Abstract: Based on the global high-precision PM2.5 satellite inversion dataset from 2001 to 2020, in conjunction with spatial statistical analysis, geographic detector, and the geographically and temporally weighted regression(GTWR), this study examined the extent of influence and spatiotemporal heterogeneity of human socioeconomic and meteorological factors on PM2.5 in the Yangtze River Delta region. The findings were as follows:(1) The GTWR considered the temporal scale and demonstrated superior fitting performance compared to traditional global regression (OLS) and geographically weighted regression (GWR) models. (2)The geographical detector's q-value provided an accurate and quantitative explanation of the influence of each driving factor on PM2.5. The q-values of different factors exhibited fluctuating patterns across various time periods. Notably, the impact of population, road freight volume, electricity consumption, and temperature at the year-end maintained  a consistent upward trend.(3)The GTWR fitting results indicated that various factors had both positive and negative effects on PM2.5. Specifically, year-end population and temperature had significant positive effects, while road freight volume, electricity consumption, and relative humidity at the year-end had significant negative effects. Other factors demonstrated varying degrees of fluctuation. Moreover, the regional variation in regression coefficients suggested distinct spatial and temporal heterogeneity among the driving factors.