Soil erosion risk assessment in Kunthipuzha sub-watershed using remote sensing and GIS

Abstract

This study mainly focused to identify the erosion prone areas in Kunthippuzha sub-basin using RUSLE as well as MMF model. The effect of spatial and temporal variations of land use-land cover on soil erosion was analysed with the help of NDVI values. The estimation was performed for the year 2000 and 2013. The mean soil erosion estimated for the year 2000 was 18.30 and 20.58 t/ha/y respectively by MMF as well as by RUSLE model. Similarly in the year 2013, it was 32.78 and 35.10 t/ha/y respectively. To find the erosion prone areas in the sub-watershed factorial scoring method was chosen, in which pixel based scoring was done based on mean soil erosion value obtained under each layers of landuse, rainfall, slope and topographic raster. From the study based on the RUSLE model, 0.15% of the area experienced very slight erosion. 40.70% of the area was with slight erosion, 28.76% area was under moderate erosion, 22.61% of the area was under severe range and 6.15% area was under very severe range. According to MMF model, the areal extent observed under slight, moderate, severe and very severe risk categories was 39.02%, 36.94%, 14.92%, 7.50% respectively. NDVI values corresponding to land use were identified, in which negative NDVI values correspond to water pixels whereas higher NDVI values represents the thick vegetation. From the analysis of the temporal variation in land use over the NDVI, it was observed that NDVI value decreased at the points where density of the canopy decreased. The spatial variation of soil erosion varies pixel to pixel according to the landuse pattern. Highest soil erosion risk was observed under built-up plus barren land. Sediment Delivery Ratio (SDR) was calculated for the entire sub- watershed, which shows that most of the eroded sediments get deposited at intermediate location before reaching the outlet. The SDR calculation enhances the importance of adopting GIS technology in soil erosion assessment.