Estimation of wave-induced contribution to Sentinel-1 Doppler shift observations

The Doppler Centroid Anomaly (DCA) registered by microwave Synthetic Aperture Radar (SAR) contains information about ocean surface motion in the radar line-of-sight direction. The recorded signal is associated with the motion induced by the total wavefield (i.e., both wind waves and swell) and underlying ocean surface currents. Hence, accurate estimates of the wave-induced contribution to the observed DCA is required in order to obtain reliable information about underlying ocean surface current. In this study, we develop an empirical geophysical model function for the estimation of the wave-induced DCA. The study is based on two months of Sentinel-1 SAR Wave mode (WV) DCA observations collocated with wind field at 10m height from the ECMWF model and sea state information from the WAVEWATCH III model.

Analysis of two months of observations acquired over land showed that thanks to the novel Sentinel-1 DCA calibration, the uncertainty in the data does not exceed 3Hz (corresponding to a radial velocity of 0.21/014 m/s in the near/far range. The relationship between the DCA and the near-surface wind is in agreement with previously reported findings under the assumption of fully developed seas; the DCA is about 24% of the range wind speed at 23° incidence angle and decreasing (up to 50%) with increasing incidence angle from 23° to 36°. However, the difference between upwind (i.e., the wind blows towards antenna) and downwind (i.e., wind blows away from the antenna) configurations is inconsistent from study to study. Reliable information about the wave field indeed helps to describe the spread in the DCA, especially at low and moderate wind speeds, and when the ocean surface is dominated by the remotely generated swell.

The CDOP model is used as a baseline for estimating the wind-wave-induced Doppler shift. Retraining of the CDOP model for the Sentinel-1 SAR observations (CDOP-S) yielded a significantly better fit. Then, we extended the GMF with parameters of the wavefield (significant wave height, mean wave period and direction) in the moment of SAR acquisition. Combining information about near-surface wind and ocean surface wave fields also considerably improves the accuracy of the wave-induced Doppler shift estimates. In turn,  the accuracy of the ocean surface current retrievals are improved as demonstrated by the promising agreement with the near-surface ocean surface current climatology based on multiyear drifter observations.