Wind vector retrieval from Gaofen-3 SAR imagery using the method of iterative line fitting in image power spectrum domain

The Gaofen-3 (GF-3) is a Chinese Synthetic Aperture Radar (SAR) satellite primarily designed for monitoring marine dynamic environments. Compared to spaceborne scatterometers, it can monitor sea surface wind fields with higher spatial resolution. Currently, the spectrum analysis-based algorithms are widely used to extract wind direction from SAR images, but they are susceptible to image noise. To address this issue, this study proposes a wind direction inversion method based on iterative linear fitting, which fully utilises wind streak information at different wavenumbers and improves wind direction inversion accuracy by iteratively removing data points with high noise levels. Twelve GF-3 SAR images were used to validate the proposed algorithm. Experimental results indicate that the root mean square error (RMSE) of the retrieved wind direction is 17.10°. Compared to the European Centre for Medium-Range Weather Forecast (ECMWF) reanalysis wind, and the mean absolute deviation is 10.93°Compared to the National Data Buoy Center (NDBC) buoy wind. Compared with the conventional local gradient algorithm, the proposed algorithm achieves higher wind direction inversion accuracy. This study provides a technical support for the operational wind field inversion using Gaofen-3 SAR images.