Evaluating the Accuracy and Spatial Agreement of Five Global Land Cover Datasets in the Ecologically Vulnerable South China Karst

Global land cover datasets show significant discrepancies in accuracy and spatial agreement, particularly in ecologically vulnerable areas like the South China Karst. A study assessing five widely used datasets (CCI-LC, MCD12Q1, GlobeLand30, GlobCover, and CGLS-LC) found that while CGLS-LC was the most accurate overall, MCD12Q1 performed best in high-elevation or steep-slope regions. Accuracy declined with increasing landscape heterogeneity and was notably lower in karst areas compared to non-karst areas. These findings highlight the challenges in land cover mapping for complex terrains and provide guidance for future dataset fusion and regional sustainable development efforts.

Liu, P., Pei, J., Guo, H., Tian, H., Fang, H., & Wang, L. (2022). Evaluating the Accuracy and Spatial Agreement of Five Global Land Cover Datasets in the Ecologically Vulnerable South China Karst. Remote Sensing, 14(13), 3090. https://doi.org/10.3390/rs14133090