Spatiotemporal variations and the driving factors of PM2.5 in Xi’an, China between 2004 and 2018

High-intensity human socioeconomic activities in Xi’an have caused fine particulate matter (PM 2.5 ) pollution. Understanding the spatial and temporal patterns and key factors influencing PM 2.5 concentration was the basic step for taking targeted measures. Thus, spatial analysis techniques are used to reveal the temporal and spatial distribution characteristics of PM 2.5 in Xi’an over a long time series; wavelet analysis and Geo-detector models are applied to assess the strength of the association between meteorological and socio-economic conditions on PM 2.5 concentrations. The results illustrated that the average PM 2.5 concentration was 40.13 μg/m 3 in 2004 and peaked at 62.06 μg/m 3 in 2011, before failing to 38.77 μg/m 3 by 2018. The PM 2.5 concentration distribution had a characteristic of high in winter and autumn but low in spring and summer, presenting a U-shaped profile. The main distribution of PM 2.5 concentrations was oriented in a northeast-southwest direction, with obvious spatial autocorrelation and spatial aggregation characteristics. The resonance cycles of the meteorological and socioeconomic elements and PM 2.5 concentrations were synchronous and divergent at different scales. U-wind was the influencing factor on PM 2.5 concentration with a positive correlation coefficient of 0.9. Before 2011, the interaction of temperature (Tem) and relative humidity (RH) had the greatest impact on PM 2.5 concentrations. Additionally, the land use and cover change (LUCC) coupled with other factors had a large influence on PM 2.5 concentrations. These relationships can shed new light on the underlying mechanisms of PM 2.5 contamination at the city level, assisting relevant departments in developing effective PM 2.5 pollution management strategies.