Methodology for Comparing Maps of Spatial Error Distribution of Digital Terrain Models

Abstract

The paper presents the usage of several methods for estimation of the spatial distribution of error in GRID 2.5D Digital Terrain Models. Most of them allow generating error and its standard deviation spatial distribution maps in many configurations depending on a number of parameters. An attempt has been made at developing a comparative algorithm to select the variant for which the best results were obtained. Moreover, it can be applied to compare different methods with each over. A prerequisite is the possession of reference data for which their faultlessness can be assumed. Research has shown that empirical formulas are the most universal method for estimating spatial accuracy of DTMs.