Model Performance Evaluation using Streamflow and Potential Evapotranspiration over Middle Tapi Basin, India

Abstract

Integrated watershed planning, management and decision-making are important for the efficient utilization of available water resources. The physics-based hydrological models enable quantitative and qualitative evaluation of water at different spatial and temporal scales. The current study aims to ascertain the reliability of hydrologic simulations using the SWAT hydrologic model. The model was developed for a climate-sensitive and agriculturally dominated catchment, i.e., the Middle Tapi basin (MTB), between Hatnur and Ukai reservoirs. The monthly inflows of the Ukai reservoir were utilized to calibrate and validate the SWAT model for 1998–2007 and 2008–2013, respectively. The simulated potential evapotranspiration (PET) at the basin scale was compared with CRU-PET data to ascertain the reliability of the simulated variable. The global sensitivity analysis shows that GWQMN.gw, REVAPMN.gw, GW_REVAP.gw, SOL_K.sol, CNCOEF.bsn, SOL_AWC.sol are the most sensitive parameters at 5% significance level. The model performance metrics, namely KGE, NSE, R² and PBIAS with respect to observed streamflow during calibration (validation) period are 0.98 (0.98), 0.98 (0.97), 0.98 (0.97) and 0.6% (−1.2%), respectively. The similar statistics for PET during simulation period (1998–2013) are 0.73, 0.84, 0.94 and −4.36%, respectively. The model performance metrics show that the hydrologic model reliably simulates monthly inflows into the Ukai reservoir and PET in MTB. Thus, the developed model can accurately forecast hydrologic changes in response to climate instability in the study area, allowing for better water management practices.