Reconstructing ocean surface current vector field from SAR doppler shift measurements

Abstract

The Doppler shift observed by single-beam synthetic aperture radar (SAR) has been widely used to retrieve the radial velocity of ocean surface currents. However, operational marine forecasting centers require full surface current vector fields for data assimilation and forecast validation. To address this need, we propose a method to reconstruct the two-dimensional surface current vector from SAR Doppler shift measurements, under the assumption that ocean mesoscale currents are quasi-geostrophic. In this approach, the known range-directed current velocity derived from SAR Doppler shift measurements is used to estimate the azimuthal component based on the geostrophic approximation. The feasibility of the proposed method is preliminarily assessed by comparing the reconstructed surface current fields with global ocean analysis and forecast products from the Copernicus Marine Environment Monitoring Service (CMEMS). The results show a bias of 0.01 m/s and a root mean square error (RMSE) of 0.1 m/s. Additionally, surface current fields in the Gulf Stream and Agulhas Current regions are reconstructed using Sentinel-1A SAR Doppler shift observations and validated against collocated drifting buoy measurements, yielding a bias of 0.05 m/s and a RMSE of 0.19 m/s. These findings suggest that the potential of the proposed method for accurately reconstructing surface current fields from SAR Doppler measurements.