Sea Ice Thickness Inversion Based on Support Vector Regression for FengYun-3E IDW Data

Due to the increasing demand for monitoring sea ice thickness (SIT) in the global environment, the SIT inversion based on global navigation satellite system-reflectometry (GNSS-R) has garnered significant attention. However, there exists a gap in utilizing the GNSS-R integral delay waveform (IDW) data for SIT inversion. In response, this letter proposes the support vector regression (SVR) method to invert SIT from the combined GNSS-R IDW datasets. The proposed method captures the nonlinear relationship between IDW and SIT, enabling adaptive optimization of the SVR-based SIT inversion model to maximize the SIT inversion accuracy. Experimental results demonstrate that the SVR method successfully inverts SIT from the combined GNSS-R IDW datasets. The validation confirms that the SVR method significantly enhances the accuracy and effectiveness of SIT inversion with the combined GNSS-R IDW-related datasets. Moreover, it consistently outperforms using IDW feature-related datasets in various aspects, including accuracy and robustness.