Digital Soil Mapping of Residual Lateritic Soils in Kerala Using Interpolation Methods

Soil is considered to be a highly heterogeneous material whose properties vary from place to place and at a particular place with respect to depth. For carrying out any kind of geotechnical analyses on a regional scale, geotechnical soil properties are required which is possible only after drilling large number of boreholes, which is both time-consuming and uneconomical. Digital soil mapping is a process by which soil properties are predicted on a regional scale from a known limited number of point observations. Digital soil map consists of pixels in a grid wherein each pixel has a unique geographic location and soil data. Different interpolation techniques can be used for the development of a digital soil map. Various studies have shown mixed conclusions about the accuracy of the different interpolation methods. In this paper, two commonly used interpolation methods namely IDW and ordinary kriging were applied to develop a digital soil map of the residual lateritic soils in Kerala. The results were validated using the commonly used leave-one-out cross-validation (LOOCV) technique. The accuracy was also checked by comparison of predicted soil properties with an independent soil dataset which was not used in the interpolation process. The results showed that ordinary kriging outperformed the IDW method, and therefore, it can be used for development of digital soil map of residual lateritic soils of Kerala.