Deriving PM2.5 from satellite observations with spatiotemporally weighted tree-based algorithms: enhancing modeling accuracy and interpretability.

Tree-based machine learning algorithms, such as random forest, have emerged as effective tools for estimating fine particulate matter (PM_2.5) from satellite observations. However, they typically have unchanged model structures and configurations over time and space, and thus may not fully capture the spatiotemporal variations in the relationship between PM_2.5 and predictors, resulting in limited accuracy. Here, we propose geographically and temporally weighted tree-based models (GTW-Tree) for remote sensing of surface PM_2.5. Unlike traditional tree-based models, GTW-Tree models vary by time and space to simulate the variability in PM_2.5 estimation, and they can output variable importance for every location for the deeper understanding of PM_2.5 determinants. Experiments in China demonstrate that GTW-Tree models significantly outperform the conventional tree-based models with predictive error reduced by >21%. The GTW-Tree-derived time-location-specific variable importance reveals spatiotemporally varying impacts of predictors on PM_2.5. Aerosol optical depth (AOD) contributes largely to PM_2.5 estimation, particularly in central China. The proposed models are valuable for spatiotemporal modeling and interpretation of PM_2.5 and other various fields of environmental remote sensing. [ABSTRACT FROM AUTHOR]