Soil loss rate estimation using a hybrid model of geographic information system coupled with fuzzy logic technique

Abstract

An attempt has been made in this study to quantify the soil loss rate and their spatial distribution using a hybrid model of geographic information system coupled with fuzzy logic technique. Shewa Semien zone, Ethiopia, characterized by steep slopes and high rainfall was chosen for this study. A total of ten thematic layers pertaining to soil erosion study, namely rainfall erosivity factor, topographic factor, drainage density factor, plane curvature factor, crop management factor, soil erodibility factor, silt percentage factor, hydraulic conductivity factor, soil texture factor, and stream power index, were analyzed. Fuzzy logic membership values were assigned to different classes within each thematic layer depending on their erosion potential. The final output raster was reclassified into four zones depending on the fuzzy membership values: slight, moderate, severe, and very severe zones with their fuzzy membership values varied from 0 to 0.05, 0.05 to 0.30, 0.30 to 0.65 and 0.65 to 1.00, respectively. The “very severe” erosion zone characterized with the soil loss rate of 67.84–96.26 t/ha/year was found to cover an areal extent of 400,524.92 ha which demands immediate conservation measures. Slope was found as a major factor determining the severity of soil loss rate spatially. The influence of slope, erosivity, stream power index and drainage density toward soil loss rate is documented well in this study. This study testifies the efficiency of the hybrid model for the demarcation of erosion hazard zones and can be applied at continental scale for landscape development.