Long-Term Trends in Dust Column Mass Density Over Central Asia: A Nighttime Satellite Sensor Analysis

Central Asia faces a significant air quality crisis due to rising PM2.5 levels from natural and anthropogenic sources. This study utilizes advanced remote sensing, including DMSP/OLS and VIIRS nighttime light data and the Compounded Night Light Index (CNLI), to analyze PM2.5 trends from 1990 to 2025, projecting to 2050. Machine learning models (Random Forest, LSTM) and statistical techniques reveal a notable annual PM2.5 increase (1.67×10⁻⁷ kg/m³), particularly in arid regions like the Karakum Desert and the Aral Sea. The CNLI and nighttime light data indicate post-Soviet economic shifts, while LSTM predicts an 8.5% rise in PM2.5 by 2050. Hotspots are identified in western Kazakhstan and Uzbekistan, driven by desertification and industrial emissions. The findings underscore the need for targeted policies, such as sustainable land management and energy-efficient urbanization, to address air quality degradation and align with global sustainability goals.