Modeling the Spatiotemporal Distribution of Livestock Grazing Density in Kazakhstan Using Machine Learning

Abstract

This article is a translation and scientific adaptation of the original English-language study focused on modeling the density of grazing livestock in Kazakhstan at a high spatial resolution (1 km²) for the period 2000–2019. Using the Random Forest model and 13 socio-environmental factors, annual maps of horse and small ruminant (sheep and goats) densities were developed for the first time, adapted to actual national statistical data. High-density hotspots were identified in the southern and southeastern regions of the country. The resulting database is open for use in agricultural policy, environmental monitoring, and spatial planning.