10.3 Sources of Remotely Sensed Data
agery derived from camera and video sensors can
be digitized, the satellite and some airborne sensors produce data that are already in digital form.
Digital images are composed of a grid of picture
elements, or pixels, representing reflectance information about objects. Reflectance is what allows us
to distinguish features in an image. It is a measure
of the amount and type of energy an object reflects
(rather than absorbs or transmits). Reflected energy, modified by the atmosphere between the
ground and the sensor, is called radiance. Radiance
is what the sensor measures; it may differ from reflectance because of haze in the atmosphere. An indepth discussion of this topic is presented in Chapter 2 of Jensen (1996).
10.3.1 Data Collection
Different sources of remotely sensed data vary in
their ability to distinguish landscape features. In the
following, we describe some of the most useful data
sources for purposes of ecological assessment, with
emphasis on digital imagery. The collection of primary data, creation of derived data in the form of
vegetation maps, and detection oflandscape change
are addressed in tum.
Advanced Very High Resolution
Radiometer (AVHRR)
A VHRR data are obtained through National Oceanic
and Atmospheric Administration (NOAA) weather
satellites. A VHRR imagery has a spatial resolution
of 1.1 km by 1.1 km pixel size with five spectral
bands. Broad area coverage is obtained with a swath
width of 2700 kilometers. Temporal resolution is
excellent, with repeat coverage over a given geographic area every 12 hours. Major applications for
these data include vegetation discrimination, vegetation biomass, snow-ice discrimination, vegetationcrop stress, geothermal mapping, and fire mapping.
This imagery has been used for fire detection and
growth monitoring in Alaska and Spain (Kasischke
and French, 1995; Pozo et al., 1997), forest classification of southeast Asia (Achard and Esteguil,
1995), and change detection in western Africa (Lambin, 1996) and to derive a relationship between
A VHRR imagery and ecoregions in Utah (Ramsey
et al., 1995).
Landsat Multispectral Scanner (MSS)
and Landsat Thematic Mapper (TM)
The Landsat Program is the longest running program in the collection of multispectral, digital data
139
of Earth's land surface from space. The temporal
extent of the collection, characteristics and quality
of Landsat data, and ability to collect new data directly comparable to that in the archive make Landsat data a unique resource for addressing a broad
range of issues in ecosystem management. The July
1997 issue of Photogrammetric Engineering & Remote Sensing paid tribute to the 25-year history of
the Landsat Program (French, 1997).
Landsat MSS imagery has a spatial resolution of
80 m, spectral resolution in four bands, revisits the
same geographic location every 16 days, and covers an area of 185 by 170 km. The Landsat program routinely gathered digital MSS data from
1972 through 1992. The resulting data, spanning
20 years, can support evaluations of change in landscapes or land cover over a longer time frame than
any available Earth observation system. MSS data
are available through the EROS Data Center (EDC)
working in cooperation with NASA. The North
American Landscape Characterization (NALC)
program has developed a data set of archived MSS
images for the conterminous United States (Lunetta
et aI., 1998). The NALC data set, sometimes referred to as MSS triplicates, consist of at least three
co-registered images acquired in the 1970s, 1980s,
and early 1990s for all the conterminous United
States.
Landsat TM imagery has 30-m spatial resolution
with seven spectral bands. The satellite covers the
same geographic location every 16 days, each scene
covering 185 km by 170 km. The Multi-Resolution
Landscape Characterization Consortium (Loveland
and Shaw, 1996) was established to purchase TM
imagery covering the conterminous United States.
Partners in the consortium include USGS Gap
Analysis Program, EPA Land Cover Program,
NOAA Coastal Change Analysis Program, USGS
National Water Quality Assessment Program,
EPA/USGS North American Landscape Characterization, USGS EDC, and USDA Forest Service.
Other ongoing large procurements of these data by
federal agencies will allow easy and economical
access to recent and future data.
Aerial Photography
Aerial photography at scales ranging from
1 : 12,000 to 1 : 24,000, sometimes referred to as
resource-scale aerial photography, has been used
for over 50 years as a tool for mapping and monitoring resource areas. Aerial photographs are typically acquired on a 5- to lO-year cycle on most federal lands. In recent years, aerial photography has
become an important source of ancillary data for
agery derived from camera and video sensors can
be digitized, the satellite and some airborne sensors produce data that are already in digital form.
Digital images are composed of a grid of picture
elements, or pixels, representing reflectance information about objects. Reflectance is what allows us
to distinguish features in an image. It is a measure
of the amount and type of energy an object reflects
(rather than absorbs or transmits). Reflected energy, modified by the atmosphere between the
ground and the sensor, is called radiance. Radiance
is what the sensor measures; it may differ from reflectance because of haze in the atmosphere. An indepth discussion of this topic is presented in Chapter 2 of Jensen (1996).
10.3.1 Data Collection
Different sources of remotely sensed data vary in
their ability to distinguish landscape features. In the
following, we describe some of the most useful data
sources for purposes of ecological assessment, with
emphasis on digital imagery. The collection of primary data, creation of derived data in the form of
vegetation maps, and detection oflandscape change
are addressed in tum.
Advanced Very High Resolution
Radiometer (AVHRR)
A VHRR data are obtained through National Oceanic
and Atmospheric Administration (NOAA) weather
satellites. A VHRR imagery has a spatial resolution
of 1.1 km by 1.1 km pixel size with five spectral
bands. Broad area coverage is obtained with a swath
width of 2700 kilometers. Temporal resolution is
excellent, with repeat coverage over a given geographic area every 12 hours. Major applications for
these data include vegetation discrimination, vegetation biomass, snow-ice discrimination, vegetationcrop stress, geothermal mapping, and fire mapping.
This imagery has been used for fire detection and
growth monitoring in Alaska and Spain (Kasischke
and French, 1995; Pozo et al., 1997), forest classification of southeast Asia (Achard and Esteguil,
1995), and change detection in western Africa (Lambin, 1996) and to derive a relationship between
A VHRR imagery and ecoregions in Utah (Ramsey
et al., 1995).
Landsat Multispectral Scanner (MSS)
and Landsat Thematic Mapper (TM)
The Landsat Program is the longest running program in the collection of multispectral, digital data
139
of Earth's land surface from space. The temporal
extent of the collection, characteristics and quality
of Landsat data, and ability to collect new data directly comparable to that in the archive make Landsat data a unique resource for addressing a broad
range of issues in ecosystem management. The July
1997 issue of Photogrammetric Engineering & Remote Sensing paid tribute to the 25-year history of
the Landsat Program (French, 1997).
Landsat MSS imagery has a spatial resolution of
80 m, spectral resolution in four bands, revisits the
same geographic location every 16 days, and covers an area of 185 by 170 km. The Landsat program routinely gathered digital MSS data from
1972 through 1992. The resulting data, spanning
20 years, can support evaluations of change in landscapes or land cover over a longer time frame than
any available Earth observation system. MSS data
are available through the EROS Data Center (EDC)
working in cooperation with NASA. The North
American Landscape Characterization (NALC)
program has developed a data set of archived MSS
images for the conterminous United States (Lunetta
et aI., 1998). The NALC data set, sometimes referred to as MSS triplicates, consist of at least three
co-registered images acquired in the 1970s, 1980s,
and early 1990s for all the conterminous United
States.
Landsat TM imagery has 30-m spatial resolution
with seven spectral bands. The satellite covers the
same geographic location every 16 days, each scene
covering 185 km by 170 km. The Multi-Resolution
Landscape Characterization Consortium (Loveland
and Shaw, 1996) was established to purchase TM
imagery covering the conterminous United States.
Partners in the consortium include USGS Gap
Analysis Program, EPA Land Cover Program,
NOAA Coastal Change Analysis Program, USGS
National Water Quality Assessment Program,
EPA/USGS North American Landscape Characterization, USGS EDC, and USDA Forest Service.
Other ongoing large procurements of these data by
federal agencies will allow easy and economical
access to recent and future data.
Aerial Photography
Aerial photography at scales ranging from
1 : 12,000 to 1 : 24,000, sometimes referred to as
resource-scale aerial photography, has been used
for over 50 years as a tool for mapping and monitoring resource areas. Aerial photographs are typically acquired on a 5- to lO-year cycle on most federal lands. In recent years, aerial photography has
become an important source of ancillary data for