SATELLITE-BASED RAINFALL DATA ASSIMILATION USING 3DVAR TECHNIQUE - CASE STUDY IN RED RIVER BASIN

Abstract

Data assimilation of rainfall as input for hydrological forecasting has been intensively researched by many scientists in hydro-meteorological fields. Especially, in areas that characterized by complex terrains and rainfall that are highly data dependent due to transboundary issues. This is true for the upstream area of Red River and Chinese territories. Of which rainfall data is not continuously measured and is not highly avaialble in some remote areas. Meanwhile, in spite of the large domain cover rainfall products from the Earth Observation Satellites (EOS), there still exist a challenge in detecting small scale convective rainfall and local rainfall. This study focuses on applying a 3DVar technique for satellite-based rainfall data assimilation using two satellite rainfall sources: GSMaP (Global Satellite Mapping) and GPM (Global Precipitation Measurement) that are collected and preprocessed by Delft-FEWS. Analyzed rainfall data was then validated using observed rainfall for August, 2018. The results showed that the analysis data could detect some local rainfall points which were not detected by the two satellite rainfall sources thus provides a more complete image of spatial rainfall distribution for the Red River Basin.