The Application of Contour Data for Generating High Fidelity Grid Digital Elevation Models

Digitized contours are increasingly being applied as a data source for generating regular grid digital elevation models (OEMs). The special characteristics of contours as a DEM data source, and some published contour-specific interpola tion algorithms, are reviewed in this paper. Test results for three algorithms are presented, employing both synthetic and real surface data. The algorithms are: linear interpo lation with four data points found in the grid axis direc tions (LIXY), linear interpolation within two data points found in the approximate direction of steepest slope (LISS), and cubic interpolation within four data points found in the approximate direction of steepest slope (CISS). The results indicate that high fidelity OEMs can be generated from cont ours, particularly when associated terrain features, such as break lines, ridges and spot heights, are incorporated into the input data. The root mean square errors of interpolated heights, relative to the contour data, range from 3% to 27% of the contour interval, depending on the interpolation al gorithm, surface structure, and composition of the input data. Error surface contours are employed to illustrate the extent of systematic errors in the interpolated OEMs. The fidelities of the OEMs interpolated by means of the steepest slope algorithms are significantly higher than those result ing from grid axis interpolation.