Rainfall Runoff Modeling Using SWAT Model for Chaliyr Basin Kerala, India

Rainfall-Runoff modeling is a mathematical representation describing the relation between rainfall and runoff to the catchment area. The rainfall runoff process is a multifaceted process as it is influenced by a number of direct and indirect factors such as precipitation distribution, evaporation, transpiration, abstraction, and topography and soil types. Moreover, detection and quantification of spatial-temporal trends and changing patterns of rainfall and hydrological modelling is essential for water resources planning, management, flood forecasting and in many other applications. There are numerous rainfall-runoff models available according to varying nature, complexity and purpose of study area. One of these hydrological model, SWAT is widely used to evaluate different parameters of the water resources namely rainfall-runoff modelling, sedimentation, eco-hydrological behavior of the watershed SWAT model has been setup to simulate rainfall-runoff for the Chaliyar river basin at Kuniyil with a catchment area of 2013.4 km. The basin was discretized into 15 sub-basin using an automated delineation routine with an input spatial data of DEM, land use/land cover, soil and slope of the Chaliyar basin. There were 103 Hydrologic Response Unit (HRUs) generated for the entire river basin. In the present study, rainfall trend analysis conducted at four different stations experienced a statistically insignificant increasing and decreasing trends for the period of 1991 to 2011. The model efficiency (R2) value obtained was 0.69, which is considered reasonably satisfactory for estimating a basin runoff. After getting good efficacy of SWAT-run, SWAT-CUP was used to calibrate and validated the model. The SWAT model has been calibrated using daily data of 4 years (2003-2007) and validated for 4 years (2008-2011) considering 2 years of data (2001-2002) for warm up period. The model yielded satisfactory and reliable results with coefficient of determination and Nash-Sutcliffe Efficiency 0.77 %& 0.75 % for calibration and 0.77 % and 0.73 % respectively for validation period.