Effects of Landscape Patterns on Atmospheric Particulate Matter Concentrations in Fujian Province, China

Abstract

Based on the observation data from 37 national ambient air automatic monitoring stations and the land use/cover data in Fujian Province, the effects of land use/cover on the annual and seasonal variations in the air quality index (AQI), PM2.5 concentration and PM10 concentration are analyzed. In addition, buffer zones with different radii are established to calculate landscape pattern metrics and the influences of landscape patterns on the AQI, PM2.5 concentration and PM10 concentration at different scales are discussed. The results show that land use/cover types have remarkable impacts on the variations in the atmospheric particulate matter concentrations. The AQI and PM10 concentration are the highest for construction land, and the PM2.5 concentration is the highest for cultivated land. The seasonal variations in all air pollutant concentrations show similar characteristics, i.e., high in spring and winter and low in summer and autumn. Different landscape metrics have diverse effects on atmospheric particulate matter concentrations at different scales. In the buffer zone with a 5000 m radius, the patch number and patch density of forest land are positively correlated with the PM2.5 concentration, while the edge density of construction land has a negative correlation with it, indicating that landscape fragmentation affects the PM2.5 concentration. More fragmented forest land has a weaker effect on the reduction in the PM2.5 concentration, and more fragmented construction land has a weaker effect on the increase in the PM2.5 concentration. Moreover, the seasonal variations in the atmospheric particulate matter concentrations are different under different land use/cover types. Except for autumn, the AQI and PM2.5 concentration are most noticeably affected by forest land in all seasons, showing negative correlations. In autumn, the impacts of cultivated land on the AQI and PM2.5 concentration are more pronounced. The PM10 concentration is substantially affected by forest land in spring and summer and is markedly influenced by construction land in autumn and winter. The analysis of the landscape metrics of forest land and construction land at different scales indicates that the optimal scale is 5000 m for studying the annual average of the AQI and PM10 concentration and is 3000 m for investigating the annual average of the PM2.5 concentration. The optimal scales to research the seasonal variations in the AQI, PM2.5 and PM10 concentrations are 4000–5000 m for forest land and construction land, while the optimal scale is 1000 m for cultivated land to research the AQI and PM2.5 in autumn. This study can provide a scientific basis for the optimization of land use/cover and landscape patterns in Fujian Province, the planning and management of green space and the selection of research scales in the future.