Author(s): Sandipan Paul; Priyank J. Sharma; Ramesh S. V. Teegavarapu
Linked Author(s): Priyank J. Sharma
Keywords: Indian Summer Monsoon Rainfall (ISMR) Gridded precipitation datasets Performance assessment Frequency-based performance measures Ranking
Abstract: Precipitation is a crucial variable in Earth system processes and determines the cyclicity of water, energy, and carbon. The characteristics of precipitation events (occurrences, magnitude, and breaks) influence the array of physical processes and socioeconomic activities. With particular relevance to India, it receives ~75-80% of its total annual rainfall through the Indian Summer Monsoon Rainfall (ISMR) from June to September (JJAS), resulting from moisture-laden winds from the SW direction in response to the land-sea pressure gradient induced by thermal contrast. The ISMR is vital for recharging groundwater, replenishing soil, sustaining river flow, and supporting Himalayan glaciers, making it crucial for India's rain-fed agrarian system, society, and economy. Hence, understanding the ISMR characteristics is essential for comprehending climatic processes, hydrometeorological forecasting, and assessing the impacts on human life and the economy (deciding the sustenance of millions of people). With the advancement of geospatial technology, different types (gauge interpolated, satellite, reanalysis, and hybrid) of precipitation datasets have emerged. Although these datasets have their advantages and limitations, there is a significant gap regarding the knowledge of representativeness of these datasets, specifically over data-scarce region like India. Therefore, this study evaluates the accuracy of gridded rainfall data over India, aiming to answer two specific questions: (a) how well do the gridded rainfall datasets capture the chronologically matched different rainfall events having distinctive magnitudes, and (b) what is the overall and regional suitability (rank) of rainfall datasets for hydroclimatic applications? This study introduces the rank score, a linear weighted combination of continuous and categorical performance measures, to evaluate each dataset's overall suitability. Moreover, the present study provides a unique and innovative perspective on the ability of data products to accurately characterise precipitation over a vast topographic, ecological, and climatic gradient region, such as India.
Year: 2025