| Summary: | Given the increasingly fragmented forest landscapes, it is necessary to map forest cover with fine spatial resolution in a large area. The European Space Agency (ESA) released the 10 m global land cover map in 2020 based on Sentinel-1 and Sentinel-2 images, and Dynamic World provides near real-time possibilities of many land cover classes based on Sentinel-2 images, but they are not designed particularly for forest cover. In this research, we aimed to develop a method to automatically estimate an accurate 10 m forest cover map in 2020 by fusing the ESA forest cover map and Dynamic World near real-time forest cover possibilities. The proposed method includes three main steps: (1) generating stable forest samples, (2) determining the threshold <i>T</i> and (3) producing the fused forest cover map. China’s Han River Basin, dominated by complex subtropical forests, was used as the study site to validate the performance of the proposed method. The results show that the proposed method could produce a forest cover map with the best overall accuracy of 98.02% ± 1.20% and more accurate spatial details compared to using only one of the two data sources. The proposed method is thus superior in mapping forest cover in complex forest landscapes.
|
|---|