A dataset of remote sensing classification for Beidagang wetlands in Tianjin from 2019 to 2022

The author used the GEE platform and combined Sentinel-1 radar images and Sentinel-2 optical images to extract information from the Beida Port wetland in Tianjin using the random forest classification algorithm. The importance of multi-source remote sensing data and time series features in the classification of the Beida Port wetland in Tianjin was explored, and the 2019-2022 Tianjin Beida Port wetland remote sensing classification dataset was finally obtained. After verification with actual test data, the overall accuracy of the classification results is 95.35%, which meets the accuracy requirements. This dataset can provide scientific basis for the conservation of species diversity, monitoring of spatiotemporal dynamic changes in wetlands, wetland management, and government decision-making in Beidagang Wetland.