Gulf Stream Detection from SAR Doppler Anomaly

Abstract

This work presents a Gulf Stream (GS) North Wall (GSNW) detection algorithm applicable to Sentinel-1 Radial surface Velocity (RVL) products derived from Doppler centroid analysis of synthetic aperture radar (SAR) data collected off the east coast of Canada from February 2017 to August 2017. Visual comparison of the extracted location of the GSNW (obtained by evaluating the peak RVL gradient in the azimuth direction), and the estimated location of the GSNW as determined by the U.S. Naval Oceanographic Office and a GSNW search region (GSNWSR), indicates that the algorithm is capable of detecting the GSNW in ∼80% of the cases evaluated. Results are dependent on the geophysical orientation of the GS across the SAR swath, such that the GSNW detector performed most reliably when the GS was oriented across the swath resulting in maximized surface velocities in the range direction. When the GS meandered, or when GS eddies appeared in the RVL data, the GSNW was often partially detected or detected multiple times, reducing confidence in the extracted location. It is anticipated that the results obtained using the Sentinel-1 RVL products will be applicable to SAR data from other platforms.