Preliminary analysis of wave retrieval from Chinese Gaofen-3 SAR imagery in the Arctic Ocean

Abstract

Synthetic aperture radar (SAR) is a powerful tool for sea surface monitoring. The 2017–2020 mission of the Chinese Gaofen-3 (GF-3) over the Arctic Ocean provided a unique opportunity to investigate the feasibility of wave retrieval in the Arctic Ocean. In this study, several GF-3 SAR images acquired in quad-polarization (QPS) mode were applied for wave retrieval using two theoretical-based algorithms, i.e. the parameterized first-guess spectrum method (PFSM) (with improved tilt modulation) and quad-polarization (Q-P). These images were collocated with wave fields simulated using the third-generation numeric wave model WAVEWATCH-III (WW3) and measurements from the Haiyang-2B (HY-2B) altimeter. A comparison of the HY-2B winds with the wind speeds generated by the geophysical model function (GMF) CMOD5.N showed a 3.36 m/s root mean square error (RMSE) with a correlation (COR) of 0.66. The retrieved significant wave height (SWH) validated against the WW3-simulated results had a 0.50 m RMSE using PFSM and 0.91 m RMSE with the Q-P algorithm. These results show that PFSM is suitable for wave retrieval from GF-3 SAR images. We also used PFSM to investigate the characteristics of the wave spectrum in the presence of sea ice. Although the analysis concludes that GF-3 SAR has the capability for wave monitoring in Arctic Ocean due to the high spatial resolution of SAR-derived wave spectra, an optimal wave retrieval algorithm needs to be developed for improving the retrieval accuracy.