Hydrological stream flow modelling on Tungabhadra catchment: parameterization and uncertainty analysis using SWAT CUP

Adequate stream flow measurement is vital for agricultural watershed management and its effect on many aspects of water balance parameters. For this reason soil water assessment tool (SWAT) has been applied for the measurement of the stream flow to the Tungabhadra catchment in India. This article describes a methodology for calibration and parameter uncertainty analysis for distributed model based on generalized likelihood measures. The sequential uncertainty domain parameter fitting algorithm (SUFI-2) and generalized likelihood uncertainty equation (GLUE) of SWAT CUP (calibration and uncertainty program) work with multiple sets of parameter values and allow to user within the slight limitation of the model structure in boundary conditions and field observations. Performance of the SUFI-2 and GLUE techniques was evaluated using five objective functions, namely P-factor, R-factor, coefficient of determination R-2, Nash-Sutcliffe (NS) and coefficient of determination divided by coefficient of regression bR(2) calculated on daily and monthly time-steps. The obtained results showed that the observed and simulated discharge were not significantly different at the 95% level of confidence (95PPU). The results show excellent correlation during monthly calibration time-steps, whereas daily calibration exhibits relatively good agreement between the observed and simulated flows.