Intelligent Wind Retrieval from Chinese Gaofen-3 SAR Imagery in Quad-Polarization

Weizeng Shao, Shuai Zhu, Xiaopeng Zhang, Shuiping Gou, Changzhe Jiao, Xinzhe Yuan, Liangbo Zhao
2019 Journal of Atmospheric and Oceanic Technology  
This study proposes the use of the artificial neural network for wind retrieval with Chinese Gaofen-3 (GF-3) synthetic aperture radar (SAR) data. More than 10 000 images acquired in wave mode and quadpolarization strip map were collected over global seas throughout the 2-yr mission. The GF-3 operated in a quad-polarization channel-vertical-vertical (VV), vertical-horizontal (VH), horizontal-horizontal (HH), and horizontal-vertical (HV). These images were collocated with winds from the European
more » ... entre for Medium-Range Weather Forecasts at a 0.1258 grid. The newly released wind retrieval algorithm for copolarization (VV and HH) SAR included CMOD7 and C-SARMOD2. We developed an algorithm based on an artificial neural network method using the SAR-measured normalized radar cross section at quad-polarization channels, herein named QPWIND_GF. Simulations using the QPWIND_GF showed that the correlation coefficient of wind speed was 0.94. We then validated the retrieval wind speeds against the measurements at a 0.258 grid from the Advanced Scatterometer. A comparison showed that the root-mean-square error (RMSE) of wind speed was 0.74 m s 21 , which was better than the wind speed obtained using state-of-the-art methods-including, for example, CMOD7 (RMSE 0.88 m s 21 ) and C-SARMOD2 (RMSE 1.98 m s 21 ). The finding indicated that the accuracy of wind retrieval from GF-3 SAR images was significantly improved. Our work demonstrates the advanced feasibility of an artificial neural network method for SAR marine applications.
doi:10.1175/jtech-d-19-0048.1 fatcat:l5g7gmmlbrbbrnc4jzd3eytkou