The Effect of Anisotropy on the Accuracy of a Digital Terrain Model

Abstract

Point interpolation is the most common method in digital terrain modeling. It is used to determine the height value of a point from the known heights of neighboring points. There are two implicit assumptions behind interpolation techniques: (a) the terrain surface is continuous and smooth and (b) there is a high correlation between the neighboring data points. The process of point interpolation is carried in two stages: a. The search for the neighborhood of point; and b. The prediction of the value of an unknown point from the searched points. Several methods have been used for searching neighborhood of point, during point interpolation process, however; the problem of correlation in data which is needed for accurate interpolation was given little attention. Geostatistical methods of Kriging provide the method for apriori analysis of data, for correlation, however the complication of the method makes it of little use. This work attempts to develop a simple graphical method (i.e. contour map) to define the search neighborhood of correlated data. The method proves to be effective and quite flexible in pinpointing area of correlation, and accurate in predicting the values of unknown points.