gridded_livestock_mongolian_plateau_2000_2020

In this study, based on the Random Forest algorithm, we integrated multi-source remote sensing data such as land cover, population density, nighttime lighting, NDVI, climate and water to construct a spatialized framework with livestock statistics, and produced a gridded high-resolution (1km) dataset of livestock (total livestock, sheep & goats, and large livestock) density at five-year intervals in the Mongolian Plateau region from 2000 to 2020, which revealed that the spatial distribution of livestock density in the Mongolian Plateau region showed significant gradient characteristics and regional differentiation patterns. Based on the high-precision spatial distribution data of livestock density obtained in this study, it can be integrated and analyzed with other natural environment data, such as quantification of grass-animal balance, water consumption of livestock, methane emission and other environmental impact assessment, etc., which is of great value for ecological environmental protection, monitoring and early warning in the Mongolian Plateau.The repository has four folders containing sample code and data for total livestock, sheep & goats, large livestock, and 2020 total livestock density spatial distribution simulations. Each file is saved in GeoTiff format with a spatial resolution of 1 km and projected as WGS_1984 UTM_zone_48N. Each file is kept under the abbreviations "Total" (for total livestock), "SG" for (sheep & goats) and "LA" for large livestock, followed by an underscore and the year. Missing data is represented by "No data".