On Removal of Sea State Contribution to Sentinel‐1 Doppler Shift for Retrieving Reliable Ocean Surface Current. Issue 9 (10th September 2020)

• Record Type: Journal Article
• Authors: Moiseev, A.; Johnsen, H.; Johannessen, J. A.; Collard, F.; Guitton, G.
• Is Part Of: Journal of geophysical research. Volume 125:Issue 9(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-09-10
• Subjects: Sentinel‐1 -- Doppler shift -- ocean surface currents -- sea state -- neural networks; Oceanography -- Periodicals; 551.4605
• DOI: 10.1029/2020JC016288
• Languages: English
• ISSNs: 2169-9275

Abstract

Abstract: The Doppler frequency shift acquired by Synthetic Aperture Radar (SAR) contains information about ocean surface motion induced by waves and the underlying ocean currents in the radar range direction. An accurate estimate of the wave‐induced contribution is therefore required to derive a reliable estimation of the ocean surface current. In this study, we developed an empirical model for estimating the wave‐induced Doppler shift based on Sentinel‐1B Wave Mode (WV) Level 2 Ocean products acquired from December 2017 to January 2018 collocated with wind field from ECMWF and wavefield from WAVEWATCH III. We found that the relationship between the wind field at 10 m height and the Doppler shift from Sentinel‐1 is in agreement with previous findings based on ASAR observations. Retraining of the conventional CDOP model for the Sentinel‐1 observations (CDOP‐S) yields distinct improvements. We speculate that the improvement is due to different sensor properties and hence biases in the data. Moreover, combing wave and wind information into the model yield considerable improvements especially for the Southern Ocean and the North Pacific. Given accurate wave bias correction, the ocean surface radial velocity maps based on 2 months of Sentinel‐1 acquisitions agree with ocean surface current climatology derived from multiyear drifter observations. This suggests that Sentinel‐1 Doppler shift observations can be used to study ocean surface currents with 20 km spatial resolutions at a monthly time scale. Plain Language Summary: The Doppler shift registered by Synthetic Aperture Radar (SAR) contains information about ocean surface motion associated with ocean waves and underlying ocean currents. An accurate estimate of the wave‐induced contribution is crucial for retrieving reliable estimation of the surface current. In this study, we use observations from the Sentinel‐1 SAR satellite collocated with wind and wave information from numerical models in order to develop an empirical model for estimating wave‐induced Doppler shift. Based on the collocated data set we found that the relationship between the wind field and observed Doppler shift is in agreement with the literature. Difference between range wind speed and range wave orbital velocity can partly explain observed spread in the relationship between wind and observed Doppler shift. (Johannessen et al., 2008, https://doi.org/10.1029/2008GL035709 ) Combining wind and wave information into the empirical model yields a considerable improvement of the wave contribution estimates (compare with conventional models based only on wind information). The SAR derived ocean surface current (using the developed model) is in agreement with current derived from the conventional ocean surface drifter observations. Therefore, given correct wave bias estimates, Sentinel‐1 observations can be used to systematically study ocean surface currents at a monthly time scale. Key Points: Combining near‐surface wind with sea state information into an empirical Doppler model improves accuracy of the wave‐induced bias estimates The CDOP empirical geophysical model function (based on the Envisat ASAR observations) should not be reused for the Sentinel‐1 observations The Doppler shift observations from Sentinel‐1 can be used to study ocean surface currents globally at a monthly timescale