efficiency of digital elevation model and its attributes for soil mapping using soil-land inference model (solim)

abstract limitations of traditional (conventional) soil surveys and improvement of information technology have lead soil surveyors to invent new methods which are generally called digital soil mapping (dsm). the aim of these methods is the prediction of soil classes or soil properties based on easily-available or measuring environmental variables. the objective of this investigation is to study the efficiency of digital elevation model and its derivates for soil mapping using soli-land inference model (solim) and credibility of its results in the borujen area, chaharmahal-va-bakhtiari province. eighteen terrain attributes including height, slope (angle), aspect, curvature, minimum curvature, maximum curvature, tangent curvature, profile curvature, planform curvature, flow direction, flow accumulation, direct radiation, diffuse duration, diffuse radiation, area solar radiation, power index, sediment index and wetness index, were derived from the dem. these derivates as well as three dominant soil subgroups and seven soil families of the region were used to construct the input data matrix of the model. results showed an accuracy of 65% and 40% for interpolation and extrapolation of the soils at subgroup level, respectively. the accuracy decreased to half when soil families were considered for credibility of the model. because of using crisp limitations in american soil taxonomy system, assessing soil survey results can be miss-leading partially, whereas using solim model shows well the reality of the soils in the field.