Estimating rainfall using soil moisture data: application and validation of the SM2RAIN in the Korean Peninsula

The Korean Peninsula experiences significant seasonal variations in rainfall, making regional monitoring critical for effective water resource management. While current rainfall monitoring is densely conducted based on in-situ measurements, the acceleration of climate change necessitates rainfall data with finer spatial resolution. This study evaluates the performance of SM2RAIN in estimating rainfall across diverse soil and rainfall conditions in the Korean Peninsula and examines its applicability in monsoon climates using both in-situ and spatial soil moisture data. The results indicate that SM2RAIN, utilizing in-situ soil moisture data, demonstrated strong rainfall estimation performance in all regions except Bosung, with statistical indicators of Nash-Sutcliffe Efficiency (NSE) ranging from 0.52 to 0.94, correlation coefficient (R) from 0.73 to 0.97, and Root Mean Square Error (RMSE) from 6.09 to 29.44 mm/hr. Conversely, rainfall estimation using SMAP L4 soil moisture data exhibited limitations due to temporal and spatial resolution constraints, with NSE values between 0.33 and 0.67, R from 0.57 to 0.83, and RMSE from 10.09 to 20.25 mm/hr, showing discrepancies in accuracy compared to in-situ data. However, SMAP L4 data proved useful for medium- to long-term rainfall forecasting. This study confirms that optimizing the parameters to SM2RAIN enables accurate and reliable rainfall estimation.