152
Geographic Information Science and Ecological Assessment
represent a model of the real world, and therefore
the GIS serves as a test bed for studying environmental processes. Clarke (1999, p. 5) notes that GIS
technology has changed our entire approach to spatial data analysis and, accordingly, Goodchild
(1992a, p. 41) states that geographic information
science should address "the generic issues that surround GIS technology, impeding its successful implementation, or emerge from an understanding of
its potential capabilities." Subsequently, one flagship journal in the field changed its title from the
International Journal of Geographic Information
Systems to IJGI Science. GIS, like its sister technology, remote sensing, can be used with scientific
or technological approaches to environmental problem solving (Curran, 1987).
Despite the growth of this literature on geographic information science, some GIS usersecologists, land-use planners, and others-might
profess (as one recently did to the author) that a
GIS is simply one of the data management tools
that they use, in the same category as spreadsheet
software. However, a number of institutional issues
related to GIS and methodological issues related to
spatial data analysis distinguish it from other data
storage support tools (spreadsheet and database
software).
11.3 Institutional Issues
Institutional issues related to GIS are primarily related to the need for investments in software, hardware, and training. Low-cost or public-domain GIS
software is available (GRASS, Idrisi, ArcView, for
example), and some packages contain powerful
tools for analyzing, or at least viewing, geographical data. High-cost software and hardware, however, are usually used to support activities requiring large amounts of data (geographical databases
for large areas at high resolution) or sophisticated
analyses (Morgan, 1987; Lauer et aI., 1991; Clarke,
1999). Although until very recently a Unix-based
(or comparable) workstation was needed to support
GIS-based analysis, a high-end personal computer
is now capable of supporting many of these activities (because of the increases in microcomputing
power that have affected all computer users). Much
of the available software requires substantial training for analysts who already have, or are concurrently acquiring, domain knowledge (geography,
ecology, forestry, conservation biology, range
management).
In addition to knowing how to use the programs
or modules in the software package and write specialized applications programs, usually in a macro
language, the GIS analysts in an organization must
also understand (1) the nature of the input data, especially their source, currency, scale, and accuracy;
(2) the fundamentals of geographic information science (Goodchild, 1992b), including coordinate
geometry and topology, data structures and data
models, algorithms and computational complexity,
and spatial statistics and spatial analysis (Goodchild, 1985); and (3) the types of analyses that potentially can be applied to a particular problem.
This is true whether the institution is a federal or
other land management agency, a university research laboratory, or an environmental consulting
firm, any of which could be involved in some capacity in an EA.
A number of other institutional issues are beyond
the scope of this paper. These include the processes
by which GIS technology is adopted by organizations and used in decision making (Lauer et aI.,
1991; Goodchild, 1992a; Obermeyer and Pinto,
1994) and ways of improving the efficacy of GIS
in supporting management and planning decisions
(Crossland et aI., 1995; Zhu et aI., 1998). Ethical
and legal issues related to privacy (Curry, 1997),
data ownership, and so forth, are also particularly
relevant to GIS. Other issues will become increasingly important in the future; these include data security and access on computer networks, institutional policies regarding digital libraries and the
global network, and cross-disciplinary community
building.
11.4 Data Sources
Agencies and organizations involved in EA must
also grapple with acquiring adequate data for their
analytical needs. There are a growing number of
sources of digital maps, as well as remotely sensed
images and derived data products, notably those
produced by the U.S. Geological Survey (USGS),
and specialized mapping programs within land
management and other agencies. The National
Aeronautics and Space Administration's (NASA)
Earth Observing System (EOS) and other programs
promise expanding availability of global and regional remotely sensed imagery and data products.
However, there is sometimes a misconception that
much of the data required to conduct an ecological
assessment or regional modeling effort already exist in an appropriate form and have only to be acquired, assembled, and analyzed. In fact, in most
inventories and assessments that rely on digital spatial data, the majority of resources are still spent on
developing the data, either because they are not
available or because existing data are out of date
Geographic Information Science and Ecological Assessment
represent a model of the real world, and therefore
the GIS serves as a test bed for studying environmental processes. Clarke (1999, p. 5) notes that GIS
technology has changed our entire approach to spatial data analysis and, accordingly, Goodchild
(1992a, p. 41) states that geographic information
science should address "the generic issues that surround GIS technology, impeding its successful implementation, or emerge from an understanding of
its potential capabilities." Subsequently, one flagship journal in the field changed its title from the
International Journal of Geographic Information
Systems to IJGI Science. GIS, like its sister technology, remote sensing, can be used with scientific
or technological approaches to environmental problem solving (Curran, 1987).
Despite the growth of this literature on geographic information science, some GIS usersecologists, land-use planners, and others-might
profess (as one recently did to the author) that a
GIS is simply one of the data management tools
that they use, in the same category as spreadsheet
software. However, a number of institutional issues
related to GIS and methodological issues related to
spatial data analysis distinguish it from other data
storage support tools (spreadsheet and database
software).
11.3 Institutional Issues
Institutional issues related to GIS are primarily related to the need for investments in software, hardware, and training. Low-cost or public-domain GIS
software is available (GRASS, Idrisi, ArcView, for
example), and some packages contain powerful
tools for analyzing, or at least viewing, geographical data. High-cost software and hardware, however, are usually used to support activities requiring large amounts of data (geographical databases
for large areas at high resolution) or sophisticated
analyses (Morgan, 1987; Lauer et aI., 1991; Clarke,
1999). Although until very recently a Unix-based
(or comparable) workstation was needed to support
GIS-based analysis, a high-end personal computer
is now capable of supporting many of these activities (because of the increases in microcomputing
power that have affected all computer users). Much
of the available software requires substantial training for analysts who already have, or are concurrently acquiring, domain knowledge (geography,
ecology, forestry, conservation biology, range
management).
In addition to knowing how to use the programs
or modules in the software package and write specialized applications programs, usually in a macro
language, the GIS analysts in an organization must
also understand (1) the nature of the input data, especially their source, currency, scale, and accuracy;
(2) the fundamentals of geographic information science (Goodchild, 1992b), including coordinate
geometry and topology, data structures and data
models, algorithms and computational complexity,
and spatial statistics and spatial analysis (Goodchild, 1985); and (3) the types of analyses that potentially can be applied to a particular problem.
This is true whether the institution is a federal or
other land management agency, a university research laboratory, or an environmental consulting
firm, any of which could be involved in some capacity in an EA.
A number of other institutional issues are beyond
the scope of this paper. These include the processes
by which GIS technology is adopted by organizations and used in decision making (Lauer et aI.,
1991; Goodchild, 1992a; Obermeyer and Pinto,
1994) and ways of improving the efficacy of GIS
in supporting management and planning decisions
(Crossland et aI., 1995; Zhu et aI., 1998). Ethical
and legal issues related to privacy (Curry, 1997),
data ownership, and so forth, are also particularly
relevant to GIS. Other issues will become increasingly important in the future; these include data security and access on computer networks, institutional policies regarding digital libraries and the
global network, and cross-disciplinary community
building.
11.4 Data Sources
Agencies and organizations involved in EA must
also grapple with acquiring adequate data for their
analytical needs. There are a growing number of
sources of digital maps, as well as remotely sensed
images and derived data products, notably those
produced by the U.S. Geological Survey (USGS),
and specialized mapping programs within land
management and other agencies. The National
Aeronautics and Space Administration's (NASA)
Earth Observing System (EOS) and other programs
promise expanding availability of global and regional remotely sensed imagery and data products.
However, there is sometimes a misconception that
much of the data required to conduct an ecological
assessment or regional modeling effort already exist in an appropriate form and have only to be acquired, assembled, and analyzed. In fact, in most
inventories and assessments that rely on digital spatial data, the majority of resources are still spent on
developing the data, either because they are not
available or because existing data are out of date