STUDY ON DATA ASSIMILATION USING SST OF SATELLITE OBSERVATIONS BY RECYCLE GREEN’S FUNCTION APPROACH IN TOKYO BAY

Abstract

 In coastal areas, numerical models have been improved in accuracy through data assimilation using water quality measurements and HF radar observations. However, until recently, the limited spatiotemporal resolution of satellite data has hindered its utilization, resulting in few application cases. In recent years, there has been significant improvement in the spatiotemporal resolution of satellite data. Therefore, the use of data assimilation with satellite data is expected to be beneficial for closed water areas.

 In this study, data assimilation was performed using satellite-observed sea surface temperature (SST) through the reuse of the Green's function method in Tokyo Bay. The aim was to optimize the parameter values related to the heat balance within the numerical model and examine the impact on the reproducibility of water temperature. As a result, the root mean square error (RMSE) between the satellite-observed SST and the calculated water temperature from the numerical model decreased by approximately 26% at maximum, suggesting the effectiveness of satellite data assimilation in coastal areas.