ABSTRACT
This chapter describes the method used to evaluate the effects of generalisation within a Geographical Information Systems (GIS). This included measuring quantitatively, across scales, the generalisation effects embedded in map data for different features, comparing manual and automated generalisation procedures, and carrying out two commonly used GIS map manipulations relying on different source-scale data. In order to measure generalisation effects, it was crucial to try to separate them out from other sources of error. The process starts by calculating the positional accuracy of the maps in order to distinguish between generalisation and non-generalisation error. An explanation of the procedure used for determining the fractal surfaces of both study areas in terms of the roughness of terrain then follows. The chapter concludes with a description of the GIS map manipulations carried out, and the methodology used for measuring generalisation effects that result from the Douglas-Peucker generalisation algorithm.
