Long-term Land Cover Dataset of the Mongolian Plateau Based on Multi-source Data and Rich Sample Annotations

The Mongolian Plateau (MP), with its unique geographical landscape and nomadic cultural features, is vital to regional ecological security and sustainable development in North Asia. Existing global land cover products often lack the classification specificity and temporal continuity required for MP-specific studies, particularly for grassland and bare area subtypes. To address this gap, a new land cover classification was designed for MP, which includes 14 categories: forests, shrubs, meadows, real steppes, dry steppes, desert steppes, wetlands, water, croplands, built-up land, barren land, desert, sand, and ice. Using machine learning and cloud computing, the novel dataset spanning the period of 1990-2020. Random Forest algorithm was employed to integrate training samples with multisource features for landcover classification, and a two-step Random Forest classification strategy to improve detail land cover results in transition regions. This process involved accurately annotating 64,345 sample points within a gridded framework. The resulting dataset achieved an overall accuracy of 83.6%. This land cover product and its approach has potential for application in vast arid and semi-arid areas.