Publications

Krishnan, S; Pradhan, A; Indu, J (2022). Estimation of high-resolution precipitation using downscaled satellite soil moisture and SM2RAIN approach. JOURNAL OF HYDROLOGY, 610, 127926.

Abstract
Soil moisture (SM) and precipitation play an essential part in land surface and atmospheric processes. Availability of high-resolution SM and precipitation data are quintessential for regional and global studies in hydro climatology. There exists a strong interconnection between the varying amounts of SM and the precipitation over a given area, giving rise to the bottom-up approach of estimating precipitation from SM2RAIN. This study examines precipitation derived from SM, downscaled using a unique cognation of land surface temperature (LST), vegetation index (VI), and SM. For this work, MODIS optical data having one km spatial resolution has been used to downscale the SMOS (25 km) and ESA CCI (25 km) SM data to generate four fine-resolution SM products SMOS-LST-NDVI, SMOS-LST-EVI, ESA CCI-LST-NDVI, and ESA CCI-LST-EVI, respectively. In this study, the Ganga basin in India is used as a case study area to assess the bottom-up SM2RAIN approach. The correlation analysis of SM products with LST, NDVI, and EVI, shows that LST has a robust negative correlation, whereas NDVI and EVI have a significant positive correlation. Results obtained from the SM2RAIN simulation from downscaled SM managed to capture the spatial variation of the high precipitation event. Validation with IMERG GPM precipitation data shows that the SMOS-LST-NDVI precipitation product best captures the spatial variation of observed precipitation. The present study results provide convincing evidence that high-resolution precipitation can be estimated using satellite-derived SM data at basin level, as long as better calibration and validation data are available.

DOI:
10.1016/j.jhydrol.2022.127926

ISSN:
1879-2707