Satellite Observation of Sea Ice Concentration Based on Chinese HY-1C/D Data

The increasingly significant global warming has severely affected the Arctic region. The melting speed of sea ice is accelerating, and its spatial coverage is gradually shrinking, causing a series of ecological problems that cannot be ignored. Therefore, conducting in-depth research on the sea ice conditions in the Arctic region is crucial. This study developed a Sea Ice Concentration Inversion model suitable for the Arctic using remote sensing data from HY-1C/D Chinese Ocean Color and Temperature Scanner (COCTS). The model's accuracy was tested using a confusion matrix and measured data. The results indicated that the model achieved an accuracy rate of 89% in sea ice extraction. Its mean absolute percentage error is lower and consistent with the other similar products. The model was applied to COCTS image data from Apr. to Sep. 2020 and 2021, revealing the monthly distribution characteristics of sea ice concentration in the Arctic region. The results indicated that sea ice concentration in the central Arctic region consistently remains high, exceeding 90% throughout the year. However, further analysis of various subregions revealed a declining trend in sea ice concentration in the Chukchi Sea, Baffin Bay, and the Laptev Sea. In contrast, the fluctuation of sea ice concentration in Greenland, Barents, and Kara was not noticeable. The findings show that the sea ice extraction model developed effectively assesses sea ice conditions in the Arctic region. It successfully demonstrates the spatial and temporal distribution of Arctic sea ice concentration and provides valuable support for ongoing research and development in sea ice inversion.