CropLayer: 2-meter resolution cropland mapping dataset for China in 2020

Existing cropland datasets in China show significant discrepancies. We created a high-resolution cropland map of China for 2020, using imagery from Mapbox and Google. By combining image quality assessments, active learning for semantic segmentation, and results integration. The accuracy achieved to 88.73%, with 30 out of 32 provincial units reporting area estimates within ±10% of official statistics. In contrast, only 9 to 1 provinces from 7 existing datasets meet the same accuracy standard. 

This dataset is formally linked to an article under review at Earth System Science Data (ESSD):How to cite. Jiang, H., Ku, M., Zhou, X., Zheng, Q., Liu, Y., Xu, J., Li, D., Wang, C., Wei, J., Zhang, J., Chen, S., and Huang, J.: CropLayer: A high-accuracy 2-meter resolution cropland mapping dataset for China in 2020 derived from Mapbox and Google satellite imagery using data-driven approaches, Earth Syst. Sci. Data Discuss. [preprint], https://doi.org/10.5194/essd-2025-44, in review, 2025.

Access the full paper and dataset via:

Preprint DOI: https://doi.org/10.5194/essd-2025-44

Link: https://essd.copernicus.org/preprints/essd-2025-44/

The dataset facilitates improved agricultural monitoring, policy evaluation, and interdisciplinary research in China’s rapidly evolving land-use context.