11.4 Data Sources
or are not appropriate due to scale, accuracy, or the
classification system used (Estes and Mooneyhan,
1994). On the other hand, existing digital imagery
and maps are acquired much more easily now than
they were even a few years ago because of access
via the World Wide Web.
Stohlgren, in Chapter 5, notes that 20% to 25%
of a budget for an EA must be devoted to data management. It could be argued, in the case of GIS
data, that this figure is reasonable subsequent to acquisition of air photos or imagery or data development by field-based mapping, spatial interpolation,
image interpretation, or modeling. Resources for
developing the input data, if they are not already
available, could equal or exceed by an order of
magnitude the other costs of an EA. For example,
detailed maps of vegetation that included canopy
cover and crown size estimates for forested stands
were produced for several million hectares of Forest Service lands in California; at 1 : 24,000 scale
and 2-ha minimum mapping unit, the cost was
about $0.30lha (Franklin and Woodcock, 1997). On
the other hand, interpolated climate surfaces might
be produced for only a few cents per square kilometer. Other existing data, such as digital terrain
models, may have been developed under federal
programs and be available at low cost unrepresentative of the actual production costs of these data
(Estes and Mooneyhan, 1994).
11.4.1 Data Sources
Digital spatial data essential to some forms of ecological assessment, especially at landscape or regional scales, include (1) remotely sensed data
(satellite imagery, digital orthophotographs, aerial
photography, or digital airborne camera data); (2)
interpolated climate data; (3) digital elevation models; (4) digital cartographic data such as we find on
a USGS I : 24,000 topographic map (transportation
routes, hydrography, land ownership); and (5) thematic maps of environmental variables, including
land use, landform, soils, geology, and vegetation
(such as those compiled by the USGS, Forest Service, National Park Service, and Gap Analysis Program). Remotely sensed data were discussed in
Chapter 10. In this chapter we will briefly discuss
the other four types of data, especially with respect
to the following: Were the spatial data created for
the first time in a digital form (through remote sensing or by interpolation), or were they digitized from
traditional maps?
Interpolated climate data are useful in ecosystem
characterization (see Chapters 22 through 29) for
predictively mapping vegetation or other ecosystem characteristics that are otherwise difficult to
153
map (see Section 11.5) and as input to ecosystem
models (discussed in Chapter 18). These may include interpolated surfaces of long-term climate
(Hutchinson, 1987, 1996; Daly et al., 1994;
Hutchinson and Gessler, 1994; Daly and Taylor,
1996; Hutchinson et al., 1996) or daily weather
variables (Running et al., 1987; Glassy and Running, 1994; Running and Thornton, 1996). Some
commercial data products are available; however,
there is still considerable debate over the most appropriate methods for interpolation (Hutchinson,
1984; Hutchinson and Gessler, 1994). These climate surfaces can be generated for a specific region from climate station data by a climatologist
with expertise in spatial analysis methods.
Digital elevation models (DEM), representing
land surface topography, are essential for a number of purposes related to EA. Elevation and simple and complex terrain attributes derived from a
DEM, such as slope and aspect, may be used directly in ecosystem characterization. More frequently, however, they are used to model the distributions of complex ecological or biophysical
patterns and processes, such as site potential, vegetation type, plant or animal species distributions,
potential solar insolation, and topographic moisture
(Moore et ai. 1991; McNab, 1993; Dubayah and
Rich, 1995; reviewed in Franklin, 1995; Franklin
et aI., 2000). DEMs also are commonly used in
ecosystem models, especially of hydrologically
mediated processes (see Chapter 18). Although
standard data products are available from the National Mapping Division of the USGS, these data
are derived from existing topographic maps that
may be more than 40 years old or are produced using photogrammetric methods that are prone to systematic and nonsystematic errors (Weibel and
Heller, 1991; Estes and Mooneyhan, 1994). Although in the future greatly expanded sources
of digital elevation models will include highresolution stereo satellite imagery, the USGS products are currently the best or only source of data
for most of the United States. Gridded (raster) products have, at best, 30-m cell size (resolution),
whereas research indicates that DEMs of 10- to 20m resolution are required to accurately depict topographic features and model topohydrological and
other biophysical processes (Hutchinson et al., 1996;
Hutchinson, 2000). A global digital elevation model
with 30 arc second resolution, GTOP030, is now
available from USGS (http://edcwww.cr.usgs.gov/
landdaac/gtop030/ gtop030.html).
In addition to the elevation contours, the other
information found on USGS topographic quadrangle maps (hydrography, roads and water bodies,
land ownership) is available from the USGS in vec-
or are not appropriate due to scale, accuracy, or the
classification system used (Estes and Mooneyhan,
1994). On the other hand, existing digital imagery
and maps are acquired much more easily now than
they were even a few years ago because of access
via the World Wide Web.
Stohlgren, in Chapter 5, notes that 20% to 25%
of a budget for an EA must be devoted to data management. It could be argued, in the case of GIS
data, that this figure is reasonable subsequent to acquisition of air photos or imagery or data development by field-based mapping, spatial interpolation,
image interpretation, or modeling. Resources for
developing the input data, if they are not already
available, could equal or exceed by an order of
magnitude the other costs of an EA. For example,
detailed maps of vegetation that included canopy
cover and crown size estimates for forested stands
were produced for several million hectares of Forest Service lands in California; at 1 : 24,000 scale
and 2-ha minimum mapping unit, the cost was
about $0.30lha (Franklin and Woodcock, 1997). On
the other hand, interpolated climate surfaces might
be produced for only a few cents per square kilometer. Other existing data, such as digital terrain
models, may have been developed under federal
programs and be available at low cost unrepresentative of the actual production costs of these data
(Estes and Mooneyhan, 1994).
11.4.1 Data Sources
Digital spatial data essential to some forms of ecological assessment, especially at landscape or regional scales, include (1) remotely sensed data
(satellite imagery, digital orthophotographs, aerial
photography, or digital airborne camera data); (2)
interpolated climate data; (3) digital elevation models; (4) digital cartographic data such as we find on
a USGS I : 24,000 topographic map (transportation
routes, hydrography, land ownership); and (5) thematic maps of environmental variables, including
land use, landform, soils, geology, and vegetation
(such as those compiled by the USGS, Forest Service, National Park Service, and Gap Analysis Program). Remotely sensed data were discussed in
Chapter 10. In this chapter we will briefly discuss
the other four types of data, especially with respect
to the following: Were the spatial data created for
the first time in a digital form (through remote sensing or by interpolation), or were they digitized from
traditional maps?
Interpolated climate data are useful in ecosystem
characterization (see Chapters 22 through 29) for
predictively mapping vegetation or other ecosystem characteristics that are otherwise difficult to
153
map (see Section 11.5) and as input to ecosystem
models (discussed in Chapter 18). These may include interpolated surfaces of long-term climate
(Hutchinson, 1987, 1996; Daly et al., 1994;
Hutchinson and Gessler, 1994; Daly and Taylor,
1996; Hutchinson et al., 1996) or daily weather
variables (Running et al., 1987; Glassy and Running, 1994; Running and Thornton, 1996). Some
commercial data products are available; however,
there is still considerable debate over the most appropriate methods for interpolation (Hutchinson,
1984; Hutchinson and Gessler, 1994). These climate surfaces can be generated for a specific region from climate station data by a climatologist
with expertise in spatial analysis methods.
Digital elevation models (DEM), representing
land surface topography, are essential for a number of purposes related to EA. Elevation and simple and complex terrain attributes derived from a
DEM, such as slope and aspect, may be used directly in ecosystem characterization. More frequently, however, they are used to model the distributions of complex ecological or biophysical
patterns and processes, such as site potential, vegetation type, plant or animal species distributions,
potential solar insolation, and topographic moisture
(Moore et ai. 1991; McNab, 1993; Dubayah and
Rich, 1995; reviewed in Franklin, 1995; Franklin
et aI., 2000). DEMs also are commonly used in
ecosystem models, especially of hydrologically
mediated processes (see Chapter 18). Although
standard data products are available from the National Mapping Division of the USGS, these data
are derived from existing topographic maps that
may be more than 40 years old or are produced using photogrammetric methods that are prone to systematic and nonsystematic errors (Weibel and
Heller, 1991; Estes and Mooneyhan, 1994). Although in the future greatly expanded sources
of digital elevation models will include highresolution stereo satellite imagery, the USGS products are currently the best or only source of data
for most of the United States. Gridded (raster) products have, at best, 30-m cell size (resolution),
whereas research indicates that DEMs of 10- to 20m resolution are required to accurately depict topographic features and model topohydrological and
other biophysical processes (Hutchinson et al., 1996;
Hutchinson, 2000). A global digital elevation model
with 30 arc second resolution, GTOP030, is now
available from USGS (http://edcwww.cr.usgs.gov/
landdaac/gtop030/ gtop030.html).
In addition to the elevation contours, the other
information found on USGS topographic quadrangle maps (hydrography, roads and water bodies,
land ownership) is available from the USGS in vec-