Remote sensing estimation of regional PM_2.5 based on GTWR model -A case study of southwest China

Air pollution in China has becoming increasingly serious in recent years with frequent incidents of smog. Parts of southwest China still experience high incidents of smog, with PM_2.5 (particulate matter with diameter ≤2.5 μm) being the main contributor. Establishing the spatial distribution of PM_2.5 in Southwest China is important for safeguarding regional human health, environmental quality, and economic development. This study used remote sensing (RS) and geographical information system (GIS) technologies and aerosol optical depth (AOD), digital elevation model (DEM), normalized difference vegetation index (NDVI), population density, and meteorological data from January to December 2018 for southwest China. PM_2.5 concentrations were estimated using ordinary least squares regression (OLS), geographic weighted regression (GWR) and geographically and temporally weighted regression (GTWR). The results showed that: (1) Eight influencing factors showed different correlations to PM_2.5 concentrations. However, the R² values of the correlations all exceeded 0.3, indicating a moderate degree of correlation or more; (2) The correlation R² values between the measured and remote sensed estimated PM_2.5 data by OLS, GWR, and GTWR were 0.554, 0.713, and 0.801, respectively; (3) In general, the spatial distribution of PM_2.5 in southwest of China decreases from the Northeast to Northwest, with moderate concentrations in the Southeast and Southwest; (4) The seasonal average PM_2.5 concentration is high in winter, low in summer, and moderate in spring and autumn, whereas the monthly average shows a "V" -shaped oscillation change.