Computing and Visualizing Morphologically Sound DEMs in a Raster Environment

The present study dealt with the problem of interpolating a continuous digital elevation model (DEM) from contour lines. Several methods were developed that make use of an interpolation algorithm. The basic idea behind the methodology has been the use of tools provided in a standard raster geographic information system (GIS) environment and the avoidance of problems related to the triangulation algorithms. The interpolation algorithm was developed using raster model operations with 3 different linear solutions. Separate morphological corrections for peaks and pits were calculated. The interpolated DEM can either be used as such for accurate spatial analysis, or generalized for coarser scale analysis, or smoothed for visualization of terrain. The generalization and visualization of DEMs forms another problem area, in which no single solution appears to achieve all the desired properties into a single DEM. Several methods for visualizing the differences between various algorithmic solutions were used to reveal the potentials and drawbacks of different interpolation methods. The interpolation algorithms were tested against 2 DEM generation algorithms provided in a common GIS software. The results show that the interpolation algorithm developed results in minimum deviations to the original data. Morphological features are represented more naturally with the algorithm, and the resulting DEM is visually pleasing since triangulation artifacts are not present.