SEASONAL VARIATION IN PM2.5 AND PM10 CONCENTRATIONS AFFECTED BY DIFFERENT LAND USE COVERS AND PRECIPITATION IN GUANGZHOU, SOUTH CHINA

Abstract

While the negative health impacts of long-term exposure to fine particulate matter (PM) have been well documented, effectively reducing PM pollution through urban land-use planning remains a complex challenge. Using continuous auto-monitoring air quality data from Guangzhou in south China, average PM2.5 and PM10 concentrations in the dry season were clearly higher than those in the wet season. During the wet season, PM2.5 and PM10 concentrations exhibited a nonlinear increase with distance from the coastline in Guangzhou. Annually, average PM2.5 and PM10 concentrations showed a negative correlation with green space cover and NDVI (Normalized Difference Vegetation Index), but a positive correlation with the extent of construction land. Notably, these correlations between average PM concentrations and NDVI/construction land cover were stronger in the dry season than in the wet season. Green space cover in the wet season did not show a significant relationship with average PM concentrations, suggesting that different urban land use/cover types exert a greater influence on PM pollution during the dry season. Furthermore, seasonal and annual average PM concentrations exhibited stronger correlations with the Euclidean distance and Kernel density of vegetable markets than with those of industrial zones. This indicates that local pollution sources, such as vehicle emissions and cooking activities, contribute substantially to elevated PM levels. The proximity of residential areas and food service establishments to vegetable markets suggests that that both traffic-related emissions and cooking fumes from daily life influence the temporal and spatial distribution of particulate pollutants.