积雪判识, 遥感, 葵花8号卫星, NDSI直接比较法

Inner Mongolia is one of significant seasonal snow-covered regions in China. Snow identification and snow depth inversion

are crucial for agricultural production, ecological assessment, and research on spring floods and snow-related disasters. In order to improve the accuracy of local snow identification, a direct comparison snow identification method based on Normalized Difference Snow Index (NDSI) is proposed in this paper, the method involves applying the NDSI difference operation between the snow map to be identified from the Himawari-8 satellite images and the snow-free base map from the current autumn to identify snow, and it is compared with the snow identification methods used in routine business. The results indicate that the Snow Mapping (SNOMAP) algorithm, based on the Normalized Difference Vegetation Index (NDVI), tends to miss some thin snow pixels, while the Fractional Snow Cover (FSC) algorithm can be affected by water bodies in snow identification and ultimately affect its accuracy. In the non-forest areas of Inner Mongolia, the accuracy of NDSI direct comparison was 3.88% higher than SNOMAP and 0.52% higher than FSC. The difference between the accuracy of NDSI direct comparison and FSC in non-forest areas was small. In forest areas, compared with FSC algorithm, NDSI direct comparison method significantly improved the identification accuracy, while the error rate decreased, indicating that NDSI direct comparison method is more suitable for snow identification in forest areas of Inner Mongolia.

snow identification, remote sensing, Himawari-8 satellite, NDSI direct comparison snow identification method