Comparison of spatial statistical methods for mapping of soil characteristics. (Case Study: rangeland of eshtehard)

Spatial correlation is the first step in the interpolation of field data and mapping of soil properties.The aim of this research was to study the efficiency of two spatial statistics methods i.e., Kriging and inverse distance weighting for mapping of soil properties. Five sampling units were selected in the region, and the location of soil profiles was so determined to cover the whole area. In each unit, six profiles and totally 30 soil profiles were dug in the whole area. Soil samples were taken from two depths of 0-20 cm and 20-80cm. Soil variables including gravel, clay, silt, lime, organic matter, pH and EC were measured in both soil depths. In the GS software, the accuracy of two spatial statistics methods was tested using cross validation with the help of two statistical parameters: MAE and MBE. According to the results, MAE and MBE, related to the Kriging method, for the majority of soil parameters, are less than that of inverse distance weighting method; therefore, Kriging is a more accurate method to interpolate soil properties.