Assessment of INSAT-based rainfall products over meteorological subdivisions of India

Abstract

Reliable and near-real-time rainfall information is essential for meteorology, agriculture, disaster preparedness, and water management. Satellite-based rainfall estimates are highly useful in continuously monitoring rainfall across a wide area. However, based on the satellite sensor capabilities and algorithms, these estimates can vary and have some limitations.  In this study, we carried out the validation of satellite-derived rainfall estimates, particularly focusing on the INSAT Multi-spectral Rainfall (IMR) and Hydro-Estimator Method (HEM), by comparing them to ground-based measurements from the India Meteorological Department (IMD) during the monsoon seasons of 2021 and 2022. Accurate rainfall data is crucial for agricultural planning, disaster management, and water resource management, making this research vital for ensuring reliable rainfall measurements across India. Our analysis explores various metrics, including correlation coefficients, root mean square error (RMSE), and bias scores, to evaluate how well IMR and HEM capture daily rainfall patterns.

The analysis reveals that IMR generally shows a stronger correlation with IMD data compared to HEM, though both products exhibit significant regional variations in performance. The RMSE analysis indicates that IMR provides more accurate rainfall estimates in central and western India. IMR and HEM display higher errors in the Himalayan region and parts of southern India. Bias analysis further reveals that IMR and HEM tend to underestimate rainfall in northern and western India and overestimate it in areas like Gangetic West Bengal and central India. These discrepancies underscore the need for continuous validation and localized calibration of satellite-derived rainfall estimates to enhance their accuracy. This research highlights the importance of improving satellite-based rainfall estimation techniques to support effective weather monitoring and dependent applications.