A study of soil-landscape patterns based upon a digital elevation model

Semi-detailed soil surveys often lack details on landscape position of specific soils. The primary objective was to investigate a method of improving survey detail, by providing soil position information derived from a digital elevation model (DEM). Slope magnitudes determined from the DEM were underestimated compared to those from the soil survey. The pattern of DEM-derived slope breaks and soil survey slope classes did not correspond along unidirectional transects, but did when the entire study area was considered. Potential depressions were located on the DEM using a combination of profile and plan curvature. Proportions of depressions were significantly different from the proportions of Gleysols and gleyed subgroups in each soil survey polygon. These depressions may be considered centre locations of actual landscape depressions. Depressions in open areas could be located with DEM data alone, but not those within treed areas. Satellite image data provided information on soil wetness in both areas. Using this information, possible depressions in treed areas were identified. Positioning soils in the landscape using both the DEM and satellite imagery was limited by the relatively coarse 25 m spatial resolution.