DEVELOPMENT OF A DATA ASSIMILATION SYSTEM BASED ON ENSEMBLE KALMAN FILTER AND ITS APPLICATION TO THE SETO INLAND SEA

Abstract

 We developed a data assimilation (DA) system aiming to improve the accuracy of the ocean numerical simulation model for the Seto Inland Sea (SIS). The Ensemble Kalman Filter (EnKF) was coupled with ROMS, leading to a new ROMS-EnKF DA system. Observing System Simulation Experiments (OSSEs) were conducted to evaluate the performance of the ROMS-EnKF system in a hypothetical scenario if a high-precision high-resolution coastal sea-surface temperature (SST) data from the latest GCOM-C satellite is available for DA. The GCOM-C SST data is demonstrated to be highly promising for better realization of the SIS circulations. On the other hand, a non-trivial accuracy deterioration occurs in the Bungo Channel, which can eventually be remedied by introducing virtual thermistor chain observations. We further argued practical and economic efficiency associated with mooring deployment designs.