Problems with standard Boolean maps produced through subjective interpretation of a phenomenon (i.e., forest or soils maps) are discussed and two alternatives based on spatial certainty are presented and discussed. One requires multiple interpretations of the phenomenon in order to construct a library of spatial uncertainties; such an Uncertainty Library can be used subsequently to estimate error across cartographic boundaries. The other requires interpreters/cartographers to identify only those map elements which are "100% certain." A spatial interpolation algorithm is subsequently applied to this information to "fill in the gaps" with certainty information for each map type. Both methods have advantages and disadvantages relative to standard Boolean maps and also to each other. These are discussed in general terms and also through the presentation of specific examples. It is concluded that though uncertainty-based cartographic representations provide more flexibility than do conventional Boolean maps, the construction of the former is not without its problems either.
@InProceedings{ lowell:1997:OIIO,
author = {Kim E. Lowell},
title = {Outside-in, Inside-out: Two Methods of Generating
Spatial Certainty Maps},
booktitle = {Proceedings, Second Annual Conference of
GeoComputation, Dunedin, New Zealand},
year = {1997},
}