Evaluation and integration of reanalysis rainfall products under contrasting climatic conditions in India

Abstract

Advancements in science and technology lead to the development of different sensors that measure precipitation, an intrinsic component of the hydrological cycle. As conventional rainfall measurement is tedious, satellite precipitation products are being developed. In the current study, the newly released ECMWF (European Centre for Medium-Range Weather Forecasts) ReAnalysis (ERA-5) along with the Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) dataset was statistically evaluated against Indian Meteorological Department (IMD) gridded data and was employed for streamflow simulations using the Soil and Water Assessment Tool (SWAT) model in five contrasting climatic regions of India. Distinct calibration scenarios were developed to test the performance of these reanalysis datasets for simulating streamflow. From categorical and continuous statistical results, ERA-5 exhibited better performance in detecting low (0–5 mm/h) and medium intensity (6–25 mm/h) rainfall, whereas CHIRPS manifested better performance in detecting high intensity rainfall events (>25 mm/h). CHIRPS proved to be effective in simulating streamflows in three out of five basins, whereas IMD exhibited better performance in the other two basins leaving ERA-5 with poor performance for streamflow simulations in all the basins. It was also observed that Satellite Precipitation Products (SPP's) are accurate in humid and tropical regions when compared to arid and semi-arid regions. The worst performance was exhibited in the Ponnaiyar river basin by SPP's and hence bias correction and integration techniques were applied to improve the performance of streamflow simulations further. The novel integration of SPP employing nudging scheme yielded better streamflow simulations when forced into SWAT hydrological model compared to the streamflow simulations obtained when loaded with bias-corrected and raw datasets. The implemented nudging scheme improved the performance of streamflow simulations and hence can be implemented in any basin that is ungauged or resulting in poor hydrological modeling performance when employed with one SPP. The adopted nudging scheme will be highly useful to generate long-term consistent precipitation records in an ungauged or poorly gauged river basin.