AHI/ATMS Fusion Dataset and Sea Surface Temperature Retrieval Results

With the rapid development of remote sensing satellites, sea surface temperature remote sensing product data with multi-scale and wide coverage features is gradually replacing traditional sea surface temperature (SST) collection methods. Thermal infrared satellite sensors have shown great potential in scientific applications due to their ability to invert SST data with high frequency and wide coverage. However, cloud coverage often leads to abnormal estimation of SST under the cloud, and some cloud coverage intervals even have missing SST data. To overcome the aforementioned shortcomings, this paper uses in-situ sea surface temperature data as a benchmark and employs spatiotemporal matching method to fuse AHI (Advanced Himawari Imager) near-infrared images with ATMS (Advanced Technology Microwave Sounder) microwave radiometer band data, establishing a matching dataset. Three machine learning methods are used to evaluate the SST inversion accuracy of the fused data. Provide important technical reference for SST inversion under all-weather conditions. (Upload data and images for readers to reproduce)