Abstract
The Surface Wave Investigation and Monitoring (SWIM) instrument is the first to adopt a rotation detection mode at multiple small incidence angles (0°-10°), which is mainly used for wave detection. SWIM is able to cover the north and south latitudes to 83°, and detect sea ice. However, the SWIM spatial resolution is relatively low (18 km), which limits its application in sea ice detection. Therefore, this study proposes, for the first time, a sea ice image reconstruction method at small incidence angles, which can improve the spatial resolution and extend the incidence angle range of image reconstruction methods. First, several image reconstruction methods are investigated and compared using SWIM data. Second, the optimal method (Scatterometer Image Reconstruction, SIR) with more sea ice detail and stronger noise suppression is selected. Third, the parameter settings of SIR are investigated and its reconstruction quality improves with increasing iteration number. Finally, based on the 6°-10° SWIM data in the Severnaya Zemlya and northeastern Kara Sea, the image reconstruction is carried out using the SIR method. The results are evaluated using the normalized standard deviation that can reach 0.55, and discussed based on sea ice charts and Sentinel-1 SAR images with good agreement.