Four-Dimensional Variational Data Assimilation of Surface Currents Measured by Oceanic Radar

Abstract

Surface current data measured by HF oceanic radar is assimilated with Regional Ocean Modeling System (ROMS) and four dimensional variational method (4D-Var). Sensitivity analyses are carried out to decide appropriate parameters for the data assimilation. 4D-Var reduces RMSE of surface velocity by 0.05 m/s at maximum. 4D-Var with the Inner-loop of 10, the radius of horizontal error covariance of 1km gets the better RMSE than other cases. The assimilated surface currents are compared with those assimilated with nudging method. While the results of nudging assimilation have discontinuous change of the currents, those of 4D-Var do not and the time series of surface velocity come closer to those of the observations than those assimilated with nudging outside of the observation range.