SOIL ATTRIBUTE PREDICTION USING TERRAIN ANALYSIS

This study is based on the hypothesis that catenary soil development o ccurs in many landscapes in response to the way water moves through an d over the landscape. Furthermore, terrain attributes can characterize these flow paths and, therefore, soil attributes. Significant correla tions between quantified terrain attributes and measured soil attribut es were found on a 5.4-ha toposequence in Colorado. Slope and wetness index were the terrain attributes most highly correlated with surface soil attributes measured at 231 locations on a 15.24-m grid. Individua lly, they accounted for about one-half of the variability in A horizon thickness, organic matter content, pH, extractable P, and silt and sa nd contents. This represents an incorporation of finer scale process-b ased information relating to soil formation patterns in the landscape. The computed and measured ranges of terrain and soil attributes, resp ectively, can be used to enhance an existing soil map, even when the e xact form of the relationship is unknown. As a first approximation, a linear relationship was assumed and the interpolated predictions of A horizon thickness and pH compared reasonably well with the observed. S uch techniques may also be applied as a first step in unmapped areas t o guide soil sampling and model development.