Problems with using overlay mapping for planning and their implications for geographic information systems

Abstract

/ As part of the planning process, maps of natural factors are often superimposed in order to identify areas which are suitable or unsuitable for a particular type of resource management. Overlay maps may also be used to identify analysis areas for predictive modeling of resource productivity and ecological response to management. Current interest in applying computer-assisted mapping technology to making overlay maps is drawing attention to geographic information systems for this purpose. The resultant maps, however, may be so inaccurate or unable to capture significant units of productivity and ecological response that they could lead to imperfect or false conclusions. Recommendations are made on how to proceed in light of these problems.An increasingly common part of the planning process for wildland areas is the preparation of map overlays to determine the suitability of an area for alternative resource management. Techniques for suitability analysis are presented in a number of papers (Varnes 1974, Larson and others 1977, Hopk]ns 1977, Ferguson 1981, FAO 19841 Steiner 1983. Maps of natural factors such as bedrock geology, soil, slope, and vegetation are superimposed in order to locate and delineate areas where particular factor combinations occur. This is done either by manually overlaying the factor maps or with an automated geographic information system (GIS). A geographic information system is a computer information system with the capability of handling spatially related data, specifically map-related data such as soil type boundaries and stream networks.An accurate overlay map would have unquestionable value in identifying types of land that will respond in a coherent and uniform way to the application of a variety of management practices. A mapping system for identifying such land types could be useful in current attempts to model the response of wildland areas. Of these models--for example, FORPLAN (Johnson and others 1986)--most utilize land types which are represented in the model by the component known as an analysis area. Because of their fundamental impor-

A field demonstration of the entropy-weighted fuzzy DRASTIC method for groundwater vulnerability assessment

“…Numerous sensitivity analysis studies have been done on DRASTIC (e.g. Bailey 1988, Heuvelink et al 1989, Lodwick et al 1990, Barber et al 1993, Merchant 1994, Babiker et al 2005, Panagopoulos et al 2006. There are two main methods of doing sensitivity analysis in DRASTIC: one is the map removal analysis and the other is single-parameter sensitivity analysis (Napolitano and Fabbri 1996).…”

A primer on evaluation and use of natural resource information for corporate data bases

“…Recently, several Forest Service authors (Bailey 1988b, Lund 1986a, Evanisko 1990) have stressed the importance of the quality of natural resource data bases, especially those intended for use in an electronic setting for corporate GIS' s. Bailey (1988a) points out that representing ecological units as uniform regions may lead to false conclusions. Such representations may not capture significant subunits of productivity or ecological response.…”

Testing of a GIS model ofEucalyptus Iargiflorens health on a semiarid, saline floodplain

“…The use of classes to aggregate spatial information has several advantages, including ease of mapping and ease of management decisions. In fact, an overlay map, whose classes identify areas that will respond in a consistent manner to various management options, would be invaluable (Bailey 1988). The use of classes is also suitable for management and display of data that exist in discrete form and for displaying data of poor accuracy or of limited extent.…”

“…The use of classes is also suitable for management and display of data that exist in discrete form and for displaying data of poor accuracy or of limited extent. However, overlaying maps to form classes may result in excessive compounding of positional or purity errors and/or lead to an inability to capture significant units of ecological response (Bailey 1988). The resultant classes may therefore not accurately represent field conditions or may simply be meaningless in terms of explaining the ecological processes in force.…”