User manual | Remote Sensing of Urban/Suburban

Review Article
Remote Sensing of
Urban/Suburban Infrastructure and
Socio-Economic Attributes*
John R. Jensen and Dave C. Cowen
Temporal and spatial resolution requirements for extracting
urban/suburban infrastructure and socio-economic attributes
from remote sen.& data are presented. The goal is to relate
the user information requirements with the current and proposed remote sensing systems to determine if there are substantive gaps in capability. Several remote sensing systems
currently provide some of the desired urban/suburban infrastructure and socio-economic information when the required
spatial resolution is poorer than 4 by 4 m and the temporal
resolution is between 1 and 55 days (e.g., Landsat MSS and
Thematic Mapper, SPOTI-4, Russian TK-350, RADARSAT, Indian IRS-ICD, NOAA AVHRR, GOES, Meteosat). Current high
spatial resolution sensor systems such as the Russian SPIN-2
KVR-1000 (2- by 2-m panchromatic; when in orbit) and proposed sensor systems (EOSAT Space Imaging IKONOS 1- by
1-mpanchromatic; Earth Watch Quickbird 0.82 by 0.82 m;
Orbview-3 1 by I m) may provide additional capability.
Large-scale metric aerial photography or digital camera
imagery with spatial resolutions ranging from I 0.25 to 1 m
will still be required to satisfy several important urban/suburban information requirements.
Introduction
Urban landscapes are composed of diverse materials (concrete, asphalt, metal, plastic, glass, shingles, water, grass,
shrubs, trees, and soil) arranged by humans in complex ways
to build housing, transportation systems, utilities, commercial buildings, and recreational landscapes (Welch, 1982;
Swerdlow, 1998). The goal of this construction is usually to
improve the quality-of-life. A significant number of professional businessmen and women and public organizations require up-to-date information about the city and suburban
infrastructure. For example, detailed urban information is required by (Cullingworth, 1997;American Planning Association, 1998):
city, county, and regional planning agencies and councils of
governments that legislate zoning regulations to improve the
* Preliminarv versions were reoorted in T.R. Tensen and D.C. Cowen
(1997),~ e & o t eSensing of ~;ban/~ub&b&Socio-economic Attributes, Proceedings, Land Satellite Information in the Next Decade
11, American Society for Photogrammetry and Remote Sensing, Bethesda, Maryland, compact disk, 19 p., and D.C. Cowen and J.R.
Jensen (1998), Extraction and Modeling of Urban Attributes Using
Remote Sensing Technology, People and Pixels: Linking Remote
Sensing and Social Science, National Academy Press, Washington,
D.C., pp. 164-188.
Department of Geography, University of South Carolina,
Columbia, SC 29208 ([email protected]).
PHOTOGRAMMETRICENGINEERING& REMOTE SENSING
city and state Departments of Commerce to stimulate development;
Tax Assessor offices that maintain legal geographic descriptions of every parcel of land, assess its value, and levy a tax
millage rate;
Departments of Transportation that maintain existing facilities, build new facilities, and prepare for future transportation demand;
private utility companies (water, sewer, gas, electricity, telephone, cable) that attempt to predict where new demand will
occur and plan for the most efficient and cost-effective method
of delivering services;
Public Service Commissions that insure that utility services
are available economically to the public;
Departments of Parks, Recreation and Tourism who improve
recreation facilities and promote tourism;
Departments of Emergency Management and Preparedness
who plan for and allocate resources in the event of a disaster;
private real estate companies attempting to find the ideal location for industrial, commercial, and residential development; and
residential, commercial, and industrial developers.
The urbanlsuburban land these professionals manage or develop is of significant monetary value. Therefore, it is not
surprising that city, county, state, and federal agencies as
well as private companies spend millions of dollars each
year obtaining aerial photography and other forms of remotely sensed data to extract the required urban information.
Much of the required information simply cannot be obtained
through in situ site surveys.
Temporal, Spectral, and Spatial Characteristlcs of Urban
Attributes and Remote Sensing Systems
Many of the detailed urbanlsuburban attributes that businesses and public agencies require are summarized in Table
1. This paper reviews how remotely sensed data may be of
value for collecting information about these attributes. To remotely sense these urban phenomena, it is h s t necessary to
appreciate the urban attributes' temporal, spectral, and spatial resolution characteristics.
Urban/Suburban Temporal Considerations
Three types of temporal resolution should be considered
when monitoring urban environments using remote sensor
data. First, urbanlsuburban phenomena progress through an
identifiable developmental cycle much like vegetation progresses through a phenological cycle. For example, Jensen
-
-
-
Photogrammetric Engineering & Remote Sensing,
Vol. 65,No. 5, May 1999,pp. 611-622.
0099-1112I99I6505-611$3.00/0
O 1999 American Society for Photogrammetry
and Remote Sensing
May 1 9 9 9
611
TABLE1. URBAN/SUBURBAN
A ~ R I B U T EAND
S THE MINIMUMREMOTE SENSING
RESOLUTIONSREQUIREDTO PROVIDE
SUCH
~NFORMAT~ON
Minimum Resolution Requirements
Attributes
Land UseILand Cover
L1-USGS Level I
LZ-USGS Level I1
L3-USGS Level III
L4-USGS Level IV
Building and Property Infrastructure
Bl-building perimeter, area, height and cadastra
information (property lines)
Transportation Inhastructure
TI-general road centerline
T2-precise road width
T3-traffic count studies (cars, airplanes, etc.)
T4-parking studies
Utility Infrastructure
U1-general utility line mapping and routing
UZ-precise utility line width, right-of-way
Us-location of poles, manholes, substations
Digital Elevation Model (DEM) Creation
Dl-large scale DEM
D2-large scale slope map
Socioeconomic Characteristcs
S1-local population estimation
SZ-regionallnational population estimation
S3-quality of life indicators
Energy Demand and Conservation
El-energy demand and production potential
E2-building insulation surveys
Meteorological Data
MI-weather prediction
M2-current temperature
M3-clear air and precipitation mode
M4-severe weather mode
M5-monitoring urban heat island effect
Critical Environmental Area Assessment
C1-stable sensitive environments
C2-dynamic sensitive environments
Disaster Emergency Response
DEl-pre-emergency imagery
DEZ-post-emergency imagery
DEB-damaged housing stock
DE4-damaged transportation
DE5-damaged utilities, services
Temporal
612
M a y 1999
Spectral
5-10 years
5-10 years
3-5 years
1-3 years
V-NIR-MIR-Radar
V-NIR-MIR-Radar
Pan-V-NIR-MIR
Panchromatic
1-5 years
Pan-Visible
1-5 years
1-2 years
5-10 rnin
10-60 rnin
1-30 m
0.25-0.5 m
0.25-0.5 m
0.25-0.5 m
1-5 years
1-2 years
1-2 years
1-30 m
0.25-0.6 m
0.25-0.6 m
Pan-V-NIR
Pan-Visible
Panchromatic
5-10 years
5-10 years
0.25-0.5 m
0.25-0.5 m
Pan-Visible
Pan-Visible
5-7 years
5-15 years
5-10 years
0.25-5 m
5-20 m
0.25-30 m
1-5 years
1-5 years
0.25-1 m
1-5 m
3-25 min
3-25 min
6-10 min
5 rnin
12-24 hr
1-2 years
1-6 months
1-5
12 hr-2
1-2
1-2
years
days
days
days
1-2 days
and Toll (1983) documented a ten-stage single-family residential housing development cycle at work in suburban Denver, Colorado that progressed from (1)rangeland to (10)
fully-landscaped residential housing, often within one year.
The image analyst must understand the temporal development cycle of the urban phenomena. If it is not understood,
embarrassing and costly interpretation mistakes can be made.
The second type of temporal resolution is how often it is
possible for a remote sensor system to collect data of the
urban landscape, e.g., every 8 days, every 16 days, or ondemand. Generally, satellite sensors that can be pointed offnadir (e.g., SPOT HRV) have higher temporal resolution than
sensors that only sense the terrain at nadir (e.g., Landsat
Thematic Mapper). Orbital characteristics of the satellite
platform and the latitude of the study area also impact the
revisit schedule. Remote sensor data may be collected on demand from sub-orbital aircraft (airplanes, helicopters),
weather conditions permitting. up-to-date remote sensor data
are critical for most urbanlsuburban applications.
Finally, temporal resolution may refer to how often land
managerslplanners need a specific type of information. For
example, local planning agencies may require population estimates every 5 to 7 years in addition to the estimates pro-
I
Spatial
Pan-v-NIR
TIR
1-8 km
1-8 km
lkm
1km
5-30 m
V-NIR-TIR
TIR
WSR-88D Radar
WSR-88D Radar
TIR
1-10 m
0.25-2 m
v-NIR-MIR
V-NIR-MIR-TIR
1-5
0.25-2
0.25-1
0.25-1
0.25-1
m
m
m
m
m
vided by the decennial census. The managerial temporal
resolution requirements for many important urban applications are summarized numerically in Table 1 and graphically
in Figure 1.
Urban/Suburban Spectral Considerations
Most image analysts would agree that, when extracting urbanlsuburban information from remotely sensed data, it is
more important to have high spatial resolution (often < 5 by
5 m) than high spectral resolution (i.e., a large number of
multispectral bands). For example, local population estimates
based on building unit counts usually require a minimum
spatial resolution of from I 0.25 to 5 m (0.82 ft to 16.4 ft) to
detect, distinguish between, andlor identify the type of individual buildings. Practically any visible band (e.g., green or
red) or near-infrared spectral band at this spatial resolution
will do. Of course, there must be sufficient spectral contrast
between the object of interest (e.g., a building) and its background (e.g., the surrounding landscape) in order to detect,
distinguish between, and identify the object from its background.
While high spectral resolution is not required, there are
still optimum portions of the electromagnetic spectrum that
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
0.3 0.5
1 rn
5
10
2030
100 m
IRS-I A6
U S S I 72.5 x 72.5
L I S P 38.25 X 38.25
23.5 x 23.5; MIR 70 x 70
W S 188 x 188
SWlR30x30m
TIRQOxQOm
2
-
,
.'
'
..,
.
lm
*
>
-.
'
2 3
5
.
:
;
:
, .:
,
i
j
:
10152030
I
I
o.;dso!s.;l.o
<
:
I I
I
~.
:. :,
i
lkm
:
100m
1000 m
I
I
1 4 1 ' ~ b 1 : i \& i i
'1A3
:
5km lOkm
:
I
do4
Nominal Spatial Resolution in meters
Figure 1. Subjective spatial and temporal resolution requirements for urban/suburban attributes overlaid on
the spatial and temporal resolution capabilities of current and proposed remote sensor systems. Refer to
Table 1for urban/suburban codes. Information presented in this type of diagram will constantly change
due to (a) the development of new remote sensing instruments and their associated temporal and spatial
resolutions, and (b) the user community continuously redefines existing data requirements and identifies
new attributes to be collected.
are especially useful for extracting certain types of urban/
suburban information (Table 1). For example, USGS Level 111
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
-
land cover is best acquired using the visible color (0.4 to 0.7
pn; V), near-infrared (0.7 to 1.1ym; NIR), middle-infrared
May 1999
613
(1.5 to 2.5 pm; MIR), andlor panchromatic (0.5 to 0.7 pm)
portions of the spectrum. Building perimeter, area, and
height information is best acquired using black-and-white
panchromatic (0.5 to 0.7 pm) or color imagery (0.4 to 0.7
pm). The thermal infrared portion of the spectrum (3 to 1 2
pm; TIR) may be used to obtain urban temperature measurements. Active microwave sensors may obtain imagery of
cloud shrouded or tropical urban areas (e.g., Japanese RRS-1
L-band, Canadian RADARSAT C-band, and European Space
Agency ERS-1, 2 C-band).
Urban/Suburban Spatial Conslderatlons
Trained image analysts utilize the tone, color, texture, shape,
size, orientation, pattern, shadow silhouette, site, and situation of objects in the urban landscape to identify and judge
their significance (Jensen, 1996). The geometric elements of
image interpretation (e.g., object shape, size, orientation, pattern, shadow silhouette) are especially useful when high spatial resolution imagery of urban environments are available.
But should we judge the usefulness of a given type of imagery (e.g., aerial photography or Landsat Thematic Mapper
imagery) for extracting very specific types of urbanlsuburban
information based solely on its spatial characteristics? One
solution might be to use the military andlor civilian versions
of the National Image Interpretation Rating Scales (NWS)developed by the Image Resolution Assessment and Reporting
Standards Committee (IRARS). The N I I R ~is the metric used by
the intelligence community to characterize the usefulness of
imagery for intelligence purposes (Leachtenaurer, 1996;
Leachtenaurer et al., 1998; Logicon, 1995; Logicon, 1997;
Pike, 1998). The NWS criteria consist of ten rating levels (0
to 9) for a given type of imagery arrived at through evaluation by trained image analysts. The IRARS committee makes
it clear that spatial resolution (ground resolved distance) is
only one of the measures of the interpretability of an image.
Other factors such as film quality, atmospheric haze, contrast, angle of obliquity, and noise can reduce the ability of a
well trained analyst to detect, distinguish between, and identify military and civilian objects in an image. While it would
be useful to use the NIIRS criteria, it is not optimum for this
review because (1) the civil NWS criteria were only recently
made available (Hothem et al., 1996; Leachtenauer et al.,
1998); (2) there has not been sufficient time for the civilian
community to familiarize itself with the concept; and, consequently, (3) the civilian community has never reported their
collective experiences in urbanlsuburban information extraction during the current and past decades in this context.
Fortunately, the civilian user community has often reported the utility of a given type of imagery for extracting
urban information based on the comparatively easy to understand concept of nominal spatial resolution. When using satellite remote sensing systems, the nominal spatial resolution
(ground resolved distance) of the sensor system is typically
used such as the Landsat Thematic Mapper's six 30- by 30-m
multispectral bands or SPOT Image's 10- by 10-m panchromatic band. Conversely, the figure of merit for measuring resolvability of a film camera system is the area weighted
average resolution (AWAR) measured in line-pairs-per-millimeter (lplrnm) (Light, 1993). A line pair is the width of one
black bar and one white space as contained on resolution
targets in an aerial photograph. Together, they form a pair
and serve as a measure of image quality for the aerial film
camera industry. The five essential elements that make up
the system AWAR are the lens, original film, image blur
(smear) on the film due to aircraft forward velocity, angular
motion, and the resolution of the duplicating film. Also,
scene contrast of the Earth and atmosphere play a role in
system resolution. Fortunately, scientists have studied the
general relationship between aerial photography scale and
AWAR. For example, Light (1993; 1996) documented that, if
we assume that the Earth is a low contrast scene, the 1:
40,000-scale National Aerial Photography Program (NAPP)
photography exhibits approximately 39 lplrnm and yields approximately 25 pm for the size of 1 lp in the image. At 1:
40,000 scale, 25 pm equates to a ground resolution of 1by 1
m for low-contrast scenes. Therefore, a minimum of a 1-m
ground resolution can be expected throughout the photographic mission. In fact, the usGs digital orthophoto quarterquad files produced from 1:40,000-scale NAPP photography
are provided at a 1-by 1-m (3.28- by 3.28-ft) spatial resolution by scanning the photography with a pixel size of 11 vm.
Light (personal communication, 1998) suggests that there is a
general linear relationship for larger scale aerial photography
obtained using metric cameras, i.e., 1:20,000-scale aerial photography equates to approximately 0.5 by 0.5 m (1.64 by 1.64
ft), 1:10,000-scale photography to 0.25 by 0.25 m (0.82 by
0.82 ft), and 1:5,000-scale photography to 0.125 by 0.125 m
(0.41 by 0.41 ft).
Another general spatial resolution rule is that there
needs to be a minimum of four spatial obsenrations (e.g., pixels) within an urban object to identify it. Stated another way,
the sensor spatial resolution should be one-half the diameter
of the smallest object-of-interest. For example, to identify
mobile homes that are 5 m wide, the minimum spatial resolution of high quality imagery without haze or other problems is I 2.5- by 2.5-m pixels (Cowen et a]., 1995).
The temporal, spectral, and spatial resolution requirements for the urban attributes summarized in Table 1 and
Figure 1 were synthesized from subjective, practical experience reported in journal articles, symposia, chapters in
books, and government and society manuals (Chisnell and
Cole, 1958; Stone, 1964; Branch, 1971; Ford, 1979; Jensen,
1983; Avery and Berlin, 1993; Light, 1993; Light, 1996;
Greve, 1996; Jensen, 1996; Philipson, 1997; Haack et al.,
1997; Keister, 1997; Cowen and Jensen, 1998; Pike, 1998;
Ridley et al., 1998; and others in the individual sections).
Ideally, there would always be a remote sensing system that
could obtain images of the terrain that satisfy the urban attributes' resolution requirements (Table 1).Practically, this is
not always the case.
Evaluation of Urban/Suburban Attributes' Spatial and Temporal
Requirements and the Availability of Remote Sensing Systems
to Provide Such Information
The relationship between temporal and spatial data requirements for selected urbanlsuburban attributes and the temporal and spatial characteristics of available and proposed
remote sensing systems is presented in Figure 1.
Land Use/Land Cover
The term land use refers to how the land is being used. Land
cover refers to the biophysical materials found on the land.
For example, a state park may be used for recreation but
have a deciduous forest cover. One method of organizing
land-uselland-cover information is to use a classification system. The most comprehensive hierarchical classification
system for urbanlsuburban land use is the Land-Based Classification Standard (LBCS) under development by the American Planning Association (1998) that updates the 1965
Standard Land Use Coding Manual (Urban Renewal Administration, 1965) which is cross-referenced with the 1987 Standard Industrial Classification (SIC) Manual (Bureau of the
Budget, 1987) and the updated North American Industrial
The LBCS requires extensive
Classification Standard (NAICS).
input from in situ site surveys, aerial photography, and satellite remote sensor data to obtain information at the parcel
level on the following five characteristics: activity, function,
PHOTOGRAMMETRICENGINEERING & REMOTE SENSING
)
land Cover Class Level
Figure 2. The general relationship
between the U.S. Geological Survey
Land Use and Land Cover Classifica
tion System class level and the
nominal spatial resolution of the remote sensing system (often referred
to as ground resolved distance in
meters).
site development, structure, and ownership (American Planning Association, 1998). The system does not provide information on land cover or vegetation characteristics in the
urban environment because it is relying on the Federal Geographic Data Committee "Standards" on these topics. The
LBCS is not complete at this time. Therefore, the following
discussion will focus on the use of the "land use and land
cover classification system for use with remote sensor data"
developed by the U.S. Geological Survey (Anderson et al.,
1976).
The general relationship between USGS land-cover classification system levels (I to IV) and the nominal spatial resolution of the sensor system (ground resolved distance in
meters) is presented in Figure 2. Generally, UsGs Level I
classes may be inventoried effectively using sensors with a
nominal spatial resolution of 2 20 to 100 m such as the
Landsat Multispectral Scanner (MSS) with 79- by 79-m
ground resolution, the Thematic Mapper (TM)at 30 by 30 m,
SPOT HRV (XS) at 20 by 20 m, and Indian LISS 1-3 (72.5 by
72.5 m, 36.25 by 36.25 m, 23.5 by 23.5 m, respectively). For
example, Plate 1depicts typical Level I urban vs. non-urban
land-cover information for Charleston, South Carolina extracted from Landsat MSS data acquired on 26 March 1981 in
red and any new urban development since 11February 1979
in yellow.
Sensors with a minimum spatial resolution of approximately 5 to 20 m are generally required in order to obtain
Level II information. The SPOT HRV and the Russian SPIN-2
TK-350 are the only operational satellite sensor systems pro-METRIC
IIIPINEERINQ & REMOTE SENSING
viding 10- by 10-m panchromatic data. RADARSAT provides
11-by 9-m spatial resolution data for Level I and 11landcover inventories even in cloud-shrouded tropical landscapes. Landsat 7 with its 15- by 15-m panchromatic band
may be launched in 1999.
More detailed Level III classes may be inventoried using
a sensor with a spatial resolution of approximately 1to 5 m
(Welch, 1982; Forester, 1985) such as IRS-lcD pan (5.8- by
5.8-m data resampled to 5 by 5 m) or large scale aerial phoace
tography. Future sensors may include ~ o s ~ ~ / S pImaging
IKONOS (1-by 1-m pan and 4- by 4-m multispectral),EarthWatch Quickbird (0.8- by 0.8-m pan and 3.28- by 3.28-m
multispectral),Orbview 3 and 4 (1-by 1-m pan and 4- by 4m multispectral),and IRS P5 (2.5- by 2.5-m). The synergistic
use of high spatial resolution panchromatic data (e.g., 1by 1
m] merged with lower spatial resolution multispectral data
(e.g., 4 by 4 m) will likely provide an image interpretation
environment that is superior to using panchromatic data
alone (Jensen, 1996).
Level IV classes and building and cadastral (property
line) information are best monitored using high spatial resolution panchromatic sensors, including aerial photography (5
0.25 to 1 m) and, possibly, the proposed EOSAT Space Imaging IKONOS (1by 1m), Earthwatch Quickbird pan (0.8 by 0.8
m), and Orbview-3 (1by 1m) data.
Urban land-uselland-cover classes in Levels I through IV
have temporal requirements ranging from 1to 10 years (Table 1and Figure 1).All the sensors mentioned have temporal
resolutions of less than 55 days, so the temporal resolution
of the land-uselland-cover attributes is satisfied by the current and proposed sensor systems.
Bulldlng and Cadastral (Property Une) Infrastructure
In addition to fundamental nominal scale land-use and landcover information (i.e., identifying whether an object is a sin-
I
Plate 1.Typical Level I urban versus non-urban land cover
information in red derived from a 26 March 1981 Landsat MSS image (79 by 79 m). The yellow areas represent
new urbanization that has taken place since analysis of
another MSS image acquired on 11February 1979. Both
the urban thematic information and the non-urban terrain
(from the near-infrared band 4 MSS data) were draped
over a usos 1:100,000-scale digital elevation model
and viewed from a position 3,000 m above the inlet to
Charleston, South Carolina Harbor. The terrain was vertically exaggerated 100x for visual effect.
May 1999
618
gle-family residence or a commercial building), transportation planners, utility companies, tax assessors, and others
require more detailed information on building footprint perimeter, area, and height; driveways; patios; fences; pools;
storage buildings; and the distribution of landscaping every 1
to 5 years. These building and property parameters are best
obtained using stereoscopic (overlapping) panchromatic aerial photography or other remote sensor data with a spatial
resolution of < 0.25 to 0.5 m (Jensen, 1995; Warner, 1996).
For example, panchromatic stereoscopic aerial photography
with 0.3- by 0.3-m (1ft) spatial resolution was used to extract the building perimeter and area information for the single-family residential area in Figure 3. With this type of data,
each building footprint, patio, outbuilding, tree, pool, driveway, fence, and contour may be extracted. In many instances,
the fence lines are the cadastral property lines. If the fence
lines are not visible or are not truly on the property line, the
property lines are located by a surveyor and the information
is overlaid onto an orthophotograph or planimetric map database to represent the legal cadastral (property) map. Many
municipalities in the United States use high spatial resolution imagery such as this as the source for some of the cadastral information andlor as an image back-drop upon which
surveyed cadastral and tax information are portrayed.
Detailed building perimeter, area, and height data can be
extracted from high spatial resolution (10.25- to 0.5-m) stereoscopic imagery (Jensen et a]., 1996). Such information can
then be used to create three-dimensional displays of the urban terrain that we can walk through in a virtual reality environment if desired (Wolff and Yaeger, 1993; Barnell, 1998).
For example, Figure 4 depicts (1) a large scale vertical aerial
photograph of downtown Columbia, South Carolina, (2) a
digital elevation model (DEM) of the same area extracted from
the stereoscopic photography depicting the height of every
building, and (3) the orthophotograph draped over the DEM
creating a virtual reality representation of a major street. Architects, planners, engineers, and real estate personnel are
beginning to use such information for a variety of purposes.
EOSAT Space Imaging (1999), OrbView (1999), and
Earthwatch (2000) plan to provide fore-aft stereoscopic images from satellite-based platforms with approximately 0.8to l-m spatial resolution. Ridley et al. (1998) conducted a
feasibility study and found that simulated 1-by 1-m satellite
stereoscopic data could "have potential for creating a national 3D building model if the processes were automated,
which would produce a much cheaper source of building
heights." The accuracy of the building maximum heights (z)
ranged between 1.5 and 3 m RMSE. Thus, the use of such imagery may not obtain the detailed planimetric (perimeter,
area) and topographic detail and accuracy (building height
and volume) that can be extracted from high spatial resolution stereoscopic aerial photography (5 0.25 to 0.5 m). Research is required using real 1- by 1-m stereoscopic data
obtained from satellite platforms (Ridley et al., 1998).
Transportation Infrastructure
Engineers often use remote sensor data to (1)update transportation network maps; (2) evaluate road, railroad, and airport runway and tarmac condition; (3) study urban traffic
patterns at choke points such as tunnels, bridges, shopping
malls, and airports; and (4) conduct parking studies (Mintzer,
1983; Haack et al., 1997). One of the more prevalent forms
of transportation data are the street centerline spatial data
(scSD]. Three decades of practice have proven the value of
differentiating between the left and right sides of each street
segment and encoding attributes to them such as street names,
address ranges, ZIP codes, census and political boundaries,
and congressional districts. SCSD provide a good example of
616
May 1999
Figure 3. Planimetric cadastral information of a singlefamily residential area extracted photogrammetrically
from panchromatic stereoscopic vertical aerial photography. Building footprints are in black, fence lines with
x's in black, driveways in white, shrubs and trees in
black, 2-f&contours in continuous black lines, and
highway right-of-way in dashed lines.
a national framework spatial data theme by virtue of their
extensive current use in facility site selection, census operations, socio-economic planning studies, and legislative redistricting (NRC, 1995; FGDC, 1997a).
Road network centerline updating in rapidly developing
areas may be performed every 1 to 5 years and, in areas with
minimum tree density (or during the leaf-off season), can be
accomplished using imagery with a spatial resolution of 1 to
30 m (Lacy, 1992). If more precise road dimensions are required such as the exact center of the road and the width of
the road and sidewalks, then a spatial resolution of 5 0.2 to
0.5 m is required (Jensen et al., 1994). Currently, only aerial
photography can provide such planimetric information (refer
to Figure 3).
Road, railroad, and bridge conditions (cracks, potholes,
etc.) are routinely monitored both in situ and using high spatial resolution remote sensor data. For example, Figure 5
presents a vertical panchromatic image of railroad and road
bridges. Careful inspection of high spatial resolution imagery
(I0.25 to 0.5 m) by a trained analyst can provide significant
information about the condition of the road and railroad
(Stoeckeler, 1979; Haack et al., 1997).
Traffic count studies of automobiles, airplanes, boats, pedestrians, and people in groups require very high temporal
resolution data ranging from 5 to 10 minutes. Even when
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
Odghd Pachromstk:Aerial P h o b g g b
Di&d Elevation Model @EM)
OrtaOpaato DrapedOver DEM
CclllPlvTrmmlc&verLoestronModel
Figure 4. (a) Panchromatic aerial photograph of Columbia, South Carolina (original at 1:6,000scale). (b)
Digital elevation model (DEM) derived from a stereopair and portrayed a s a shaded-relief model. (c)An orthophotograph of the area draped over the DEM and displayed in a three-dimensional perspective projectlon. (d) The DEM and a viewshed model were used to identify the optimum building on which to place a
cellular phone transceiver.
'
such timely data are available, it is difficult to resolve a car
or boat using even 1-by 1-m data. This requires high spatial
resolution imagery from 5 0.25 to 0.5 m. Such information
can only be acquired using aerial photography, digital cameras, or video sensors that are (1) located on the top edges of
buildings looking obliquely at the terrain, or (2) placed in
aircraft or helicopters and flown repetitively over the study
areas. When such information is collected at an optimum
time of day, future parking and traffic movement decisions
can be made. Parking studies require the same high spatial
resolution (I0.25 to 0.5 m) but slightly lower temporal resolution (10 to 60 minutes). Doppler radar has demonstrated
some potential for monitoring traffic flow and volume.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
UtHity lmructure
Urbanlsuburban environments are enormous consumers of
electrical power, natural gas, telephone service, and potable
water (Haack et al., 1997). In addition, they create great quantities of refuse, waste water, and sewage. The removal of
storm water &om urban impervious surfaces is also a serious
problem (Schultz, 1988). Automated mappinglfacilities management (AMIFM) and geographic information systems (GIS)
have been developed to manage extensive right-of-way corridors for various utilities, especially pipelines uadkowski et
al., 1994; Jensen et al., 1998). The most fundamental task is
to update maps to show a general centerline of the utility of
interest such as a powerline right-of-way. This is relatively
\
I
617
applied to the commercial high spatial resolution remote
sensor data.
Terrain elevation does not change very rapidly. Therefore, a DEM of an urbanized area need only be acquired once
ever 5 to 10 years unless there is significant development
and the analyst desires to compare two different date DEMs
to determine change in terrain elevation, identify unpermitted additions to buildings, or determine changes in building
heights.
Soci&conomic Charactedstics
Selected socio-economic characteristics may be extracted directly from remote sensor data or by using surrogate information derived from the imagery. Two of the most important
attributes are population estimation and quality-of-life indicators.
Figure 5. High resolution vertical panchromatic imagery of
railroad and highway bridges. Road and railroad condition
as well as bridge condition can be monitored using such
imagery.
I
straightforward if the utility is not buried or obscured by
trees and if 1-to 30-m spatial resolution remote sensor data
are available. It is also often necessary to identify Prototype
utility (e.g., pipeline) routes (Feldman et al., 1995). Such
studies require geographically extensive imagery such as
SPOT (20 by 20 m) or Landsat 'Ilematic Mapper data (30 by
30 m) that have relatively large scene dimensions. The majority of the actual and proposed rights-of-way may be observed well on imagery with 1 to 30 m spatial resolution
obtained once every 1 to 5 Years. But*when it is necessary to
inventory the exact location of transmission tower footpads,
utility poles, and manhole covers; the true centerline of the
utility; the width of the utility right-of-way; and the dimensions of buildings, pumphouses, and substations, then it is
necessary to have a spatial resolution of from I 0.25 to 0.6 m
(~adkowskiet al., 1994). Ideally, new facilities are inventoried every 1 to 2 years.
Digital Elevation Model (DEMI Creation
Most GIS used for socio-economic or environmental planning
include a digital elevation model (DEM) (Cowen et al., 1995).
The public often forgets that digital elevation models are derived primarily from analysis of stereoscopic remote sensor
data (Jensen, 1995). It is also possible to extract relatively
coarse z-elevation information using SPOT 10- by 10-m data,
SPIN-2 data (Lavrov, 1997), and Landsat Thematic Mapper 30by 30-m data (Gugan and Dowrnan, 1988). Any DEM to be
used in an urbanlsuburban application should ideally have
z-elevation and x, y coordinates that meet Geospatial Positioning Accuracy Standards (FGDC, 1997b). A sensor that
can provide such information at the present time is stereoscopic large-scale metric aerial photography with a spatial
resolution of I 0.25 to 0.5 m. A DEM of downtown Columbia,
South Carolina, derived from aerial photography, was previously shown in Figure 4. The DEM data can be modeled to
compute slope and aspect statistical surfaces for a variety of
applications. It can also be used to identify the optimum location for placing various utilities as shown in Figure 4d.
Ridley et al. (1998) found that simulated 1-by 1-m stereoscopic satellite data when processed using standard off-theshelf DEM generation packages, yielded a z-elevation RMSE
ranging from 1.5 to 2 m after editing and showed considerable potential for creating DEMs for use in national mapping.
Digital desktop softcopy photogrammetry is revolutionizing
the creation and availability of DEMs (Petrie and Kennie,
1990; Jensen, 1995) and should be of significant value when
I
618
M a y 1999
Population Estimation
Knowing how many people live within a specific geographic
area or administrative unit (e.g., city, county, state, country)
is very powerful information. In fact, it has been suggested
that the global effects of increased population density on ecosystem land-cover conversion and human well-being may be
much more sigdicant than those arising from climate change
(Skole, 1994). Population estimation can be perfomed at the
local, regional, and national level based on (1)counts of individual dwelling units, (2) measurement of urbanized land
areas (often referred to as settlement size), and (3) estimates
derived from land-uselland-cover classification (Lo, 1995;
Sutton et a]., 1997).
Remote sensing techniques may provide population estimates that approach the accuracy of traditional census methods if sufficiently accurate in situ data are available to calibrate the remote sensing model. Unfortunately, ground-based
population estimations may be inaccurate (Clayton and Estes,
1979). many instances in developing
remote
sensing methods may be superior to ground-based methods.
The most accurate remote sensing method for estimating
the population of a local area is to count individual dwelling
units based on the following assumptions (Forester, 1985;
Lindgren, 1985; Lo, 1986; Lo, 1995; Holz, 1988; Haack et al.,
19971:
the imagery must be of sufficient spatial resolution to identify
individual structures even through tree cover and whether
they are residential, commercial, or industrial buildings;
some estimation of the average number of persons per dwelling unit must be available;
some estimate of the number of homeless, seasonal, and migratory workers is required; and
it is assumed that all dwelling units are occupied, and that
only n families live in each unit (calibrated using in situ investigation).
This is usually performed every 5 to 7 years and requires high spatial resolution remotely sensed data (< 0.25 to
5 m). For example, individual dwelling units in a section of
Irmo, South Carolina were extracted hom 2.5- by 2.5-m aircraft multispectral data (Cowen et al., 1995). Correlation of
the remote sensing derived dwelling unit data with U.S. Bureau of the Census dwelling unit data for the 32 census block
area yielded an P = 0.81 (correlation coefficient of 0.91).
These findings suggest that the new high spatial resolution
panchromatic sensors may provide a good source of information for monitoring the housing stock of a community on a
routine basis. This will enable local governments to anticipate and plan for schools and other services with data that
have a much more frequent temporal resolution than does
the decennial census. These data will also be of value for
real estate, marketing, and other business applications (Lo,
1995). Unfortunately, the dwelling unit approach is not suitPHOTOGRAMMETRICENGINEERING & REMOTE SENSING
2. URBAN/SUBURBAN
ATTRIBUTES THAT MAYBE EXTRACTED
FROM
REMOTE SENSOR
DATAUSINGTHE FUNDAMENTAL
ELEMENTSOF IMAGE
INTERPRETATION
AND USEDTO ASSESS HOUSING
QUALITYAND/OR QUALITY-OFLIFE
IABLE
Attributes
Site
Situation
Building
single or multiple-family
size (sq. ft.)
height (ft.)
carport or garage (attached, detached)
age (derived by convergence of evidence)
Lot
size (sq. ft.)
front yard (sq. ft.)
back yard (sq. ft.)
street frontage (ft.)
driveway (paved, unpaved)
fenced
pool (in-ground, above-ground)
patio, deck
out-buildings (sheds)
density of buildings per lot
percent landscaped
health of vegetation (e.g. NDVI greenness)
property fronts paved or unpaved road
abandoned autos
refuse
Adjacency to Community Amenities
schools
churches
hospitals
fire station
library
shopping
open space, parks, golf courses
Adjacency to Nuisances and Hazards
heavy street traffic
railroad or switchyard
airports and/or flightpath
freeway
located on a floodplain
sewage treatment plant
industrial area
power plant or substation
overhead utility lines
swamps and marsh
steep terrain
able for a regional/national census of population because it is
too time consuming and costly (Sutton et a]., 1997). Broome
(personal communication, 1998) suggests that this method requires so much i n situ data to calibrate the remote sensor
data that it can become operationally impractical. Research
is required to document the utility of the method in a variety
of cultures and population densities.
Scientists have known for some time that there is a relationship between the simple urbanized built-up area (settlement size) extracted horn a remotely sensed image and
settlement population (Tobler, 1969; Olorunfemi, 1984),
where r = a x Pb and r is the radius of the populated area
circle, a is an empirically derived constant of proportionality, P is the population, and b is an empirically derived exponent. Estimates of these parameters are fairly consistent at
regional scales but the estimate of the a parameter varies between regions. For example, Sutton et al. (1997) used Defense Meteorological Satellite Program Operational Linescan
System (DMSP-OLS)
visible near-infrared nighttime 1-by 1-km
imagery to inventory urban extent for the entire United
States. When the data were aggregated to the state or county
level, spatial analysis of the clusters of the saturated pixels
predicted population with an rZ = 0.81. Unfortunately,
"DMSP imagery underestimates the population density of urPHOTOGRAMMETRICENGINEERING 81REMOTE SENSING
ban centers and overestimates the population density of suburban areas" (Sutton et al., 1997).
Another widely adopted population estimation technique is based on the use of the Level I through I11 land-use
information previously described. This approach assumes
that land use in an urban area is closely correlated with population density. Researchers establish a population density
for each land use by field survey or census data. Then, by
measuring the total area for each land-use category, they estimate the total population for that category. Summing the estimated totals for each land-use category provides the total
population projection (Lo, 1995). The urban built-up area
and land-use data method can be based on more coarse spatial resolution multispectral remote sensor data (5 to 20 m)
every 5 to 15 years.
Quality-of-Life Indicators
Lo and Faber (1998) suggest that adequate income, decent
housing, education and health services, and good physical
environment are important indicators of social well-being
and quality-of-life. Evaluating the quality-of-life of a population on a continuing basis is important because it helps planners and government agencies involved with the delivery of
human services to be aware of problem areas.
In the past, most quality-of-life studies made use of census data to extract socio-economic indicators. Only recently
have factor analytic studies documented how quality-of-life
indicators (such as house value, median family income, average number of rooms, average rent, and education) can be estimated by extracting the urban attributes summarized in
Table 2 from relatively high spatial resolution (10.25 to 30
m) imagery (Monier and Green, 1953; Green, 1957; McCoy
and Metivier, 1973; Tuyahov et al., 1973; Henderson and
Utano, 1975; Jensen, 1983; Lindgren, 1985; Holz, 1988; Avery and Berlin, 1993; Haack et al., 1997; Lo and Faber,
1998). Note that the attributes in Table 2 are arranged by site
(building and lot) and situation. The site may be situated in
positive and negative surroundings.
Onsrud et al. (1994), Curry (1997), and Slonecker et al.
(1998) point out that scientists must exercise wise judgment
when using remotely sensed data to extract socio-economic
and/or quality-of-life information so that they do not infringe
on an individual's right to privacy. The misuse of the high
spatial resolution remote sensor data will likely be the impetus for future restrictive legislation.
Energy Demand and Production Potentlal
Local urbanlsuburban energy demand may be estimated using
remotely sensed data. First, the square footage of individual
buildings is determined. Local ground reference information
about energy consumption is then obtained for a representative sample of dwellings in the area. Regression relationships
are derived to predict the energy consumption anticipated
for the region. This requires imagery with a spatial resolution
of from 10.25 to 1 m. Regional and national energy consumption may be predicted using DMSP imagery (Welch,
1980). Unfortunately, DMSP imagery of urbanized areas are
recorded at 6-bits radiometric resolution (0 to 63), causing
most of the urban, energy consuming areas to saturate at a
brightness value of 63 (Elvidge et al., 1997; Sutton et al.,
1997).
It is also possible to predict how much solar photovoltaic energy potential a geographic region has by modeling
the individual rooftop square footage, slope, and orientation
(e.g., north or south) with known photovoltaic generation
constraints. This requires very high spatial resolution imagery ( 1 0.25 to 0.5 m) (Clayton and Estes, 1979; Angelici et
a]., 1980).
Studies have documented how high spatial resolution (1
May 1999
619
severe weather mode, every 6 minutes in precipitation mode,
and every 10 minutes in clear air mode.
High spatial resolution (5 to 30 m) day- and night-time
thermal-infrared data may be used to obtain detailed quantitative spatial information on the urban heat island effect (Lo
et al., 1997). Some have used AVHRR thermal-infrared data
for this application with mixed results (e.g., Roth et al.,
1989).
Figure 6. Overturned tanker and associated spill in Anchorage, Alaska. The high spatial resolution panchromatic vertical aerial photograph was obtained shortly
after the accident (courtesy of AeroMap US.; Schweitzer
and McLeod, 1997).
to 5 m) pre-dawn thermal infrared imagery (8 to 1 2 km) can
be used to inventory the relative quality of housing insulation if (1) the rooftop material is known (e.g., asphalt versus
wood shingles), (2) moisture is not present on the roof, and
(3) the orientation and slope of the roof are known (Colcord,
1981; Eliasson, 1992). If energy conservation or the generation
of solar photovoltaic power were important, these variables
would probably be collected every 1 to 5 years.
Meteorological Data
Daily weather in urban environments affects people, schools,
businesses, and telecommunication and transportation systems. Great expense has gone into the development of nearreal-time monitoring of frontal systems, temperature, precipitation, and severe storm warning systems. These important
meteorological parameters are monitored almost exclusively
by sophisticated airborne and ground-based remote sensing
systems. For example, two Geostationary Operational Environmental Satellites (GOES) are positioned at 35,800 km
above the equator in geo-synchronous orbits. GOES West obtains information about the western United States and is
parked at 135' west longitude. GOES East obtains information
about the Caribbean and eastern United States and is parked
at 75" west longitude. Every day millions of people watch
the progress of frontal systems that sometimes generate
deadly tornadoes and hurricanes. Full hemispheric disk images may be obtained every 25 minutes. Intense storms in
relatively smaller regions (3000 by 3000 km)may be imaged
every 3.1 minutes. The spatial resolution is 1 by 1 km for the
visible band and 4 to 8 krn for the thermal infrared bands
(Kidder and Haar, 1995). European nations use Meteosat
with visible near-infrared bands obtained at 2.5 by 2.5 km
and thermal infrared data collected at 5 by 5 km every 25
minutes. Early hurricane monitoring and modeling based on
these data have saved thousands of lives in recent history.
For example, in 1989 Hurricane Hugo caused approximately
one billion dollars in damage to residential, commercial, and
industrial facilities but no lives were lost because of remotesensing assisted early warning and evacuation.
The public also relies on ground-based National Weather
for precipitaService Weather Surveillance Radar (WSR-88~)
tion mapping and timely severe storm warning. The Doppler
radar "composite reflectivity" product is projected onto a
Cartesian geographical map with a 1-by 1-km resolution out
to 230 km or at a 4- by 4-km resolution out to 460 km (Crum
and Alberty, 1993). The data are obtained every 5 minutes in
620
May 1 9 9 9
Critical Environmental Area Assessment
Urbanlsuburban environments often include very sensitive
areas such as wetlands, endangered specie habitat, parks,
land surrounding treatment plants, and the land in urbanized
watersheds that provides the runoff for potable drinking water. Relatively stable sensitive environments only need to be
monitored every 1 to 2 years using a multispectral remote
sensor collecting 1-to 10-m data. For extremely critical areas
that could change rapidly, multispectral remote sensors (including a thermal infrared band) should obtain < 0.25- to 2m spatial resolution data every 1 to 6 months.
Disaster Emergency Response
Recent floods (Mississippi River in 1993; Albany, Georgia in
1994), hurricanes (Hugo in 1989, Andrew in 1991, Fran in
1996), tornadoes (every year), fires, tanker spills, and earthquakes (Northridge, California in 1994) demonstrated that a
rectified, pre-disaster image database is indispensable. The
pre-disaster data only need to be updated every 1 to 5 years.
It should be high spatial resolution (1-to 5-m) multispectral
data if possible.
When disaster strikes, high resolution (< 0.25- to 2-m)
panchromatic and/or near-infrared data should be acquired
within 1 2 hours to 2 days (e.g., Figure 6; Schweitzer and
McLeod, 1997). If the terrain is shrouded in clouds, imaging
radar might provide the most useful information. Post-disaster images are registered to the pre-disaster images, and manual and digital change detection takes place (Jensen, 1996). If
precise, quantitative information about damaged housing
stock, disrupted transportation arteries, the flow of spilled
materials, and damage to above-ground utilities are required,
it is advisable to acquire post-disaster < 0.25- to 1-m panchromatic and near-infrared data within 1 to 2 days (Jensen
et al., 1998). Such information was indispensable in assessing damages and allocating scarce clean-up resources during
Hurricane Hugo, Hurricane Andrew (Davis, 1993), Hurricane
Fran (Wagman, 1997), and the recent Northridge earthquake.
Observations
Table 1 and Figure 1 reveal that there are a number of remote sensing systems that currently provide some of the desired urban infrastructure and socio-economic information
when the required spatial resolution is poorer than 4 by 4 m
and the temporal resolution is between 1 and 55 days. However, very high spatial resolution data (< 1 by 1 m) is required to satisfy several of the data requirements. In fact, as
shown in Figure 1,the only sensor that currently provides
such data on-demand is aerial photography (5 0.25 to 0.5 m).
E O S A T / ~Imaging
~~C~
lKONOS (1999) with its 1-by 1-m panchromatic data, OrbView 3 (1999) with its 1-by 1-m panchromatic data, and Earthwatch Quickbird with its 0.8- by
0.8-m panchromatic data (2000) may satisfy some of these
urban data requirements, but not all. It may be necessary to
develop higher spatial resolution (< 0.25- to 0.5-m) satellite
remote sensor data to provide some of the detailed urban/
suburban infrastructure and socio-economic information, or
utilize aerial photography. None of the sensors can provide
the 5- to 60-minute temporal resolution necessary for traffic
and parking studies except for (1)repetitive aerial photography (very costly), or (2) the placement of digital or video
PHOTOGRAMMETRE ENGINEERING & REMOTE SENSING
cameras on the top edge of buildings to obtain an oblique
view. The GOES satellite constellation (east and west) and the
European Meteosat provide sufficient national a n d regional
weather information at reasonable temporal resolution (3 to
25 minutes) and spatial resolutions (1to 8 km and 2.5 to 5
km,respectively). Ground-based National Weather Service
Weather Surveillance Radar provides sufficient spatial resolution (1 by lo) and temporal resolution (5 to 1 0 minutes) for
precipitation and intense storm tracking in urban environments.
References
American Planning Association, 1998. Land-Based Classification
Standards: Dmft Classification, American Planning Association
Research Department, http://www.planning,org/lbcs.
Anderson, J.R., E. Hardy, J. Roach, and R. Witmer, 1976. A Land Use
and Land Cover Classification System for Use with Remote Sensor Data, USGS Professional Paper #964, Washington, D.C., 28
-P.
Angelici, G.L., N.A. Bryant, R.K. Fretz, and S.Z. Friedman, 1980. Urban Solar Photovoltaics Potential: An Inventory and Modeling
Study Applied to the San Fernando Valley Region of Los Angeles, JPL Report #80-43, Pasadena, California, JPL, 55 p.
Avery, T.E., and G.L. Berlin, 1993. Fundamentals of Remote Sensing
and Airphoto Interpretation, Macmillan, New York, pp. 377404.
Barnell, B., 1998. ArcView 3D Analyst Used in Virtual Reality Models, ARC News, 20(2):1-3.
Branch, M.C., 1971. City Planning and Aerial Information, Harvard
University Press, Cambridge, Massachusetts, 283 p.
Bureau of the Budget, 1987. Standard Industrial Classification Manual, Government Printing Office, Washington, D.C.
Chisnell, T.C., and G.E. Cole, 1958. Industrial Components - A Photo
Interpretation Key on Industry, Photogmmmetric Engineering,
24:590-602.
Clayton, C., and J.E. Estes, 1979. Distributed Parameter Modeling of
Urban Residential Energy Demand, Photogrammetric Engineering & Remote Sensing, 45:106-115.
Colcord, J.E., 1981. Thermal Imagery Energy Surveys, Photogmmmetric Engineering b Remote Sensing, 47:237-240.
Cowen, D.J., and J.R. Jensen, 1998. Extraction and Modeling of Urban Attributes Using Remote Sensing Technology, People and
Pixels: Linking Remote Sensing and Social Science, National
Academy Press, Washington, D.C., pp. 164-188.
Cowen, D.J., J.R. Jensen, G. Bresnahan, D. Ehler, D. Traves, X.
Huang, C. Weisner, and H.E. Mackey, 1995. The Design and Implementation of an Integrated GIs for Environmental Applications, Photogmmmetric Engineering 6 Remote Sensing, 61:13931404.
C m ,T.D., and R.L. Alberty, 1993. The WSR-88D and the WSR-88D
Operational Support Facility, Bulletin of the American Meteorological Society, 74(9):1669-1687.
Cullingworth, B., 1997. Planning in the USA: Policies, Issues and
Processes, Routledge, London, 280 p.
Curry, M.R., 1997. The Digital Individual and the Private Realm, Annals of the Association of American Geographers, 87(4):681-699.
Davis, B.A., 1993. Mission Accomplished, NASA Tech Briefs, 17(1):
14-16.
Eliasson, I., 1992. Infrared Themography and Urban Temperature
Patterns, International Journal of Remote Sensing, 13:869-879.
Elvidge, C.D., K.E. Baugh, E.A. Kihn, H.W. Kroehl, and E.R. Davis,
1997. Mapping City Lights with Nighttime Data from the DMSP
Operational Linescan System, Photogmmmetric Engineering 6.
Remote Sensing, 63:727-734.
Feldman, S.C., R.E. Pelletier, E. Walser, J.R. Smoot, and D. Ahl,
1995. A Prototype for Pipeline Routing Using Remotely Sensed
Data and Geographic Information System Analysis, Remote
Sensing of Environment, 53:123-131.
FGDC, 1997a. Fmmework: Introduction and Guide, Federal Geographic Data Committee, Washington, D.C., 105 p.
PHOTOGRAMMETRICENGINEERING& REMOTE SENSING
, 1997b, DmftGeospatial Positioning Accuracy Standards,
Federal Geographic Data Committee, Washington, D.C., 32 p.
Ford, K., 1979. Remote Sensing for Planners, State Univ. of New Jersey, Rutgers, 219 p.
Forester, B.C., 1985. An Examination of Some Problems and Solutions in Monitoring Urban Areas from Satellite Platforms, International Journal of Remote Sensing, 6:139-151.
Green, K., D. Kempka, and L. Lackey, 1994. Using Remote Sensing to
Detect and Monitor Land-Cover and Land-Use Change, Photogmmmetric Engineering b Remote Sensing, 60:331-337.
~ r e e iN.E.,
,
1957. Aerial ~hotogra~hic
Interpretation and Social
Structure of the City, Photogmmmetric Engineering, 23:89-96.
Greve, C.W., 1996. Digital Photogmmmetry: An Addendum to the
Manual of Photogmmmetry, American Society for Photogrammetry & Remote Sensing, Bethesda, Maryland, 247 p.
Gugan, D.J., and I.J. Dowman, 1988. Topographic Mapping from
SPOT Imagery, PhotogmmmetricEngineering 6 Remote Sensing,
54:1409-1404.
Haack, B., S. Guptill, R. Holz, S. Jampoler, J. Jensen, and R. Welch,
1997. Chapter 15: Urban Andvsis and Planning, Manual of Photographic ?nterpretation, American Society for Fhotogra&etry
& Remote Sensing, Bethesda, Maryland, pp. 517-553.
Henderson, F.M., and J.J. Utano, 1975. Assessing General Urban Socioeconomic Conditions with Conventional Air Photography,
Photogmmmetria, 31:81-89.
Holz, R.K., 1988. Population Estimation of Colonias in the Lower Rio
Grande Valley Using Remote Sensing Techniques, paper presented at the Annual Meeting of the Association of American
Geographers, Phoenix, Arizona.
Hothem, D., J.M. Irvine, E. Mohr, and K.B. Buckely, 1996. Quantifying Image Interpretability for Civil Users, Proceedings, ASPRS/
ACSM Annual Convention, American Society fur Photogrammetry and Remote Sensing, Bethesda, Maryland, pp. 292-298.
Jadkowski, M.A., P. Convery, R.J. Birk, and S. Kuo, 1994. Aerial Image Databases for Pipeline Rights-of-way Management, Photogmmmetric Engineering 6 Remote Sensing, 60:347-353.
Jensen, J.R. (editor), 1983. Urban/Suburban Land Use Analysis, Manual of Remote Sensing, Second Edition (R.N. Colwell, editor),
American Society of Photogrammetry, Falls Church, Virginia,
pp. 1571-1666.
, 1995. Issues Involving the Creation of Digital Elevation Models and Terrain Corrected Orthoimagery Using Soft-Copy Photogrammetry, Geocarto International, 10:l-17.
, 1996. Introductory Digital Image Processing: A Remote Sensing Perspective, Prentice-Hall, Saddle River, New Jersey, 318 p.
Jensen, J.R., and D.L. Toll, 1983. Detecting Residential Land-Use Development at the Urban Fringe, PhotogmmmetricEngineering 6
Remote Sensing, 48:629-643.
Jensen, J.R., and D.C. Cowen, 1997. Remote Sensing of UrbanlSuburban Socio-economic Attributes, Proceedings Land Satellite Information in the Next Decade II, American Society for Photogrammetry & Remote Sensing, Bethesda, Maryland, Compact Disk, 19
PJensen, J.R., D.C. Cowen, J. Halls, S. Narumalani, N. Schmidt, B.A.
Davis, and B. Burgess, 1994. Improved Urban Jnfrastructure
Mapping and Forecasting for BellSouth Using Remote Sensing
and GIs Technology, PhotogrammetricEngineering 6 Remote
Sensing, 60:339-346.
Jensen, J.R., X. Huang, D. Graves, and R. Hanning, 1996. Cellular
Phone Transceiver Site Selection, Raster Imagery in Geogmphic
Information Systems (S. Morain and S. Baros, editors], OnWard
Press, Santa Fe, New Mexico, pp. 117-125.
Jensen, J.R., J. Halls, and J. Michel, 1998. A Systems Approach to
Environmental Sensitivity Index (ESI) Mapping for Oil Spill
Contingency Planning and Response, Photogmmmetric Engineering b Remote Sensing, 64(10):1003-1014.
Kidder, S.Q., and T.H.V. Haw, 1995. Satellite Meteorology, Academic Press, New York, pp. 87-144.
Lacy, R., 1992. South Carolina Finds Economical Way to Update
Digital Road Data, GIS World, 5:58-60.
Lavrov, V.N., 1997. Space Survey Photocameras for Cartographic
Purposes, Proceedings, Fourth International Conference on ReMay 1999
621
mote Sensing for Marine and Coastal Environments, ERIM, Lansing, Michigan, 7 p.
~eachteiauer,J.c., 1996. National Imagery Interpretability Rating
Scales Overview and Product Description, Proceedinrrs. ASPRSACSM Annual Convention, Americ& society for ~hltogrammetry and Remote Sensing, Bethesda, Maryland, 1:262-272.
Leachtenauer, J.C., K. Daniel, and T. Vogl, 1998. Digitizing Satellite
Imagery: Quality and Cost Considerations, Photogrammetric Engineering & Remote Sensing, 64:2%-34.
Leberl, F.W., 1990. Radargrammetric Image Processing, Artech
House, Norwood, 595 p.
Light, D.L., 1993. The National Aerial Photography Program as a Geographic Information System Resource, Photogmmmetric Engineering 6 Remote Sensing, 59(1):61-65.
-,
1996. Film Cameras or Digital Sensors? The Challenge Ahead
for Aerial Imaging, Photogrammetric Engineering b Remote
Sensing, 62(3):285-291.
Lindgren, D.T., 1985. Land Use Planning and Remote Sensing, Martinus Nijhoff Inc., Boston, 230 p.
Lo, C.P., 1986. The Human Population, Applied Remote Sensing,
Longman, New York, pp. 40-70.
-,
1995. Automated Population and Dwelling Unit Estimation
from High-Resolution Satellite Images: A GIS Approach, International Journal of Remote Sensing, 16:17-34.
Lo, C.P., and B.J. Faber, 1998. Interpretation of Landsat Thematic
Mapper and Census Data for Quality of Life Assessment, Remote
Sensing of Environment, in press.
Lo, C.P., D.A. Quattrochi, and J.C. Luvall, 1997. Application of Highresolution Thermal Infrared Remote Sensing and GIS to Assess
the Urban Heat Island Effect, International Journal of Remote
Sensing, 18(2):287-304.
Logicon, 1995. Multispectml Users Guide, Logicon Geodynamics,
Fairfax, Virginia, 102 p.
-,
1997. Multispectml Imagery Reference Guide, Logicon Geodynamics, Fairfax, Virginia, 100 p.
McCoy, R.M., and E.D. Metivier, 1973. House Density vs. Socioeconomic Conditions, Photogmmmetric Engineering, 39:43-49.
Mintzer, 0.W. (editor), 1983. Engineering Applications, Manual of
Remote Sensine. Second Edition [R.N. Colwell. editor). American Society of ~hotogrammetry,Falls Church, Virginia, pp.
1955-2109.
Monier, R.B., and N.E. Green, 1953. Preliminary Findings on the Development of Criteria for the Identification of Urban Structures
from Aerial Photographs, Annals of the Association of American
Geographers, special issue.
Montesano, A.P., 1997. Roadrnap to the Future, Earth Observation
Magazine, 6:16-19.
NRC,1995. A Data Foundation for the National Spatial Data Infrastructure, Mapping Science Committee, National Research Council, Washington, D.C., 55 p.
Olorunfemi, J.F., 1984. Land Use and Population: A Linking Model,
Photogmmmetric Engineering b Remote Sensing, 50:221-22 7.
Onsrud, H.J., J.P. Johnson, and X.R. Lopez, 1994. Protecting Personal
Privacy in Using Geographic Information Systems, Photogmmmetric Engineering b Remote Sensing, 60(9):1083-1095.
Petrie, G., and T.J.M. Kennie, 1990. Terrain Modeling in Surveying
and Civil Engineering, Whittles Publishing, London, 351 p.
Philipson, W., 1997. Manual of Photographic Interpretation, American Society for Photogrammetry & Remote Sensing, Bethesda,
Maryland, 830 p.
622
May 1 9 9 9
Pike, J., 1998. Space Policy Project of the Fedemtion of American
Scientists, Federation of American Scientists, Washington, D.C.,
http://www.fas.org/irp/imint/niirs.htm.
Ridley, H.M., P.M. Atkinson, P. Aplin, J. Muller, and I. Dowrnan,
1998. Evaluating the Potential of the Forthcoming Commercial
U.S. High-Resolution Satellite Sensor Imagery at the Ordnance
Survey, Photogmmmetric Engineering b Remote Sensing, 63(8):
997-1005.
Roth, M., T.R. Oke, and W.J. Emery, 1989. Satellite-derived Urban
Heat Islands from Three Coastal Cities and the Utilization of
Such Data in Urban Climatology, International Journal of Remote Sensing, 10(11):1699-1720.
Schultz, G.A., 1988. Remote Sensing in Hydrology, Journal of Hydrology, 100:239-265.
Schweitzer, B., and B. McLeod, 1997. Marketing Technology that is
Changing at the Speed of Light, Earth Observation Magazine, 6:
22-24.
Skole, D.L., 1994. Data on Global Land-Cover Change: Acquisition,
Assessment, and Analysis, Changes in Land Use and Land
Cover: A Global Perspective (W.B. Meyer and B.L. Turner, editors), Cambridge University Press, Cambridge, pp. 437472.
Slonecker, E.T., D.M. Shaw, and T.M. Lillesand, 1998. Emerging Legal and Ethical Issues in Advanced Remote Sensing Technology,
Photogmmmetric Engineering 6 Remote Sensing, 64(6):589-595.
Stone, K.H.,
1964. A Guide to the Interpretation and Analysis of Aerial Photos, Annals of the Association of American Geographers,
54:318-328.
Stoney, W.E., and J.R. Hughes, 1998. A New Space Race is On! GIS
World, 11(3):4446.
Stoeckleler, E.G., 1979. Use of Aerial Color Photography for Pavement Evaluation Studies, Highway Research Record, 319:40-57.
Sutton, P., D. Roberts, C. Elvidge, and H. Meij, 1997. A Comparison
of Nighttime Satellite Imagery and Population Density for the
Continental United States, Photogmmmetric Engineering 6 Remote Sensing, 63:1303-1313.
Swerdlow, J.L., 1998. Making Sense of the Millennium, National
Geogmphic, 193:2-33.
Tobler, W., 1969. Satellite Confirmation of Settlement Size Coefficients, Area, 1:30-34.
Tuyahov, A.J., C.S. Davies, and R.K. Holz, 1973. Detection of Urban
Blight Using Remote Sensing Techniques, Remote Sensing of
Earth Resources, 2:213-226.
Urban Renewal Administration Department of Commerce, 1965.
Standard Land Use Coding Manual: A. Standard System for
Identifying and Coding Land Use Activities, Government Printing Office, Washington, D.C.
Wolff, R.S., and L. Yaeger, 1993. Visualization of Natuml Phenomena, Springer-Verlag, Santa Clara, 374 p.
Wagman, D., 1997. Fires, Hurricanes Prove No Match for GIS, Earth
Observation Magazine, 6:2 7-29.
Warner, W.S., R.W. Graham, and R.E. Read, 1996. Chapter 15: Urban
Survey, Small Format Aerial Photogmphy, Wittles Publishing,
Scotland, pp. 253-256.
Welch, R., 1980. Monitoring Urban Population and Energy Utilization Patterns from Satellite Data, Remote Sensing of Environment, 9:l-9.
-,
1982. Spatial Resolution Requirements for Urban Studies. International Journal of Remote Sensing, 3:139-146.
(Received 03 March 1998; accepted 29 April 1998; revised 13 August
1998)
PHOTOGRAMMETRICENGINEERING & REMOTE SENSING

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