Mapcurves: a quantitative method for comparing categorical maps

We present Mapcurves, a quantitative goodness-of-fit (GOF) method that unambiguously shows the degree of spatial concordance between two or more categorical maps. Mapcurves graphically and quantitatively evaluate the degree of fit among any number of maps and quantify a GOF for each polygon, as well as the entire map. The Mapcurve method indicates a perfect fit even if all polygons in one map are comprised of unique sets of the polygons in anothermap, if the coincidence amongmap categories is absolute. It is not necessary to interpret (or even know) legend descriptors for the categories in themaps to be compared, since the degree of fit in the spatial overlay alone forms the basis for the comparison. This feature makesMapcurves ideal for comparing maps derived from remotely sensed images. A translation table is provided for the categories in each map as an output. Since the comparison is category-based rather than cell-based, the GOF is resolution-independent. Mapcurves can be applied either to entire map categories or to individual raster patches or vector polygons. Mapcurves also have applications for quantifying the spatial uncertainty of particular map features. W. W. Hargrove (&) Environmental Science Division, Oak Ridge National Laboratory, P.O. Box 2008, M.S. 6407, Oak Ridge, TN 37831-6407, USA E-mail: [email protected] Tel.: +1-800-2412748 F. M. Hoffman Computer Science and Mathematics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA E-mail: [email protected] P. F. Hessburg USDA Forest Service, PNW Research Station, Wenatchee, WA 98801, USA E-mail: [email protected] J Geograph Syst (2006) 8: 187–208 DOI 10.1007/s10109-006-0025-x