SOIL EROSION AND SEDIMENT YIELD ASSESSMENT OF KUNTHIPPUZHA SUB-WATERSHED USING SWAT AND RUSLE

Abstract

The two essential fundamental natural resources for the existence of life are soil and water. Soil erosion is an environmental crisis in the world today that threatens the natural environment and also agriculture. Erosion removes the top fertile soil, degrades soil fertility, water quality and soil productivity. Almost 119.2 M-ha area out of 328 M-ha of geographical area is severely eroded in India (NBSS & LUP, 2005). The soil erosion risk assessment is helpful for land evaluation in the regions where soil erosion is a major threat for sustainable agriculture. Field studies for prediction and assessment of soil erosion are expensive and time-consuming. Soil erosion models are able to take into account many of the complex interactions that affect the rate of soil loss or erosion and are capable of simulating erosion processes in watersheds. There are several types of models like empirical, semi-empirical, and physical models. One of the extensively used empirical models to forecast soil erosion caused by water loss is the Revised Universal Soil Loss Equation (RUSLE) model. Among different physical models, Soil and Water Assessment Tool (SWAT) is widely used to predict soil erosion and sediment yield in the watersheds. The main objective of the study was to find the water balance components, soil erosion and sediment yield of Kunthipuzha sub-watershed using SWAT, soil erosion and erosion prone areas using RUSLE and to compare model results with observed data and to estimate the Sediment Delivery Ratio (SDR). SWAT model was simulated for a period of 32 years (1990-2021). Global sensitivity analysis of the model was done using the Sequential Uncertainty Fitting (SUFI-2) algorithm in SWAT-CUP. SWAT model was calibrated for discharge as well as sediment yield. The calibration was done for the period of 21 years from 1990 to 2014, and validation was done for a period of 3 years from 2015 to 2017. Nash Sutcliffe Efficiency (NSE) and R² for the calibration period was 0.77 and 0.86, and for the validation period it was 0.75 and 0.81 respectively for discharge. In case of sediment yield, NSE and R² for the calibration period were 0.79 and 0.88, whereas for the validation period it was 0.76 and 0.82 respectively. During the years of study, the outflow from the watershed is mainly in the form of surface runoff (ranges between 43 to 54%) and ground water flow (24 to 34%).From the soil erosion estimation using RUSLE and SWAT models, average annual soil loss was estimated to be 8.865 and 8.1 t ha −1 y −1 respectively. Sediment yield estimated from the outlet (Sub-basin 27) is 3.9 t ha −1 y −1 . Both in RUSLE and SWAT, major area of the basin is in slight erosion category (5-10 t ha −1 y −1 ). In SWAT, sub-Basins 2, 3, 4 and 7 in the North-eastern region comes under very severe erosion category (> 40 t ha −1 y −1 ). Since most of the watershed (around 70% area) comes under slight erosion category, soil erosion can be controlled by practicing agronomical measures. In the moderate erosion risk areas (around 9.35%), contour bunds and terraces are suggested. The SDR obtained for the entire watershed was very low (0.01 to 0.036), which indicate that even though more amount of soil gets eroded from some of the sub-basins, it gets deposited at intermediate locations before reaching the Pulamanthole gauging station. Analysis using both the models shows that the North-Eastern area of the watershed (Mannarkkad, Pottasserry, Puthupariyaaram etc.) experience more erosion, and hence more soil conservation measures need to be adopted in this region. Both models gave similar trend of spatial variation of soil erosion qualitatively and quantitively and less deviation from the observed sediment data. The result obtained is helpful for giving recommendations for proper soil conservation measures in the area.