EUMETSAT ASCAT Wind Product User Manual

ASCAT Wind Product User Manual
Ocean and Sea Ice SAF
Version 1.4
March 2008
SAF/OSI/CDOP/KNMI/TEC/MA/126
DOCUMENT SIGNATURE TABLE
Name
Prepared by :
Date
O&SI SAF Project
Signature
March 2008
Team
Approved by :
O&SI SAF Project
March 2008
Manager
DOCUMENTATION CHANGE RECORD
Issue / Revision
Version 0.1
Version 1.0
Version 1.1
Version 1.2
Date :
Feb 2007
Mar 2007
Sep 2007
Oct 2007
Change :
Minor
Minor
Minor
Version 1.3
Oct 2007
Minor
Version 1.4
Mar 2008
Minor
Description :
Draft version
First external version
Added release note (sec. 1.4)
Single transponder calibrated L1 data
(version 5.2.1)
Take some comments of ORR-A part 1 into
account
Three-transponder calibrated L1 data
(version 5.5.0)
KNMI, De Bilt, the Netherlands
Reference: SAF/OSI/CDOP/KNMI/TEC/MA/126
Cover illustration: ASCAT winds over the Atlantic on 12 February 2007, around 21 hour UTC (red arrows). In this
picture, only half of the ASCAT wind vectors, with 25-km cell spacing, are plotted. The blue and purple arrows
show a 3-hour forecast of the winds by the KNMI High-Resolution Limited Area Model (HiRLAM). In the
background, a METEOSAT infrared image is shown in black and white.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
1
SAF/OSI/CDOP/KNMI/TEC/MA/126
Contents
Introduction.......................................................................................................................3
1.1.
Overview...................................................................................................................3
1.2.
Access to data and helpdesk....................................................................................3
1.3.
Disclaimer .................................................................................................................4
1.4.
References ...............................................................................................................4
1.5.
Useful links ...............................................................................................................5
1.6.
Limitations of the 25-km ASCAT winds ....................................................................6
2. The ASCAT scatterometer instrument..............................................................................7
3. Algorithms.........................................................................................................................8
3.1.
Wind definition ..........................................................................................................8
3.1.1.
Geophysical model function..............................................................................9
3.2.
Wind retrieval..........................................................................................................10
3.2.1.
Ambiguity removal ..........................................................................................10
3.2.2.
Quality control.................................................................................................10
4. Processing scheme ........................................................................................................11
4.1.
Backscatter calibration ...........................................................................................11
4.2.
NWP collocation .....................................................................................................11
4.3.
Validation................................................................................................................11
4.4.
Quality control and monitoring ................................................................................11
5. Dissemination .................................................................................................................13
6. Data description..............................................................................................................14
6.1.
Wind product characteristics...................................................................................14
6.2.
File formats .............................................................................................................15
7. Data quality.....................................................................................................................17
7.1.
Accuracy .................................................................................................................17
7.2.
Reliability and data use...........................................................................................18
7.3.
Ambiguity selection.................................................................................................18
8. Glossary .........................................................................................................................19
9. ASCAT BUFR data descriptors ......................................................................................20
1.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
2
SAF/OSI/CDOP/KNMI/TEC/MA/126
1. Introduction
1.1.
Overview
The Ocean and Sea Ice Satellite Application Facility (OSI SAF) produces a range of air-sea
interface products, namely: wind, sea ice characteristics, Sea Surface Temperatures (SST)
and radiative fluxes, Surface Solar Irradiance (SSI) and Downward Long wave Irradiance
(DLI).
KNMI is involved in the OSI SAF and the EUMETSAT Advanced Retransmission Service
(EARS) ASCAT service as the centre where the Level 1b to Level 2 processing is carried
out. This document is the Product User Manual to the ASCAT wind products. Quality
monitoring information on this product and more general information on the whole OSI SAF
project, is available on the OSI SAF web site at the following address: http://www.osisaf.org/. The user is strongly encouraged to register on this web site in order to receive the
service messages and the latest information about the OSI SAF products.
The wind products will be distributed in two resolutions: a 50-km resolution product with 25km cell spacing and a 25-km resolution product with 12.5-km cell spacing. Both resolutions
will be available as a regional EARS product with a timeliness of approximately 30 minutes
from sensing time and as a global OSI SAF product with a timeliness of approximately 2.5-3
hours from sensing time.
The scatterometer is an instrument that provides information on the wind field near the ocean
surface, and scatterometry is the knowledge of extracting this information from the
instrument’s output. Space-based scatterometry has become of great benefit to meteorology
and climate in the past years, see e.g. [Ref-1].
KNMI has a long experience in scatterometer processing and is developing generic software
for this purpose. Processing systems have been developed for the ERS, NSCAT, SeaWinds
and ASCAT scatterometers. Scatterometer processing software is developed in the
EUMETSAT Numerical Weather Prediction Satellite Application Facility (NWP SAF), whereas
wind processing is performed (semi-)operationally in the Ocean and Sea Ice SAF (OSI SAF)
and in the EARS project.
EUMETSAT makes available near real-time Level 1b scatterometer products from the MetOp
satellite through EUMETCast (global data) and through a private network (regional data).
These products are used as basis for further processing at KNMI.
The OSI SAF and EARS products are delivered on request through the KNMI FTP server to
all users and through the EUMETCast system. See also http://www.knmi.nl/scatterometer/
for real-time graphical examples of the products and up-to-date information and
documentation.
This user manual outlines user information for the OSI SAF and EARS products based on
the ASCAT scatterometer. Section 2 presents a brief description of the ASCAT instrument,
section 3 the processing algorithms, and section 4 gives an overview of the data processing
configuration. Section 5 provides details on how to access the products. Detailed information
on the file content and format is given in section 6, while in section 7 the product quality is
elaborated.
1.2. Access to data and helpdesk
These wind products are disseminated through the EUMETCast system; for access please
contact EUMETSAT (http://www.eumetsat.int/). The products are also available on a
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
3
SAF/OSI/CDOP/KNMI/TEC/MA/126
password protected FTP server. Contact the KNMI scatterometer team ([email protected]) for
access.
For a swift response management procedure, user requests on the OSI SAF data products
should be issued at the Ocean and Sea Ice SAF website (http://www.osi-saf.org/). User
requests on the EARS products can be sent to [email protected].
1.3.
Disclaimer
All intellectual property rights of the OSI SAF and EARS products belong to EUMETSAT.
The use of these products is granted to every interested user, free of charge. If you wish to
use these products, EUMETSAT's copyright credit must be shown by displaying the words
"copyright (year) EUMETSAT" on each of the products used.
1.4.
References
[Ref-1]
Isaksen, L., and A. Stoffelen, 2000, “ERS-Scatterometer Wind Data Impact on
ECMWF’s Tropical Cyclone Forecasts”, IEEE-Transactions on Geoscience and
Remote Sensing (special issue on Emerging Scatterometer Applications) 38 (4), pp.
1885-1892
[Ref-2]
Valenzuela, G. R., Theories for the interaction of electromagnetic and ocean waves
- a review, Bound. Layer Meteor., 13, 612-685, 1978
[Ref-3]
Thesis “Scatterometry” by Ad Stoffelen, 1998, available on
http://igitur-archive.library.uu.nl/dissertations/01840669/inhoud.htm
[Ref-4]
Hersbach, H., Stoffelen, A. and de Haan, S., CMOD5 - An improved geophysical
model function for ERS C-band scatterometry, J. Geophys. Res. , accepted, 2007.
[Ref-5]
Donelan, M. A., and W. J. Pierson, Radar scattering and equilibrium ranges in windgenerated waves with application to scatterometry, J. Geophys. Res., 92, 49715029, 1987
[Ref-6]
Donelly, William J., James R. Carswell, and Robert E. McIntosh, Revised ocean
backscatter at C and Ku band under high wind conditions, J. Geophys. Res., 104,
11,485-11,497, 1999
[Ref-7]
Pierson, W.J., Probabilities and statistics for backscatter estimates obtained by a
scatterometer, J. Geophys. Res., 94, 9743-9759, 1989; correction in J. Geophys.
Res., 95, 809, 1990
[Ref-8]
Stoffelen, Ad, Siebren de Haan, Yves Quilfen, and Harald Schyberg, ERS
Scatterometer Ambiguity Removal Comparison, OSI SAF report, 2000, available on
http://www.knmi.nl/scatterometer/publications/
[Ref-9]
Portabella, M., Stoffelen, A., Quality Control and Wind Retrieval for SeaWinds,
EUMETSAT fellowship report, 2002, available on
http://www.knmi.nl/scatterometer/publications/
[Ref-10] de Vries, J., Stoffelen, A. and Beysens, J., Ambiguity Removal and Product
Monitoring for SeaWinds, NWP SAF report NWPSAF_KN_TR_001, available on
http://www.knmi.nl/scatterometer/publications/
[Ref-11] EUMETSAT, ASCAT Level 1 Product Generation Function Specification,
EUM.EPS.SYS.SPE.990009
[Ref-12] Bonekamp, H., ASCAT BUFR File name conventions, EUM/MET/TEN/06/0377
[Ref-13] ADDITIONS TO BUFR/CREX TABLES FOR PRE-OPERATIONAL
IMPLEMENTATION ENDORSED BY CBS for full operational status on 7 November
2007 (updated 04/01/07), pp55-60, available on
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
4
SAF/OSI/CDOP/KNMI/TEC/MA/126
http://www.wmo.int/web/www/WMOCodes/Updates/BUFRCREX/Preoperational050
107.doc
[Ref-14] Verspeek, J., Portabella, M., Stoffelen, A. and Verhoef, A., ASCAT Calibration and
Validation, OSI SAF Technical Report, SAF/OSI/CDOP/KNMI/TEC/TN/163
[Ref-15] Figa, J., and Stoffelen, A., 2000, “On the Assimilation of Ku-Band Scatterometer
Winds for Weather Analysis and Forecasting”, IEEE-Transactions on Geoscience
and Remote Sensing (special issue on Emerging Scatterometer Applications) 38
(4), pp. 1893-1902
[Ref-16] Graham, R., Anderson, D. Hollingsworth, A. and Böttger, H., Evaluation of ERS-1
wind extraction and ambiguity removal algorithms: meteorological and statistical
evaluation, ECMWF report, ECMWF, Reading, England, 1989
[Ref-17] Mastenbroek, K., “Wind-Wave Interaction”, thesis at the Delft University of
Technology, Delft, the Netherlands, 12 December 1996
[Ref-18] Stoffelen, A., A Generic Approach for Assimilating Scatterometer Observations,
ECMWF seminar, 2000, available on http://www.knmi.nl/scatterometer/publications/
[Ref-19] Thesis “Wind Field Retrieval from Satellite radar systems” by Marcos Portabella,
available on http://www.knmi.nl/scatterometer/publications/
[Ref-20] Portabella, M. and Stoffelen, A., Development of a Global Scatterometer Validation
and Monitoring, Visiting Scientist report for the Ocean & Sea Ice SAF, January
2007.
1.5.
Useful links
KNMI scatterometer web site: http://www.knmi.nl/scatterometer/
- Information on OSI SAF activities at KNMI: http://www.knmi.nl/scatterometer/osisaf/
- Information on EARS activities at KNMI: http://www.knmi.nl/scatterometer/ears/
- ASCAT visual products: http://www.knmi.nl/scatterometer/ascat_osi_25_prod/,
http://www.knmi.nl/scatterometer/ascat_ear_25_prod/,
http://www.knmi.nl/scatterometer/ascat_osi_12_prod/,
http://www.knmi.nl/scatterometer/ascat_ear_12_prod/ (not all products and links are
available yet)
Information on EARS and EUMETCast: http://www.eumetsat.int/
O&SI SAF wind product documentation on http://www.osi-saf.org/:
Scientific documents:
- 2D Variational Ambiguity Removal
Technical documents
- Science Plan
- OSI SAF User Requirements Document
- OSI SAF Project Plan
- OSI SAF Report on Algorithm Development and Prototyping Activities
- OSI SAF top-level Configuration Management Plan
- OSI SAF Output Products Format Document
- OSI SAF Software Requirements Document
- ICD for the wind production of the OSI SAF
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
5
SAF/OSI/CDOP/KNMI/TEC/MA/126
- SVVP for the wind production of the OSI SAF
NWP SAF website: http://www.metoffice.gov.uk/research/interproj/nwpsaf/index.html
EUMETSAT ASCAT documentation:
http://www.eumetsat.int/Home/Main/What_We_Do/Satellites/EUMETSAT_Polar_System/Spa
ce_Segment/SP_1139325735344?l=en and
http://www.eumetsat.int/Home/Main/Publications/Technical_and_Scientific_Documentation/T
echnical_Notes/SP_1126189367518?l=en
1.6.
Limitations of the 25-km ASCAT winds
Currently, although we have good confidence in the quality of the products, the ASCAT
instrument is still in its cal/val stage and small adjustments in the products can be expected.
Product quality parameters are available in near-real time if you click on the ‘Monitoring
information’ link on (http://www.knmi.nl/scatterometer/ascat_osi_25_prod/).
The following restrictions and limitations hold:
1) Although the Level 1 product is fully calibrated now using three transponders on the
ground, the backscatter values are corrected using a calibration table, derived from a
geophysical wind model. The backscatter values in the Level 2 wind product are however
plain copies of those in the Level 1 product. The corrections are very small and expected to
be incorporated into a new version of the Geophysical Model Function.
2) The Sigma-0 Usability Flag in the Level 1 product is set to “unusable” in all Wind Vector
Cells. Therefore, this flag is neglected in the wind processing.
3) The normalisation of the Maximum Likelihood Estimator (MLE), which is reflected in
“Backscatter Distance” in the output product, is to be adjusted and will change after the
commissioning phase.
4) The Quality Control (threshold for setting of the KNMI quality control flag) needs to be
tuned.
5) The Product Monitoring flag is not yet implemented (see section 7.2).
These restrictions and limitations are subject to further study during the ASCAT
commissioning phase in the next few months.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
6
SAF/OSI/CDOP/KNMI/TEC/MA/126
2. The ASCAT scatterometer instrument
The Advanced SCATterometer (ASCAT) is one of the instruments carried on-board the
Meteorological Operational (MetOp) polar satellites launched by the European Space
Agency (ESA) and operated by the EUropean organisation for the exploitation of
METeorological SATellites (EUMETSAT). MetOp-A, the first in a series of three satellites,
was launched on 19 October 2006 and two successors are planned to be launched
approximately 4-5 years after each other.
ASCAT is a real aperture radar using vertically polarised antennas. It transmits a long pulse
with Linear Frequency Modulation (‘chirp’). Ground echoes are received by the instrument
and, after de-chirping, the backscattered signal is spectrally analysed and detected. In the
power spectrum, frequency can be mapped into slant range provided the chirp rate and the
Doppler frequency are known. The above processing is in effect a pulse compression, which
provides range resolution.
Two sets of three antennas are used to generate radar beams looking 45 degrees forward,
sideways, and 45 degrees backwards with respect to the satellite's flight direction, on both
sides of the satellite ground track. These beams illuminate 550 km-wide swaths (separated
by about 700 km) as the satellite moves along its orbit, and each provide measurements of
radar backscatter from the sea surface on a 25 km or 12.5 km grid, i.e. each swath is divided
into 21 or 41 so-called Wind Vector Cells (WVCs). The result is three independent
backscatter measurements for each WVC, obtained using the three different viewing
directions and separated by a short time delay. As the backscatter depends on the sea
surface roughness as a function of the wind speed and direction at the ocean surface, it is
possible to calculate the surface wind speed and direction by using these 'triplets' within a
mathematical model.
Figure 1: ASCAT wind
scatterometer geometry (source:
EUMETSAT web site).
The instrument operates at a frequency of 5.255 GHz (C-band), which makes it rather
insensitive to rain.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
7
SAF/OSI/CDOP/KNMI/TEC/MA/126
3. Algorithms
Scatterometry was developed heuristically. It was found experimentally that the sensitivity to
wind speed and direction describe well the changes in backscatter over the ocean at
moderate incidence angles due to changes in surface roughness, as depicted in figure 2
[Ref-2]. In return, backscatter measurements can be used to determine the wind speed and
wind direction in a WVC.
A schematic illustration of the processing is given in figure 3. After defining the wind output
and motivating the Geophysical Model Function that is used, the algorithms developed at
KNMI are described.
Figure 2: Schematic
representation of microwave
scattering and reflection at a
smooth (a), rough (b) and very
rough (c) ocean surface. As the
roughness increases more
microwave power is returned
towards the direction of the
microwave source.
3.1.
Wind definition
A scatterometer measurement relates to the ocean surface roughness (see figure 2), while
the scatterometer product is represented by the wind at 10m height over a WVC. It is
important to realize that in the approach followed here the radar backscatter measurement
σ 0 is related to the wind at 10 meter height above the ocean surface, simply because such
measurements are widely available for validation. This means that any effect that relates to
the mean wind vector at 10 meter height is incorporated in the backscatter-to-wind
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
8
SAF/OSI/CDOP/KNMI/TEC/MA/126
relationship. As such, air stability, the appearance of surface slicks, and the amplitude of
gravity or longer ocean waves, depend to some degree on the strength of the wind and may,
to the same degree, be fitted by a geophysical model function, GMF ([Ref-3]; Chapter I).
Stoffelen ([Ref-3]; Chapter IV) discusses a unique method to determine the accuracy of
scatterometer, buoy, and NWP model winds.
3.1.1. Geophysical model function
For the OSI SAF and EARS ASCAT wind products the CMOD5.5 geophysical model function
(GMF) is used. This model function enables the calculation of wind speeds meeting the
product requirements between 0 and 25 m/s. CMOD5.5 is based on CMOD5 [Ref-4]. It is
known from extensive validation work on ERS that a CMOD5 bias of 0.5 m/s against buoys
persists for all wind speeds [Ref-20] and this is compensated in CMOD5.5. The CMOD5.5
inverted winds are thus 0.5 m/s higher than CMOD5 winds.
At low wind speeds the wind direction and speed may vary considerably within the WVC.
Locally, below a speed of roughly 2 ms−1 calm areas are present where little or no
backscatter occurs, perhaps further extended in the presence of natural slicks that increase
the water surface tension [Ref-5]. However, given the variability of the wind within a footprint
area of 25 km it is, even in the case of zero mean vector wind, very unlikely that there are no
patches with roughness in the footprint. As the mean vector wind increases, the probability of
a calm patch will quickly decrease, and the mean microwave backscatter will increase. Also,
natural slicks quickly disappear as the wind speed increases, and as such the occurrence of
these is correlated to the amplitude of the mean vector wind over the footprint, as modelled
by the GMF. Low scatterometer wind speeds are thus providing useful information.
NWP collocation
Figure 3: Overview of wind retrieval algorithm
Quality Control
Wind computation by GMF Inversion
Fieldwise
Ambiguity Removal
At high wind speeds wave breaking will further intensify, causing air bubbles, foam and spray
at the ocean surface, and a more and more complicated ocean topography. Although
theoretically not obvious, it is empirically found that σ 0 keeps increasing for increasing wind
speed from 25 m/s to 40 m/s, and that a useful wind direction dependency remains [Ref-6],
albeit gradually weakening.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
9
SAF/OSI/CDOP/KNMI/TEC/MA/126
3.2.
Wind retrieval
The GMF has two unknowns, namely wind speed and wind direction, so, if more than two
backscatter measurements are available then these two unknowns may be estimated using a
maximum-likelihood estimator (MLE) as the objective function for determining wind vector
solutions [Ref-7]. The MLE is defined by ([Ref-3]; Chapter II)
J = ( z oi − z m (u , χ i )) 2 ,
where z = (σ 0)0.625 are the transformed backscatter data, zoi are the backscatter
measurements, zm(u,χi) are the model backscatter values corresponding to the
measurements. The local minima of J correspond to wind vector solutions. The three
independent measurements (fore, mid and aft beam) well sample the azimuth variation of the
GMF in order to resolve the wind direction, albeit ambiguously.
3.2.1. Ambiguity removal
ASCAT scatterometer winds have a multiple ambiguity and there are up to two wind
solutions in each WVC on the earth’s surface. These ambiguities are removed by applying
constraints on the spatial characteristics of the output wind field, such as on rotation and
divergence. Several ambiguity removal (AR) schemes were evaluated for ERS data [Ref-8]
in the OSI SAF Development Phase. In addition to the subjective comparison of AR
schemes, a method for the objective comparison of AR performance among the different
schemes was used. In [Ref-8] it is shown that this way of comparison is effective to evaluate
the shortcomings of AR schemes, but also reveals a more general way forward to improve
AR, which is followed up by tuning 2D-VAR. For ASCAT this tuned version of 2D-VAR is
used.
3.2.2. Quality control
Since the scatterometer wind retrieval problem is over-determined, this opens up the
possibility of quality control (QC) by checking the inversion residual J. If J is normalised by
the expected isotropic error variance then it is in theory inversely proportional to the log
probability that a WVC is affected solely by a uniform wind. If Var(σm)i = (Kpi2σoi2) are the
measurement variances then the norm for the inversion residual is √3 times the RMS of
(Var(σm)i)0.625 (Stoffelen, 1998). Generally this normalised MLE is substantial and, as a
consequence, the inferred probability low, when there is substantial wind or sea state
variability within the WVC.
As such, Stoffelen [Ref-3] and Portabella and Stoffelen [Ref-9] found that the inversion
residual is well capable of removing cases with extreme variability (at fronts or centres of
lows), or with other geophysical variables affecting the radar backscatter.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
10
SAF/OSI/CDOP/KNMI/TEC/MA/126
4. Processing scheme
KNMI has a processing chain running in near real-time with ASCAT data, including
visualisation on the internet. This processor is based on the NWP SAF software and runs in
the KNMI operational environment. The processing includes monitoring and archiving
functionalities. A global overview of the modules of the ASCAT scatterometer processor is
given below.
4.1.
Backscatter calibration
During the calibration/validation period, the backscatter values in the Level 1 product are
calibrated by adding a WVC and beam dependent bias in dB to the incoming σ 0s. The
calibration table was obtained by fitting the actual measurements to the theoretical GMF
function. More details are provided in [Ref-14]. Note that the calibrated backscatter values
are only available within the wind processing software; the σ 0 data in the wind product are
identical to those in the Level 1 product.
4.2.
NWP collocation
KNMI receives NWP model data from ECMWF twice a day through the RMDCN.
NWP model sea surface temperature and land-sea mask data are used to provide
information about possible ice or land presence in the WVCs. WVCs with a sea surface
temperature below 272.16 K (-1.0 °C) are assumed to be covered with ice and no wind
information is calculated. Land presence within each WVC is determined by using the landsea mask available from the model data. The weighted mean value of the land fractions of all
model grid points within 80 km of the WVC centre is calculated. The weight of each grid point
scales with 1/r2, where r is the distance between the WVC centre and the model grid point. If
this mean land fraction value exceeds a threshold of 0.02, no wind retrieval is performed.
Also the land fractions present in the beam information in the level 1b product are
considered: if any land fraction in the fore, mid or aft beam exceeds 0.02, no wind retrieval is
performed.
NWP forecast wind data are necessary in the ambiguity removal step of the processing.
Wind forecasts are available twice a day (00 and 12 GMT analysis time) with forecast time
steps of +3h, +6h, …, +36h. The model wind data are linearly interpolated with respect to
time and location and put into the level 2 information part of each WVC (see section 6.2).
4.3.
Validation
Each step in the processing is validated separately by a quality control and monitoring
scheme. The product validation step is controlled by visual inspection, and a statistical
analysis is performed to control the validation steps. The inversion step is controlled in the
same way. For ambiguity removal schemes an objective scheme exists that relies on
initialisation with a one-day lead NWP forecast and validation of the ambiguity selection
against NWP analyses, as in [Ref-8]. Moreover, de Vries et al [Ref-10] describe subjective
comparison of the 2D-VAR and PreScat schemes by routine operational meteorologists.
4.4.
Quality control and monitoring
In each WVC, the σ 0 data is checked for quality and completeness and the inversion
residual (see section 3.2.2) is checked. Degraded WVCs are flagged; see section 6.2 for
more details.
The quality of the delivered products is controlled through an ad hoc visual examination of
the graphical products and the automatic production of control parameters.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
11
SAF/OSI/CDOP/KNMI/TEC/MA/126
The examination of the products is done at KNMI by experts. Specific tools have been
developed to help this analysis. User queries obviously lead to the inspection of suspect
products. The ad hoc and user-queried inspections are used for quality assurance.
An information file is made for each product. The content of the file is identical whatever the
product and results from a compilation of all the global information concerning this product.
From these files, various graphs are produced to visually display the confidence levels of the
products and their evolution with time. These graphs are available on the KNMI website.
Data quality is also available to the users within the products; see section 6 and 7 for a
description of quality flags.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
12
SAF/OSI/CDOP/KNMI/TEC/MA/126
5. Dissemination
The ASCAT BUFR L2 products will be disseminated on EUMETCast. Please consult
http://www.eumetsat.int/, under ‘Access to Data’ for more information on EUMETCast
dissemination and how to receive these and other EUMETSAT meteorological satellite
products.
The products are also made available on a password-protected ftp site. This password is
provided to new users by Email request. Please send your requests to [email protected].
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
13
SAF/OSI/CDOP/KNMI/TEC/MA/126
6. Data description
6.1.
Wind product characteristics
Physical definition
Horizontal wind vector at 10m height.
Units and range
Wind speed is measured in m/s and wind direction in meteorological (WMO) convention
relative to North: 0 degrees corresponds to a wind flowing to the South with a clockwise
increment. The wind speed range is from 0-50 m/s, but wind speeds over 25 m/s are
generally less reliable [Ref-6].
Input satellite data
ASCAT level 1b data from EUMETSAT are described in their technical documentation
[Ref-11]. The global ASCAT data are acquired in Svalbard (Norway) and transmitted to the
EUMETSAT central processing facilities in Darmstadt, where they are processed up to level
1b. Regional EARS data are acquired locally at various stations, transmitted to the
EUMETSAT central processing facilities, processed up to level 1b and then transmitted to
KNMI via a dedicated private network. The product contains geo-located measurement
triplets on a satellite swath WVC grid of 25 km or 12.5 km size.
Geographical definition
The MetOp satellite flies in a near-polar sun synchronous orbit at 98 degrees inclination at
approximately 800 km orbit height. The two satellite swaths are located to the left and to the
right of the satellite ground track. The swath width is either 21 25-km size WVCs,
corresponding to 525 km or 41 12.5-km size WVCs, corresponding to 512.5 km. Products
are organised in messages of 42 or 82 WVCs, respectively (one row of data).
Coverage
The global OSI SAF product has a global coverage. For the regional EARS product, data is
available only when the satellite is in sight of a ground station. Most ground stations are in
the North Atlantic region. Actual coverage is available on the OSI SAF and EARS ASCAT
product visualisation websites (see http://www.knmi.nl/scatterometer/).
Output product
The input product in BUFR is processed into a BUFR output product including a unique wind
solution (chosen), its corresponding ambiguous wind solution and quality information
(distance to cone, quality flag). The 25-km cell spacing product contains two wind solutions,
whereas the 12.5-km cell spacing product contains only one wind solution (the chosen one).
Delivery time
A wind product is available for distribution within 10 minutes after the input product reception
at KNMI. For the global product, the delivery time between acquisition of the data and
availability for the user ranges from 100 to 130 minutes. The regional product will be
available within 30 minutes after data acquisition.
Expected accuracy
The expected accuracy is defined as the expected bias and standard deviation of the primary
calculations. The accuracy is validated against in situ wind measurements from buoys,
platforms, or ship, and against NWP data. Even better, the errors of all NWP model winds, in
situ data, and scatterometer winds are computed in a triple collocation exercise [Ref-3]. The
performance is pretty constant over the globe and depends mainly on the sub footprint wind
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
14
SAF/OSI/CDOP/KNMI/TEC/MA/126
variability. The performance of the products issued by the OSI SAF and EARS is
characterised by a wind component RMS error smaller than 2 m/s and a bias of less than
0.5 m/s in wind speed.
6.2.
File formats
Wind products are in BUFR. A complete description of BUFR can be found in WMO
publication No 306, Manual on Codes. The graphical displays of the wind products are
available and explained on the web: see the links on http://www.knmi.nl/scatterometer/.
The file name convention for the Level 2 product is
ascat_YYYYMMDD_HHMMSS_metopa_ORBIT_SRV_T_SMPL(_CONT).l2_bufr, where
ƒ
ascat denotes the instrument
ƒ
YYYYMMDD denotes the date of the first data in the file
ƒ
HHMMSS denotes the time (UTC) of the first data in the file
ƒ
metopa denotes the satellite name
ƒ
ORBIT is the orbit number (00000-99999)
ƒ
SRV is the service (eps for global OSI SAF or ear for regional EARS)
ƒ
T is the processing type (o for (pre)operational, t for test)
ƒ
SMPL is the WVC sampling (cell spacing): 250 or 125
ƒ
CONT (optional field) refers to the product contents: ovw for a product containing only
Ocean Vector Winds and no soil moisture information, CONT is omitted if the product
contains both winds and soil moisture.
ƒ
l2_bufr (l is the lowercase L) denotes Level 2 product in BUFR
Examples of file names are
ascat_20070213_021503_metopa_01653_eps_t_250.l2_bufr for a global test product
containing soil moisture information, or
ascat_20070213_021503_metopa_01653_ear_o_125_ovw.l2_bufr for a regional operational
product containing no soil moisture information.
The filename convention follows the EUMETSAT proposal as described in [Ref-12]. The wind
product is stored in the BUFR format as proposed for ASCAT and described in [Ref-13], a list
of descriptors (fields) contained in each WVC is provided in section 9.
The BUFR data contain three main sections: level 1 information (fields 1-62), level 2 Soil
Moisture information (fields 63-82) and level 2 wind information (fields 83 and up). The L1
information is simply copied into the L2 data. Currently, the Soil Moisture information is not
yet available and the corresponding fields are set to missing values. The Soil Moisture
information will become available at a later stage.
Field 84 (‘Generating Application’) contains the value 91 which means that first guess model
winds are used for ambiguity removal. The interpolated model winds are in the fields 85-86.
The Wind Vector Cell Quality Flag (field 89, table 021155) has the following definitions:
Description
Not enough good sigma-0 available for wind retrieval
Poor azimuth diversity among sigma-0 for wind retrieval
Any beam noise content above threshold
Product monitoring not used
O&SI SAF/EARS
BUFR bit
1
2
3
4
Fortran bit
22
21
20
19
ASCAT Wind Product User Manual version 1.4 March 2008
15
SAF/OSI/CDOP/KNMI/TEC/MA/126
Description
Product monitoring flag
KNMI quality control fails
Variational quality control fails
Some portion of wind vector cell is over land
Some portion of wind vector cell is over ice
Wind inversion not successful for wind vector cell
Reported wind speed is greater than 30 m/s
Reported wind speed is less than or equal to 3 m/s
Rain flag for the wind vector cell is not usable
Rain flag algorithm detects rain
No meteorological background used
Data are redundant
Distance to GMF too large
Reserved
Missing value
BUFR bit
5
6
7
8
9
10
11
12
13
14
15
16
17
18-23
All 24 set
Fortran bit
18
17
16
15
14
13
12
11
10
9
8
7
6
5-0
All 24 set
In Fortran, if the Wind Vector Cell Quality is stored in an integer I then use BTEST(I,NDWNB-1) to test BUFR bit NB, where NDW=24 is the width in bits of the data element in BUFR.
If the ‘monitoring not used’ bit (BUFR bit 4) is set to zero, the product is monitored. If the
product is monitored and the ‘monitoring flag’ bit (BUFR bit 5) is set to zero, the product is
valid; otherwise it is rejected by the monitoring, see section 7.2. The monitoring bits are set
to the same value in all WVCs in one BUFR output file.
If the KNMI QC flag (BUFR bit 6) is set in a WVC this means that the backscatter information
is of poor quality for various reasons, such as a too large inversion residual, or a too high
noise value in the input product. WVCs, in which the KNMI QC flag is set, are not used in the
calculation of the analysis field in the ambiguity removal step. However, after the ambiguity
removal the wind solution closest to the analysis field is chosen (if wind solutions are present
in the WVC). This means that such a WVC may contain a selected wind solution, but it is
suspect.
Land presence flag is set if a land fraction (see section 4.2) larger than zero is calculated for
the WVC. As long as the land fraction is below the limit value, a reliable wind solution may
however still be present.
Ice presence flag is set if the SST calculated for the WVC (see section 4.2) is below 272.16
(-1.0 °C).
If the variational QC flag is set, the wind vector in the WVC is rejected during ambiguity
removal due to spatial inconsistency. A wind solution is present, but it may be suspect.
It is recommended not to use WVCs with the monitoring flag, the KNMI quality control flag or
the variational quality control flag set. See section 7.2 for more information on product
reliability.
The ‘likelihood computed for solution’ (descriptor 021104) actually contains the log10 of the
calculated likelihood for the wind solution. This is done since otherwise values close to zero
will be rounded to zero in the BUFR encoding. In order to recalculate the probability, the user
should compute 10 to the power <value from BUFR>.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
16
SAF/OSI/CDOP/KNMI/TEC/MA/126
7. Data quality
7.1.
Accuracy
As introduced in section 6.1, the accuracy should be better than 2 m/s in wind component
RMS with a bias of less than 0.5 m/s in wind speed.
20
15
10
5
0
270
180
90
0
0
5
10
15
20
Wind speed (m/s) − ECMWF
0
N=3267229
mx= 7.66 my= 7.66
m(y−x)= −0.00 s(y−x)= 1.26
cor_xy= 0.94
360
Node: all
V comp. (m/s) − Z0.625
20
10
0
−10
−20
−20
−10
0
10
20
U comp. (m/s) − ECMWF
N=3267229
mx= −0.15 my= −0.26
m(y−x)= −0.12 s(y−x)= 1.45
cor_xy= 0.98
90
180
270
Wind dir (deg) − ECMWF
N=2792250
mx= 177.72 my= 178.57
m(y−x)= 0.86 s(y−x)= 14.93
cor_xy= 0.99
Node: all
U comp. (m/s) − Z0.625
Figure 4: Contoured
histograms of the 25-km
ASCAT wind product.
Node: all
Wind dir (deg) − Z0.625
Wind speed (m/s) − Z0.625
Node: all
20
10
0
−10
−20
−20
−10
0
10
20
V comp. (m/s) − ECMWF
N=3267229
mx= 0.78 my= 0.73
m(y−x)= −0.05 s(y−x)= 1.63
cor_xy= 0.95
Figure 4 shows two-dimensional histograms of the retrieved winds versus ECMWF 10m wind
background for the 25-km wind product, with the backscatter calibration (see section 4.1)
applied. The data for these plots are from consecutive orbits from 23 August 2007 until 31
August 2007.
The top left plot corresponds to wind speed (bins of 0.4 m/s) and the top right plot to wind
direction (bins of 2.5°). The latter are computed for ECMWF winds larger than 4 m/s. The
bottom plots show the U and V wind component statistics. N is the number of data; mx and
my are the mean values along the x and y axis, respectively; m(y-x) and s(y-x) are the bias
and the standard deviation with respect to the diagonal, respectively; and cor_xy is the
correlation value between the x- and y-axis distributions. The contour lines are in logarithmic
scale: each step is a factor of 2 and the lowest level (outer-most contour line) is at N/64000
data points.
From these results, it is clear that the spread in the distributions is small. The wind speed
bias is quite small and it is clear from the bottom plots that the standard deviations in the
components are well below 2 m/s. Note that the data in the plots include WVCs that are
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
17
SAF/OSI/CDOP/KNMI/TEC/MA/126
quality controlled, i.e. have a too large inversion residual. This means that in the end product,
the average quality will be even better, since WVCs of poor quality are flagged and can be
filtered out.
7.2.
Reliability and data use
For global coverage products, it is possible to generate a product monitoring flag, based on a
multi-step check. If in one product the number of WVC Quality Control rejections, the mean
residual, or the wind speed bias with respect to the NWP background is above certain
threshold values, then the monitoring event flag is raised since the product is suspicious. The
threshold values are based on evaluation of the product statistics over a long period [Ref-10].
Because of the granular nature of the EARS product, where files containing only a few
minutes of data are generated and disseminated in real-time, this mechanism raises
problems if there are not enough data in one pass for a statistically valid check. The multistep monitoring check is sensitive to noise which is larger when mean values are calculated
over fewer WVCs. If, accordingly, the thresholds are set high, many bad products will pass
the check. On the other hand, if the threshold values are set too low, too many false alarms
will be raised. This problem can be solved probably by evaluating not only the data of the last
processed pass, but by evaluating the last 60 or 70 minutes of data present, although
originating from typically 4 or 5 passes of 15 minutes. In case of instrument degradation or
other problems, the monitoring event flag will be raised with some delay, inherent to the
discontinuous nature of the data stream. Sensitivity tests will provide guidance on this issue.
It is planned to implement the monitoring event flag in a future version of the global and
regional products.
7.3.
Ambiguity selection
A version of 2D-VAR is used with minimal regional performance differences [Ref-10]. This
improved version was obtained after taking into account the findings of [Ref-8]. A variational
QC step is performed to reject a few WVCs, which are in meteorological unbalance with their
neighbours. The variational QC flagged WVCs are flagged in the output product.
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
18
SAF/OSI/CDOP/KNMI/TEC/MA/126
8. Glossary
AR
Ambiguity Removal
ASCAT
Advanced Scatterometer
BUFR
Binary Universal Format Representation
EPS
Eumetsat Polar System
ERS
European Remote-Sensing Satellite
EUMETCast EUMETSAT’s Digital Video Broadcast Data Distribution System
HDF
Hierarchical Data Format
JPL
Jet Propulsion Laboratory (NASA)
KNMI
Royal Netherlands Meteorological Institute
MetOp
Meteorological Operational satellite
NASA
National (US) Air and Space Agency
NOAA
National (US) Oceanic and Atmospheric Administration
NSCAT
NASA Scatterometer
NWP
Numerical Weather Prediction
OSI SAF
Ocean and Sea Ice SAF
QC
Quality Control
QuikSCAT
US dedicated scatterometer mission
RMDCN
Regional Meteorological Data Communication Network
SAF
Satellite Application Facility
SeaWinds
US scatterometer on-board QuikSCAT platform
SM
Soil Moisture
SOS
Scatterometer Ocean Stress
SST
Sea Surface Temperature
SVVP
Software Verification and Validation Plan
U
West-to-east wind component
V
South-to-north wind component
WVC
Wind Vector Cell
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
19
SAF/OSI/CDOP/KNMI/TEC/MA/126
9. ASCAT BUFR data descriptors
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Descriptor
001033
001034
025060
001007
002019
001012
004001
004002
004003
004004
004005
004006
005001
006001
005033
005040
006034
010095
021157
021150
008085
002111
002134
021062
021063
021158
021159
021160
021161
021162
021163
021164
021165
021166
008085
002111
002134
021062
021063
021158
021159
021160
021161
021162
021163
021164
021165
021166
008085
002111
002134
O&SI SAF/EARS
Parameter
Identification Of Originating/Generating Centre
Identification Of Originating/Generating Sub-Centre
Software Identification
Satellite Identifier
Satellite Instruments
Direction Of Motion Of Moving Observing Platform
Year
Month
Day
Hour
Minute
Second
Latitude (High Accuracy)
Longitude (High Accuracy)
Pixel Size On Horizontal-1
Orbit Number
Cross Track Cell Number
Height Of Atmosphere Used
Loss Per Unit Length Of Atmosphere Used
Beam Collocation
Beam Identifier
Radar Incidence Angle
Antenna Beam Azimuth
Backscatter
Radiometric Resolution (Noise Value)
ASCAT Kp Estimate Quality
ASCAT Sigma-0 Usability
ASCAT Use Of Synthetic Data
ASCAT Synthetic Data Quality
ASCAT Satellite Orbit And Attitude Quality
ASCAT Solar Array Reflection Contamination
ASCAT Telemetry Presence And Quality
ASCAT Extrapolated Reference Function
ASCAT Land Fraction
Beam Identifier
Radar Incidence Angle
Antenna Beam Azimuth
Backscatter
Radiometric Resolution (Noise Value)
ASCAT Kp Estimate Quality
ASCAT Sigma-0 Usability
ASCAT Use Of Synthetic Data
ASCAT Synthetic Data Quality
ASCAT Satellite Orbit And Attitude Quality
ASCAT Solar Array Reflection Contamination
ASCAT Telemetry Presence And Quality
ASCAT Extrapolated Reference Function
ASCAT Land Fraction
Beam Identifier
Radar Incidence Angle
Antenna Beam Azimuth
Unit
Code Table
Code Table
Numeric
Code Table
Code Table
Degree True
Year
Month
Day
Hour
Minute
Second
Degree
Degree
M
Numeric
Numeric
m
dB/m
Flag Table
Code Table
Degree
Degree
dB
%
Code Table
Code Table
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
Code Table
Degree
Degree
dB
%
Code Table
Code Table
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
Code Table
Degree
Degree
ASCAT Wind Product User Manual version 1.4 March 2008
20
SAF/OSI/CDOP/KNMI/TEC/MA/126
Number
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Descriptor
021062
021063
021158
021159
021160
021161
021162
021163
021164
021165
021166
025060
025062
040001
040002
021062
021151
021152
021153
021154
021062
021088
040003
040004
040005
040006
040007
020065
040008
040009
040010
025060
001032
011082
011081
020095
020096
021155
021101
021102
031001
011012
011011
021156
021104
011012
011011
021156
021104
Parameter
Backscatter
Radiometric Resolution (Noise Value)
ASCAT Kp Estimate Quality
ASCAT Sigma-0 Usability
ASCAT Use Of Synthetic Data
ASCAT Synthetic Data Quality
ASCAT Satellite Orbit And Attitude Quality
ASCAT Solar Array Reflection Contamination
ASCAT Telemetry Presence And Quality
ASCAT Extrapolated Reference Function
ASCAT Land Fraction
Software Identification
Database Identification
Surface Soil Moisture (Ms)
Estimated Error In Surface Soil Moisture
Backscatter
Estimated Error In Sigma0 At 40 Deg Incidence Angle
Slope At 40 Deg Incidence Angle
Estimated Error In Slope At 40 Deg Incidence Angle
Soil Moisture Sensitivity
Backscatter
Wet Backscatter
Mean Surface Soil Moisture
Rain Fall Detection
Soil Moisture Correction Flag
Soil Moisture Processing Flag
Soil Moisture Quality
Snow Cover
Frozen Land Surface Fraction
Inundation And Wetland Fraction
Topographic Complexity
Software Identification
Generating Application
Model Wind Speed At 10 m
Model Wind Direction At 10 m
Ice Probability
Ice Age (A-Parameter)
Wind Vector Cell Quality
Number Of Vector Ambiguities
Index Of Selected Wind Vector
Delayed Descriptor Replication Factor
Wind Speed At 10 m
Wind Direction At 10 m
Backscatter Distance
Likelihood Computed For Solution
Wind Speed At 10 m
Wind Direction At 10 m
Backscatter Distance
Likelihood Computed For Solution
Unit
dB
%
Code Table
Code Table
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
%
%
dB
dB
dB/Degree
dB/Degree
dB
dB
dB
Numeric
Numeric
Flag Table
Flag Table
%
%
%
%
%
Numeric
Code Table
m/s
Degree True
Numeric
dB
Flag Table
Numeric
Numeric
Numeric
m/s
Degree True
Numeric
Numeric
m/s
Degree True
Numeric
Numeric
Note that descriptor numbers 93-96 can be repeated 1 to 144 times, depending on the value
of the Delayed Descriptor Replication Factor (descriptor number 92)
O&SI SAF/EARS
ASCAT Wind Product User Manual version 1.4 March 2008
21
">