Simulation of floods using merged satellite and gauge precipitation data

Four merging techniques: linearised weighting (LW), multiple linear regression (MLR), inverse root-mean-square error (IRMSE), weighting and optimal interpolation (OI) were applied to merge widely used satellite precipitation datasets (SPDs). Statistical evaluation was done using: root mean square error (RMSE), Pearson correlation coefficient (r) and bias. While Nash-Sutcliffe efficiency (NSE), coefficient of determination (R2) and percent bias (PBIAS) were used to assess the SWAT model performance. Peak events were evaluated using peak rate discharge errors (ENS and ERP). The statistical evaluations proved OI, the most suitable data merging technique. Hydrological assessment showed that the observed and simulated flows have a strong correlation for both the gauge and the merged data. For peak evaluation, the values of ENS and ERP were close to their optimal values of 1 and 20, respectively. Peak rate discharge error (ERP) has lesser values for the merged data as compared to the gauge data, advocating its use for hydrological modelling in data scares regions.