Evaluation of quantitative precipitation forecast in five Indian river basins.

Quantitative precipitation forecast (QPF) obtained from a numerical weather prediction model is used for streamflow forecasting. This study systematically evaluates the performance of a deterministic QPF obtained from a national agency, the National Centre for Medium Range Weather Forecasting (NCMRWF), compared with that obtained from three international agencies. For observation/reference datasets, we used satellite, raingauge, and satellite-gauge merged data products. The forecast skill is evaluated for 177 sub-basins in Ganga, Narmada, Mahanadi, Tapti, and Godavari river basins. Our results indicate that the forecast accuracy of NCMRWF is closely comparable with that of QPF obtained from a European agency. We conclude that selecting accurate observation/reference data is critical for forecast performance evaluation. We determine the preferred forecast and reference datasets in the selected river basins. Our results suggest that for estimating the parameters of a post-processor, a comparable dataset can be used for which data are available for a longer duration. [ABSTRACT FROM AUTHOR]