Spatiotemporal variations and mechanism of PM2.5 pollution in urban area: The case of Guiyang, Guizhou, China

PM2.5 has been a hot concern in the recent decade. Many studies have focused on metropolises or those areas with poor air quality, but the PM2.5 of more widespread areas is less considered. Considering the challenges of rapid economic growth and environmental problems against a developing region, we took Guiyang as a study case to assess the spatiotemporal variations and mechanism of PM2.5 pollution in an urban area from 2000 to 2020 in an extended sense. Based on PM2.5 concentration data from 14 monitoring points in Guiyang, spatiotemporal variations and formation mechanism were assessed using wavelet, moving maximal information coefficients, and spatial correlation analysis. The urban Nighttime light data was selected to evaluate the impacts of socioeconomic factors on PM2.5 concentration using spatial correlation analysis. Further, wavelet and statistical analysis were adopted to analyze multi-dimensional temporal variations of PM2.5 hourly concentration and the relationship with pressure, temperature, vapor pressure, relative humidity, wind, and visibility. The PM2.5 hourly concentration was obtained from the monitoring points in downtown Guiyang according to data continuity and availability. PM2.5 had different temporal variations at daily, monthly, seasonal, and annual levels, and interannual variation was the most obvious. The temperature was the main factor leading to the interannual temporal variation of PM2.5. Wind and pressure were more significant for the responses of a shorter period variation with -0.76 and -0.80 of the minimum of correlation coefficient, respectively. Meanwhile, human activities significantly influenced spatiotemporal variations of PM2.5. A spatial correlation analysis between PM2.5 and the related influencing factors from 2000 to 2018 was implemented based on a geographic information system. Besides, the landcovers within a buffer zone with a radius of 1 km on 14 monitoring points were visually interpreted to analyze the relationship between PM2.5 and landcovers. Moreover, multivariate wavelet coherence analysis revealed the PM2.5 interaction among monitoring points. The PM2.5 concentration in Guiyang dropped from 49 μg/m3 in 2012 to about 27 μg/m3 in 2018, and the air quality greatly improved. As in most cities, Guiyang has a significant PM2.5 pollution island effect, with traffic and building land density contributing to higher PM2.5 concentrations. There were some typical nonlinear spatiotemporal variations between PM2.5 and its influencing factors, and these variations varied with the selected scale.