Assessing the performance of reanalysis climatic data in reproducing past precipitation and temperature patterns in India

ABSTRACT Many reanalyses have been used in climate-related studies globally, but these products often contain significant biases, especially when it comes to precipitation and temperature (P&T) data. The potential of these reanalyses to reproduce P&T patterns changes significantly at the regional level, particularly in areas with considerable spatio-temporal variability, like India. Five global reanalyses, namely the Climate Forecast System Reanalysis (CFSR), the European Centre for Medium-Range Weather Forecasts (ECMWF) fifth generation Reanalysis (ERA-5), the Modern-Era Retrospective Analysis for Research and Application (MERRA-2), the Japan Meteorological Agency (JMA) 55-year Reanalysis (JRA-55), and the National Centre for Medium-Range Weather Forecasting (NCMRWF) Reanalysis (IMDAA) estimate the annual and seasonal variability in spatio-temporal patterns of P&T data with respect to the IMD gridded data for 30 years from 1991 to 2020. The reanalyses were evaluated for spell analysis, and statistical and categorical indices, for seasonal and annual daily values. MERRA-2 followed by JRA-55 exhibited the best performance for precipitation across India, as gauge-based precipitation observations were used with updated forcing. Similarly, JRA-55 followed by MERRA-2 perform better than other datasets for temperature data. The reliability and relevance of reanalyses for climatic and hydrological studies in India are better understood as a result of these findings.