A Holistic GIS-based Approach for Thematic Extraction of the Soil Erosion Estimates Using Predictors as a Function of Climate,Land Cover,Relief,Soil and Topography

Abstract

Present claim for update of existing soil information has taken a heavy toll to fit the needs of the current environmental modelling data demands. The information derived from the age old data of 1960's and 1970's that are being used in most cases at present situation are losing its relevance to represent the reality of now existing soil status. Due to various transformations that have undergone in the land use, crop management practices, intensive cultivation integrated with unscrupulous fertilization (imbalanced fertilization), certain fertile soils of the past have reached a status of degraded lands or unproductive lands. Henceforth, present focus is visualized on developing modelling approaches through exploitation of the new GIS and remote sensing techniques as a feasible option and to cut down the cost factor that would be a certain unaffordable demand through conventional approaches. In this study, "SEIMS network" (Soil and Environment based Mapping System) approach was adopted to update information on the soil loss due to water erosion. Conceptually, this approach is based on the principles of Data Mining and Knowledge Discovery (KDD) method. The spatial relationships among the independent variable related to the soil erosion process (predictors) are accounted to estimate soil erosion through spatial modelling. In this study, about four climatic variables (temperature, rainfall, potential evapotranspiration and rainfall seasonality), one for land cover (derived from MODIS spectral bands), three variables for soil attributes (soil crusting, soil erodibility, top soil organic carbon content) and two terrain parameters (altitude and slope) were chosen as predictors for modelling soil erosion process. The reclassified soil erosion map derived through SEIMS network scheme exhibited a better correlation (r $^2$ = 0.891) with the published class-based soil erosion map of Tamil Nadu (NBSS & LUP, 1997). Thereby, holistic GIS-based approach was found to be efficient in transforming the useful subjective, qualitative and categorical information into objective and quantitative information serving the present demands of soil information update.