Temporal and spatial distribution of PM2.5 in the Xinjiang region based on AOD fusion data and the GTWR model

Abstract

ABSTRACT By integrating MODIS aerosol optical depth data and combining it with ground-based PM2.5 measurements and meteorological data, a geographically and temporally weighted regression (GTWR) model was used to estimate the spatiotemporal distribution of PM2.5 in Xinjiang in 2022. Findings include the following: (1) The GTWR model effectively represents PM2.5 distributions, with a correlation coefficient of 0.92, surpassing the geographically weighted regression model by 5.7% and the land use regression model by 9.5%. (2) Monthly PM2.5 levels in Xinjiang exhibit a pattern of initial decline followed by an increase. Elevated coal-fired heating in February and high air humidity in July contribute to higher atmospheric particulate concentrations. The peak monthly average concentration reaches 79.05 µg m−3 in February, and the lowest concentration drops to 33.86 µg m−3 in July. (3) Influenced by dust-prone conditions and atmospheric diffusion in the Taklimakan Desert region, PM2.5 levels in Xinjiang show a notable southwest-to-northeast gradient.