conclusion from this study was that the suitability of various digital elevation data is primarily a function of the research objectives and scale of application (Miller et al., 1999).
In a study conducted by Garbrecht et al. (1994), the accuracy of drainage features extracted from DEMs as a function of DEM resolution is evaluated. The horizontal resolution of a DEM with an original grid spacing of 30 meters is decreased by cell aggregation. Selected drainage features for several hypothetical channel network configurations were extracted for a range of DEM resolutions using TOPAZ software.
The study by Garbrecht et al. (1994) concluded that the dependency of physical characteristics on grid resolution “was introduced by the inability of a DEM to accurately reproduce drainage features that are at the same scale as the spatial resolution of the DEM.” In other words, the number of channels, the size of the drainage area, and the channel network pattern from a low resolution DEM may depart considerably from the one obtained by a high resolution DEM. For sinuous channels, the use of a low resolution DEM results in shorter channel lengths. For networks with a high drainage density, the use of a low resolution DEM leads to channel and drainage area capturing, This being the point at which the DEM resolution can no longer resolve the separation between channels or drainage boundaries. If small drainage features are important, then resolution must be selected relative to the size of these features (Garbrecht et al., 1994).
Garbrecht et al. (1999b) discuss the extraction of drainage properties from DEMs. This study compares methods of extracting length and slope values using