 This study investigated the saturation levels of landsat imagery for different vegetation types in a subtropical region of eastern China. It found that pine forest and mixed forest had the highest saturation levels, 159 and 152 mg per hectare, respectively, while Chinese fir and broadly forest had lower saturation levels, 143 and 123 mg per hectare, respectively. Additionally, it determined that stratifying the data according to vegetation types and slope aspects provided more accurate estimates than non-stratified data. Finally, it developed AGB estimation models based on stratification of both vegetation types and slope aspects, which provided the most accurate estimation with the smallest root mean squared error, RMSE, of 24.5 mg per hectare. This article was authored by PanPanzhao, Dengshan Lu, Guangxing Wang and others.