as given in the S-file. The apparent velocity is thus approximate and affected by the
location.
The purpose of the program is to calculate Vp/Vs values for individual events and calculate the
average for a group of events. In addition, the program can calculate the apparent velocity for each
event based on P or S-times. Wadati diagrams with plot can also be calculated directly from EEV.
The information can be used to obtain a first impression of crustal parameters.
For each calculation, events can be selected based on: Minimum number of stations, maximum rms
of the fit (S-P vs P, or arrival times), and minimum correlation coefficient of the fit. For the apparent
velocity calculation, the data can also be selected in distance and azimuth ranges.
The output gives:
T0 :
N :
VPS :
NP :
NS :
AVSP:
Wadati calculated origin time
Number of stations used for Vp/Vs
Vp/Vs ratio
Number of stations for P- velocity
Number of stations for S-velocity
Average S-P times with sd
182
AVDI:
Average distance with sd
The average Vp/Vs is calculated for the whole data set. Individual Vp/Vs values outside the range
1.53 to 1.93 are excluded. An output file wadati.out is generated.
Example of a run to calculate Vp/Vs
Input file name
collect.out
Wadati (1), apparent velocity (2) or both (3) ?
1
Wadati parameters:
Minimum number of stations
3
Maximum rms
1
Minimum correlation coefficient
0.9
1994 616 1841 28.3 No data for Wadati
1994 10 4 1322 55.8 No data for Wadati
1995 822 0141 5.3 T0: 141
8.1 N:
8 VPS:
1995 1120 0401 58.9 T0: 4 1 59.5 N:
7 VPS:
1996 6 3 1955 35.6 No data for Wadati
1996 6 3 1955 35.6 No data for Wadati
1996 6 3 1955 40.1 No data for Wadati
1996 6 6 0648 29.8 T0: 648
1.1 N:
22 VPS:
1996 6 6 0648 30.6 T0: 648 38.7 N:
3 VPS:
1996 6 6 0648 29.8 No data for Wadati
1996 6 7 1325 28.5 T0: 1325 28.1 N:
9 VPS:
1996 6 7 1325 29.1 T0: 1325 28.1 N:
9 VPS:
1996 610 1 4 47.0 No data for Wadati
1996 610 1 4 47.0 No data for Wadati
1996 623 0117 57.8 No data for Wadati
1996 623 0117 58.1 No data for Wadati
1996 625 0337 31.7 T0: 337 33.1 N:
21 VPS:
1996 7 5 0220 46.5 T0: 220 45.8 N:
6 VPS:
1996 713 0556 46.0 No data for Wadati
1996 718 0946 51.4 No data for Wadati
1996 718 2255 6.0 No data for Wadati
1996 726 0742 12.0 T0: 742 11.8 N:
6 VPS:
1.84 RMS:
1.76 RMS:
1.19 CORR:
0.87 CORR:
0.997
0.995
1.51 RMS:
2.16 RMS:
11.00 CORR:
0.80 CORR:
0.380
1.000
1.72 RMS:
1.72 RMS:
1.06 CORR:
1.06 CORR:
0.973
0.973
1.75 RMS:
1.76 RMS:
1.36 CORR:
0.28 CORR:
0.999
0.999
1.74 RMS:
0.47 CORR:
0.993
Number of events for which vp/vs were calculated
Number of events selected for average
Average VP/VS =
1.75
SD=
0.01 N=
9
3
3
Example of a run to calculate apparent velocity
Input file name
collect.out
Wadati (1), apparent velocity (2)
2
Apparent velocity parameters:
Distance range
50 200
Azimuth range
0 180
Minimum number of stations
2
Maximum rms
1
AVSP: 30.212.7 AVDI: 152.0 30.7
AVDI: 143.3 31.1
AVSP: 23.0 9.5 AVDI: 158.7 26.6
AVDI: 90.5 35.8
AVDI: 89.0 35.6
AVSP: 50.536.0 AVDI: 145.8 40.8
AVDI: 116.3 38.2
AVDI: 118.8 45.6
AVSP: 14.4 7.5 AVDI: 106.0 39.2
AVDI: 111.7 37.7
AVSP: 14.4 7.5 AVDI: 176.7
1.5
AVDI: 178.0
2.9
AVSP: 22.0 5.7 AVDI: 150.0
4.2
Output file is wadati.out
or both (3) ?
NP:
NS:
NP:
NS:
NS:
NP:
NS:
NS:
NP:
NS:
NP:
NS:
NP:
4
3
3
6
6
6
3
4
5
6
3
4
2
VP
VS
VP
VS
VS
VP
VS
VS
VP
VS
VP
VS
VP
: 6.95
: 3.82
: 6.76
: 3.72
: 3.69
: 7.60
: 3.97
: 3.61
: 6.75
: 3.82
: 5.13
: 2.80
: 10.17
RMS:
RMS:
RMS:
RMS:
RMS:
RMS:
RMS:
RMS:
RMS:
RMS:
RMS:
RMS:
RMS:
0.20
0.25
0.00
1.79
2.78
0.42
0.04
0.66
0.94
0.97
0.07
1.06
0.00
CORR:
CORR:
CORR:
CORR:
CORR:
CORR:
CORR:
CORR:
CORR:
CORR:
CORR:
CORR:
CORR:
1.000
1.000
1.000
0.998
0.996
1.000
1.000
1.000
0.999
0.999
1.000
0.859
0.000
183
6.23 Calculating spectra, the SPEC program
The SPEC program is used for making spectra of many seismic signals in a semiautomatic manner. It
can be used for several investigations:
A: Making a large series of signal spectra, which can be corrected for instrument and path.
Average spectra are calculated. There are two options for further processing the calculated spectra:
Option (1): Calculate acceleration density spectra which are plotted compared to the Peterson noise
model.
Option(2): Using the slope of the flat part of the displacement spectra to calculate the near surface
attenuation kappa (see 6.2.11 and application note qspec.pdf in INF)
B: Making relative spectra of seismic events or background noise in order to determine the soil
response. When using relative spectra of horizontal versus vertical components, this is referred to as
the Nakamura method (Nakamura, 1989).
C: Making relative spectra of signals from two stations in order to determine Q.
The program makes output files for generating GMT plots in addition to standard SEISAN plots.
Note: Parameter file has changed between SEISAN 7.2 and 8.0 (number of windows and overlap has
been added).
The program can technically operate in two ways: (1) Making relative spectra of a series of pairs of
stations terminated by the average spectra, (2) Making a series of spectra for a number of stations
and events. The spectra can be corrected for distance, Q, and instrument response. In addition, the
spectral levels can be expressed in moment or moment magnitude calculated in the same way and
with the same units as in MULPLT. All relevant parameters are taken from the CAT files, the CAL files
and the input parameter file for SPEC. Window selection for the spectra can be specified to be related
to the P, S arrival times or the earthquake origin time and it is thus possible to automatically make e.g.
S-wave spectra of a large set of stations and events. Optionally, noise spectra, can be calculated
together with the signal spectra. The noise window is selected at the start of the waveform file.
Before the program is started up, the input files must be prepared. The program need two input files.
The parameter file (default spec.par) gives the parameters to use and the list of stations to process.
The event file (default spec.inp) is a CAT file with events to use or a filenr.lis type file with waveform
file names (can only be used if no readings are needed, like for Nakamura studies). An example of a
spec.par and spec.inp file is found in DAT. These files can be used immediately with the test data set.
The program produces several output files. The main output is in spec.out with the parameters used,
the station event combinations used and error messages. The other files are giving output of most
graphs shown. These ASCII output files can be used in other plotting programs, however they have
been specifically formatted for the SEISAN GMTXY plotting script. The number of files depends on
number of stations used. Examples of files could be
spec_all_ASK__S__Z.out
spec_all_BER__S__Z.out
spec_all_gmt.out
spec_ave_ASK__S__Z.out
spec_ave_BER__S__Z.out
All spectra from ASK, S Z
All spectra from BER, S Z
All spectra from ASK and BER
Average spectrum from ASK, S Z
Average spectrum from BER, Z Z
In order to plot these files with GMTXY (only Unix), give e.g. command
gmtxy spec_all_ASK__S__Z.out
Limitations of amount of data: The program is set up to handle 100 spectra of up to 30000 points
each for one run. The dimensions can be increased in spec.for, however the program must then be
recompiled. The spectral windows are 10% tapered. The analyzed signals will be checked for clipping
and rejected if clipped. A message is then given in spec.out
The spec.par file
The file contains alternate lines of parameter names and parameter values, and must contain the
184
number of lines shown in the example below.
selection criteria 1: P, 2: S, 3: S from P, 4: abs
2
start
1
window length, # of windows,overlap
5.0,1,1
number of times to smooth
5
gain factor of channel 1
1
noise spectrum 0: n 1: y
1
make relative spectras 1: y, 0; n
1
plot pics
1
frequency band to use
1.0 7.0
response removal: 0: none, 1: displ., 2: vel., 3: accel. 4. noise pow. 5. kappa
0
rotate: 0: no, 1: yes
0
q0, qalpha and kappa
-1.0,0,0
distance correction
1
minimum correlation and minimum sn for kappa
0.5 2.0
velocity and density
0.0 0.0
magnitude spectrum
0
stations and components, format a5,1x,a4,1x,a5,1x,a4
FOO
S Z SUE
S Z
The parameters:
Selection criteria: Determines how the start of the time window is selected. 1: Start with the P-arrival
time, 2: Start with the S-arrival time, 3: Start with the S-arrival time calculated from the P-arrival time
assuming a P to S velocity ratio of 1.78, 4: Start with 'start' (see next parameter) seconds after the
origin time as given in the CAT file header. This option can be used if no readings are available in the
CAT file. When using a P or S-time for start of window, the program uses the first P or S phase found
in the CAT file for a given station. Component is of no importance here, so there is only a need for
e.g. one P-time for the station being processed if 3 component data is used. This is also the case
when rotating the signal, see below. However, on the trace plots, only readings on those components
shown will be seen on the plots.
Start: If the selection criterion is 1,2 or 3, this is the number of P or S travel times (from the origin)
used to find start time of window. Use 1.0 if the window shall start exactly at the phase time picked. If
selection criteria is 4, start is the number of seconds after the origin time.
Window length, #of windows, overlap:
- Window length: Window length in secs for both signal and noise (if selected) .
- # of windows: If more than 1, spectra will be made in several windows following the first window
and average spectra will be made. This option can only be used if selection criteria is 4. Used for
noise studies or Nakamura studies.
- Overlap: Windows can overlap (factor < 1.0) exactly follow each other (factor=1.0) or have gaps
(factor >1.0). E.g. 0.9 is equal to 10 % overlap.
Number of times to smooth: Number of times to smooth, 0 means no smoothing.
Gain factor of channel 1: Factor that the spectral level for channel 1 is multiplied with. This can be
used if the response shape is the same for the two channels and only the levels are different. If the
shape is also different, set factor to 1 and use response removal below.
Noise spectrum: If 0, no noise spectrum, if 1, make noise spectrum. The noise window is taken from
the beginning of the trace and the window length is the same as given above.
185
Make relative spectra: If zero, no relative spectra, if 1, make relative spectra. The relative spectra
will appear one on each page, and the average relative spectrum on the last plot (see Figure 34). If
no relative spectra are chosen, only one trace and one spectrum is shown per page and the average
spectrum is shown on the final plot. MUST BE SET to 1 to calculate Q, see below.
Plot pics: If 1, the phase pics in the CAT file spec.inp will be plotted.
Frequency band used: Lower and upper frequency bands for the spectral plots.
Response removal: If 0, no response is removed, else 1: displacement, 2: velocity, 3: acceleration
(units is nm, nm/s and nm/s*s), 4: Power spectral density in dB relative to ((1m/s**2)**2)/Hz. This
option is used for seismic background noise studies., 5: Determine kappa. The flat part of the
spectrum (frequency below corner frequency) is approximated by a straight line and kappa calculated
for each event and the average at the end (in spec.out file and on final plot). The spectrum will
normally be corrected for Q, BUT NOT kappa. For more details, see application note qspec.pdf in
INF. Make sure to set appropriate frequency limits and correct distance corrections. Can be used for
both P and S-spectra.
A cal file for each channel must be available in the CAL directory (see section 4.6). For relative
spectra, the response removal has no importance if the response is the same for the channels
compared. A simple correction can be made with "Gain factor of channel 1" parameter above. NOTE:
If moment or magnitude spectrum is made, response removal MUST be 1.
Rotate components: If 1, the horizontal components are rotated. This means that if the user has
specified N or E, radial or transverse respectively will be used instead. The original data remain
unchanged. If start time of spectra are chosen by using P or S, there must be a reading from those
components if the pics are to be plotted. If the parameter is zero, no rotation is done. See also
MULPLT for more details of rotation.
Q0, qalpha and kappa:
Q-correction:
Parameters in Q-relation Q = Q0**qalpha used for spectral correction (see also section on MULPLT
for standard attenuation relations). Only used if response is removed. If first 2 parameters are 0,0, no
Q-correction. New from SEISAN7.2 is that a kappa correction also can be used (see MULPLT
spectral section).
Calculation of Q:
If Q0 and qalpha is set to –1,0, the relative spectra will be used to calculate q as a function of f (see
standard relations in MULPLT section) and the plots will show q as a function of f. This can be used
for both P-waves and S-waves. The distance correction MUST be set, S-velocity must be given (see
below) and it is recommended to assume body wave spreading (amplitude proportional to 1/distance,
factor is 1.0 below). If the response of the 2 stations is not identical, correction for response must also
be made. There must be an origin time and phase readings must be available in order to calculate Q.
Q is calculated as ln(A2(f)/A1(f)) * f * pi * (t2-t1) + alpha * ln(R2/R1), where A1 and A2 are spectral
levels at frequency f for the two stations, t1 and t2 are travel times, alpha is geometrical spreading
exponent (1.0 is body wave spreading) and R1 and R2 are hypocentral distances (km). Q values
lower than 1 and higher than 5000 are not used, the Q(f) plot might then display a long straight line.
The Q=Q0*f**qalpha is calculated from the 'good' values'.
Distance correction alpha: The spectral amplitudes are multiplied by R**(distance correction) if
different from zero. This option MUST be set if moment or moment magnitude options (see below) are
selected as well as calculation of Q. However, it can be used without instrumental correction. For
body waves, use 1. Note that the geometrical spreading use here is simpler than used in MULPLT.
Minimum correlation coefficient and minimum signal to noise ratio for kappa: The minimum
correlation coefficient and signal to noise ratio for an event to be included in average kappa. The
coefficients are from the linear fit to the flat part of the spectrum.
186
Velocity and density: Velocity (km/sec) and density (g/cm*cm) used for calculating moment spectra.
If set to 0,0, no moment spectra are calculated. See section on MULPLT for details of calculation.
Magnitude spectrum: If 1, the spectral level is converted to moment magnitude, see MULPLT for
details of calculation.
Stations and components: Station-component pairs used, one pair per line, format
(a5,1x,a4,1x,a5,1x,a4). If no relative spectrum is used, the first station-component on the line is used.
Averaging in spec:
Q: For each frequency, the average linear 1/Q and corresponding sd is calculated. The upper and
lower bounds are calculated by subtracting and adding the sd. These values are then converted back
to Q and finally the log is taken. Only the ‘good’ individual values are used. There is a possibility that
the lower bound becomes negative. In that case, the log Q is set to zero. Because the average is
made in 1/Q, the upper and lower bounds curves will not be symmetric around the average Q-curve.
Power spectrum: For each frequency the dB values are averaged and upper and lower curves
should be symmetric.
Kappa: Same as for Power spectrum.
Other spectra: The linear spectra or relative spectra are averaged. The sd used in the log spectra are
calculated by subtracting log average spectrum from log(average spectrum + sd).
Running the program:
The program gets the first pair of stations (or one station) from spec.par, calculates the spectra using
the list of events in spec.inp and at the end of the station list, calculates the average spectral ratios for
all pairs (max 100). All spectra are then shown on one plot together with averages and standard
deviation. Then the next pair of stations is processed in the same way and the program continues
until the end of file spec.par. Each pair of stations with signals and spectra is plotted on one page. If
no relative spectra are made, the plots look similar except that only one station is shown.
Hard copy plots are made for each page and sent to the printer if specified (see below). The hard
copy postscript file is called spec.eps and when the program finishes, a file with the last plot is
available on the disk.
For each spectrum (relative or single), the average spectrum (or Q) is calculated both as an average
of the log spectrum and as an average of the linear spectrum. There is no frequency weighting and
since all values shown on the plot are used, the average value will be more representative of the high
frequency part of the spectrum since there are more values. This can be regulated by choosing
another frequency range. The average spectra shown on the last plot are log-averages. If option to
calculate Q is used, the plots show 1/Q as a function of frequency instead of relative spectra
(proportional to relative spectra). For each event, Q0 and qalpha are calculated.
When calculating kappa, the average spectrum do not have much physical meaning since the
averages are made from absolute spectra of events that might have very different moments. So the
kappa calculated from the average spectrum is not to be used.
Interactive output of level and frequency:
With a spectral ratio (or Q) plot on the screen, position the cursor at the point of interest on the
spectrum and click. The level and frequency will now be displayed on the right side of the plot.
The output file spec.out gives details of the run like averages and missing data. The output file
spec_ave.out gives the x and y-values of the average spectrum IF IT HAS BEEN PLOTTED ON THE
SCREEN. File spec_rel.out gives the values of the relative spectra.
There are 4 interactive input options:
0: All spectra are calculated but not sent to the plotter or screen except the last plot with the average
spectra (sent to both screen and printer). Used for checking the files or making a final run. If no
relative spectrum is chosen, no final plot is made. For each station and event combination, check
187
lines are written out on the screen.
1: All plots are shown on the screen, but not sent to the laser printer.
2: All plots are shown on the screen and at the same time sent to the laser printer.
3: No plots are shown on the screen, all are sent to the laser printer. For each station event
combination, check lines are written to the screen.
How to run the program with only waveform files available:
Since the program requires S-files for input, these must first be generated.
Step 1: Generate S-files in your local directory with AUTOREG,
Step 2: Make the spec.inp file with COLLECT.
With only waveform files and no readings in the spec.inp file, it is only possible to use option 4
(absolute time) for start criteria. Since the events have not been located, the "origin time" read from
the S-files will be identical to the waveform file start time, so the parameter "start" can then be set to
number of seconds after waveform file start time. Figure 34 shows an example.
188
Figure 34. An example of using the SPEC program. On top the original traces are shown with
windows chosen, in the middle the spectra of each channel and at the bottom, the relative
spectrum. Lower right shows the input parameters used. In some cases (kappa and Q) the
values calculated for this case are also shown.
189
Figure 35. An example of a GMT plot. The figure shows an example of making noise spectra of
several traces.
6.24 Seismic risk related programs
This section is written by K. Atakan. Extensive testing of the programs was done over the years by
many users. A.Ojeda, performed testing and prepared input files for the CRISIS99 and CRIATT
programs.
Introduction
Currently, the SEISAN package includes a series of stand-alone programs that can be used in a
number of tasks that are needed to perform seismic hazard analysis. The basic requirements for
performing a probabilistic seismic hazard analysis may be summarized as follows:
Homogenize the earthquake catalogue and assess the completeness
Define the seismic source zones.
Prepare input parameters from the earthquake catalogue for each source zone.
Prepare attenuation relations for the region.
Compute hazard in terms of peak ground acceleration (PGA).
Assess site effects.
Prepare response spectra.
Following is a list of programs that constitutes the part of the SEISAN analysis package, which deals
with seismic hazard and related problems. Most of these programs are described in more detail in
different sections of the SEISAN manual.
SELECT: Select a subset of earthquake data according to given criteria.
STATIS: Statistical information about the database is computed and can be used in the analysis.
CATSTAT: Program to compute and plot the yearly, monthly and daily number of events from a given
catalogue.
CAT_AGA: Program to reorder the hypocenter lines in a CAT-file according to hypocenter agency in
order to put the prime estimate in the beginning.
190
CLUSTER: Program that searches for the dependant events in time and distance in a given
earthquake catalogue.
EXFILTER: Identifies the probable explosions, based on the user-defined parameters involving timeof-day distribution and the mining locations. It can be used for catalogue clean up and discrimination
between the earthquakes and man-made explosions.
MAG: Magnitude regression and conversion program. Prepares also a plot showing the scatter data
and the best-fitted line. Magnitude conversions are then performed after a user defined priority list.
EPIMAP: Plots coastlines, national boundaries and earthquake epicenters. It can also contour the
produced output map file from hazard programs such as EQRISK, and overlay on the epicenter map.
It is also possible to select a subset of earthquakes from a chosen polygon on the epicenter map.
BVALUE: Prepares magnitude-frequency of occurrence diagrams and computes a- and b-values with
maximum likelihood and least square approximation. In addition, the threshold magnitude and the
maximum observed magnitude can be obtained.
CODAQ: Computes the Q value from a given set of seismograms. This can be used later in the
CRIATT program to create the attenuation table.
CRIATT: Computes attenuation tables for a given set of parameters using the random vibration
theory.
CRISIS99: Computes seismic hazard in terms of the probability of exceedance vs earthquake
intensity measures such as peak ground acceleration (PGA) or any other spectral ordinate. It can also
compute hazard for a given grid of map co-ordinates corresponding to user-defined different return
periods. (SUN and PC). The Windows program must be installed separately, look for ZIP file in SUP.
EQRISK: Program to compute seismic hazard in terms of probabilities of exceedances vs earthquake
intensity measures such as peak ground acceleration (PGA), for a given site or a grid of sites for up to
eight different return periods. Currently 1975 version is used.
EQRSEI: Converts the output file from the EQRISK program "eqrisk.out", to individual contour files
corresponding to each return period specified. These files can later be used directly as an input to
EPIMAP to plot the PGA contour maps.
SPEC: Computes amplitude spectra for a given set of earthquake records and plots spectral ratios. It
can be used to assess local site effects.
Probabilistic earthquake hazard computations can be done, using the two alternative programs
CRISIS99 or EQRISK. In addition, the programs listed above and a number of other programs that
manipulates earthquake data within the SEISAN package, are useful tools to assess the parameters
that are needed to perform a seismic hazard analysis for an area of interest. The two main programs,
CRIATT for computing the attenuation tables and CRISIS99 (modified version 1999) to compute
seismic hazard are explained in more detail in the following. Both programs are written by Mario
Ordaz of the Institute of Engineering, UNAM (Ordaz, 1991; 1999). The well-known hazard program
EQRISK, on the other hand, is written by Robin K. McGuire and the original manual is distributed
through United States Department of the Interior, Geological Survey (McGuire, 1976).
The two alternative hazard programs CRISIS99 and EQRISK have a number of features that are
present in both. However, there are some advantages and disadvantages with both programs. In
terms of the computing time and parameter input both programs require the same time. In the case of
EQRISK, earthquake source zones are defined as arbitrary polygons (quadrilaterals). CRISIS99, on
the other hand, operate with completely arbitrary polygons for the definition of the source zones and
dipping planes may also be defined. In the MS-Windows 95 version, the source zones and the input
parameters can be checked interactively through a user-friendly interface. In terms of the attenuation
relations, CRISIS99 uses a table created by a separate program (CRIATT) and is therefore flexible (it
also allows different attenuation relations for different source zones), whereas the attenuation relation,
in the case of EQRISK, is given through a pre-determined mathematical formulation. Finally,
CRISIS99 is superior to EQRISK, as it takes into account the uncertainties through the standard
deviations introduced on several input parameters.
Step by step procedure for seismic hazard analysis
Following is a summary of the steps that have to be completed in order to produce a seismic hazard
map.
1. Compile a catalogue for the area of interest from local, regional and global sources.
191
2. Evaluate the preliminary catalogue completeness by plotting histograms showing the distribution of
events in time for different magnitude intervals. It may be necessary to divide your catalogue into two;
(i) pre-instrumental and (ii) instrumental. Programs SELECT and CATSTAT can be used for this
purpose.
3. Convert magnitudes into one uniform magnitude, preferably to moment magnitude MW. To do this,
regression curves must be prepared for different magnitude scales. Program MAG can be used for
this purpose.
4. Clean up the catalogue for dependant events (i.e. induced seismicity, non-earthquakes,
foreshocks, aftershocks, earthquake swarms). Here a search has to be made for clusters of events
both in time and space. Plots of histograms for specific sequences of time and space will reveal this.
Program CLUSTER can be used for this purpose. The probable explosions may be removed by using
the program EXFILTER.
5. The evaluation of the catalogue completeness is dependent upon the clean-up process and the
magnitude unification. It is therefore necessary that steps 2-4 be repeated until a reliable catalogue is
prepared.
6. Select the set of earthquakes from your catalogue from the part, which is complete for the chosen
threshold magnitude and uniform in magnitude scale. Program SELECT can be used with different
criteria for this purpose. Note the catalogue time span.
7. Prepare a seismicity map for the area of interest with the selected data, using EPIMAP. Delineate
the earthquake source zones. Here, zooming and the area selection procedures of EPIMAP may be
used.
8. Use additional information from geology, geophysics, seismotectonics, paleoseismology etc. to
improve the source zonation.
9. For each earthquake source zone select the subset of events that fall in the chosen area. This can
be done by using the EPIMAP program, which enables to draw polygons interactively on the screen
and put the subset of events within this polygon into a file. Alternatively SELECT program can be
used to extract the subsets of data corresponding to the defined source zones.
10. If the hazard is to be computed using CRISIS99 or by EQRISK, note the x, y (longitude, latitude),
co-ordinates for each corner of the polygon.
11. The seismicity within each source zone is assumed to be uniform following a Poissonian
occurrence. In order to define this, a set of critical parameters has to be assessed for each source.
These are:
Number of earthquakes above a threshold magnitude: This is the a-value for the lower bound
magnitude. Catalogue time span: This is the time span of your catalogue where it is complete.
Beta (bvalue * ln (10)) and its standard deviation: The b-value is the slope of the best-fitted line to the
cumulative curve for the magnitude frequency of occurrence distribution (Gutenberg-Richter relation).
Maximum expected magnitude with its standard deviation: This is usually inferred through other
available information, such as geology, palaeoseismicity, or subjective judgement of the scientist. It is
usually set to half a magnitude higher than the maximum observed when no information is available.
Maximum observed magnitude: This is the largest magnitude observed within the catalogue time
span.
Threshold magnitude: The so-called lower bound magnitude, which is chosen, based on the
engineering considerations. Usually magnitudes less than 4.0 are not considered engineering
significant.
In order to obtain each of the above critical parameters, a thorough evaluation of the earthquake
catalogue is needed. BVALUE program can be used to obtain some of these parameters. However,
while running the program, choosing the magnitude interval and the magnitude increment has to be
done critically, taking into account the catalogue completeness and the detection threshold. These
parameters will later be used in the input for the seismic hazard analysis program CRISIS99.
Alternatively, the same input parameters are also needed for the EQRISK program. For each source
zone, plot the magnitude- frequency of occurrence curves.
12. Try to assess whether there are characteristic earthquakes in your region. This can be done with
a careful examination of your catalogue and the active faults in the area. Studying the magnitudefrequency of occurrence through the BVALUE program will help assessing this.
13. Try to establish an acceptable attenuation relation for your area. This can be done through
empirical estimations or theoretically based on the random vibration theory (RVT). CRIATT program
can be used to create the attenuation table. Alternatively, if you have an already established
attenuation relation, this can be directly used in the EQRISK program. In this case, you can skip the
steps 13-16, and continue from step 17 and onwards.
192
14. Establish a reliable Q factor by using the CODAQ program. This will be used in the attenuation
program CRIATT to create the attenuation tables necessary for the hazard analysis.
15. Create the necessary input file for the CRIATT by modifying the sample-input file ‘criatt.inp’. or
use program CRIPAR.
16. Run CRIATT to create the attenuation table necessary for the CRISIS99.
17. Create the input file for the CRISIS99 program by modifying the example-input file ‘crisis99.inp’.
Make sure that the critical parameters are reliable and the geometry of the source zones are correct
(see the program description).
18. Run the CRISIS99 program with the input file you have created and the output attenuation table
from CRIATT. The program will generate the output files with the probability of exceedance rate vs
earthquake intensity (e.g. PGA), for the required return periods. Alternatively, if you have prepared
the input for the EQRISK program, hazard can be computed by running the EQRISK program, for a
given set of return periods (up to eight), for selected sites or for a grid of sites.
19. Repeat stages 6 to 17 to refine your model and the corresponding results.
20. Convert the output hazard "map" file from CRISIS99 for the computed return periods to individual
contour files. Alternatively, if you have used EQRISK to compute hazard, the output file "eqrisk.out"
can be converted using EQRSEI program, into individual contour files for previously defined return
periods.
21. Plot the hazard maps for the desired return periods. Contouring option from EPIMAP can be used
for this purpose (only for the EQRISK). Plot also the graphs for probability of exceedance rates vs
PGA for selected critical sites.
22. Try to assess the local site effects for the critical sites. SPEC program can be used to obtain the
amplification factors due to unconsolidated sediments. These factors can be used later to adjust the
response spectra.
Many of the programs mentioned above are described individually throughout this manual at different
sections. In the following the programs that are directly relevant to hazard computations and not
described in other sections of the manual are explained in detail.
CRISIS99:
CRISIS99 is a computer program to compute seismic hazard in extended regions. It was developed
at the Institute of Engineering, UNAM, Mexico, by Mario Ordaz ([email protected]),
Armando Aguilar and Jorge Arboleda.
Basic input data are: geometry of the sources, seismicity of the sources, and attenuation relations.
Source geometry can be modeled as: 1) area sources, using a polygon with at least three vertex;
longitude, latitude and depth must be given for each vertex, so this type of source can be used to
model, for instance, dipping plates or vertical strike-slip faults; 2) fault sources, using polylines; and 3)
point sources, included essentially for academic purposes.
Seismicity of the sources can be modeled either as Poisson or characteristic earthquake process. In
the first, magnitude frequency relations are smoothly truncated Gutenberg-Richter curves, whereas
for the second, the program assumes a Gaussian distribution of the magnitudes.
Hazard computations can be performed simultaneously for several intensity measures, for instance,
PGA, PGV, and several spectral ordinates. Required attenuation laws are given in the form of tables
containing the median values of the intensity measures as a function of magnitude (the rows of the
table) and focal distance (the columns of the table). Several attenuation models can be used in the
same run, assigning an attenuation pattern to each source.
Using a recursive triangularization algorithm, spatial integrations are performed optimizing the number
of calculations, so CRISIS99 will integrate with more points for the nearest sources and less (or none)
for distant sources.
CRISIS99 considers two different kinds of earthquake occurrence processes: Poisson process and
characteristic earthquake process. CRISIS99 is oriented to computing hazard in extended regions.
Hazard estimations are made for points in a grid that is not necessarily rectangular. The program can
run under SunSOLARIS, SunOS and on PC (Windows95 or higher). Sun versions are to be used as a
stand-alone program. The Windows version, on the other hand, also contains a windows interface for
193
visual inspection of the input data as well as the results. Data validation options are available (only for
the Windows version) and parameters can be given in a user-friendly graphic environment. CRISIS99
contains also a post-processing module that can be used to visualize the results, given in terms of
maps of intensity measures for an arbitrary return period or exceedance rate curves for a selected
site, not necessarily a point in the original grid of sites. Also, if several intensity measures are
included in the computations, uniform-hazard spectra can be produced. The main results of a run are
also written to ASCII files, so the user can use his/her own post-processing techniques/software.
For the Windows version, a separate compressed file ‘crisis99.zip’ is included with sample-input data
in SUP. Instructions on how to install the Windows version are included in the file ‘crisis99.txt’ in the
INF directory. The Sun UNIX versions, are part of the standard SEISAN distribution and need not be
installed specifically.
Detailed description of the input and output files is given in the pages below.
Input files for the CRISIS99
There are basically two input files that are required. First is an attenuation table (or several tables),
and second is the major input parameter file where the file name for the attenuation table is also
given. The input file can be prepared based on the format descriptions given below or modifying the
example input file. An example-input file is included in the DAT directory with the file name
"crisis99.inp".
There are some limitations in the input parameters. Following is a summary of the maximum values
set in the program:
Attenuation Models :
Intensity levels*:
Structural periods:
Number of regions:
Magnitudes in attenuation model:
Distances in attenuation model:
Number of sub-sources per region:
5
20
15
200
10
21
4000
(* the term ’intensity’ here should not be mixed with macroseismic intensity. In this context ’intensity’ is
meant as any chosen ground motion measure, such as PGA, PGV or any other spectral ordinate).
In the following the input file is described in more detail (by Mario Ordaz).
I. GENERAL DATA FILE
Format is free unless indicated otherwise.
1. General title of the run. 1 line
TITGEN (A80)
2. Global parameters of the run. 1 line.
NREG, NMOD, NT, NA
NREG: Total number of regions (sources) in which the seismogenic area is divided.
NMOD: Number of different attenuation models.
NT: Number of spectral ordinates (or, in general, measures of intensity) for which seismic hazard
is to be computed.
NA: Number of levels of intensity for which seismic hazard will be computed.
3. Parameters for each spectral ordinate. NT lines. Free format
T(I), AO(I), AU(I)
T(I): Structural period of i-th spectral ordinate. It is used only for identification purposes, so in the
cases in which structural period has no meaning, it can be just a sequential number.
AO(I): Lower limit of intensity level for i-th spectral ordinate.
AU(I): Upper limit of intensity level for i-th spectral ordinate.
194
Exceedance rates for the i-th intensity will be computed at NA values, logarithmically spaced
between AO(I) and AU(I)
4. More Global parameters
RMAX, TR1, TR2, TR3, TR4, TR5
RMAX: Parameter controlling the spatial integration process. Sources at distances greater than
RMAX kilometers from a site will be ignored.
TR1,…,TR5: CRISIS-99 will generate a file containing intensity levels for fixed return periods
TR1,…,TR5. See below for the description of this output file. Five values must be always given.
5. Parameters defining the basic grid of points in which hazard is to be computed. 1 line
LOI, LAI, DLO, DLA, NLO, NLA
LOI, LAI: Longitude and latitude, respectively, of the origin of the grid.
DLO, DLA: Longitude and latitude increments
NLO, NLA: Number of lines of the grid in the longitude and latitude directions, respectively.
Results will be given for points (LO(I ),LA(I)), where
LO(I) = LOI + (J-I)*DLO , J=1, NLO
LA(I) = LAI + (I-1)*DLA, I=1, NLA
6. Number of polygons to be used to reduce the initial rectangular grid. 1 line.
NPOLGRID
Introducing one or more boundary polygons can reduce the initial rectangular grid of points. If
polygons are given (NPOLGRID>0) the computation of hazard will be performed only for those
points of the grid, which are inside one of the polygons. If NPOLGRID=0 computations will be
made for all points in the rectangular grid. NPOLGRID<=10.
If NPOLGRID>0 then the following lines must be given for each polygon:
7. Definition of the k-th boundary polygon.
NVERGRID (K)
LONG (K, 1), LAT (K, 1)
NVERGRID(K) lines
LONG (K, 1), LAT (K, 1)
NVERGIRD(K): Number of vertex of polygon k. NVERGRID(K)<=30.
LONG (K, I), LAT (K,I), I=1,…,NVERGRID(K): Co-ordinates of the polygon’s vertex. The
polygon must be described counter clockwise.
8. Files of attenuation tables. NMOD lines
MODELO (I) (A20)
MODELO (I): Name of the file containing the i-th attenuation table (including path). The
format of attenuation tables is explained below.
9. Data defining seismicity in each region. NREG blocks.
TITULO (N) (A80)
IC(N), IE(N), IMO(N)
NV(N)
LONG(1),LAT(1),PROF(I)
…
NV lines
LONG(NV), LAT(NV), PROF(NV)
Poisson model: (IC(N)=1)
LAMBDA0(N), EB(N), CB(N), EMU(N), SMU(N), MMAX(N),M0(N)
Characteristic model: (IC(N)=2)
EMT(N), T00(N), D(N), F(N), SMT(N), M0(N), MU(N)
TITULO(N): Identification name for source N
IC(N): Flag defining the type of occurrence model assumed for N-th source. IC(N)=1 for Poisson
195
model, IC(N)=2 for characteristic-earthquake model.
IE(N): Defines type of source. IE(N)=0 for area source, IE(N)=1 for line source and IE(N)=2 for
point source.
IMO(N): Number of the attenuation model that will be used with this source. Must be between 1
and NMOD.
NV(N): Number of vertex defining source N.
LONG(I), LAT(I), PROF(I), I=1,…,NV(N): Co-ordinates of vertex I of source N. LONG(I)
and LAT(I) are geographical coordinates of point i, whereas PROF(I) is the depth of the point,
in km, which must be positive.Sources can be of three types: areas (polygons), polylines or points.
Polylines and points can be given in any order. In general, in the case of an area source, CRISIS99
will divide the polygon into triangles. It first checks if triangulation can be made in the XY plane.
Numbering of the vertex of the polygon must be done counter-clockwise in this plane when looked
from above
the surface of the Earth. If there are vertical planes, CRISIS99 will try to triangulate the area in
the XZ plane, so numbering of vertex must be done counter-clockwise in this plane. Finally,
CRISIS99 will try to triangulate in the YZ plane. There are some bizarre source geometries
that cannot be well resolved by CRISIS-99, for instance, an L-shaped vertical plane. In these
cases, an error will be reported.
Poisson model:
LAMBDA0(N): Exceedance rate of magnitude M0(N). The units are earthquakes/year.
EB(N), CB(N): Expectation and coefficient of variation, respectively, of the “b-value” for the
source, given in terms of the natural logarithm.
EMU(N), SMU(N): Expected value and standard deviation, respectively, of the maximum
magnitude for the source.
MMAX(N): Maximum observed magnitude in this source.
M0(N): Threshold magnitude for source N. The catalogue of earthquakes is assumed to be
complete for M>M0. Earthquakes with M<M0 are absolutely ignored.
Characteristic model:
EMT(N): Median value of the times between characteristic earthquakes with M>M0. This is the
inverse of the exceedance rate for M>M0.
T00(N): Time elapsed since the last occurrence of a characteristic earthquake.
D(N), F(N): Parameters defining the expected magnitude as a function of time, as in the slippredictable model. It is assumed that
E(M|t)=max(M0(N),D(N)+F(N)*LN(t))
Of course, if F(N) is set to zero, then D(N) becomes the expected time-independent magnitude of
the characteristic earthquake.
SMT(N): Standard deviation of the magnitude of the characteristic earthquake. It is assumed
independent of time.
M0(N): Minimum possible magnitude of a characteristic earthquake. Earthquakes with M<M0 are
absolutely ignored
MU(N): Maximum magnitude of the characteristic earthquake to be used in the integration
process.
10. Name of the map file. 1 line
File name (including path) containing the base map to be used in post-processing with CRISIS99 for
windows. This name does not have any influence in the hazard computations. However, CRISIS99
expects a line here.
11. Name of the file of cities. 1 line
File name (including path) containing the co-ordinates of cities, to be used in post-processing with
CRISIS99 for windows. This name does not have any influence in the hazard computations.
However, CRISIS99 expects a line here.
12. ATTENUATION TABLES
NMOD attenuation tables must be given each one in a different file.
196
The tables give to CRISIS99 the relations between magnitude, focal distance and median
intensities. CRISIS99 expects the following parameters in the i-th attenuation file, I=1,…,NMOD:
1. Parameters defining the magnitude limits. 1 line
MINF(I), MSUP(I), NMAG(I)
MINF(I): Lower limit of magnitude given in the table.
MINF(I): Upper limit of magnitude given in the table.
NMAG(I): Number of magnitudes for which intensity is given.
CRISIS99 assumes than intensities are given for magnitudes M(K),
where M(K)=MINF(I)+(K-1)*DMAG, where DMAG=(MSUP(I)-MINF(I))/(NMAG(I)-1).
2. Parameters defining the distance limits. 1 line
RINF(I): Lower limit of distance given in the table.
RINF(I): Upper limit of distance given in the table.
NRAD(I): Number of distances for which intensity is given.
CRISIS-99 assumes than intensities are given for distances R(K),
where log(R(K))=log(RINF(I))+(K-1)*DLRAD, where DLRAD=(log(RSUP(I))-log(RINF(I)))/(NRAD(I)-1).
That is, distances are supposed to be logarithmically spaced.
3. For each of the NT different intensity measures, the following block of lines:
T(I,J), SLA(I,J), AMAX(I,J)
SA(I,1,1,1), SA(I,1,1,2),…,SA(I,J,K,L),….,SA(I,NT,NMAG(I),NRAD(I)
T(I,J): Structural period of j-th spectral ordinate. It is used only for identification purposes, so in
the cases in which structural period has no meaning, it can be just a sequential number.
SLA(I,J): Standard deviation of the natural logarithm of the j-th measure of intensity in the i-th
model.
AMAX(I,J): Maximum possible value of the j-th intensity in model I. The integration process will be
truncated, regarding as impossible (zero probability) values larger than AMAX(I,J). If AMAX(I,J) is
set to zero, then integration with respect to possible values of intensity will be performed from 0 to
∞.
SA(I,J,K,M): Median value of the intensity in model I, for the J-th spectral ordinate, the K-th
magnitude and the L-th distance.
For each attenuation model, given in a separate file, CRISIS99 reads the above mentioned
parameters in the following form:
D0 J=1,NT
READ(8,*) T(I,J),SLA(I,J),AMAX(I,J)
DO K=1,NMAG(I)
READ(8,*) (SA(I,J,K,L),L=1,NRAD(I))
ENDDO
ENDDO
Output files from CRISIS99
CRISIS99 generates several output files, whose names begin with the base name requested at the
beginning of the run. The output files are:
1. Main results file. This file with - .res - extension contains a printout of the name of the run, the
values assigned to the variables, characteristics of the attenuation models, geometrical and
seismicity description of the sources, the data defining the computation grid, etc. It also gives the
final results, that is, exceedance rates for each site and type of intensity. It also gives a brief
summary of the computations for each site, indicating which sources are of interest to the site and
which sources were skipped.
2. Graphics file. The principal graphics file with - .gra - extension contains a brief identification
header, and the exceedance rates for the type and levels of intensity requested. This file can be
used as input file to plot intensity versus exceedance rate curves. CRISIS99 generates also a
197
binary file with the exceedance rates for each structural period, so CRISIS99 will generate NT
binary files. These binary files will be used only in the Windows System version of CRISIS99 to
make hazard maps. The names of these files are base_name.b1, base_name.b2,…,
base_name.bNT.
3. Map file. This file with - .map - extension contains intensity levels for fixed return periods
(TR1…TR5) for each type of intensity and site. It also gives the co-ordinates of each site. This file
can be used to generate contour or 3d maps of intensity levels associated to constant
exceedance rates.
Example output files are included in the DAT directory (crisis99.res, crisis99.gra, crisis99.map).
CRIATT:
In this program, an earthquake source model and results from Random Vibration Theory (RVT) (e.g.
Boore, 1983; 1989), are used to estimate attenuation of ground motion parameters as a function of
moment magnitude, MW, and hypocentral distance, R. Ground motion is assumed to be band-limited,
stationary and of finite duration.
For estimating the Fourier acceleration spectra, a(f), it is assumed an omega square constant stress
drop source model given by Brune (1970). The expression for a(f) is:
a(f) = CG (R) S(f) D(f)
(18)
where
2
3
C = (4π RVφFV) / (4πρβ )
2
2
(19)
2
S(f) = M0f / (1+f /f0 )
(20)
and
D(f) = P(f)e π
- Rf/βQ(f)
(21)
Thus the spectrum a(f) is the multiplication of a constant C (independent of frequency), geometrical
spreading term G(R), source function S(f), and diminution function D(f). In C, Rvø is equal to average
radiation pattern (0.55), F is free surface effect (2.0), V is partition of a vector into two horizontal
3
components (0.707), ρ is density in gm/cm , and β is shear wave velocity in km/sec.
In S(f), M0 is the seismic moment and f0 is the corner frequency, given by (Brune, 1970)
6
1/3
f0 = 4.9 x 10 β(Δσ / M0)
(22)
where β is in km/sec, Δσ is the stress drop in bars, and M0 is in dyne-cm. The diminution factor D(f),
accounts for loss of energy due to internal friction and scattering.
At distances less than a certain critical value of Rc, the strong motion records are dominated by S-waves.
1/2
Thus for R < Rc, G(R) = 1/R is the geometrical spreading. For R > Rc, G(R) = 1/(R Rc) .
The diminution function D(f) in equation (21) requires Q(f) and P(f), where the quality factor defined by the
regional attenuation is expressed by Q(f) = Q0 f ε (f is frequency and ε ≤1.0), and whereas
-
f
P(f) = e πκ
(23)
P(f) reconcile an additional attenuation term which may be related to near-surface loss of energy
where kappa is a high frequency decay factor (Singh et al., 1982).
Input file for CRIATT
198
The standard input file for the CRIATT program can be created by modifying the example input file. A
total of 23 parameters provide the necessary input for calculating the attenuation tables, which is
based on equation 18, described earlier. The user should define the magnitude and the distance
limits. It is important to note here that some combinations of parameters may result in 0 values for
large distances in the table, which creates problems for the CRISIS99 program. In order to avoid this,
the distance ranges are set to Rmin<10 km (CRISIS99 requires one-digit only) and Rmax < 500 km.
Usually the regional attenuation term and the site factor are the most critical factors in the definition of
a(f). The effect of the high-frequency decay factor can only be seen when the combination of the
kappa parameters (kappa0 and kappa1) are chosen correctly (e.g. increasing kappa1 with kappa0
kept constant, would result in low ground motion values). An example input file is included in the DAT
directory with the file name "criatt.inp".
Output file from CRIATT
The output of the CRIATT program is a file containing the attenuation tables for the selected spectral
ordinates (i.e. as a default only PGA corresponding to a period of 0.005 sec is computed). For each
spectral ordinate, the file will contain a set of values (e.g. PGA) for different distances. This file is then
used as one of the inputs to the CRISIS99 program. The file name is user defined. An example output
file is included in the DAT directory with the file name "criatt.tab".
CRIPAR
The program was used earlier to generate input for both crisis and criatt but now it is only used with
criatt due a format change for crisis99.
EQRISK:
This popular program for computing seismic hazard is written by McGuire (1976), and the complete
manual is published as an open file report. The following is a short summary of the program operation
and a full description of the input parameters as well as format of the input file. These descriptions are
as they are given in the original manual (McGuire, 1976).
The program EQRISK evaluates risk (hazard) for each site-source combination and intensity level
and calculates the total annual expected number of occurrences of intensity greater than those levels
of interest at a site by summing the expected numbers from all sources. Seismic source areas are
specified as a set of arbitrarily shaped quadrilaterals. For ease of use, gross sources may be divided
into sub-sources, which are a string of quadrilaterals, each two adjacent subsources having two
common corners. A Cartesian co-ordinate system is used and the location of the origin is arbitrary.
Input file for EQRISK
The default input file is named "DATA" and is hardwired into the program (for the SUN version file
name should be uppercase). An example input file is included in the DAT directory with the name
"eqrisk.inp", which should be renamed to "DATA" before running. Following is the description of the
individual parameters and their format as described in the original manual (McGuire, 1976).
Card 1 (Format 20A4): Title. Any 80 characters can be used to describe the problem.
Card 2 (Format 3I10): NSTEP, JCALC, JPRINT.
NSTEP is the number of integration steps used in integrating over distance for each site-source
combination.
JCALC is the flag indicating how integration on magnitude is to be performed (JCALC=0 is used
for analytical integration, and the form of the attenuation function is described in the original
manual. JCALC=1 is used for numerical integration on magnitude. The user must supply own
attenuation function in subroutine RISK2.)
JPRNT is the flag indicating the desired output (JPRNT=0 is used to print only total expected
numbers and risks at a site which is normally used when a grid of sites being examined.
JPRNT=1 is used to print expected numbers from each site-source combination, normally used
when examining a single site).
199
Card 3 (Format I5, 12F5.3): NLEI, TI(1), TI(2), ..... TI(NLEI).
NLEI is the number of intensities to be examined.
TI(1), TI(2) and so on, are intensities for which expected numbers and risks are calculated at
each site. Note, that the values for TI(i) may be Modified Mercalli Intensity or the natural logarithm
of ground acceleration, velocity, displacement or spectral velocity. In printing results, the program
prints both TI(i) and its antilogarithm. Values for array TI must be specified in increasing order.
Card 4 (Format 8F10.2): RISKS(1), RISKS(2), ..... RISKS(8).
RISKS(1), RISKS(2), and so on are risks (probabilities of exceedance) for which the
corresponding intensities are desired. These intensities are calculated by interpolation on a
logarithmic scale, between intensities (in the list of examined intensities, TI) having larger and
smaller risks. Both the corresponding intensity and its antilogarithm are printed. Values for array
RISKS must be specified in order of decreasing risk. If fewer than eight values are desired, leave
succeeding spaces on the card blank. To avoid large errors and subsequent misinterpretation,
the program will not extrapolate to calculate intensity values corresponding to risk levels
specified; it is the user's obligation to choose values for array TI which will result in risks which
bound those specified in array RISKS. This is of course, a matter of judgement and experience.
The user must be cautioned that in a grid site system appropriate values for array TI may vary
considerably for the different sites examined. The intensities interpolated for levels specified in
RISKS will be most accurate for closely spaced values of TI.
Card 5 (Format 8F10.2): C1, C2, C3, SIG, RZERO, RONE, AAA, BBB.
C1, C2, C3 and RZERO are parameters in the attenuation equation for mean intensity discussed
in the original manual (McGuire, 1976):
mI(S,R)=C1+C2*S+C3*ALOG(R+RZERO)
SIG is the standard deviation of residuals about the mean. If no dispersion of residuals are
desired, insert a very small value for SIG (rather than exactly 0.0).
RONE is the limiting radius inside of which no attenuation of motion is desired, for values of focal
distance closer than RONE, the mean intensity is calculated using RONE in place of R in the
attenuation equation above. If this feature is not desired, insert zero for RONE.
AAA and BBB are parameters in the equation limiting the mean intensity:
max mI(s)=AAA+BBB*S
The value specified for BBB must be between zero and C2 for this limiting equation to make
sense. If it is not, an error message will result and program operation will terminate.
Card 6 (Format I10, 6F10.2): NGS, NRS(1), NRS(2), .... NRS(NGS).
NGS is the number of gross sources to be specified.
NRS(1), NRS(2), and so on are the number of subsources in gross source 1,2, etc. See the
original manual (McGuire, 1976), for a general description of the source specification.
Card (set) 7 (Format I10, 6F10.2): LORS(I), COEF(I), AM0(I), AM1(I), BETA(I),
RATE(I), FDEPTH(I).
There must be NGS+1 of these cards, one for each gross source and one for background
seismicity.
LORS(I) is a flag indicating whether the source area has a loose or strict lower bound (LORS=0
implies a loose lower bound and LORS=1 implies a strict lower bound).
COEF(I) is a coefficient modifying the expected number of exceedances from gross source I. Its
most common value is +1.0.
AM0(I) is the loose or strict bound lower magnitude or intensity for gross source I.
AM1(I) is the upper bound magnitude or intensity for gross source I.
BETA(I) is the value of ß for gross source I. It is equal to the natural logarithm of 10, times the
Richter b-value for the source.
RATE(I) is the rate of occurrence of events having magnitudes of intensities greater than AM0(I).
If a discrete distribution on intensities has been used to calculate the rate, the user may wish to
specify AM0(I) as one-half intensity unit lower than the lowest intensity used to establish the rate.
Note that for gross sources RATE(I) is in units of number per year; for background seismicity it is
in units of number per year per 10,000 km.
FDEPTH(I) is the focal depth of events in gross source I, in km. If epicentral distances are
required for all sources and for background seismicity for the attenuation function, insert zero for
200
FDEPTH(I).
If no background seismicity is desired, leave the last card in this set completely blank.
Card 8 (Format 4F10.2): X1, Y1, X2, Y2.
There must be NRS(1)+NRS(2)+ .... +NRS(NGS)+NGS of these cards. The first NRS(1)+1 cards
specify co-ordinates of subsources in gross source 1, the next NRS(2)+1 cards specify coordinates of subsources for gross source 2, and so on. Internally, the point X1, Y1 is connected to
X2, Y2, as well as both to the previous and the subsequent points designated as X1, Y1, as long
as these are both in the same gross source. Point X2, Y2 is connected similarly. An example is
elucidating. The following points define two gross sources having two subsources.
0.0
0.0
10.0
0.0
0.0
5.0
8.0
8.0
-5.0
10.0
6.0
15.0
10.0
20.0
11.0
20.0
15.0
15.0
16.0
15.0
15.0
0.0
16.0
0.0
Card 9 (Format 2I5, 4F10.2): NX, NY, XZERO, YZERO, XDELTA, YDELTA.
There can be any number of these cards, one for each site or grid of sites to be examined.
NX and NY are the number of grid points in the X (East-West) and Y (North-South) directions;
that is, they are the number of columns and rows in a grid of sites to be examined. For
specification of a single site, NX and NY must have values of unit. Zero or negative values for NX
and NY are meaningless and will cause program to terminate.
XZERO and YZERO are the co-ordinates of the site to be examined, or are the lower left corner
of the grid if NX and/or NY are greater than one.
XDELTA and YDELTA are the grid spacing in the X and Y directions. When the grid option is not
used, these variables may be left blank or set equal to zero.
Final card: Insert one blank card at the end of the input deck.
Output file from EQRISK
There is only one standard output file generated by EQRISK which has a default file name of
"eqrisk.out". This file contains the results of the hazard computations for each site for the specified
exceedance probabilities. This output file can easily be converted to individual intensity (e.g. PGA)
contour files (one for each level of exceedance probability), using the program EQRSEI. The resulting
contour maps from these output files may then be plotted by EPIMAP.
The detailed format of this output file is described in the original manual (McGuire, 1976), and is not
repeated here. A test set of input and output files are given in DAT.
EQRSEI:
The program EQRSEI converts the output file "eqrisk.out" from the EQRISK program into individual
intensity contour files for the previously defined return periods. There may be up to eight such files
(eqrsei1.out, eqrsei2.out ...., eqrsei8.out). These files can then be used as input to the EPIMAP
program to plot the contours of PGA values on the epicenter maps. Each file contains also some
header information, where the individual contours and the contour intervals are given. In addition, the
color codes are also given. The individual contours and the contour intervals can be modified by
editing the header lines of these files.
CLUSTER:
This is a program that searches for the dependant events in a given catalogue (compact file) with
respect to time and distance. It is written by Juan Pablo Ligorría and Conrad Lindholm. The input is a
standard Nordic file with header lines only (compact file). The user has to give the number of days to
be searched before and after the main event, and the distance limits in km. The magnitude of the
main event over which the search will be performed is also user defined. The output is a repetition of
the input catalogue with "?" placed at the end of each dependant event which falls within the limits
defined in the interactive input in time and distance. The default file name is "cluster.out". The user
201
should then work systematically through these events and decide whether they should be cleaned or
not. This process, we feel, should be done manually, because deleting events from the catalogue
(especially the historical part), may have serious implications later in the hazard computations.
Clusters of foreshocks, aftershocks or other dependant events such as earthquake swarms can be
delineated by this program.
6.25 Magnitude relations, MAG
The MAG program calculates simple magnitude relations. The program has three functions: (1)
Calculate parameters for a magnitude scale (Ml or Mc), (2) Calculate relation between two different
magnitudes and/or spectral parameters and (3) Calculate a new magnitude as a function of an
existing magnitude, a natural step following function (2). All three functions can be done at the same
time. Function (3) can also be used for moving a particular magnitude type and/or agency to the first
magnitude position in line 1 to be plotted with EPIMAP.
ALL HEADER LINES ARE SEARCHED FOR MAGNITUDE INFORMATION
Input: The data input is a CAT-file like one made with SELECT or COLLECT or it can be a compact
file if only magnitude comparison is made. Optionally there can be a parameter file, which MUST be,
called mag.par and reside in the working directory. An example of the parameter file is found in DAT
and also shown below. The parameter file is not needed for all operations, see details below.
1: Magnitude scales
Coda magnitude Mc: The coda magnitude scale used is
Mc = A * log(coda) + B * dist + C
where Mc is the coda magnitude, coda is the coda length in secs, dist is the hypocentral distance in
km (calculated from epicentral distance and depth in CAT file) and A, B and C are constants to be
determined. This is done in two ways
3d regression
m = A * log (coda) + B * dist + C
2d regression
m = A * (log (coda) + dist_coff * dist) + C
with B = A * dist_coff where dist_coff is given in the parameter file and m is the reference magnitude.
SO B AND dist_coff ARE DIFFERENT. The CAT-file must contain coda readings, epicentral
distances and a magnitude in the header line. A linear regression is then made between the known
magnitude from a given agency and the observed coda lengths following the relations above. The
user has the option to choose the type of magnitude to use in the regression. Usually Ml or Mb are
used. All station-event combinations are used to determine simultaneously the 3 constants A, B and
C. Since the data often is too bad to determine all 3 parameters at the same time, the program will
also calculate just A and C using a fixed user supplied value for the distance correction to the coda.
The constant dist_coff is given in the mag.par file as the second parameter under MAG_PAR_COF
(see below). IN ORDER FOR THE CODA SCALE OPTION TO WORK, THE DISTANCE
COEFFICIENT MUST BE DIFFERENT FROM ZERO.
Output: On the screen the constants will be printed out and a file mag_coda.out will contain pairs of
values m and (log(coda) + dist_coff*dist), which can be used to plot the distance corrected coda
relation. If results from the 3D is to be plotted, dist_coff must be calculated as dist_coff=B/A, put into
mag.par and mag run again.
A typical coda magnitude relation is Mc = 2.0*log(coda) + 0.0035*dist – 0.87 (Lee et al, 1972.)
202
Local magnitude Ml:
The local magnitude scale is calculated by determining an amplitude attenuation scale using
amplitudes and distances in CAT file. The parameters in the Ml magnitude scale are computed in two
ways: (1) for every event individually, parameters are determined as averages of all events; (2)
simultaneous inversion for all events, also inverting for site factors.
(1) For each event (only type L and R are used) a,b,c are calculated if at least 3 stations are available
using least squares regression as follows:
log(amp) = a * log(dist) + b * dist + c
(2) The simultaneous inversion for all events is done using a singular value decomposition algorithm.
The inversion is done as in the equation above, but in addition the inversion is also done for the site
terms. The parameter c is fixed to 2. The output is written to the file mag_ml_inv.out. The magnitudes
given in the output file are based on c=2.
Note: The simultaneous inversion has not been tested much.
The relation above can be derived from the standard geometrical spreading and attenuation relations:
amp = (dist ** a) * exp(pi * f * dist/(v*q ))
where f is the frequency, v is the velocity and q = q0*f**qalpha. The relation can be rewritten
log(amp) = a * log(dist) + (pi * f * dist) /(v * q0*f**qalpha * 2.3)
Since qalpha often is close to 1.0, the relation can be simplified to the frequency independent relation:
log(amp) = a * log(dist) + (pi * dist) /(v * q0 * 2.3)
If body wave spreading is assumed (a=1), q0=100 and v=3.5 km/sec, the relation is
log(amp) = 1.0 * log(dist) + 0.004 * dist
which is comparable to the relation shown below for California.
Similarly to the coda relation, a 2D relation is also calculated
log(amp) – b*dist = a * log(dist)
where b=dist_coff is fixed to the value given in the mag.par file (same parameter as used for coda).
This gives a more stable solution, however b = dist_coff must be determined by trial and error or fixed
using known values from e.g. q-studies.
The amplitudes are assumed to be ground displacements (in SEISAN they are ground displacements
highpass filtered at 1.25 Hz to resemble Wood Anderson seismograms, see MULPLT). The distance
ratio between stations with the maximum distance and minimum distance must be more than 3 for the
event to be selected for analysis.
It is assumed that a and b will be the same for all events, while c is different (magnitude dependent).
At the end, the average constants a and b are calculated of all values a and b which are not deviating
too much (a must be in the range 0 to -5, hardwired). Distance attenuation coefficients a and b are
supposed to be negative since amplitude decrease with distance. To get the local magnitude scale
Ml = log(amp) - a * log(dist) - b * dist - C
the constant C must be determined by fixing the magnitude at some reference distance like the
original Wood Anderson definition with Ml = 3 at dist= 100 km and amp=1/2200 mm = 454 nm
203
(assuming gain of the Wood Anderson seismograph to be 2200, Hutton and Boore, 1987). The
determination of a and b does not work well unless the observations are very good. The relation for
California is (Hutton and Boore, 1987)
Ml = log(amp) + 1.1 *log(dist) + 0.00189 * dist - 2.09
Output: On the screen the constants will be printed out and a file mag_amp.out will contain the values
of a, b and c.
2: Magnitude relations and/or spectral parameter relations
Linear regression (maximum likelihood) can be made between any two magnitudes and/or spectral
parameters on any of the header lines of an event in a CAT-file or a compact file. The user is
interactively prompted for the magnitude type and/or spectral parameters and agencies to compare. If
none is given, no magnitude comparison will be made. If several magnitudes/spectral parameters fit
the requirement, the last one is used. If e.g. the first header line has a BER Ml and the last header
line also has a BER Ml, the last one will be used. Maximum likelihood linear fitting is used. It is
assumed that both variables have normal and correlated errors. See subroutine maxlik.for in LIB for
more info.
The following parameter can be selected:
Any magnitude and agency
Seismic moment(log)
Stress drop (log)
Corner frequency (log)
Source radius(log)
Spectral decay
Omega zero level (log)
If any of the spectral parameters are selected, or moment magnitude is without agency, there will be
an additional question about which station and component. A blank return means the average will be
used. With these parameter selections, it is possible to compare spectral parameters from any two
channels, compare the average spectral parameter with the parameter from one channel etc.
Output: A plot will be shown on the screen with the observations and the least squares fit and the
values are also printed out on the screen. A file mag_mag.out contains the pairs of magnitudes used.
3: Magnitude conversions
If a relation between two magnitude scales is known, e.g. by using option 2 above, an output file can
be made with the converted magnitudes. The relation to use is specified in the mag.par file. Several
different input magnitude types and agencies can be used and the relation-agency used is given in a
priority list in the mag.par file, see example below. It is here shown that if a BER Mc is available, this
will be the first choice. If no BER Mc then BER Mb will be the next choice etc. The new magnitude will
have type X and agency NEW.
Output: The output file is mag_new.out and has the same format as the input file. On the header line,
the old magnitudes are removed and in the first magnitude position will be the converted magnitude
(NEW) while in the second magnitude position, the magnitude selected for conversion will be given.
The third magnitude position is blanked out.
Summary of output files:
mag_amp.out: Details each event for amplitude regression.
mag_coda.out: Magnitude vs coda, see text.
mag_mag.out:
Pairs of magnitudes used for regression.
mag_new.out:
Events with converted magnitudes only.
mag_newa.out: All events, both converted and non converted (due to no correct input
magnitude available).
204
mag_spec.out: Summary of normal header line, all associated magnitudes
and spectral parameters.
mag_ml_inv.out From Ml inversion
In DAT there is an example mag.par file.
Figure 36. Example of using the MAG program. Relation between NORSAR and Bergen local
magnitudes.
205
An example of the mag.par parameter file:
This file is for parameters for MAG and called MAG.PAR. The name must
be in lower case on Sun. The following shows the parameters which can be set.
The file can contain any lines in any order, only the lines with
recognized keywords and a non blank field under Par 1 will be read. The
comments have no importance. The text fields are left justified, the
real numbers can be anywhere within the 10 columns of the parameter.
-MAGAGA is the magnitude type and agency to use for the converted magnitude.
-MAGREL gives the magnitude and agency to use for conversion e.g. LBER and
the parameters 2 and 3 gives the relationship:
magnitude out = magnitude in * par2 + par3
The magnitude conversion uses one of the MAGREL relationships, where the
priority is in the same order as found in the list.
SCREENOUT can be (Y)ES or (N)O, indicates if a line is printed on the screen
for each event.
When doing a magnitude regression on coda or amplitude, BAD STATION indicates
stations not to be used. MAG_TYP_COF is the magnitude type and agency to correlate
coda readings with and the second parameter is the distance correction
term used when calculating a coda magnitude relation with a fixed distance
term. The same parameter is also used for amplitude regression.
The input file can be either Nordic or compact Nordic, however if coda or
amplitudes are to be used, it must be NORDIC.
KEYWORD........Comments...........................Par.1.....Par.2.....Par.3.....
BAD STATION
Station not used for mag relation BER
BAD STATION
Station not used for mag relation XXX
BAD STATION
Station not used for mag relation BER
MAGAGA
Magnitude type and agency, new mag XNEW
MAGREL
Mag type, agency, a,b for new mag CBER
1.0
2.0
MAGREL
Mag type, agency, a,b for new mag BBER
1.0
2.0
MAGREL
Mag type, agency, a,b for new mag SBER
1.0
2.0
MAGREL
Mag type, agency, a,b for new mag WBER
1.0
2.0
SCREENOUT
Header line printed each event
N
MAG_TYP_COF
Mag. type for corr., dist coff.
LNAO
0.002
6.26 Explosion filtering, EXFILTER
The program EXFILTER is used to identify probable explosions in a catalog of seismic events.
Man-made seismic events like quarry blasts, mining explosions and other explosions show a certain
distribution in time and space. Therefore the method of explosion identification here is based on
normalizing the time of day distribution of seismic event occurrence as a function of area. The
program works on the following principle: Areas where explosions occur are defined. If an event is
located in one of these areas, with a magnitude below a given maximum magnitude, with a depth less
than a given maximum depth, within a given time of day interval and within a given year interval, it is
identified and marked as probable explosion. The areas are defined by polygons of any shape. For
definition of the filter areas, a list of mine locations (with consideration of location accuracy), locations
of explosions and locations of event clusters (they might be clearly related to mine locations, but
others might indicate unknown explosion sites) can be used. The next step is to define the
parameters for each area to get a normal time of day distribution. They can be determined following
the steps:
1)
- get the time of day distribution of events (program CATSTAT)
- select a time window of probable explosions
- select events within time window of probable explosions
2)
- get the distribution of magnitudes of events within time window of probable explosions
(program BVALUE)
- select the maximum magnitude
3)
- test parameters defined with program EXFILTER for the defined area and adjust the
206
parameters if the time of day distribution is not normal.
For more details, see Ottemöller (1995).
The program uses a parameter file, EXFILTER.PAR which MUST be located in the DAT directory.
An example of the parameter file EXFILTER.PAR
------------- Parameter file for program EXFILTER ------------------------------------------------------------------------------------- This file must follow the following format rules: __________
1. Any number of comment lines
2. Any line with first character # defines the parameters of that area
3. Any line with first character * defines the parameters of exceptions within
an already defined area. I.e. you can define an area around a volcano,
and make an exception for that very small area.
4. In the first line of each parameter set the parameters are:
Maxdepth: Events with depth above this value are not explosions
Maxmag : Events with magnitude above this value are not explosions
Lhour
: Lower limit of time interval for explosion time
Hhour
: Upper limit of time interval for explosion time
Btime
: Lower limit of yearly interval
Etime
: Upper limit of yearly interval
N
: Number of latitude-longitude pairs in polygon
5. From the second line the lat-long pairs are given
-------------------------------------------------------------Area----------- Maxdepth Maxmag
Lhour
Hhour
Btime
Etime
N
# area 1
# area 2
# area 3
# area 4
100.0
64.0
70.50
68.50
3.6
23.8
20.0
28.50
11
66.70
71.20
67.00
15
23.80
26.00
29.50
197001
68.80
70.80
64.30
200012
20.20
29.50
29.50
12
69.80
70.00
64.00
18.30
27.50
34.00
100.0
67.90
3.8
17.80
15
68.80
20
20.20
197001
66.70
200012
23.80
4
66.50
21.00
100.0
67.90
3.5
17.80
0
68.80
1
20.20
197001
66.70
200012
23.80
4
66.50
21.00
100.0
68.50
4.0
28.50
2
68.50
17
36.00
197001
67.00
200012
36.00
4
67.00
29.50
The EXFILTER program searches for probable explosions using a catalog-file as input and marks
events that might be explosions with 'P' as Event ID in the output file exfilter.out.
Example of program run
<exfilter>
NUMBER OF AREAS:
55
FILENAME... ?
june.cat
************************************
Number of probable explosions found:
Output written in file: exfilter.out
************************************
90
Figure 37 show how the filter works for events in Scandinavia.
207
Figure 37. Explosion filtering
The top figure shows the filter areas used for Scandinavia. The bottom right figure shows the time
of day distribution for a 10 year Scandinavian catalog before filtering (made with CATSTAT) and
the figure bottom left shows the distribution after filtering.
208
6.27 Inversion of travel time data and joint hypocenter determination
6.27.1 VELEST
Introduction
The program VELEST is used to solve the coupled hypocenter velocity model problem for local
earthquakes. It performs a simultaneous inversion for hypocenters and velocity model. The inversion
is limited to first arriving phases. A detailed program description is given in the ‘VELEST USER’S
GUIDE’ (Kissling et al 1995). A recipe for preparing data and use of the inversion routine is presented
in ‘Initial reference models in local earthquake tomography’ (Kissling et al, 1994). The two documents
are available in one Postscript file in the INF directory, the filename is ‘velest.ps’. The derived model
can be used as an improved model for earthquake location or as a starting model for 3-D inversion.
For a fixed velocity model and constant station corrections, VELEST in simultaneous mode performs
the Joint-Hypocenter-Determination (JHD).
The original version of VELEST by Kissling is included in the Sun and Linux versions. A version modified
to compile VELEST under Windows has been provided by Freddy Aldersons (e-mail:
[email protected]). This Windows package is included in the file velest_pc_3.3.zip, which is
located in the SUP directory. The files have to be extracted to the PRO directory.
The implementation of VELEST to SEISAN is given by the program VELMENU.
VELMENU provides:
- automatic format conversion to VELEST
- generation of parameter files using the SEISAN system
- execution of VELEST
- conversion back to SEISAN format
After preparing a dataset of local earthquake data, VELMENU can be used to work with the VELEST
inversion routine. The first time VELMENU is used, all input files for the inversion with default parameters
can be generated. These parameter files then can be changed interactively and the inversion started with
VELMENU.
Running VELMENU
The program is started with ‘velmenu’. After entering the filename of the earthquake data the menu of
VELMENU appears.
Example of program run :
velmenu
File name of earthquake data in Nordic Format :
select.out
VELEST MENU
----------1.
2.
3.
4.
5.
6.
A.
B.
C.
Q.
Create VELEST command file (velest.cmn)
Edit/change VELEST command file (velest.cmn)
Create station select file (selstat.lis)
Edit/change station select file (selstat.lis)
Create model file
Edit/change model file
RUN VELEST
Edit inversion output file
Convert VELEST output to Nordic format and make diff-file
End
Choice ?
209
The complete inversion-process of earthquake data in SEISAN format, including all conversions and
preparation of parameter files, can be done with VELMENU. The steps are as follows:
1. Create VELEST command file (velest.cmn)
The user is asked for inversion or JHD and the appropriate parameters are set. The file ‘velest.cmn’ is
the central VELEST parameter file. To create it, the file of earthquake data is read to determine the
parameters that depend on the data. These are the number of events and the center of Cartesian
coordinate system, which is simply determined as the average of latitude and longitude of epicenter
locations. The remaining parameters are set to default values.
3. Create station select file (selstat.lis)
For the inversion, VELEST will use phases from stations with an epicentral distance below a
maximum distance only. In addition in VELMENU a selection of stations has to be used, only phases
from stations given in the file ‘selstat.lis’ will be used for inversion. When generating the file, the
maximum distance between station and hypocenter (parameter ‘dmax’) is read from ‘velest.cmn’ and
the input data are scanned to get a list of stations, which are within the limit to any epicenter. Editing
the file, stations can be added or removed. If all stations should be used for inversion, the parameter
‘dmax’ in the file ‘velest.cmn’ has to be increased.
Example of ‘selstat.lis’:
#
# STATION SELECT FILE FOR PROGRAM VELEST
#
# STATIONS WILL BE USED IN THE VELEST
#
INVERSION PROGRAM
#
# COMMENT LINES START WITH #
#
KONO
BER
NRA0
...
NOTE: The order of the stations is as given by the input data file. VELEST uses the last station as
reference station, so you may want to change the order.
5. Create model file
The input model file ‘model.inp’ is created using the model as given in the ‘STATION0.HYP’ file. The
‘STATION0.HYP’ file, if available, will be read from the local directory, otherwise from the DAT
directory. This might be a reasonable starting model, but of course the model file has to be changed.
A. RUN VELEST
Once the parameter files are created the inversion program can be started. The inversion study
requires interactive changing of parameters, which is supported by VELMENU. All input parameter
files can be changed from VELMENU. NOTE:, ‘... please accept the warning: To calculate a Minimum
1-D model a single or even a few VELEST runs are useless, as they normally do not provide any
information on the model space!’ (Kissling et al, 1995). The conversions and the inversion programs
are started as one process.
Before the inversion routine is started the station locations will be converted from the STATION0.HYP
file and the earthquake data in Nordic format will be converted to CNV (hypocenters and travel times)
format. NOTE: VELEST does not support 5 character station codes, therefore in the conversion to
VELEST, only the first 4 characters are used if the station code has 5 characters. In the conversion of
the earthquake data only phase readings from stations included in the station selection file will be
used. Arrivals with a time residual, given in the Nordic input file, above five seconds are omitted. Only
the first arriving phase of P and S respectively are used. The hypocenter location given by the
inversion will be determined by first arrivals only. The original data might include more phases like Pg,
Sg or Lg. Therefore, to get a comparison of hypocenter locations between the HYP location program
and VELEST, a Nordic file including the same data as the CNV file is created and the HYP program
run on this file before VELEST is started. The HYP program can be skipped by pressing ‘CTRL+C’,
while it is running.
210
The results of the inversion will be given in a text file that can be viewed within VELMENU. VELMENU
provides an option to convert the VELEST output file with final hypocenter locations in CNV format
back to Nordic format and to write a file that shows differences (velout.dif) in location and time
between the two location routines, HYP and VELEST, based on the same input data.
Example of ‘velout.dif’:
differences: inversion output - SEISAN input
first line input, second line output, third line difference
1984
1984
time:
8 5 0235 22.7 L
8 5 235 22.1 L
-0.6 latitude:
59.449
4.968 0.0
59.458
5.140 0.1
0.009 longitude:
BER 9 1.3 1.9CBER 3.7BBER
BER 9
0.172 depth:
0.1
1
1
1984
1984
time:
...
8 6 0406 26.6 L
8 6 4 6 26.6 L
0.0 latitude:
59.538
5.678 1.9 BER 9 1.8 1.7CBER 3.6BBER
59.505
5.655 6.2 BER 9
-0.033 longitude:
-0.023 depth:
4.3
1
1
Files will be overwritten, when VELMENU is started again. To work with different datasets or
parameter files it is recommended to work on different directories or to change the filenames, but note
that the default filenames (see below) will be used in VELMENU.
Problems: VELEST skips events without phase readings and therefore the number of events read by
VELEST will be different from the number given in the velest.cmn file. If this is the case VELEST
stops with the message STOP: ...end...(VELEST was running with the SINGLE-EVENT-OPTION).
Events without phase readings will not be listed in the invers.out file, and should be deleted from the
input file.
Joint-Hypocenter-Determination (JHD)
VELEST for fixed velocities and station corrections can be used as a JHD routine. For JHD,
VELMENU is used in the same way as described above for inversion. The only difference is that
when generating the ‘velest.cmn’ you have to choose JHD. The appropriate file for JHD is then
generated. Some parameters in the ‘velest.cmn’ file are different, compared to the inversion. These
are dmax, nsinv and invertratio, see ‘VELEST USER’S GUIDE’ for details. The output of final
hypocenter locations as described above can be converted to Nordic format, but note that the JHD
will be based on first arriving phases only.
Example of JHD:
File name of earthquake data in Nordic Format :
select.out
VELEST MENU
----------1. Create VELEST command file (velest.cmn)
...
Choice ?
1
Inversion or JHD (I / J) ?
j
Origin of cartesian coordinates :
number of events:
61
59.394
-5.707
Potential problem: We have seen cases where in JHD mode the depth parameter in the inversion is
sensitive to invertratio, which when set to 1. in JHD means that VELEST inverts for station correction
in every iteration. VELEST in this case worked better with an invertratio of larger than 1. See VELEST
manual for details.
List of files generated by VELMENU / VELEST
data.cnv
earthquake data in CNV format, VELEST input, generated by VELMENU
211
data.nor
fin_hyp.cnv
hyp.out
hypsum.out
input.mod
invers.out
nor1.dat e
print.out
selstat.lis
sta_cor.out
station.sta
velout.dif
velout.nor
earthquake data in Nordic format, HYP input, generated by
VELMENU
final hypocenter locations in CNV format, VELEST output
earthquake data in Nordic format, HYP output
HYP output file
input model, VELEST input, generated by VELMENU
documentation of inversion, VELEST output
earthquake data in Nordic format, VELMENU input
HYP output file
selection of stations, generated by VELMENU
station corrections, VELEST output
station locations, VELEST input, generated by VELMENU
difference file between HYP and VELEST location routine, VELMENU
output
final hypocentre locations, same as fin_hyp.cnv, in Nordic format,
VELMENU output
velest.cmn
VELEST control file, VELEST input, generated by VELMENU
6.27.3 NOR2DD
(contributed by Brian Baptie)
This program produces input for the Double Difference Program HYPODD (Waldhauser and
Ellsworth, 2001; Waldhauser 2000) from Nordic and STATION0.HYP files. The Nordic file has to be
given as argument when running the program, example:
nor2dd select.out
The files created are:
phase.dat – phase input data
station.dat – station coordinates
HYPODD is available from: http://geopubs.wr.usgs.gov/open-file/of01-113/
6.27.3 NOR2JHD_PUJOL
(contributed by Brian Baptie)
This program produces input for Pujol’s JHD program (Pujol, 2003) from Nordic and STATION0.HYP
files. The Nordic file has to be given as argument when running the program, example:
nor2jhd_pujol select.out
The files created are:
syn.times – phase input data
syn.vel – velocity model
stalist.elev – station list
The inversion tool is available from:
ftp://beagle.ceri.memphis.edu/pub/pujol/JHD
http://www.orfeus-eu.org/links/softwarelib.htm
6.28 Analysis of volcanic earthquakes
212
By Brian Baptie, BGS
Background
An important part of volcanic seismology and the seismic monitoring of active volcanoes is the correct
recognition of the different types of seismic event generated by the volcanic activity. The principal event
types include, volcano-tectonic events, caused by shear or tensile failure of rocks; long period events,
generated by a volumetric source in a liquid; hybrid events; and volcanic tremor.
To be of value for volcanic monitoring, any database of seismic events should include the type or
sub-class of individual events. This should allow users to then extract phase and location information over
a selected time period for individual event types and calculate hourly and daily rates of event. Simple
histogram plots showing the distribution of subclasses over time can be generated with the program
VOLCSTAT (Unix only).
Initialization
The user should create a text file in the DAT directory called VOLCANO.DEF (an example is already
in the directory). The format of this file will be one line of text (80A) followed by successive lines with
the format "i2,1x,6A,1X,40a" for number, sub-class code and description. An example of the file is
shown below. Comments are preceded with '!'.
Current volcano sub-classes:
1 vt
volcano-tectonic
2 hybrid hybrid
3 lp
long-period
4 tremor volcanic tremor
5 rf
rockfall
6 un
unknown
7
QUIT
! Comment line 80 characters
! Individual sub-class line
! The last line should contain
this entry
Registering volcanic sub-classes
Registration should be carried out as normal in MULPLT. From multi-trace mode enter 'p' to create a
new s-file for the event in the database. Answering 'LV' to the prompt for event type marks the event
as a local volcanic in the headers. If the VOLCANO.DEF file has been set up correctly in the DAT
directory, the information on the different sub-classes will be printed to the terminal. Choosing an
appropriate number selects the volcanic sub-class. The sub-class code is then entered in the s-file.
Modification of the s-file to incorporate volcanic sub-classes
The volcanic sub-class information is stored in a type 3 line within the s-file, e.g.
VOLC MAIN tremor
Columns 2:10 'VOLC MAIN'
Columns 12:17
a6
Column 80
'3'
3
: Header identifier
: Sub-class flag
: line type identifier
This allows the use of a maximum 6-character sub-class identifier, e.g. 'hybrid', which can then be
searched for and selected.
VOLCSTAT: Creating histogram plots
The program reads S-files directly from the database, and creates input files as well as a GMT script
to produce histogram plots of the distribution of subclasses over time. The user needs to enter
database name, start and end time, and the subclasses that are to be plotted. An example of a plot is
shown in Figure 38. The program supports 1-char subclass names only.
The following output files are created:
213
volcstat.batch - c-shell script to generate Postscript output using GMT
volcstat_counts.ps - Postscript output file
volcstat_counts_<type>.out - for each event, the Julian date is written out, one file per subclass
volcstat_daily_<type>.out - number of events per day, files written for each subclass
volcstat_counts_total.out - total event counts for each subclass
counts r
132
88
44
0
counts l
36
24
12
0
counts h
33
22
11
0
Figure 38. Bar diagrams showing distribution of events of different subclasses over time.
RSAM
1-minute RSAM data can be created with WAVETOOL.
Future Extensions:
It is intended that additional parameters can be included in the above structure to included routine
measurements of the volcanic earthquakes. For example, signal duration, peak amplitude and mean
frequency can be calculated for individual stations and included on additional type 3 lines with a
volcanic identifier. Parameters on each channel can then be averaged an inserted on the volcanic
header line.
214
The proposed format for these lines is as follows
column
2:5
7:10
12:15
19:20
22:29
31:32
34:41
43:44
46:53
55:56
58:65
66:79
80
format
a4
a4
a4
a2
g8.3
a2
g8.3
a2
g8.3
a2
g8.3
a1
description
'VOLC'
Volcanic identifier
station
component
'PA' field identifier
peak amplitude
'DU' field identifier
signal duration
'MF' field identifier
mean frequency
'SB' field identifier
signal bandwidth
blank
'3' line type identifier.
For example
VOLC
VOLC
VOLC
VOLC
MAIN
KTK1
KTK1
KTK1
tremor
S Z
PA .152E-06
S N
PA .167E-06
S E
PA .141E-06
DU
DU
DU
1.325 MF
1.997 MF
1.543 MF
2.472
2.067
1.998
3
3
3
3
This method of inclusion of volcanic parameters should allow for future flexibility such as incorporation
of an additional parameter fields in columns 66 to 79. Also the use of type 3 lines means that existing
software, such as the update program, are unaffected by these lines.
6.29 FK Analysis
The FK routines were provided by Tormod Kværna from NORSAR and implemented into SEISAN by
Andrius Pacesa.
Some basics
The FK-analysis, more strictly slowness analysis, is a standard tool in seismic array processing. It is
used to find the apparent velocity and back azimuth of an incoming wavefront. Apparent velocity can
be used to identify the type of wave (P, S, Lg and etc.) and the approximate distance to the source
can be determined for teleseimic events. Utilizing azimuth and distance to the source, one can define
the approximate location of the signal source.
A description of frequency-wavenumber analysis - “f-k analysis” - may be found in Capon (1969). This
method has been further developed to include wide-band analysis and maximum-likelihood estimation
techniques - see Kværna and Doornbos (1986).
The principle of slowness analysis is beamforming in the frequency domain for a number of different
slowness values and calculating the power for each beam. The beam power will be a maximum in
case the slowness of the beam coincides with the slowness of the wavefront crossing an array. So
the beam having the maximum power will indicate the slowness of the incoming signal.
Running the program
The FK program can be started directly with command ‘fk’ or from MULPLT. The program expects that the
file ‘waveform.out’ with the seismic traces as input data, is available in the current directory. If the program
is invoked directly, this file has to be created before using mulplt, selecting a window and creating the
‘waveform.out’ file.
In general it is more useful, to start the FK program from MULPLT since the input file needs to be created
by mulplt. The result of the fk analysis can be saved to the S-file.
The steps are:
-
start MULPLT
select channels and a time window
215
-
use option fk to start FK program (this option creates file ‘waveform.out’ and starts FK program),
accept maximum or pick value with mouse
The options in FK are:
R-Redo: Repeat fk analysis with different parameters
M-Mouse: ‘m’ or mouse click to pick values different from maximum
S-Save and quit: save picked value to file and quit
Q-Quit: quit
-
use option ‘save and quit’ to save your result, so that it can be used by MULPLT
back in MULPLT: pick phase on the first trace used, to store back azimuth and apparent velocity in
the S-file
in case of teleseismic events, the apparent velocity can be used for location, the fk analysis has to be
done on the P phase
Note: The FK program only works by default with station file ‘STATION0.HYP’. If coordinates are in e.g.
STATIONt.HYP, the user will be asked to specify another station file letter, in this case ‘t’.
Example
Input:
Enter lower frequency,<ENTER> for default value 1 Hz
Enter higher frequency,<ENTER> for default value 5 Hz
Enter max slowness, <ENTER> for default value 0.4 s/km
Enter No of grid points, <ENTER> for default value 51
Do you want to plot level numbers, N/Y, <ENTER>=Y
It will take some time ...
APPARENT VELOCITY 10.26
NORM. POWER MAX
0.99
MAX X SLOWNESS
0.06
LOW FREQUENCY
0.94
QUALITY (1=best, 4=worst)
Plot file is fk.eps
AZIMUTH
POWER MAX IN dB
MAX Y SLOWNESS
HIGH FREQUENCY
1
140
72.44
-0.07
5.00
Example of output file fk.out:
DATE, TIME AND WINDOW LENGTH
99 256 9 13 12 0 33.843
7.62
APPARENT VELOCITY 10.26
AZIMUTH
140
NORM. POWER MAX
0.99
POWER MAX IN dB 72.44
MAX X SLOWNESS
0.06
MAX Y SLOWNESS
-0.07
LOW FREQUENCY
0.94
HIGH FREQUENCY
5.00
QUALITY (1=best, 4=worst) 1
VALUES TO SAVE
APPARENT VELOCITY 10.26
AZIMUTH
140
Station
Long
Lat
Elev
Xcoord
Ycoord Zcoord
NRA0 S Z
11.54150 60.73533
300
0
0
300
NRA1 S Z
11.54233 60.73650
291
45
129
291
NRA2 S Z
11.54333 60.73433
311
99
-111
311
…
NRD4 S Z
11.56333 60.72717
379
1186
-907
379
NRD5 S Z
11.54750 60.72217
348
326
-1463
348
NRD6 S Z
11.52883 60.72334
352
-688
-1334
352
NRD7 S Z
11.51617 60.73017
337
-1377
-574
337
NRD8 S Z
11.51667 60.73900
301
-1349
407
301
XSLOW
YSLOW POWER
0.40
0.40
0.05
0.38
0.40
0.05
0.37
0.40
0.05
0.35
0.40
0.05
...
216
Figure 39. The FK program can be started from MULPLT. The traces shown were selected and
used as input to the FK program. The result of the FK analysis is shown in Figure 40. The
event shown here is part of the testdata set.
Figure 40. Output from the FK program. Contours and values are the normalized maximum
power.
217
6.30 Surface wave analysis (SUN)
The programs by Robert Herrmann (Herrmann, 1996) to estimate the shear wave velocity of the earth
by inversion of surface wave group velocities are distributed with Seisan. The programs are part of
Herrmann’s package ‘Computer Programs in Seismology’ . For more information check:
http://www.eas.slu.edu/People/RBHerrmann/ComputerPrograms.html
The programs have been implemented with SEISAN by Vunganai Midzi, who has written a guide on how
to use the programs. This document is given as the Word file ‘surface.doc’ in the INF directory. Also see
section 6.2.11 for details on output files that can be generated with MULPLT. The programs are included
in the tar file ‘surface.tar’ in the PRO directory, but are not installed as part of the standard installation. If
you wish to use the programs, you need to extract the programs:
cd <SEISAN_TOP>/PRO
tar xvf surface.tar
You can recompile the programs with the commands:
make -f Makefile.sur clean
make -f Makefile.sur
6.31 Instrument response
In SEISAN the instrument response can be stored as pairs of frequency, amplitude and phase or as
poles and zeros. The formats that can be used include GSE2, SEISAN and SEED. The SEISAN and
GSE response formats are described in Appendix 3. For a detailed description of the GSE format, the
reader is referred to GSETT-3 (1997). The program RESP creates response files in SEISAN and
GSE format. SEED format response files can be extracted from a SEED volume.
6.31.1 Create instrument response files, RESP
Introduction
The purpose of this program is to (1) Make Seisan or GSE2 response files, (2) Provide the engineer
maintaining seismic instrumentation with a practical tool for calculating and checking response
functions of the most common elements of a seismic system. The program can calculate response
functions of velocity transducers, accelerometers, filters and amplifiers, input poles and zeros or
tabulated values and multiply the combinations together to get complete system response functions.
The program produces a table with the response function and a simple graphical expression of the
response curve. For the purpose of checking measured values, a file with these values can be used
as input and will be plotted together with the theoretical values. The program can calculate,
acceleration, velocity or displacement response. Program PR_ RESP can make a table of many
response files.
The instrument response
The seismic recording system can consist of seismic sensor, analog-digital converter, amplifier and
filters. For a detailed discussion the user is referred to Scherbaum (1996). The combined response
can be given in the frequency domain as frequency response function or in the Laplace domain as
transfer function. The frequency response is given in pairs of frequency amplitude phase (FAP), while
the transfer function is given as poles and zeros (PAZ). The combined frequency response is
obtained through multiplication of the response from the individual components, while the transfer
function is obtained by combing the PAZ from the components. Amplifiers and accelerometers are
specified simply by a constant gain. Filters are assumed to be Butterworth. RESP can be used to
218
write finite impulse response (FIR) coefficients (Scherbaum, 1996) that are used as anti-alias filters in
most modern digitizers if GSE is used as output format. SEISAN has no capability to read the FIR
filters or to correct for them. However, the FIR filters are part of a full description of the instrument
response and should be at least included for information if possible.
The seismometer is assumed to have the following velocity frequency response:
T (ω ) =
ω2
ω 02 − ω 2 + i 2ωω 0 h
Which corresponds to the transfer function:
T (s) =
− s2
ω 02 + s 2 + 2sω 0 h
where s=iω, ω is the angular frequency 2πf in Hz, ω0 the resonance frequency of the seismometer,
i = − 1 and h the damping (normally around 0.7).
NOTE: In the equation for the frequency response, the sign "+ 2*i*h.." was "-" before March 2000, so
old parameter files may have to be regenerated. The sign depends on the definition of the signs in the
Fourier transform and therefore may be different in different text books. It may even be wrong
although it looks right, if a wrong Ansatz is done. Due to the wrong sign, the FAP values in the
SEISAN response files were wrong, however the programs use the constants given in the files and
the correct response is generated. If you have the instrument constants in your old response files and
not just FAP, the old response files can be used.
The transformation from displacement to velocity or back is done by multiplying with i*ω.
In addition to or instead of using the equation above, values can also be entered as discrete values or
as poles and zeros.
The SEISAN response function is calculated for 60 frequencies between 0.01 and 100 Hz and the
steps between the frequencies are approximately logarithmic. The response function is normalised at
1.0 Hz ( see Table 1) and the gain at 1.0 Hz is given separately.
NOTE: ALL UNITS ARE IN METERS, SECONDS OR G (9.8ms-2)
NOTE: It seems that although the GSE format is clearly defined, there has been different
interpretations. This has also led to changes in SEISAN since the GSE response was introduced with
SEISAN. For more details, see Appendix 3.
Which format to use
SEISAN, since version 7.1, supports the GSE2 calibration format in addition to the SEISAN response
file format. We recommend that you use the GSE2 format, since it presents one of the most widely
used calibration formats. Storage of the response in terms of PAZ is recommended over FAP, since
the PAZ representation describes the continuous transfer function. You may continue using existing
SEISAN response files and add new files in GSE2 format, or replace the old SEISAN response files
with new GSE2 files.
How to run the program
The program has quite a few options, which easily may lead to confusion. Before you start you should
know which format you want to use (GSE2 or SEISAN) and whether you want to describe the
response in terms of FAP or PAZ. The recommended choice is to use GSE2 and PAZ.
219
Type RESP to start the program. You will then get a series of questions as indicated below in upper
case letters. All input is format free. A sample run is shown below.
CHOSE OUTPUT FORMAT:
0: NO OUTPUT FILE
1: SEISAN FAP
2: SEISAN PAZ
3: GSE2 FAP
4: GSE2 PAZ
Answer with 0-4, options 1-4 will create respective response files in selected format, option 0 will only
calculate and show the response on the screen. SEISAN PAZ can only be used if number of poles
and number of zeros are less than 38. If more are input, a table will be generated automatically in
FAP format.
TYPE OF SENSOR:
1: NONE
2: SEISMOMETER
3: ACCELEROMETER
Answer with 1, 2 or 3. Number 1 is used when only calculation of filters or amplifiers are desired, 2 is
a standard velocity transducer and 3 a standard accelerometer. If a seismic sensor is used, you will
get additional questions on the constants of the sensor. If a seismometer is chosen, the following
questions must be answered:
SEISMOMETER NATURAL PERIOD ?
This is measured in seconds. For most short period systems the value would be 1.0 second.
SEISMOMETER DAMPING RATIO ?
The damping ratio should ideally be 0.7. This depends on the damping resistance.
For both the seismometer and accelerometer, the following question is given:
SENSOR LOADED GENERATOR CONSTANT (V/M/S OR V/G) ?
This is the generator constant of the sensor in terms of volt per unit of of ground motion
(meter/second or g). It is important to note that this is the loaded constant, which means the effective
output of the sensor taking into account amplifier input and damping resistances.
Now comes questions about amplifier, filter and recording unit.
RECORDING MEDIA GAIN (COUNT/V OR M/V) ?
If you have a recording media, the gain can be given here, otherwise just enter 1.0
If the output format is GSE, the response is always calculated in displacement units, while for
SEISAN output and seismometer or accelerometer, the following options appear:
TYPE OF RESPONSE:
1: DISPLACEMENT
2: VELOCITY
3: ACCELERATION
Normally for a seismometer, one wants to calculate the displacement response and for an
accelerometer, the acceleration response. However it might sometimes be interesting to look at e.g.
the velocity response for a seismometer (after all, the seismometer is normally a velocity transducer
!!). Enter the appropriate number.
AMPLIFIER GAIN (DB) ?
220
This is the amplifier gain in dB. Since this question is only asked once, this gain must include gain of
all units except the recorder (asked below). This could e.g. include gain of the VCO system.
NUMBER OF FILTERS (0-10), RETURN FOR NONE ?
Up to 10 filters can be specified. If you answer 0, no filters are used and no more questions on filters
will appear. Otherwise one line of input must be given for each filter as follows:
FREQUENCY AND NUMBER OF POLES FOR EACH FILTER,
POLES NEGATIVE FOR HIGH PASS
Each line requires two numbers, the corner frequency of the filter and the number of poles. A high
pass filter is given by letting number of poles be negative. It is not always easy to know whether a
filter is e.g. one 2 pole or two 1 pole filters, the user needs to experiment with this.
FILE NAME FOR FILE WITH POLES AND ZEROS, RETURN FOR NO FILE
Here a file with poles and zeros can be entered. If seismometer constants have been chosen above,
the values calculated with poles and zeros are multiplied with the values previously calculated. The
free format file contains:
1. line: NP: Number of poles, NZ: Number of zeros, Norm: Normalization constant
Following NP lines contain one pair each of real and imaginary poles
Following NZ lines contain one pair each of real and imaginary zeros
The next 2 options are only shown if the output file is selected to be FAP:
FILE NAME FOR TABULATED VALUES, RETURN FOR NO FILE
Here a file with tabulated values are entered. If seismometer constants or poles and zeros have been
chosen above, the tabulated values will be interpolated and multiplied with the values previously
calculated for from above. The free format file contains:
1. line: N: Number of tabulated values, Norm: Normalization constant
Following N lines contain one each frequancy, amplitude and phase(deg)
GIVE FILE NAME FOR MEASURED VALUES, RETURN FOR NONE
Give file name for measured values. In most cases you have none so just make a return. The format
of the input file is as follows:
frequency, amplitude, phase
frequency, amplitude, phase
etc.
e.g.
0.2,0.7,200
0.7,0.8,100
10.0,0.1,33
The file has no blank lines and can contain up to 60 data sets. It is important to note that the
amplitude values should be NORMALIZED at 1.0 hz.
Now there is no more input to the response parameters, and the output is:
GAIN FACTOR AT 1.0 HZ: 12345.6
221
This is the gain of the system at 1.0 hz and is also the value for normalizing the response curve, that
is, all calculated values are divided by this number. There is no unit for gain of an amplifier and for
displacement response using a seismometer and drum recording. If the recording is digital, the unit
would be counts/meter and for a velocity response counts/meter/second etc. If a file with poles and
zeros is used without any other information, the normalization constant must have the unit of count/m,
similar for the tabulated input.
Further output is given in a file called resp.out, see Table 1 for an example.
The response curves (amplitude and phase) are now printed/plotted on the screen. First comes the
amplitude response (amplitude in db versus log frequency). By pushing return, the phase response is
shown (phase shift (deg) versus log frequency). After the plots, the SEISAN calibration file can
optionally be made, follow instructions, see example below. The response file MUST be calculated for
the displacement response, and all calculation in SEISAN assume that response is calculated in
counts/m.
After the SEISAN response file is made, the current parameters will be displayed and one or several
can be changed without entering all again. Like if the gain has changed at a certain date, only change
date and gain. This feature (new in SEISAN7.2) has been put in to be able to quickly make many
similar response files, like when all files have to be put in for a network.
Comments to data for response files
Station and channel codes
It is important that the station and channel codes are made exactly as they appear in the waveform
files. If not, SEISAN is not able to identify the channel.
Date
The date given here corresponds to the date from which the calibration information is valid. The
SEISAN system will always look for the most recent calibration file relative to the date of the
earthquake.
Latitude, longitude and elevation
These data are for information only, it is not used anywhere in SEISAN, so it does not have to be
entered, however there is room for it in the SEISAN waveform file headers.
Comment
No information used by the system.
Plot
After the response file has been written out, a plot is made with PRESP of the file. There will also be a
plotfile, presp.eps, which can be sent to the printer.
The response file can store the response in different ways:
1: Parameters used for calculating the response: Generator constant, filters etc. In addition, the
response (amplitude and phase) at 30 frequencies are listed. In this case the response is calculated
from the parameters.
2: Incomplete set of parameters or no parameters and the response at 30 frequencies. In this case
the response is calculated by interpolation of the 30 values.
3: Poles and zeros: No discrete values are given and the response is calculated directly from the
poles and zeros.
See also Appendix 2 for the SEISAN waveform file format and section 4.6.
IMPORTANT: PUT RESPONSE FILE IN CAL DIRECTORY OR ONE OF ITS STATION
SUBDIRECTORIES
222
Example of running the program:
oxygen-(larso)23: resp
RESP - PROGRAM TO CREATE RESPONSE FILES IN SEISAN
OR GSE FORMAT. THE RESPONSE CAN BE CREATED
AS POLES AND ZEROS (PAZ) OR FREQUENCY
AMPLITUDE AND PHASE (FAP). THE SAME
TRANSFER FUNCTION AND FILTERS ARE USED
IN BOTH CASES.
CHOSE OUTPUT FORMAT: 0:
1:
2:
3:
4:
NO OUTPUT FILE
SEISAN FAP
SEISAN PAZ
GSE2 FAP
GSE2 PAZ
4
TYPE OF SENSOR:
1: NONE
2: SEISMOMETER
3: ACCELEROMETER
2
SEISMOMETER NATURAL PERIOD ?
1
SEISMOMETER DAMPING RATIO ?
.7
SENSOR LOADED GENERATOR CONSTANT (V/M/S OR V/G) ?
300
RECORDING MEDIA GAIN (COUNT/V OR M/V) ?
2048
AMPLIFIER GAIN (DB) ?
40
NUMBER OF FILTERS (0-10), RETURN FOR NONE ?
1
FREQUENCY AND NUMBER OF POLES FOR EACH FILTER,
POLES NEGATIVE FOR HIGH PASS
10 2
FILE NAME FOR FILE WITH POLES AND ZEROS, RETURN FOR NO FILE
FILE NAME FOR MEASURED VALUES, RETURN FOR NO FILE
223
AMPLITUDE RESPONSE
SEISMOMETER
DISPLACEMENT
--------------------------------------------------------------I.
.
.
.
. ++++++++++
.I
I
++++++
++++++
I
I.
.
.
. +++
.
.
+++I
I
+ ++
I
I. . . . . . . . . . . . . .++ . . . . . . . . . . . . . . . .I
I
++
I
I.
.
. ++
.
.
.
.I
I
++
I
I.
.
++
.
.
.
.I
I. . . . . . . . .++ . . . . . . . . . . . . . . . . . . . . .I
I.
.
++
.
.
.
.
.I
I
++
I
I.
.++
.
.
.
.
.I
I
++
I
I. . . .++ . . . . . . . . . . . . . . . . . . . . . . . . . .I
I
++
I
I. ++
.
.
.
.
.
.I
I ++
I
I+
.
.
.
.
.
.I
--------------------------------------------------------------FREQ
0.01
0.03
0.14
0.71
3.68
19.19
100.00
GAIN FACTOR AT 1 HZ: 0.276E+09
RETURN FOR PHASE RESPONSE
AMPL
9.88
3.66
1.36
0.504
0.187
0.694E-01
0.257E-01
0.955E-02
0.354E-02
0.131E-02
0.488E-03
0.181E-03
0.671E-04
0.249E-04
0.924E-05
0.343E-05
0.127E-05
0.472E-06
0.175E-06
PHASE RESPONSE
SEISMOMETER
DISPLACEMENT
--------------------------------------------------------------I.
.
.
. +
.
.
.I
I
++
I
I.
.
.
.
+
.
.
.I
I
++
I
I. . . . . . . . . . . . . . . . . . . .++ . . . . . . . . . .I
I
++
I
I.
.
.
.
. ++
.
.I
I
+
I
I.
.
.
.
.
+
.
.I
I. . . . . . . . . . . . . . . . . . . . . . . .++ . . . . . .I
I.
.
.
.
.
+.
.I
I
+++
I
I.
.
.
.
.
. ++++
.I
I++++++++++++++
++++I
I. . . . . . . ++++++++++. . . . . . . . . . . . . . . . . . .I
I
+++
I
I.
.
.
++ .
.
.
.I
I
+ +
I
I.
.
.
. +
.
.
.I
--------------------------------------------------------------FREQ
0.01
0.03
0.14
0.71
3.68
19.19
100.00
GSE RESPONSE FILE (Y/N=default)?y
Enter station code. e.g. BERGE, max 5 chars
TEST
Enter component (4 chars) e.g. SL Z
First character is type, must be one of the following:
S: Short period, L: Long period
B: Broad band,
A: Accelerometer
Last character must be Z,N or E
Character 2 and 3 can be anything
S Z
Enter date as YYYYMMDDHHMMSS, at least up to the day (e.g. 19880123):20000101
Latitude (Up to 4 decimal places and - for south), return for none:
Longitude (Up to 4 decimal places and - for west), return for none:
Enter elevation in meters (integer), return for none:
Enter comments, one line. e.g. amp type, sensor type return for none
PHAS DEG
163.
144.
125.
106.
87.1
68.0
49.0
29.9
10.8
-8.21
-27.3
-46.3
-65.4
-84.4
-103.
-123.
-142.
-161.
-180.
Response file name is: TEST_S__Z.2000-01-01-0000_GSE
RESPONSE CURVE IS IN FILE resp.out
Examples of response files is given in Appendix 3
224
Table 1 Example of resp.out:
SENSOR TYPE: SEISMOMETER
RESPONSE: DISPLACEMENT
SEISMOMETER PERIOD=
1.00000
GENERATOR CONSTANT=
300.000
DAMPING RATIO
=
0.700000
AMPLIFIER GAIN(DB)=
40.0000
RECORDING GAIN=
2048.00
FILTER CONSTANTS
F= 10.00
POLES= 2
GAIN AT 1 HZ=
2.75728E+08
F=
F=
F=
F=
F=
F=
F=
…
F=
F=
F=
F=
F=
F=
F=
F=
F=
F=
F=
F=
F=
F=
0.0050
0.0059
0.0070
0.0083
0.0098
0.0120
0.0140
T=
T=
T=
T=
T=
T=
T=
200.00
169.49
142.86
120.48
102.04
83.33
71.43
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
0.000000
0.000000
0.000000
0.000001
0.000001
0.000002
0.000004
AMPDB=-135.1
AMPDB=-130.8
AMPDB=-126.4
AMPDB=-121.9
AMPDB=-117.6
AMPDB=-112.3
AMPDB=-108.3
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
-90.4
-90.5
-90.6
-90.7
-90.9
-91.1
-91.2
0.3900
0.4600
0.5500
0.6500
0.7700
0.9100
1.1000
1.3000
1.5000
1.8000
2.1000
2.5000
2.9000
3.5000
T=
T=
T=
T=
T=
T=
T=
T=
T=
T=
T=
T=
T=
T=
2.56
2.17
1.82
1.54
1.30
1.10
0.91
0.77
0.67
0.56
0.48
0.40
0.34
0.29
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
AMP=
0.082352
0.133868
0.224204
0.356744
0.554684
0.820676
1.198877
1.580098
1.933016
2.420457
2.877005
3.460298
4.027073
4.855642
AMPDB= -21.7
AMPDB= -17.5
AMPDB= -13.0
AMPDB= -9.0
AMPDB= -5.1
AMPDB= -1.7
AMPDB=
1.6
AMPDB=
4.0
AMPDB=
5.7
AMPDB=
7.7
AMPDB=
9.2
AMPDB= 10.8
AMPDB= 12.1
AMPDB= 13.7
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
PHAS=
-125.9
-133.0
-142.3
-152.9
-165.6
-179.7
163.3
148.6
137.0
123.6
113.5
103.0
94.6
84.1
FOR MORE DETAILS ON HOW TO UNDERSTAND GSE AND SEED RESPONSE PARAMETERS,
SEE Havskov and Gerardo (2001), chapter 6. The book is either with the SEISAN CD or on our home
page (File instrument.pdf)
6.31.2 SEED response
SEISAN can directly read SEED responses, which is poles and zeros, given as velocity response and
transfer function types A (Laplace Transform in Rad/sec) and B (Analogue in 1/sec). Storage of response
in one of these is the most common. The resp files can be created with rdseed from a full or dataless
SEED volume (rdseed –R –f seed_volume). RDSEED creates files with the pattern
‘RESP.NC.STAT.LC.CHC’, where NC=network code, STAT=station code, LC=location code (not used by
SEISAN) and CHC=channel code. The resp files need to be stored in the CAL directory and SEISAN will
find the correct file. The resp file can contain response information from several time intervals. SEISAN
uses the date and time of the waveform data to find the corresponding instrument response.
SEED response files are given in stages, for example seismometer, digitizer and FIR filters are stored
as individual stages. The overall response is made by combining all the stages. SEISAN uses the
following blockets from the SEED resp file (for more details see IRIS Consortium 1993):
B052F22 - start date
B052F23 - end date
B053F03 - transfer function type, A=Laplace Transform (Rad/sec), B=Analog (1/sec)
B053F07 - A0 normalization factor (A0 is checked against poles and zeros at normalization frequency
and changed if not correct). The product of poles and zeros at the normalization frequency and A0
gives 1.
B053F08 - Normalization frequency
B053F10-13 - zeros, if transfer function type is B, normalization factor A0 is changed to (A0)/(2 pi) for
225
each zero
B053F15-18 - poles, if transfer function type is B, normalization factor A0 is changed to (A0)*(2 pi) for
each pole
B058F04 - gain
The overall gain factor is given by the product of normalization factors and gain factors from all
stages. One zero is added to convert to displacement response. It is assumed that input units are V/m
and output units are counts, no checks are done on input and output units.
6.31.3 SEED response to GSE, SEEDRESP2GSE
SEEDRESP2GSE converts SEED resp files as written out by rdseed to GSE format. The program
only supports poles and zeroes and transfer function type Laplace Transform. The program asks for
station and component names and a time. This is because the resp file could have data from several
channels and cover several time intervals with different instrument configuration.
6.32 Macroseismic data in SEISAN
Macrocosmic information in SEISAN are of 2 kinds. The summary information with maximum
intensity, macroseismic epicenter etc has a special line in the S-file (see Appendix1) and the SELECT
program can search for felt events. In addition, from SEISAN, version 8.1, a format has been defined
to store macroseismic observations used to create e.g. maps with isoseismals. The observations files
are stored in a local ISO format. For a format description and suggestion for file names, see Appendix
1. There are currently no SEISAN programs that generates these files so they have to be made
manually from the observations. The files in ISO are linked to the events as given by the S-file data
base structure in the same way as waveform files are linked to the events. Thus information about
event source parameters and felt information is available together.
Program EPIMAP can plot the new files. The requirement is that the filename ends with .mac. The
intensities will be plotted as number on the map. A new Unix program can also be used with the data.
Program MACROMAP can use the macroseismic observation file as input to create a map of the
observations using GMT. The program generates a GMT script file, macromap.gmt, which then is
executed from within the program to create a PostScript output file, macromap.ps. This file is then
displayed, from within the program, with Unix command gv (GhostView).
The input can also be from a file made with macroquest (web based interactive program for input from
the public, to be distributed with SEISAN CD). In addition to making the plot, a conversion from the
web format to SEISAN format is made (output file macromap.out). This option requires an input file
with postal codes in order to get location of the observations.
MACROMAP can be executed directly or from eev. When executed directly from the prompt line, the
options are:
-macroinput
-placename
-postfile
file with macroseismic observations, SEISAN format, abs path or in ISO
optional additional file with place names, to be shown on map, abs path or in DAT,
epimap format is used
optional file with postal code, abs path, used with web option
If used with eev, the place name file must have name place_names.macro
An example of the postal code file is
1300
1310
1312
1314
59.880
59.880
59.870
59.900
10.520
10.550
10.510
10.460
SANDVIKA
BLOMMENHOLM
SLEPENDEN
SKUI
226
The content is postal code, latitude, longitude and location. The format is a10,ff10.3,2x,a30. the postal
code does not have to be a number, but can be any string.
6.33 Correlation of waveform signals, CORR
The cross-correlation function provides a measure of similarity between signals. In seismology crosscorrelation can be used to measure the similarity of waveforms between seismic events and in case
of similarity to determine relative arrival times of a seismic wave between two events. Waveform
similarity is caused by proximity in hypocenter location and similarity in focal mechanism between two
events. The program CORR can be used to either identify groups of similar events within a data set or
to determine relative arrival times. It also can be used to determine correlation of a master event with
continuous data and extract event based files. The cross-correlation function of signals x and y is
computed in the time domain as:
n
rxy (i ) =
∑x
j =1
n
∑x
j
2
j
j =1
y( j +i −1)
n
∑y
2
( j +i −1)
j =1
Group identification mode
The first step when running in this mode is to compute cross-correlation between the same stations
n
for all pairs of events. The resulting cross-correlation matrix (for each station containing
∑ (i − 1)
i
values, where n is the number of events) is then used to identify groups of similar events. This is done
in a rather simple approach. Groups of events for a particular station are formed starting with one
event and adding to the group events that have a cross-correlation above the threshold, and that
have no higher correlation with any other event. In case some event has the highest correlation with
an event that is grouped already, it is added to the same group, although correlation with the first
event in the group was not the highest. The grouping is given in corr.out. Alternatively the crosscorrelation matrix can be used with cluster analysis algorithm outside SEISAN (cluster software not
part of SEISAN). Visual inspection of the waveforms is highly recommended to confirm the correlation
results.
Phase detection mode
Phase arrivals of similar events can be determined accurately through cross-correlation. P and S
arrivals can be determined independently. The procedure starts by selecting and picking phases for a
master event that is representative for the group of events. Analysis of this event needs to be done
accurately as it is the basis for the subsequent analysis. Phases of the other events in the group can
be determined through cross-correlation of either the complete trace or a selected phase window with
the master event. To pre-select a time window, manual identification of the phase for subsequent
events is required prior to running CORR. This may be necessary for example if the waveform file
contains several events. The phase arrival time is given by the maximum of the cross-correlation
function that needs to exceed a minimum threshold. The arrival time is written out as absolute time.
Filtering is applied to the signals if selected for both master and subsequent events, this may be
necessary especially when dealing with events of different size. The filtering introduces a phase shift,
which is applied to both signals. However, the absolute phase arrival for the subsequent event is
consistent with the master event picked time.
Continuous mode
The main objective of running CORR in this mode is to identify a master waveform signal in a
227
continuous data stream, given by waveform data files. The times when correlation is higher than the
selected threshold level are written out, and can be visualized by splitting the corr.out file and using
EEV and MULPLT. In addition, it is possible to cut out individual event files (see CONTINUOUS
EXTRACT parameter).
Input to the program is given through the file corr.inp. A sample file is given in the DAT directory; the
data used in the example are part of the test data set (TEST database 2003/06, see training
document). The program is run by command corr in the same directory as corr.inp and the s-files. The
waveform files can be in any SEISAN standard place. All standard waveform formats can be used.
The parameters in corr.inp are as follows:.
SFILE MASTER: sfile name of master event, remove or comment out this parameter to run program
in group identification mode to determine cross-correlation matrix between all events and
identify group of similar events
SFILE EVENT: sfile name of events that will be either cross-correlated among themselves, or
compared to the master event, there can be several of these.
MIN CORR: minimum correlation required either for grouping or phase identification
INTERACTIVE: set to 1. for interactive use where graphics are displayed on the screen, which is
useful for testing; or set to 0. for non-interactive run
FILTER: this parameter allows to enable (1.) or disable (0.) filtering as defined for each channel with
parameter line STATION
PRE SIGNAL: duration of signal to include before phase arrival if used
EVENT SELCRIT: cross-correlation with the master signal can be computed either for the complete
trace of the subsequent event (0.) or the same part of the signal (either P or S) as for the
master event (1.) including the pre signal part and of the same duration as defined in
STATION line
N DOUBLE SRATE: if sampling rate is to be increased give factor n to double sampling rate n times,
0. for none, this makes it possible to get phase reading with resolution greater than
sampling interval
CC MATRIX WINDOW: time interval in seconds to use in computation of cross-correlation matrix
instead of the duration given for each station, if different from 0.; full trace is used if EVENT
SELCRIT is set to 0.
START LATITUDE and START LONGITUDE: these can be set to write values to corr.out, which is in
Nordic format and can be used as input for location programs; this can be used if all events
analysed belong to one cluster and the same starting location is to be used for all of them;
set to 999. to disable
FIX DEPTH: allows to fix depth to given value in corr.out, which can be useful if data is input to
location program; set to 999. to disable
WAVE CORR OUT: CORR can write out cross-correlation function and input traces to waveform
output files of the selected duration, to disable set this parameter to 0., 1. for full data or 2.
for reduced data, where 1 for data > MIN CORRELATION, otherwise 0
WAVENAME OUT: CORR writes out waveform filenames to corr.out, it is possible to either keep
original waveform names (0.) or put corr output file names (1.), which after SPLIT of
corr.,out allows inspection of the results using eev and mulplt
SINGLE BIT: the data can be reduced to 1 bit; reduce to 0 (-) and 1 (+) if set to 1, full data if set to 0.
CONTINUOUS MODE: write out all detection times for correlation above threshold if set to 1.;
otherwise only phase for maximum correlation
CONTINUOUS EXTRACT: extract event waveform files for correlation above threshold, 0. for no
extract, 1. to extract single channel used, 2. extract all channels
STATION: one line for configuration of each channel
STAT, COMP: station and component codes
SELCRIT:
1=P, 2=S, 4=full trace
DURATION: signal duration in seconds if (selcrit<4) starting from either P or S
FLOW, FHIGH:
filter limits for bandpass filter, can be; can be disabled by FILTER
(see above)
Example of STATION line:
228
KEYWORD...STAT......COMP......SELCRIT...DURATION..FLOW......FHIGH.....
--- p --STATION
PCA
S Z
1.
6.
3.
8.
--- s --STATION
EDI
S E
2.
5.
3.
8.
Ouput files:
corr.out: This is the main output file. The file is in Nordic format and contains the phase readings if run
in phase detection mode and can be used with the SEISAN location programs directly. In
continuous mode, the file can contain more than one phase reading per channel. In group
identification mode the file contains the event list, cross-correlation matrix and suggested
groups of similar events.
corr.trace: This files gives details of program run and can provide information on cause of errors.
229
7. Programming in SEISAN and list of files in SEISAN distribution
This chapter gives a bit more technical details of SEISAN starting with a short programmers guide
with description of sample and test programs.
7.1 Programmers guide and some test programs
SEISAN is conglomerate of programs and subroutines and it can be difficult to find out which routines
to use and how to start a new SEISAN program. The most common method is to use an existing
program and modify it. The intention with this section is to make it easier by providing a few sample
programs which then can be modified to do specific tasks.
The compilation of existing SEISAN programs has been described in section 3.7 and details of the
commands are found in the Makefiles. In this distribution, 3 sample programs have been included,
which each illustrate the used of some SEISAN features. All 3 programs are included in the Makefiles
and can therefore be compiled directly, modified and recompiled.
Reading and writing S-files
A basic operation is to be able to read and write S-files, since all parameters are contained in the S-files.
Starting with version 7.2, a new library (rea.for) and include block (rea.inc for definition of variables) has
been included to make it easier to read and write data into S-files. Earlier, S-files were only read and
written as text strings and individual parameters were then read/written to the text strings. Now the new
routines do it all. These routines are now used in a few programs, but will be included whenever a
program is substantially modified. The sample program is called sample_read_write_s.for. The program
illustrates how to read all parameters in an S-file, make modifications and write out the file again. The
program can be useful, if the user needs a program where special parameters are needed for a particular
analysis or for output in another format.
Reading waveform files
In SEISAN, waveform files can be in SEISAN, SAC (Unix only) or GSE format. SEISAN format is slightly
different depending on which compute platform it is written and byte swapping has to be done in some
cases. In order to automatically handle the reading of waveform files, irrespective of format and computer
platform, a set of standard routines are used (waveform.for) and an include block where all parameters
and data end up (waveform.inc).
The sample program is called sample_read_wav.for. The program illustrates how to read many waveform
files belonging to one event as if it was one file, irrespective of format. It also demonstrates how to read
just one waveform file. There is no detail on how to write a SEISAN binary file in this program, but some
info is given under the format description in Appendix 2 and the program tsig.for described below
illustrates a simple write.
The sample program sample_read_cont illustrates how to extract out a time segment of the
continuous data base. The program also shows how to write a Seisan file with all headers. The
program is started from the command prompt:
sample_read_cont
start_time interval
where start time is yyyymmddhhmmss and interval is interval in minutes.
There is also a routine in Java available to read all SEISAN binary formats. The program is called
SFORMAT (written by T. Utheim). Similarly there is a sample program to read all SEISAN binary
formats in Perl (written by Angel Rodriguez). The program is called seibinasc.pl and you need a Perl
interpreter to run it. Before starting the program, a DIRF must first be made of waveform input files.
230
The output is identical to a SEISAN ASCII file as made by SEIASC.
Graphics in SEISAN
SEISAN uses a set of graphics routines, which are identical in call on all 3 platforms . These routines
then call low level routines which are platform dependent (X on Unix and Windows calls on PC). The
programmer only has to use the high level routines. The routines also generate a PostScript output if
a given parameter is set. The program is called sample_grapichs.for. The program illustrates how to
initiate graphics, make a few simple calls, get up and use the mouse and make a hard copy file. Most
of the general graphics general routines are located in file seiplot.for and common variables in
seiplot.inc. The program can be useful for testing functionality of the mouse.
Program LSQ is a simple example of how to make xy-graphics. It also shows how to make the output
files for gmtxy. In order to find more info (apart from manual in INF) on gmtxy, see file gmt.for in LIB
and gmt_xy_par in INF.
Program to make test signals, TSIG:
It is often useful to be able to work with controlled waveform data so a program making test signals is
included. The program makes several traces, all with same length and sample rate and trace 1 is the sum
of all traces. For each trace selected, the parameters selected are: Frequency, amplitude (remember this
is integer numbers in file so use at least 1000), phase, delay (delay time when the signal appears on trace
relative to start of trace, the data before is zero) and damping. The damping is used to simulate
seismometer damping or simple a damped signal and has a similar physical meaning as the
seismometer damping constant, but period is not recalculated to simulate changing period with damping.
Zero damping is no damping.
An additional trace can be made with a Brune displacement pulse generated with parameters corner
frequency (f0), Q and kappa (see MULPLT) and travel time. Travel time is used for Q-correction and also
places the pulse at travel time distance from the origin (start of trace), so length of trace must be longer
than travel time. If zero q and kappa, no attenuation is used. The program also write an S-file with relevant
parameters. The program illustrates a simple writing of a SEISAN waveform file.
Java programs in SEISAN
The Java programs are each given as a Jar-file like jseisan.jar. the jar-files are located in the PRO
directory. The jar-file contains all the Java source code, the Java classes and the project file so a user can
decompress the jar file, change the script and make a new version of the program. The programs are
started using a script file in the COM directory and no classpath has to be set, when SEISAN has been
correctly installed.
7.2 CONTENTS OF PRO, LIB, INC, INF, COM, DAT, SUP and PIC DIRECTORIES
The PRO, LIB, INC and COM directories contain software, the DAT directory parameter files for
operating the SEISAN system and INF contains documentation and manuals. All files are listed and
explained in the file ‘seisan.all’ in the INF directory.
The program CHECK can check if a distribution is complete. Run CHECK and use option ‘basic’. The
content of the distribution is compared to the ‘seisan.all’ file in the INF directory.
The seisan.all file also list programs no longer compiled, but with source code included in case there
should be a future need for these programs
231
8. Acknowledgments
A large number of people have contributed to the SEISAN system. From the British Geological
Survey, Jim Bolton has spent several months cleaning up the software, putting in error checking and
put in the new version of EPIMAP. Jane Exton has also been involved in several of the SEISAN
database problems. A main contributor has been Barry Lienert who has spent several months at our
institute to modify and adopt the HYP program, he has also written the complete HYP manual. Bent
Ruud has contributed with the core of the AUTOPIC software and has helped out in practical aspects
of installing it. Kuvvet Atakan has written the seismic hazard part. Ronald Arvidson has tested large
parts of the system and done work on several programs and contributed with the modified version of
FOCMEC. The help of Arne Sjursen has been essential for implementing X. Vunganai Midzi did the
complete testing of version 6.0. Mario Villagran with programs, many suggestions and bug reports
has interacted with the development of SEISAN. Ezra Twesigomwe, Berit Storheim, K. Atakan and
Alice Walker have debugged the manuscript.
Version 7: Bladimir Moreno has made the Windows graphics, made SEISAN run under Linux, written
several other programs and has thus enormously contributed to this version. The hazard part has
been updated by Kuvvet Atakan and Anibal Ojeda. Andrius Pacesa has implemented the FK routine.
The programs have been tested and the manual been checked by Margaret Grandison, Waldo
Taylor, Vunganai Midzi, Berit Storheim, Anne Lise Kjærgaard, Anibal Ojeda, Ileana Boschini and
Cecilie Langeland.
Version 7.1: This version was tested by Anne Lise Kjærgaard, Margaret Grandison and Vunganai
Midzi.
Version 7.2: Several contributions, including changes to MULPLT and implementing HYPO71, were
made by Brian Baptie from the BGS. W_EMAP is a new program that was provided by Fernando
Carrilho. The CPLOT program was written by Susanne Lund Jensen from KMS. Susanne has also
checked this version of the manual.
Version 8.0: The graphics part on Unix has been improved thanks to Freya Cromarty and Frederik
Tillmann. The on-site-inspection group at the CTBTO has financially supported the development of
SEISCONF and JSEISAN, as well as the modification of other programs. JSEISAN and SEISCONF
were written by Bladimir Moreno. A new version of W_EMAP was provided by Fernando Carrilho.
Mario Ordaz contributed the DEGTRA A4 software. Angel Rodriguez contributed the Perl reading
routine and Terje Utheim the sfile Java program. Mathilde Bøttger Sørensen has revised the manual
and tested the distribution. Dieter Stoll has provided information on how to compile on MacOSX and
tested the software on Mac. RefTek has provided the rt_seis program.
Version 8.1: Rodrigo Luciano Pereira Canabrava made a major contribution to SEISAN by writing and
implementing routines to read and write SEED data. Richard Luckett implemented the ISC location
program. Mathilde Bøttger Sørensen wrote scripts for the new macroseismic part. Brian Baptie
contributed tools to convert event data for use with travel time conversion programs.
Finally we will thank all the patient users who have suffered from the "bugs" and have given useful
feedback.
232
9. References
Anderson, D. (1982). Robust earthquake location using M-estimates. PEPI, 30, 119-130.
Banfill, R. (1996). PC-SUDS Utilities. A collection of tools for routine processing of seismic data stored in
the seismic unified data system for DOS (PC-SUDS), Version 2.5, Small Systems Support.
Boore,D.M. (1983). Stochastic simulation of high frequency ground motions based on seismological
models of the radiated spectra. Bulletin of the Seismological Society of America, 73, 18651884.
Boore,D.M. (1989). Quantitative ground motion estimates. In: Earthquake Hazards and the Design
of
Constructed Facilities in the Eastern United States. K.H.Jacob and C.J.Turkstra (Eds.),
Annals of the New York Academy of Sciences, 558, 81-94.
Bouchon, M. (1981). A simple method for calculating Green's functions for elastic layered media. Bull.
Seism. Soc. Am. 71, 959-972.
Brune,J.N. (1970). Tectonic stress and spectra of seismic shear waves. Journal of Geophysical
Research, 75, 4997-5009.
Capon, J., High-resolution frequency-wavenumber spectrum analysis. Proc. IEEE 57, 1408-1418, 1960.
Chiu, J, B. L. Isacs and R. K. Cardwell (1986). Studies of crustal converted waves using short-period
seismograms recorded in the Vanatu Island arc, Bull. Seism. Soc. Am. 76,177-190.
Chapman, C.H., 1978. A new method for computing synthetic seismograms, Geophys.
J. R. astr. Soc., 54, 481-518.
Chapman, C.H. and Orcutt, J.A., 1985. The computation of body wave synthetic seismograms
in
laterally homogeneous media, Reviews of Geophysics, 23, 105-163.
Chapman, C.H., Chu Jen-Yi, and Lyness, D.G., 1988. The WKBJ seismogram algorithm, in: D.J.
Doornbos (ed.), Seismological algorithms, Academic Press, London, pp47-74.
Dey-Sarkar, S.K. and Chapman, C.H. (1978). A simple method for computation of body-wave
seismograms, Bull. Seismo. Soc Am., 68, 1577-1593.
Draper, N.R. and Smith, H. (1966). Applied regression analysis, John Wiley and Sons, New York.
Ebel, J. E. and K. P. Bonjer (1990). Moment tensor inveriosn of small earthquakes in southwestern
germany for fault plane solution. Geophys. J. Int. 101, 133-146.
Goldstein, P. (1999). SAC user’s manual, Lawrence Livermore Laboratory, University of California.
GSETT-3, (1997). Provisional GSE 2.1, Message Formats & Protocols, Operations Annex 3.
Havskov, J and G. Alguacil (2004). Instrumentation in earthquake seismology. Springer 358 pp.
Havskov, J, S. Malone, D McCloug and R. Crosson (1989). Coda Q for the state of Washington. Bull.
Seism. Soc. Am.,79, 1024-1038.
Herrmann, R. B. (1985). An extension of random vibration theory estimates of strong ground motion
to
large distances, Bull. Seism. Soc. Am.,75, 1447-1453.
Herrmann, R., B. and A. Kijko (1983). Modelling some empirical vertical component Lg relations, Bull.
Seism. Soc. Am. 73,157-171.
Herrmann, R. B. (1996). Computer programs in seismology. Manual, Saint Louis University.
Hutton, L. K. and D. Boore (1987). The Ml scale in Southern California. Bull. Seism. Soc. Am. 77,20742094.
nd
IRIS Consortium(1993). Standard for the Exchange of Earthquake Data – Reference Manual, 2 Edition.
Kanamori, H (1977). The energy release in great earthquakes. Journal of Geophysical Research, 82,
1981-1987.
Kissling, E., W.L. Ellsworth, D. Eberhart-Phillips and U. Kradolfer, (1994). Initial reference model in local
earthquake tomography, Journal of Geophysical Research, Vol. 99, No. B10, 19 635-19
646.
Kissling, E., U. Kradolfer and H. Maurer, (1995). Program VELEST USERS GUIDE - Short Intrduction.
Klein, F. W. (1984) Users guide to HYPOINVERSE, a program for Vax and PC350 computers to solve for
earthquake locations. USGS open file report 84-000.
Kværna, T. and D.J. Doornbos, An integrated approach to slowness analysis with arrays and treecomponent stations, in: NORSAR Semiannual Technucal Summary, 1 October 1985 - 31
March 1986, Scientific Report No. 2-85/86, NORSAR, Kjeller, Norway, 1986.
Lee, W. H. K., R. E. Bennett and L. Meagher, 1972. A method for estimating magnitude of local
earthquakes from signal duration, U.S.G.S Open file report.
233
Lee, W.H.K. and Lahr, J.C. 1975. HYPO71 (revised): a computer program for determining hypocenter,
magnitude and first motion pattern of local earthquakes. Open-file report, U.S. Geological
Survey, 75-311.
Lienert, B.R.E.,E. Berg and L. N. Frazer (1986). Hypocenter: An earthquake location method using
centered, scaled, and adaptively least squares, BSSA, Vol 76, 771-783.
Lienert, B. R. E. (1991). Report on modifications made to Hypocenter. Institute of Solid Earth Physics,
University of Bergen.
Lienert, B.R.E and J. Havskov (1995). A computer program for locating earthquakes both locally and
globally, Seismological Research Letters, 66, 26-36.
McGuire,R.K. (1976). EQRISK. Evaluation of earthquake risk to site. United States Department of the
Interior, Geological Survey, Open File Report 76-67, 90p.
Nakamura, Y. (1989). A method for dynamic characteristics estimation of subsurface using microtremors
on the ground surface. Q. Rep. Railway Tech. Res. Inst., 30, 1989.
Roberts, R.G., Christoffersson, A., and Cassidy, F., 1989. Real time events detection, phase identification
and source location estimation using single station component seismic data and a small
PC, Geophysical Journal, 97, 471-480.
Ordaz,M. (1991). CRISIS. Brief description of program CRISIS. Institute of Solid Earth Physics, University
of Bergen, Norway, Internal Report, 16p.
Ottemöller, L. (1995). Explosion filtering for Scandinavia, Norwegian National Seismic Network technical
report # 2, IFJF, University of Bergen. 209 pp.
Ottemöller, L. and J. Havskov (1999). SeisNet: A General Purpose Virtual Seismic Network. SRL 70, 5,
522-528.
Ottemöller, L., N.M. Shapiro, S.K. Singh and J.F. Pacheco (2002). Lateral variation of Lg wave
propagation in southern Mexico, J. Geophys. Res. 107.
Ottemöller, L. and J. Havskov (2003). Moment Magnitude Determination for Local and Regional
Earthquakes Based on Source Spectra, Bull. Seism. Soc. Am., 93, 203-214.
Peterson, J. (1993). Observation and modeling of seismic background noise, U.S. Geol., Survey OpenFile report 93-322, 95p.
Pujol, J. Software for joint hypocentral determination (2003). In "International Handbook of Earthquake
and Engineering Seismology", Part B, Vol. 81, W.H.K. Lee, P. Jennings, H. Kanamori, C.
Kisslinger, Eds.
Ruud, B.O., E.S. Husebye, S.F. Ingate and A. Christoffersen (1988). Event location at any distance using
seismic data from a single, three-component station. Bull. Seism. Soc. Am. 78, 308-325.
Ruud, B.O., and Husebye, E.S., (1992). A new three-component detector and automatic single
station bulletin production, Bull. Seism. Soc. Am., 82, 221-237.
Scherbaum, F. (1996). Of Poles and Zeros – Fundamentals of Digital Seismology, Kluwer Academic
Publishers.
Singh, S.K., Apsel,R.J., Fried, J. and Brune,J.N. (1982). Spectral attenuation of SH-waves along the
Imperial fault. Bulletin of the Seismological Society of America. 72, 2003-2016.
Snoke, J. A., J. W. Munsey, A. G. Teague and G. A. Bollinger (1984). A program for focal mechanism
determination by combined use of polarity and SV-P amplitude ratio data. Earth quake
notes, 55, p15.
Waldhauser F. and W.L. Ellsworth (2000). A double-difference earthquake location algorithm: Method and
application to the northern Hayward fault, Bull. Seism. Soc. Am., 90, 1353-1368.
Waldhauser, F., hypoDD: A computer program to compute double-difference earthquake locations (2001).
USGS Open File Rep., 01-113.
Veith, K. F. and G. E. Clawson (1972). Magnitude from short period P-wave data. Bull. Seism. Soc. Am.
62, 435-440.
234
Appendix 1 - The Nordic format
Free columns are included for two purposes:
1. To obtain a readable format
2. To have some space for possible future extensions
Here are examples, top 3 lines for positioning only.
1
2
3
4
5
6
7
1234567890123456789012345678901234567890123456789012345678901234567890123456789
.
.
.
.
.
.
.
.
------------------------------------------------------------------------------1984 1022 2102 23.2 LE 69.330 27.440 11.0F NAO 34 5.2 3.8LNAO 4.0BPDE 3.2SISC1
NORTHERN FINLAND
3
NRSA SZ IPN 1 D 2244 13.44 0345 1234.6 1.33 245.2 08.6 841022 120.2 3 5 12345
NRSA SZ ILG 1 D 2244 13.44 0345 1234.6 1.33 265.0 03.6 841022 120.2 3 5 12345
1985
510 21 5 16.1 LE 60.240
6.170 30.0F BER 6 2.3 3.8LNAO 4.0BPDE 3.2SISC
1.5
0.5
0.9
5.0
0.4
5
8505210425.WNN
6
NORTHERN HORDALAND
F 3.5 61.22 0.5
5.33 0.8 23456 2
456 2 99 11BER1
STAT SP IPHASW D HRMM SECON CODA AMPLIT PERI AZIMU VELO SNR
7
BER SZ IPG 2 U 2105 25.41 200
HYA SZ ISG 1
2105 33.1
ODD SZ IP
3
2105 20.1
250
ODD SZ EPG
2105 22.9
ODD SZ LG
2105 55.8
--------------------------------------------------------------------------------
Below are examples of how the last free columns of type 4 lines are used in the Nordic Databank in
Helsinki and in Bergen:
1985
510 21 5 16.1 LE 60.240
6.170 30.0F BER 6 2.3 3.8LNAO 4.0BPDE 3.2SISC1
1.5
0.5
0.9
5.0
0.4
5
8505210425.WNN
6
ACTION:UPD 93-07-09 09:40 OP:jens STATUS:
ID:19920101080359
I
STAT SP IPHASW D HRMM SECON CODA AMPLIT PERI AZIMU VELO SNR AR TRES W DIS CAZ7
NRSA SZ IPN 1 D 2105 13.44 0345 1234.6 1.33 245.2 08.6 5.5 2 -0.7 9 555 235
BER SZ IPG 2 U 2105 25.41 200
HYA SZ ISG 1
2105 33.1
ODD SZ IP
3
2105 20.1
250
ODD SZ EPG
2105 22.9
ODD SZ LG
2105 55.8
Note in this example the fault plane solution line(F) and the HYP error line(E)
1993 1028 0800 26.4 L 57.518
7.119 18.8 BER 6 .6 2.6CBER
1
GAP=201
1.20
6.4
7.0 6.8
.3359E+01 -.2719E+00
.3054E+02E
93.2
74.8
-48.2
2
F
ACTION:SPL 95-01-08 09:40 OP:jh
STATUS:
ID:19931028080019
I
9310-28-0800-19S.NSN_17
6
STAT SP IPHASW D HRMM SECON CODA AMPLIT PERI AZIMU VELO SNR AR TRES W DIS CAZ7
BLS5 SZ EP
D 8 0 56.80 129
-.110 216 349
BLS5 SZ ESG
8 1 23.59
-.910 216 349
BLS5 SZ EP
8 0 56.80 129
-.110 216 349
BLS5 SZ ESG
8 1 23.59
-.910 216 349
Location parameters:
AR : Azimuth residual when using azimuth information in locations
TRES: Travel time residual
W
: Actual weight used for location ( inc. e.g. distance weight), i2
DIS : Epicentral distance in kms
CAZ : Azimuth from event to station
----------------------------------------------------------------------------Note: Type 1 line must be the first, all type 4 lines should be together and
the last line must be blank
--------------------------------------------------------------------------------
235
FORMAT DESCRIPTION:
Type 1 Line:
Columns Format Description
1
2- 5
6
7- 8
9-10
11
I4
I2
I2
Free
Year
Free
Month
Day of Month
Fix o. time
Comments
Normally blank, an F fixes origin
time
12-13
14-15
16
17-20
21
22
23
I2
I2
A1
A1
Hour
Minutes
Free
Seconds
Location model indicator
Distance Indicator
Event ID
24-30
31-38
39-43
44
45
F7.3
F8.3
F5.1
A1
A1
Latitude
Longitude
Depth
Depth Indicator
Locating indicator
46-48
49-51
52-55
A3
Hypocenter Reporting Agency
Number of Stations Used
RMS of Time Residuals
56-59
60
61-63
64-67
68
69-71
72-75
76
77-79
80
F4.1
A1
A3
F4.1
A1
A3
F4.1
A1
A3
A1
Magnitude No. 1
Type of Magnitude
L=ML, B=mb, S=Ms, W=MW, G=MbLg, C=Mc
Magnitude Reporting Agency
Magnitude No. 2
Type of Magnitude
Magnitude Reporting Agency
Magnitude No. 3
Type of Magnitude
Magnitude Reporting Agency
Type of this line ("1"), can be blank if first
line of event
F4.1
Any character
L = Local, R = Regional, etc.
E = Explosion, etc.
P = Probable explosion
V = Volcanic
Q = Probable volcanic
Degrees (+ N)
Degrees (+ E)
Km
F = Fixed, S = Starting value
----------------------------, * do
not locate
If more than 3 magnitudes need to be associated with the hypocenter in the first line, a
subsequent additional type one line can be written with the same year, month, day until event
ID and hypocenter agency. The magnitudes on this line will then be associated with the main
header line and there is then room for 6 magnitudes.
Type 2 line (Macroseismic information)
1-5
6-20
21
22
a
a1
23
a1
24
a1
25
a1
26
a1
27
28-29 i2
30
a1
31-32 a2
Blank
Any descriptive text
Free
Diastrophism code (PDE type)
F = Surface faulting
U = Uplift or subsidence
D = Faulting and Uplift/Subsidence
Tsunami code (PDE type)
T = Tsunami generated
Q = Possible tsunami
Seiche code (PDE type)
S = Seiche
Q = Possible seiche
Cultural effects (PDE type)
C = Casualties reported
D = Damage reported
F = Earthquake was felt
H = Earthquake was heard
Unusual events (PDE type)
L = Liquefaction
G = Geysir/fumerol
S = Landslides/Avalanches
B = Sand blows
C = Cracking in the ground (not normal faulting).
V = Visual phenomena
O = Olfactory phenomena
M = More than one of the above observed.
Free
Max Intensity
Max Intensity qualifier
(+ or - indicating more precicely the intensity)
Intensity scale (ISC type defintions)
MM = Modified Mercalli
236
RF = Rossi Forel
CS = Mercalli - Cancani - Seberg
SK = Medevev - Sponheur - Karnik
33
34-39
40
41-47
48
49-51
52
f6.2
f7.2
f3.1
a1
53-56 f4.2
57-61 f5.2
62-63 i2
64-68 f5.2
69-70
71
72
73-75
76-79
80
i2
a1
a3
a1
Free
Macroseismic latitude (Decimal)
Free
Macroseismic longitude (Decimal)
Free
Macroseismic magnitude
Type of magnitude
I = Magnitude based on maximum Intensity.
A = Magnitude based on felt area.
R = Magnitude based on radius of felt area.
* = Magnitude calculated by use of special formulas
developed by some person for a certain area.
Further info should be given on line 3.
Logarithm (base 10) of radius of felt area.
Logarithm (base 10) of area (km**2) number 1 where
earthquake was felt exceeding a given intensity.
Intensity boardering the area number 1.
Logarithm (base 10) of area (km**2) number 2 where
earthquake was felt exceeding a given intensity.
Intensity boardering the area number 2.
Free
Quality rank of the report (A, B, C, D)
Reporting agency
Free
Type of this line ("2")
Type 3 Line (Optional):
Columns Format Description
1
2-79
80
A
A1
Free
Text
Type of this line ("3")
Comments
Anything
Type 4 line:
Columns Format Description
1
2- 6
A5
Free
Station Name
Comments
Blank = End of readings = end of
event
S = SP, I = IP, L = LP etc
Z, N, E
7
8
9
10
11-14
15
16
17
18
A1
A1
19-20
I2
Hour
21-22
23-28
29
30-33
34-40
41
42-45
46
47-51
52
53-56
57-60
61-63
64-68
69-70
71-75
76
77-79
80
I2
F6.0
Minutes
Seconds
Free
Duration (to noise)
Seconds
Amplitude (Zero-Peak)
Nanometers
Free
Period
Seconds
Free
Direction of Approach
Degrees
Free
Phase Velocity
Km/second
Angle of incidence (was Signal to noise ratio before version 8.0)
Azimuth residual
Travel time residual
Weight
Epicentral distance(km)
Free
Azimuth at source
Type of this line ("4"), can be blank, which it is
most often
A1
A2
I1
A1
I4
g7.1
F4.0
F5.0
F4.0
F4.0
I3
F5.1
I2
F5.0
I3
A1
Instrument Type
Component
Free or weight, see note below
Quality Indicator
I, E, etc.
Phase ID
PN, PG, LG, P, S, etc. **
Weighting Indicator (1-4)
0 = full weight (as in HYPO)
Free or flag A to indicate automartic pick, removed when picking
First Motion
C, D
Note: Currently 15 to 18 can also be used for phase assuming
column 11-14 is not blank. See note ** below.
Hour can be up to 48 to
indicate next day
NB: Epicentral distance: Had format I5 before version 7.2. All old lines can be read with
format F5.0 with same results, but now distance can also be e.g. 1.23 km which cannot be read
by earlier versions. However, an UPDATE would fix that.
** Long phase names: An 8 character phase can be used in column 11-18. There is then not
room for polarity information. The weight is then put into column 9. This format is recognized
by HYP and MULPLT.
237
Type 4 cards should be followed by a Blank Card (Type 0)
Type 5 line (optional): Error estimates of previous line, currently not used
by any SEISAN programs.
Columns Format Description
1
2-79
80
A1
Comments
Free
Error estimates in same format as previous line, normally
type 4
Type of this line ("5")
Type 6 Line (Optional):
Columns Format Description
1
2-79
80
A
A1
Comments
Free
Name(s) of tracedata files
Type of this line ("6")
Type 7 Line (Optional):
Columns Format Description
1
2-79
80
A
A1
Comments
Free
Help lines to place the numbers in right positions
Type of this line ("7")
Type F Line (Optional): Fault plane solution
Columns
1:30
31:36
71:76
78:78
79:79
Format
3F10.0
I6
A6
A1
A1
Description
Strike, dip and rake, Aki convention
Number of bad polarities
Method or source of solution, seisan mames INVRAD or FOCMEC
Quality of solution, A (best), B C or D (worst), added manually
Blank: Prime solution, overwritten when focmec or invrad
makes a new solution, non blank: remain in file, cannot be
plotted
O: Remain in file and can be plotted
Type E Line (Optional): Hyp error estimates
Columns Format Description
1
Free
2 - 5 A4
The text GAP=
6 - 8 I3
Gap
15-20 F6.2 Origin time error
25-30 F6.1 Latitude (y) error
31-32
Free
33-38 F6.1 Longitude (x) error (km)
39-43 F5.1 Depth (z) error (km)
44-55 E12.4 Covariance (x,y) km*km
56-67 E12.4 Covarience (x,z) km*km
68-79 E14.4 Covariance (y,z) km*km
Type I Line, ID line
Columns
1
2:8
9:11
Format description
Free
Help text for the action indicator
Last action done, so far defined SPL:
REG:
ARG:
UPD:
UP :
REE:
DUB:
NEW:
12
Free
13:26
Date and time of last action
27
Free
28:30
Help text for operator
31:34
Operater code
35
Free
36:42
Help text for status
43:56
Status flags, not yet defined
57
Free
58:60
Help text for ID
61:74
ID, year to second
75
If d, this indicate that a new file
one or more seconds different from an existing ID
76
Indicate if ID is locked. Blank means
Split
Register
AUTO Register, AUTOREG
Update
Update only from EEV
Register from EEV
Duplicated event
New event
id had to be created which was
to avoid overwrite.
not locked, L means locked.
238
Type H line, High accuracy hypoenter line
Columns
1:55
16
17
23
24:32
33
34:44
44
45:52
53
54:59
60:79
80
As type 1 line
Free
Seconds, f6.3
Free
Latitude, f9.5
Free
Longitude, f10.5
Free
Depth, f8.3
Free
RMS, f6.3
Free
H
Type E13 and EC3 line, explosion information
Example
1980 0124 0927
CHARGE(T): 0.5
LE
HAA
Haakonsvern, underwater explosion
E13
EC3
E13
Information on explsion site, time and agency, same format as a type 1 line, no magnitudes
used, last 3 columns must be E13
EC3
Information on charge and site
Columns
2:11
11:12
13:22
23:77
78:80
Info text
Blank
Charge in tons, f10.3
Any information, a
EC3
Type MACRO3 line: File name of macroseismic observations in ISO directory
Example:
1980-03-14-0456-05.MACRO
MACRO3
An example of the file is:
Sunnfjord
Comment
60.500
60.560
60.570
1980
5.270
5.260
5.050
1.0 EMS
1.0 EMS
1.0 EMS
314
5088
5100
5112
456
5 GMT 1980
314
556
5 Local time
MJOELKERAAEN
ISDALSTOE
ROSSLAND
1. Line
Location, GMT time, Local time. Format a30,i4,1x,2i2,1x,2i2,1x,i2,’
GMT’,1x,i4,1x,2i2,1x,2i2,1x,i2,1x,’Local time’
2. Line
Comments
3. Line
Observations: Latitude, Longitude,intensity, code for scale, postal code or similar,
location,Format 2f10.4,f5.1,1x,a3,1x,a10,2x,a
Type 3 line giving xnear/xfar
Definition of xnear and xfar to be used with HYPOCENTER.
Example
XNEAR 1000.0 XFAR 2000.0
Columns
8-13: xnear value
20-25: xfar value
3
239
Appendix 2 - The Seisan waveform file format
The file is written from Fortran as an unformatted file. This means that the file contains additional
characters (not described below, see end of this Appendix) between each block, which must be taken
into account if the file is read as a binary file. If read as Fortran unformatted, the content will appear
as described below. However, the internal structure is different on Sun, Linux and PC. SEISAN
automatically corrects for these differences. The SEISAN ASCII format has identical headers to the
binary files, however the binary samples are written as formatted integers, one channel at the time
just like the in the binary format.
EVENT FILE HEADER
-----------------CONTAINS MINIMUM 12 ASCII STRINGS OF 80 BYTES.
ALL FORMATS I OR A UNLESS OTHERWISE SPECIFIED.
line 1
1 2 -
1: FREE
30: NETWORK NAME,
COULD E.G. BE WESTERN NORWAY NETWORK
31 - 33: NUMBER OF STATIONS, MAX 999
34 - 36: YEAR-1900, e.g. 101 for 2001 (I3)
37
38 - 40: DOY
41
42 - 43: MONTH
44
45 - 46: DAY
47
48 - 49: HR
50
51 - 52: MIN
53
54 - 59: SEC, FORMAT F6.3
60
61 - 69: TOTAL TIME WINDOW (SECS), FORMAT F9.3
70 - 80: FREE
71
72
73 - 80: FREE
-------------------------------------------------------------|
line 2
1 - 80: FREE
-------------------------------------------------------------|
line 3
1
2 5: STATION CODE (A4), first 4 characters
6 9: COMPONENT (A4)
10
: STATION CODE (A1), LAST CHARACTER IF 5 CHARACTER STATION CODE
11 - 17: START TIME RELATIVE TO EVENT FILE TIME (SECS) F7.2
18
: BLANK
19 - 26: STATION DATA INTERVAL LENGTH (SECS) F8.2
27 - 52: SECOND CHANNEL
53 - 78: THIRD CHANNEL
79 - 80: BLANK
|
------------------------------------------------------------|
line 4-XX, where XX depends on number of channels, however, XX
is at least 12 so there might be some blank lines.
1 - 80: THREE MORE CHANNELS (SAME FORMAT AS line 3)
-------------------------------------------------------------|
EVENT FILE CHANNEL HEADER
------------------------HEADER IS 1040 BYTES LONG, WRITTEN AS ONE VARIABLE DEFINED AS
CHARACTER*1040
THE PARAMETERS ARE WRITTEN FORMATTED WITH INTERNAL WRITE INTO
1040 BYTE TEXT STRING.
FORMAT IS ALWAY I-FORMAT UNLESS OTHERWISE SPECIFIED
1 6 10
13
14
17
18
20
21
23
24
5: STATION CODE (A5)
9: COMPONENT (A4), E.G. SZ L : SZ LOW GAIN, AN: ACCELERATION NS
-
12: YEAR – 1900, e.g. 101 for 2001, (I3)
-
16: DOY
-
19: MONTH
-
22: DAY
-
25: HR
240
26
27 29
30 36
37 44 51
52 60
61 70
71 76
28: MIN
: TIMING INDICATOR, BLANK: TIME IS OK, E: UNCERATIAN TIME
35: SECOND (F6.3)
43: SAMPLE RATE (F7.2 or any f-format)
50: NUMBER OF SAMPLES (I7)
59: LATITUDE (F8.4), optional
69: LONGITUDE (F9.4), optional
75: ELEVATION (METERS), optional
: Indicate gain factor: Blank: No gain factor, G: Gain factor in
column 148 to 159
77
: 2 OR 4 FOR 2 OR 4 BYTE INTEGER, BLANK IS 2 BIT
78
: P: Poles and zeros used for response info, blank: Seismometer
period etc used for response info. See below for details.
T: Use up to 30 tabulated values irrespective of what is given
below. If less than 30, blank characters must be given.
79
: C: a combination of table, poles and zeros or instrument
constants have been used, for information only. Value in 78
must then be T.
F: Force use of header response, e.g. generated by MULPLT. Only
gain at 1 hz is correct and 78 must be set to T.
80 - 80: FREE
148- 159: Normally comment, if 76 set to G, this is a gain factor,
format G12.7. All samples read from channel are multipled
by this factor when read by routine seisinc. Used when data
is stored in units of e.g. nm where values can be less than 1.
Currently only generated by MULPLT when option OUT is used to
extract part of a waveform file.
81 - 160: COMMENT LINE DESCRIBING THE SYSTEM RESPONSE (A80)
If character 78 is blank, option 1:
161 - 240: (10G8.3) 1. SEISMOMETER PERIOD
2. FRACTION OF CRITICAL DAMPING
3. SEISMOMETER GENERATOR CONSTANT (V/m/s) or
ACCELEROMETER SENSITIVITY (V/G)
4. AMPLIFIER GAIN
5. RECORDING MEDIA GAIN (I.E. 2048 COUNTS/VOLT)
6. GAIN AT 1.0 HZ, UNITS: COUNTS/METER
7. CUTOFF FREQUENCY FOR FILTER1 (HZ)
8. # OF POLES FOR FILTER1 (NEGATIVE FOR HIGHPASS)
9. CUTOFF FREQUENCY FOR FILTER2 (HZ)
10. # OF POLES FOR FILTER2 (NEGATIVE FOR HIGHPASS)
241 - 320: (10G8.3) FREQUENCIES AND #'S OF POLES FOR FIVE MORE FILTERS
321 -1040: RESPONSE CURVES (9(10G8.3) FREQ., AMPL. (REL. 1.0 HZ) AND PHASE,
WRITTEN IN GROUPS OF 10 FREQUENCIES, 10 AMPLITUDES AND 10 PHASES
If character 78 is P, option 2:
161 - 182 (1X,2I5,G11.4) 1. NUMBER OF POLES
2. NUMBER OF ZEROS
3. NORMALIZATION CONSTANT, COUNTS/M
183 - 240 (5G11.4)
2 Poles in pairs of real and imaginary parts
241 -1040 (G11.4)
Remaining poles and zeros. 7 values are written
and then 3 spaces are left blank, see example
below.
For each pole or zero, there are two real numbers representing the real and imaginary part of the pole
or zero, thus the number of poles is half the number of values written. First all the poles are written in
pairs of real and imaginary parts, then follow the zeros. There is room for a total of 37 poles and zeros
(74 pairs). The poles and zeros are written in a simulated line mode to make it easier to read, thus the
3 blanks after writing 7 values. It is assumed that the response is in displacment with units of
counts/m.
SLR
L
11
.0000
-.3280
.0000
.0000
E 86 199
5
7 18 15
6 35.960
.2760E+11 -.3770
-.2320
.0000
.0000
-.3280
-.2140E-01 .0000
.0000
.0000
1.000
.1830
-.2320
.0000
.0000
.0000
1320
-.3770
.0000
-.3280
.0000
.0000
4P
-.1830
-.2320
.0000
.0000
.0000
-.6540
.0000
-.2140E-01
.0000
NOTE: The component information in character 6 IS VERY IMPORTANT. It MUST be A if an
accelerometer is used, any other character assumes a velocity transducer. This is only relevant
however if option 1 is used where response values will be calculated from the free period etc. If option
1 with discrete values or poles and zeros are used, the first component character can be anything.
-------------------
241
| EVENT FILE
|
|
HEADER
|
------------------|
------------------| EVENT FILE
|
| FIRST CHANNEL |
|
HEADER
|
------------------|
------------------|
DATA
|
| FIRST CHANNEL |
------------------|
------------------| EVENT FILE
|
| NEXT
CHANNEL |
|
HEADER
|
------------------|
------------------|
DATA
|
| NEXT
CHANNEL |
------------------|
|
|
|
------------------| EVENT FILE
|
| LAST
CHANNEL |
|
HEADER
|
------------------|
------------------|
DATA
|
| LAST
CHANNEL |
-------------------
at least 12* 80 BYTES
1040 BYTES
1040 BYTES
1040 BYTES
To write a SEISAN file: If main headers are called mhead, channel header chead, data is data
(integer), there is nchan channels and each has nsamp samples, then the file is written as
Do i=1,12
Write(1) mhead(i)
Enddo
Do k=1,nchan
Write(1) chead
Write(1) (data(i),i=1,nsmap)
Enddo
This example only works up to 30 channels when writing main header. For more channels, see e.g.
program SEISEI how to do it.
Details of binary file structure
When Fortran writes a files opened with "form=unformatted", additional data is added to the file to
serve as record separators which have to be taken into account if the file is read from a C-program or
if read binary from a Fortran program. Unfortunately, the number of and meaning of these additional
characters are compiler dependent. On Sun, Linux, MaxOSX and PC from version 7.0 (using Digital
Fortran), every write is preceded and terminated with 4 additional bytes giving the number of bytes in
the write. On the PC, Seisan version 6.0 and earlier using Microsoft Fortran, the first 2 bytes in the file
are the ASCII character "KP". Every write is preceded and terminated with one byte giving the number
of bytes in the write. If the write contains more than 128 bytes, it is blocked in records of 128 bytes,
each with the start and end byte which in this case is the number 128. Each record is thus 130 bytes
long. All of these additional bytes are transparent to the user if the file is read as an unformatted file.
However, since the structure is different on Sun, Linux, MacOSX and PC, a file written as unformatted
on Sun, Linux or MacOSX cannot be read as unformatted on PC or vice versa. . The files are very
easy to write and read on the same computer but difficult to read if written on a different computer.
To further complicate matters, the byte order is different on Sun and PC.
This means that version 7.0 can read all earlier waveform files on all platforms from all platforms.
However, files written on version 7.0 PC cannot be read by any earlier versions of Seisan without
modifying the earlier seisan version.
242
In SEISAN, all files are written as unformatted files. In order to read the files independently of where
they were written, the reading routine (buf_read in seisinc, in LIB) reads the file from Fortran as a
direct access file with a record length of 2048 bytes. The additional bytes are thrown away, the
relevant bytes fished out and swapped if the file is written on a different computer than where it is
read.
Since there is no information stored in the header of the file giving the byte address of each channel,
the routine must read the first file-header, calculate how many bytes there are down to where the next
channel starts, jump down and repeat the process until the desired channel is reached (this is also
how SUDS files are read). However, compared to reading the file as unformatted, only a fraction of
the file is read to fish out a particular channel. Once the channel header has been read, the start
address is stored in the subroutine so any subsequent access to that channel is very fast. Overall,
random access to SEISAN waveform files is much faster with the binary read than the previous
(version 5.0 and earlier) unformatted read. Only in the case where the whole file is read is the
unformatted read faster.
PC file structure
Up to and inluding version 6.0
---------------------------------------one byte: K indicates start of file
---------------------------------------one byte: # of bytes following
---------------------------------------128 bytes or less of data
---------------------------------------one byte: # of bytes in previous record
---------------------------------------one byte: # of bytes in following record
---------------------------------------128 bytes or less of data
.....
.....
Sun and Linux file structure
PC structure from version 7.0
----------------------------4 bytes: # of bytes following
----------------------------one block of data
----------------------------4 bytes: # bytes in prev. write
------------------------------4 bytes: # of bytes following
------------------------------one block of data
------------------------------........
........
........
From version 7.0,the Linux and PC file structures are exactly the same. On Sun the structure is the
same except that the bytes are swapped. This is used by SEISAN to find out where the file was
written. Since there is always 80 characters in the first write, character one in the Linux and PC file
will be the character P (which is represented by 80) while on Sun character 4 is P.
243
Appendix 3 – Response file formats used in SEISAN
SEISAN can use either SEISAN response file format or GSE response file format. The detailed
SEISAN format is given here as well as the parts of the GSE format used in SEISAN.
SEISAN response file format
There are three SEISAN response formats. The instrument response can either be presented as (1)
instruments constants, (2) pairs of frequency amplitude and phase or as (3) poles and zeros.
Line
1
6
10
11
14
18
21
24
27
30
52
61
71
78
1
-
5:
9:
:
12:
16:
19:
22:
25:
28:
35:
59:
69:
75:
:
STATION CODE (A5)
COMPONENT (A4), E.G. SZ L : SZ LOW GAIN, AN: ACCELERATION NS
CENTURY, 0: 1900, 1:2000
YEAR
DOY
MONTH
DAY
HR
MIN
SECOND (F6.3)
LATITUDE (F8.4), optional
LONGITUDE (F9.4), optional
ELEVATION (METERS), optional
P: Poles and zeros
Blank: Seismometer period etc used for response info. See below for
details.
T: Use up to 30 tabulated values irrespective of what is given
below. If less than 30, blank characters must be given.
: C: a combination of table, poles and zeros or instrument
constants have been used, for information only. Value in 78
must then be T.
F: Force use of header response. Only
gain at 1 hz is correct and 78 must be set to T.
-
79
Line 2
1 - 80: COMMENT LINE DESCRIBING THE SYSTEM RESPONSE (A80)
OPTION 1
If character 78 is blank, option 1:
Line 3
1 - 80: (10G8.3)
1. SEISMOMETER PERIOD
2. FRACTION OF CRITICAL DAMPING
3. SEISMOMETER GENERATOR CONSTANT (V/m/s) or
ACCELEROMETER SENSITIVITY (V/G)
4. AMPLIFIER GAIN
5. RECORDING MEDIA GAIN (I.E. 2048 COUNTS/VOLT)
6. GAIN AT 1.0 HZ, UNITS: COUNTS/METER
7. CUTOFF FREQUENCY FOR FILTER1 (HZ)
8. # OF POLES FOR FILTER1 (NEGATIVE FOR HIGHPASS)
9. CUTOFF FREQUENCY FOR FILTER2 (HZ)
10. # OF POLES FOR FILTER2 (NEGATIVE FOR HIGHPASS)
Line 4
1 _ 80: (10G8.3) FREQUENCIES AND #'S OF POLES FOR FIVE MORE FILTERS
Lines
1 _
1 _
1 _
5-13: RESPONSE CURVE in 3 blocks of 3 lines
80: (10G8.3) FREQ.
80: (10G8.3) AMPL. (REL. 1.0 HZ)
80: (10G8.3) PHASE,
The are only for info in option 1
OPTION 2
If character 78 is T, use tablated values, format as in option 1
OPTION 3
If character 78 is P, option 2:
Line 3
1 – 22:
(1X,2I5,G11.4) 1. NUMBER OF POLES
2. NUMBER OF ZEROS
244
3. NORMALIZATION CONSTANT, COUNTS/M
2 Poles in pairs of real and imaginary parts
23 - 80 (5G11.4)
Lines 4-13
1 – 80: (G11.4) Remaining poles and zeros. 7 values are written
and then 3 spaces are left blank.
Example of SEISAN FAP response file
KBS
B
Z100
1
1
1
0
0
0.000
360.
.700
.260E+04 0.
.419E+06.684E+10 0.
0.000 0.
0.000
0.
0.00 0.
0.00 0.
0.00 0.
0.00 0.
0.00
.500E-02.700E-02.980E-02.140E-01.190E-01.270E-01.370E-01.520E-01.730E-01.100
.480E-02.694E-02.978E-02.140E-01.190E-01.270E-01.370E-01.520E-01.730E-01.100
138.366 123.400 113.340 106.128 101.813 98.283 96.034 94.289 93.054 92.229
.140
.200
.280
.390
.550
.770
1.10
1.50
2.10
2.90
.140
.200
.280
.390
.550
.770
1.10
1.50
2.10
2.90
91.592 91.114 90.796 90.571 90.405 90.289 90.203 90.149 90.106 90.077
4.10
5.80
8.10
11.0
16.0
22.0
31.0
43.0
60.0
85.0
4.10
5.80
8.10
11.0
16.0
22.0
31.0
43.0
60.0
85.0
90.054 90.038 90.028 90.020 90.014 90.010 90.007 90.005 90.004 90.003
Example of SEISAN PAZ response file using the same constants as above
KBS
B
Z100
2
0.
0.
0.
1
1
1
0
0
0.000
P
3 0.1089E+10-0.1222E-01 0.1246E-01-0.1222E-01-0.1246E-01
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
GSE response file format
Below is an example of a GSE response file generated by RESP for the same parameters as used for
the SEISAN file above. The GSE response format is rather complex, and can contain parts that will
not be understood by SEISAN.
Below follows an example of the GSE response using FAP. The first line gives station and sensor
type. The numbers following are gain in nm/c (0.15) at reference period (1. seconds), sample rate
(20) and date. The following line (FAP2) gives a gain factor (1.) and the output units (V for Volts).
Finally follows the frequency, gain and phase triplets.
CAL2 KBS
B
FAP2 1 V
0.00500
0.00590
0.00700
0.00830
0.00980
0.01200
0.01400
…
60.00000
71.00000
85.00000
100.00000
Z
STS-1
0.15E+00
60
0.48047733E-02
0.57846815E-02
0.69361408E-02
0.82667302E-02
0.97841555E-02
0.11995634E-01
0.14000172E-01
138
130
123
117
113
108
106
0.59999992E+02
0.70999992E+02
0.84999992E+02
0.99999985E+02
90
90
90
90
1.
20.00000 2003/ 1/ 1
0: 0
In the poles and zero representation, the file looks like:
CAL2
KBS
PAZ2
1
B
V
Z
STS-1
0.15E+00
0.26000000E-05
1.
20.00000
2
2003/
3
1/
Laplace
1
0:
0
transform
-0.12217305E-01 0.12464144E-01
-0.12217305E-01 -0.12464144E-01
0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00
DIG2 2 0.41900000E+06
20.00000 quanterra
The first line is the same as before. The PAZ2 lines has the following meaning: Normalization
constant of seismometer and filter (0.26e-5), number of poles and zeroes (2 and 3), and the type of
response (Laplace transform). The DIG2 line has the gain of amplifier and AD converter combined
(0.419e6 c/V) and sample rate.
245
In the simplest case, the response is given by the PAZ and a scaling factor. It is common (like in
SEED) to have two scaling constants, one that normalizes the PAZ to amplitude 1 at a calibration
period, and another constant that gives the amplitudes in the physical units. This is NOT the case with
the GSE2 format. The GSE2 response for PAZ normally contains at least two parts, the CAL2 line
and a PAZ2 line. The scaling factor should scale the PAZ to output/input units, NOT normalize. In the
CAL2 line, the system sensitivity at a calibration period is given in units input/output, but is generally
not needed. The total response is given by the PAZ, multiplied with the PAZ2 scaling factor, or the
product of several stages.
This is how SEISAN reads the response, however, if it finds that the PAZ2 gives normalized values at
the calibration period, the response is multiplied with the sensitivity given in the CAL2 line (this is
done because such GSE files have been seen).
FIR filters can be specified in GSE as an additional stage and can be written out by the RESP
program. An example is given below. The FIR filter coefficients are required to completely describe
the instrumentation. However, they are not used in SEISAN.
FIR2 3 0.10E+01 3 0.030 A 180
0.18067658E-06 0.88848173E-06 0.24195760E-05 0.37699938E-05 0.32037496E-06
-0.18581748E-04 -0.69605186E-04 -0.16990304E-03 -0.32499805E-03 -0.51408820E-03
-0.68242848E-03 -0.75194239E-03 -0.65436028E-03 -0.37627667E-03 0.94138085E-05
0.35409257E-03 0.49653835E-03 0.35531260E-03 -0.29224902E-05 -0.37382543E-03
…
246
INDEX
.
.SEISAN
.SEISAN_RES
11
11
3
3-component stations
59, 90, 155
A
Acceleration
89
Add waveform format
6
AGA
5
Agency
12, 65, 138, 203
Agency, select for
120
AMP
87
Amplitude
88
Amplitude attenuation
201
Amplitude, automatic
88
Andalucia
137, 143, 148
Andalusia
149
Angle of incidence
161
Apparent velocity
67, 88, 90, 181
APPEND
130
Append another event to current event
35
ARCFETCH
144
ARCSEI
144
Arcsei.def
144
Array processing
214
Array stations
59
ASCII files
134, 149
ASCII waveform format
142
ASCII, convert to
149
ASCII, convert to SEISAN
146
ASCII, SEISAN to
134
ASCSEI
146
ASSOCI
171
Associ_rest.out
171
Associate events in time
38
Attenuation equation
199
Attenuation relations
189
AUTO
155
AUTO_LOCATE
106
AUTO_PROCESS
106
Automate EPIMAP
113
Automatic amplitude
88
Automatic coda length
88, 106
Automatic phase picking
39, 155
AUTOPIC
17
AUTOPIC.INP
155
Autopic.out
155
AUTOREG
130
Autoscale depth profile
111
Autoscaling
92
Autosig
157
Autosig.par
A-value
Average station residual
Azimuth
Azimuth error
159
191
67
88, 155
67
B
Backazimuth
91
Background noise
99, 185
Background seismicity
199
Bandpass filter, default in mulplt
106
BASE
31
Bash
11
BGISEI
146
Binary SEISAN waveform file structure 241
Bisquare weighting
67
BOUCH
171
Bouch.out
176
BOUSEI
28, 171
Brune spectrum
230
BUL
125
BUL.INP
17, 126
Bul.ps
126
Bullen table
73
Bulletin
125
Bulletin program
17
B-value
152, 191, 199
BVALUE
152
Bvalue.eps
153
Bvalue.out
153
C
CAL directory
55, 148
Calibration file
221
CAT directory
5
CAT_AGA
137, 138
Cat_aga.par
138
Catalog work 6, 29, 56, 57, 123, 129, 191, 205
Catday.out
129
CAT-file
5, 8, 26, 68
Cathour.out
129
Catmonth.out
129
CATSTAT
129
Catyear.out
129
Change filters
80, 104
Change of day
69
Channel order
85
Channel selection
74, 79
Check
230
Check data base
31
CHECK_BASE
31
CHECKRE
31
China
146
Chronological order
124
247
CITSEI
146
CityShark
146
CLUSTER
200
CNVSSR and CNVSSA
147
Coda length
66, 87, 88, 106, 201
Coda magnitude
60, 66, 119
CODAQ
29, 165
Codaq.inp
165
Codaq.par
165
COLLECT
122
COLOR.DEF
17
Command to the operating system
37
Commands in EEV
33
Common tasks in SEISAN
26
Compact file
8, 123
Compile programs
18
Componen.for
88
Component codes
7
Component names in S-file
88
Composite fault plane solution
36, 163
Compression
7
Conrad interface
66, 67
Consistency check
67
Console window
54
Cont
86
Continuous data
8, 76, 86, 134
Continuous plotting
75
Contour file
112
Contour maps
200
Conversion of component codes
142
Conversion of station codes
142
Conversion programs
137
Convert angles
161
Convert to ASCII
149
Converted magnitude
203
Copy events
35
Copy_wav.out
133
Corner frequency
97, 197
Correct polarity
136
Correlation
226
Crash, check base
31
CRIATT
197
CRIPAR
192, 198
CRISEI
112
CRISIS99
190, 192
Crustal model
16
Crustal parameters
181
Crustal structure determination
63
CSS format
146
D
Daily number of events
Damped signal
DAT directory
Data base content
Data base error
Data base, check
Database generation
Database name
129
230
16
31
36
31
16
16
Database security
Database structure
Dataseis
DC level
Default channel
Default filter
Defaults
DEL directory
DELET database
Delete automatic picks
Delete event
Delete ID-line
Delete many S-files
Delete S-file
Delete waveform file
Depth as function of rms
Depth profile
Depth, fixing
Depth, select for
Depth, start value
Different databases
Digital data
Dimension
Diminution function
Dip
DIRF
Displacement
Distance id
Distance indicator
Distance order
Distance range, select for
Distance weighting
DR3024 and DR3016
DRSEI
Drum recording
Duplicate event
Duplicate ID
Duration magnitude
30
41, 57
147
81
105
106
105
4
35, 41
86
31, 35
124
125
86
86, 91
38
111
62
119
62
41
30
13, 88
95, 197
162
125
89
119
86
85
120
62
146
146
75
35
30, 123
66
E
Earthquake catalogue
189
Earthquake statistics
127, 129
EDBPICK picking program
139
Editor
12
EEV
26, 32
EEV over several months
36
EEV print command
38
EEV, jump to another month and database 37
EEV.OUT
35
End time
16
Environmental variable
11
Epicenter area selection
113
EPIMAP
16, 45, 109, 127
Epimap.cor
113
Epimap.inp
113
Epimap.num
113
EQRISK
190, 198
EQRSEI
112, 190, 200
Error estimate, HYP
68
248
Error in S-file
36
Errors in hypocenter
120
EUROPE.MAP
16
Event felt, select for
119
Event ID
86
Event type
5, 86
Event, duplicate ID
123
EXFILTER
205
Explosion
36, 56, 86, 119
Explosion filtering
205
Extract data in JSEISAN
49
Extract data in MULPLT
92
Extract from waveform file
92
Extract waveform files
131, 133
Extracting events from the database
122
F
Fault plane solution
36, 42
Fault plane solution, composite
162
Fault plane solution, plot
116
Fault plane solution, plotting several
161
FDSN
149
Federation of Digital Seismographic Networks
149
Felt information
225
File conversion and modification programs137
File number
74, 75
File types
8
FILENR.LIS
9, 56, 125
Filter and spectral analysis
80
Filter in WAVETOOL
132
Filter, high pass
76
Filter, instability
76
Filter, low pass
76
Filters in MULPLT
80
Fixed scaling
92
Fixing depth
62, 66, 67
Fixing location
62, 66
Fixing origin time
62, 66
FK analysis
90, 214
Focmec.eps
162
Focmec.inp
161
Focmec.lst
162
Focmec.out
162
Format
7
Format, mulplt
75
Format, Nordic
234
Format, volcanic information
212
Format, waveform
239
Fourier acceleration spectra
197
Free surface effect
197
G
Gain factor
Gap range, select for
GBV recorder
GCF format
Generic Mapping Tools
Geometrical spreading
240
120
147
147
141
95, 197, 202
GeoSig
GeoSys
Geotech Instruments
GET_WAV
Get_wav.out
GhostView
GIINOR
GIISEI
Global event location
Global phases
GMT
GMTMAP
GMTXY
Go to event by number or date
Graphics
Graphics problem on PC
Graphics problem on UNIX
Ground displacement
Ground motion seismogram
GSE
GSE format
GSE response file format
GSERESP
GSN
GSRSEI
Güralp
GURSEI
147
147
148
133
133
15
138
147
61
88
57, 141, 212
36
57, 130, 230
35
57, 230
16
13
202
89
6, 55
140
245
147
147
147
147
147
H
H_model.par
H_model_out
Hard copy files
Hardcopies
Hardcopy in single trace mode
Hardwired constants
Header times
Help in SEISAN
Herkij_distance
Herrmann
Herrmann’s package
HERSEI
High accuracy
High frequency decay factor
High pass filter
HINNOR
Hints on modelling
Homepage
Horizontal component
HYP output in S-file
HYP, do not locate
HYP, multiple model mode
Hyp.cat
Hyp.out
HYP_ISC
HYPINV
HYPNOR
HYPO style weight
HYPO71
HYPO71 files to Nordic files
63
63
12
75
85
104
136
51
20, 105
28, 171
217
171
20, 31
197
80
138
176
3
90
65
61
63
124
162
73
72
138
88
36, 65, 69
138
249
Hypocenter program
Hypocenter, add to S-file
Hypocenters, more than one
Hypoinverse
Hypoinverse format
Hypoinverse to Nordic
Hypsum.out
59
36
56
36, 72
140
138
68
I
IASP
180
Iasp.out
180
IASP91.HED
9, 179
IASPEI phases
59, 67
IASPEI Seismic Format
140
IASPEI software library
148
IASPEI91
9, 61, 67, 88
Iceland
150
ID, lock
124
Identifying seismic phases
88
ID-line
5, 86, 124
IL
36
IMS1.0
140
Index file
9, 33, 35, 41, 57, 120
Index.out
120
Indexeev.out
35, 166
Input file for CRIATT
197
Input file for EQRISK
198
Input files for the CRISIS99
193
INPUTEPI
36, 88
INPUTONE
36
INPUTONE.
88
Inserting events in the database
122
Instrument response
54, 217
Interactive work
32
International Monitoring System
140
Inversion
208
inversion of surface wave group velocities217
Invertratio
210
INVRAD
37, 41, 42
IRIS
147, 149, 150
IRISEI
147
ISC
61, 138, 139, 140
ISC location
36, 73
ISC to Nordic
138
Iscloc.def
73
ISCNOR
138
ISCSTA
139
ISF format
140
ISFNOR
140
ISMSEI
147
Israel
137, 147
J
Java
Java, start command
JHD
Joint hypocenter determination
jseisan
JSEISAN
229, 230
12
208, 210
208, 210
12
43
JSEISAN, configuration parameters
JSEISAN, configure
51
44
K
K2SEI
KACSEI
Kappa, determine
Kinemetrics
KINSEI
KW2ASC
147
147
183
139, 147
147
147
L
LD_LIBRARY_PATH
Least squares
Lennartz
LENPCQ
LENSEI
Lg phase
Lg phase velocity
Lg waves
Linking programs
Linux file structure
Linux, compiler options
Local database, update
Local index file
Local magnitude
Local magnitude parameters
LOCAL_CAL
Locate event
Locate event in MULPLT
Locate two events together
Locate with Bullen table
Locate with ISC program
Location code
Location programs
Location with one station
Location, max distance
Location, min # of stations
Lock ID
LOG directory
Log file
Long phase names
Low pass filter
Lower bound magnitude
LSQ
Lsq.eps
Lsq_gmt.out
11, 12
130
148, 149
148
148
59
67
180
18
242
18
124
41, 57
59, 119, 202
67
12, 55
37, 50
89
37
73
36, 73
8
59
59
67
68
124
5
123, 124
61, 89, 237
80
191
130, 230
130
130
M
M88SEI
MacOSX
Macroseismic information
MAG
Mag_amp.out
Mag_coda.out
Mag_mag.out
Mag_new.out
Mag_newa.out
Mag_spec.out
148
13
37, 225
98, 127
203
203
203
203
203
203
250
Magnitude
59, 66, 67, 201
Magnitude conversion
203
Magnitude correction
68
Magnitude in epimap
109
Magnitude order
127
Magnitude relation
200
Magnitude residuals
65
Magnitude weight
66
Magnitude, fixed
60
Magnitude, move from 2 header to first 127
Magnitude, select for
120
Magnitudes without type
119
Magnitudes, more than 3
56, 60
Mailing lists
3
Majordomo
3
Makefile
18
MAKEREA
16, 130
MAP
45
Map files
16, 141
MARS88
148
Maximum expected magnitude
191
Maximum likelihood b-value
153
Maximum no of iterations
66
Maximum number of traces
13
Maximum observed magnitude
191
Mb
60, 90, 119
Mc
60, 201
Merge catalogs
56
Merge events near in time
171
Merge waveform files
91, 135
MERGE_WAVEFORM
92
Merging files from different networks
135
Ml
59, 67, 201
MODEL.DEF
63
Modelling parameters, Q
66
Modes of operation
26
Moho
66, 67
Moment
97, 183, 197
Moment tensor inversion
41
Monthly epicenter files
123
Monthly number of events
129
Mount Unix disk on PC
16
Mouse, test
230
Moving files between Sun and PC
9
Ms
60, 119
MULPLT
16, 30, 37, 73
Mulplt, format
75
Mulplt, option cont
86
MULPLT.DEF
17, 85, 104
MULPLT.OUT
99
Multiple event S-file
122
Multiplexed files
30
Multitrace mode
75
Mw
98
N
Nakamura
Name of the local Index file
Nanometer
183
57
59, 89
Nanometrics
148
NANSEI
12, 148
NDC
140
Nearest station
62
NEIC
148
NEISEI
148
Network code, put in waveform file
142
New events into SEISAN
86
NEWEVE
30
Noise spectrum
96, 99, 183, 184
Noise study
99, 185
Nordic format
5, 8, 234
Nordic to HYPO71
140
NORGSE
140
Norhin
72
NORHIN
140
NORHYP
127, 140
NORIMS
140
Normalization constant
221
O
Omega square
Operator ID
Order of magnitudes
ORFEUS
OS9SEI
Other databases
Output file from CRIATT
Output file from EQRISK
Output files from CRISIS99
Output, corrected trace
197
123
127
143
148
5
198
200
196
92
P
P phase
Parameter files
PCEQ format
PCK files
PCQSEI
PCSUDS
PDAS files to SEISAN
PDASEI
PDE
PDE bulletin file
PDE e-mail
PDENOR
Peak ground acceleration
Peak ground velocity
Period
Perl
Peterson noise model
Pg phase
PGA
Phase AMP
Phase mouse key
Phase name
Phase name key
Phase name length
Phase picking
Phase picking mode
59
16, 17
148
139
148
150
148
148
140
137
141
140
190
190
88
230
99, 183
59
190
87
105
89, 236
105
236
27, 48, 73, 87
85
251
Phase reading indicator
74
Phase weight key
105
Phase, compare theoretical
36
Phase, select in data base
120
Picking phases, use of filters
80
PITSA
149, 150
Pitsa program
37
Place names
111
Plot contours
109
Plotting continuous data
86
Plotting epicenters
116
Plotting place names
111
Plotting response files
56
Plotting traces
37
Pn phase
59
Polarity
80, 88, 136, 160
Polarity, correct
136
Polarity, select for
119
Poles and zeros
148, 220, 240
Polygon
205
Polygon, select
120
Polygon, select in
45, 111
Postscript
12, 75
Power spectral density
185
Power spectrum
96, 99
P-phase picking
157
PR_RESP
55, 217
Preprocessing of data
87, 106
Presp
56
Presp.eps
221
Print S-file
38
Print.out file
68
Printer
12
Printing
56
Printing on PC
15
Probabilistic seismic hazard analysis
189
Probable explosion
119
Problem HYP, location
69
Problem HYP, Moho not found
66
Problem MULPLT extract
133
Problem UPD
125
Problem WAVETOOL, zero output
133
Problem, backup files in EEV
36
Problem, Bouchon
173
Problem, EPIMAP
113
Problem, file name length
8
Problem, filenr.lis
125
Problem, filter
76
Problem, Herrmann
172
Problem, HYP
61, 63, 65
Problem, instrument correction
90
Problem, MULPLT area
111
Problem, MULPLT spectra
99
Problem, no location
61
Problem, PC
16
Problem, picking amplitude
88
Problem, q-correction
98
Problem, reading synthetic phase
89
Problem, RESP
218
Problem, rotation
100
Problem, Rotation and removing response 91
Problem, select
122
Problem, single station location
59
Problem, UPD
125
Problem, UPDATE
124
Problem, VELEST
210
Profile, hypocenters
111
Profile.num
113
PSNSEI
148
Public Seismic Network
148
Q
Q, determine by spectral fitting
98
QLg
180
QNX
148
QNXSEI
148
Quality factor
165, 197, 201
Quality factor, correct for
183
Quality factor, determine by spectra
183
Quality factor, determine with coda
165
Quanterra
147
R
Radiation pattern
197
Rake
162
Rdseed
133
RDSEED
149
REA directory
5
REA.LOG
31
Read S-file
229
Recompile programs
18
Recover files from the DELET database
35
REFTEK
144, 150
REG_KEEP_AUTO
86
Regional attenuation
197
Register event
130
Register event, keep phases
20
Registration
74
Registration and preprocessing
87
Relative spectra
184
Rename event type
38
Reorder Hypocenters
137
REPORT
98, 126
Report.inp
127
Report.out
127
Report_n.out
127
Reporting agency
66
RESAMP
134
Resampling
134
RESET TEST
65, 66
Residual weight
68
Resolution
106
RESP
55, 217
Response file
9, 54, 89, 240, 243
Response in header
240, 243
Response removal
48, 184
Response, from where
90
Response, GSE
147
252
Response, show curves
Rg phase velocity
RMS as a function of depth
RMSDEP
Rotated seismograms
RSAM
RSASEI
RT_SEIS
RTU.INI
90
67
38
38
90
133, 213
148, 149
144
149
S
S phase
59
SAC
6, 150
SAC libraries
18
SACAUX
12
SACLIB
12
SACSEI
149
SAMBA
16
Sample rate
134
SAMPLE_GRAPHICS
230
SAMPLE_READ_CONT
229
SAMPLE_READ_WAV
229
Search
38
Search by day and hour
35
Searching headers in waveform files
136
Searching in the data base
44, 117
SEED
6, 149
SEED data, time gap
75
SEED response
55
SEED, extract data
133
SEIASC
134
SEIASC format
239
SEIBINASC
230
SEICUT
135
SEIDEL
135
Seidim.inc
13
SEIGMT
141
SEIM88A
149
SEIPITSA
149
SEISAF
150
SEISAN ASCII format
239
Seisan extension
13
SEISAN shortcut
15
SEISAN waveform file
8
SEISAN waveform files to or from ascii 134
SEISAN.DEF
18, 44, 51, 104
Seisan.ini
15
SEISAN_EDITOR.
12
SEISAN_TOP
12
SEISANPR
15
SEISARCH
12
seisconf
12
SEISEI
135
SeisGram
149
SEISLOG
148
Seismic design spectra
189
Seismic hazard analysis
190
Seismic moment, average
69
Seismic noise
99
Seismic risk related programs
189
Seismic source zones
189
Seismogram
75, 76
Seismometer constants
219
SEISNET
28, 86
SELECT
6, 41, 45, 117
Select epicenters in an area
113
Select events from database
35
Select for phase
120
Select.inp
121
Select.out
117
Selection in polygon
111
Selection on errors
120
SELMAP
141
SELSEI
136
Selstat.lis
208
Serial number, instrument
142
SESAME
150
S-file
5, 8, 26, 29, 122
S-file, error
36
S-file, parameters
229
S-file, print
38
S-file, update
38
S-files, collecting
122
sformat
12
SFORMAT
229
Sg phase
59
SGRSEI
149
Short user guide
26
Signal duration
157
Signal, damped
230
Signal.out
92
SILSEI
150
Simulate drum recording
75
Single channel files
135
Single files
5
Single station location
59
Single trace mode
85
Sismalp
150
SISSEI
150
Site effects
189
Sn phase
59
Soil amplification
183
Sonic boom
119
Source displacement spectrum
95
Source function
197
Source parameters
95
Source radius
96
Source time function
176
S-P and L-S differences
61
Spain
141, 148
SPEC
182
Spec.par
183
Spectral analysis
182
Spectral fitting
98
Spectral parameters 105, 127, 157, 183, 203
Spectral parameters, update
124
Spectral ratio
183
Spherical harmonic station corrections
66
253
SPLIT
Split an event
Sprengnether
SSA, Kinemtrics
SSR, Kinemetrics
STA/LTA
Stand-alone programs
Start and end date
Start depth
Start time
Starting location
Station and channel codes
Station azimuth
Station coordinates
Station listing in print.out
Station selection
Station, only 4 characters
STATION0.HYP
STATIS
Statistics
STATSTAT
Stress drop
Strike
Subset of database
SUD2ASC
SUDS
SUDS format
SUDSEI
Sun and PC differences
Surface wave analysis
Surface wave dispersion
SUSE Linux
Swarm, identify
Swarm,seismic
Swarm.out
SYNT
Synthetic seismograms
System response
30, 122
35
146
146
146
157
57
57
62
16
62, 67
221
91
16, 139
68
139
138
16, 65
127
127, 129
129
96, 97, 197
162
57
150
150
242
150
7, 241
96, 217
95
18
129
129
129
173
171
54
T
T and P-axis
Terra Technology
TERSEI
Test graphics
Test mouse
Test programs
Test signals, make
Theoretical phases
Threshold magnitude
Time correction
Time delay
Time gap
Time, uncertain flag in waveform file
Timing error
Timing indicator
TMP
Top directory
T-phase
Trace plotting
161
150
150
230
230
229
230
36, 89
191
136
136
75
148
76, 80
240
7
11
67
27, 73, 107
TRACE_PLOT
Travel time error
TSIG
TTLAYER
107
67
230
179
U
Uncertain time
80
Unformatted read and write
241
Units
96, 218
Unknown type
119
UPD
125
UPDATE
32, 38, 41, 57, 98, 123, 124
Update header line
60
Update spectral parameters
124
Update without relocation
124
Updated S-files
38, 123
Upper and lower case
9
Upper bound magnitude
199
UPPER case
88
USERCOM
38
USGS
141
USGSNOR
141
Using azimuth phases
67
V
VELEST
Velest for Windows
Velest.cmn
VELMENU
Velocity
Velout.dif
Version
Volcanic
Volcanic event
Volcanic tremor
VOLCANO.DEF
VOLCSTAT
Vp/Vs ratio
Vp/Vs, calculate
208
208
208
208
89
210
18
119
86
212
212
212
66
181
W
W_EMAP
Wadati
WADATI
WAV data base
WAV directory
Waveform conversion programs
Waveform file format
Waveform file name
Waveform file structure
Waveform file, cut
Waveform file, delete
Waveform file, extract from
Waveform file, fix headers
Waveform file, split
Waveform files
Waveform files, join
Waveform files, merge
Waveform.for
117
39
181
7
6
142
239
136
241
135
91
92
136
135
6, 120
130
91
229
254
Waveform.inc
Waveform_names.out
WAVETOOL
WAVFIX
Wavfix.tim
WAVFULLNAME
WEB options
Weight
WGSSEI
Wildcard
Windows
Windows configuration
Windows NT
Windows95
WKBJ
Wood Anderson
Working directory
WORLD.MAP
Write S-file
229
120
131
130
136
137
121
62, 67, 88
150
125
52
15
15
15, 26
171, 172
89, 202
53
16
229
X
X Windows
Xnear and xfar
11
66
Y
Y5DUMP
Yearly number of events
148
129
Z
Zoom in JSEISAN
Zoom in map
Zoom in MULPLT
48
45
79
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement