30-Meter Resolution Dataset of Abandoned and Reclaimed Croplands in Inner Mongolia, China (2000-2022)

At the end of the last century, the expansion of agricultural land in northern China's arid and semi-arid regions intensified the conflict between agricultural development and ecological protection. Accurately mapping the distribution of abandoned cropland is crucial for balancing these competing interests. This dataset provides high-resolution (30-meter) mapping of abandoned and reclaimed croplands across Inner Mongolia, China, covering the period from 2000 to 2022. Using an innovative temporal segmentation approach developed based on Google Earth Engine, the dataset integrates ground sample collection of major crops and inactive cropland with an analysis of Normalized Difference Vegetation Index (NDVI) characteristics during key growth stages. This method enables precise classification of cultivation status and adopts a binary classification strategy with adaptive optimization, improving the efficiency of sample generation for the Random Forest algorithm. The resulting cropland status maps for Inner Mongolia exhibit an annual accuracy of 97% to 99%. By applying the Temporal Segmentation of Abandoned and Reclaimed Cropland (TSARC) method, time series maps of abandoned and reclaimed cropland were created with an overall accuracy of 87.61%. This dataset provides crucial information for understanding the spatiotemporal trends in abandonment rates from arid to semi-humid regions, supporting agricultural and environmental management in Inner Mongolia.