Natural and anthropogenic contributions to desertification in Central Asia

Abstract

Desertification is one of the most critical global environmental challenges, with significant ecological and socio-economic impacts. Understanding its dynamics and underlying drivers is crucial for developing effective prevention and control strategies. This study provides a comprehensive spatiotemporal analysis of desertification dynamics in the drylands of Central Asia (CA) from 1982 to 2020, employing the Desertification Difference Index (DDI). A non-linear Granger causality framework was utilized to quantify the relative contributions of natural and anthropogenic drivers to desertification across various land cover types. The trend analysis revealed that approximately 14.81% of the drylands in Central Asia experienced desertification during the study period. Pixel-wise fitted random forest Granger causality models identified Snow Water Equivalent (SWE) as a dominant factor influencing dryland changes in the region, particularly in cropland, grassland, shrubland, and bare land. In contrast, temperature emerged as the dominant driver of forest degradation. Overall, 69.1% of desertification in CA was attributed to natural drivers, while anthropogenic drivers accounted for 30.9%. These findings offer pivotal insights into the processes driving desertification in Central Asia's drylands, highlighting the need for region-specific land management and ecological conservation strategies.