Wave retrieval from quad-polarized Chinese Gaofen-3 SAR image using an improved tilt modulation transfer function

An accurate Modulation Transfer Function (MTF) is essential for Synthetic Aperture Radar (SAR) wave spectra retrieval. This study aimed to investigate the performance of a quad-polarized retrieval algorithm based on fully polarimetric SAR image data using improved tilt MTF and considering influence wind speed. The key factor in scheme from (Vertical–Vertical (VV), Horizontal–Horizontal (HH), Vertical–Horizontal (VH), Horizontal–Vertical (HV)) images. In this study, waves were inverted more than 1300 Gaofen-3 (GF-3) images acquired quad-polarization strip mode with spatial resolution 16 m swath coverage 50 km. winds retrieved Geophysical Model (GMF) C-band model (CSARMOD-GF), which suitable re-calibrated GF-3 VV-polarization. comparison speed yielded Root Mean Square Error (RMSE) 1.73 m/s Correlation Coefficient (COR) 0.94. validation Significant Wave Height (SWH) simulated Waves Nearshore (SWAN) against Haiyang-2B (HY-2B) altimeter an RMSE 0.56 COR 0.87. results reveal that SAR-derived SWAN-simulated SWH are analysis full technique was applied collected images, statistical 0.51 m, 0.75, Scatter Index (SI) 0.44 compared SWHs simulations SWAN model. non-polarized contribution Normalized Cross Section (NRCS; unit: dB) caused by breaking σ0wb calculated theoretical approach employs VV-polarized calibrated NRCS σ0vv HH-polarized NRCS. effect studied. bias (SAR-derived minus SWH) increased as (σ0wb/σ0vv) ratio (>0.4) increased, accuracy when less 0.4. behavior reasonable since inevitably affects modulation. Therefore, should be considered proposed paper under extreme sea states, such typhoons hurricanes.