Analysis of the impact of multiscale green landscape on urban PM2.5

PM2.5 is a serious threat to human health. Greenspace landscape patterns, which have a major impact on urban air quality, have recently attracted much attention, but most recent studies have focused only on — area configurations or structures, and there is a lack of research on the factors that influence PM2.5 concentrations and their spatial differentiation characteristics at seasonal and spatial scales. Based on the data of air quality monitoring stations in Beijing and remote sensing data, this paper selects indicators and data on five kinds of landscape variables and explores the response characteristics of PM2.5 to surrounding environmental factors from multiple perspectives through the least absolute shrinkage and selection operator (LASSO), multiple regression, Spearman grade correlation, and geographical detector techniques. The results show that the response of PM2.5 to landscape patterns in autumn is more obvious than that in other seasons. The PM2.5 concentration is positively correlated with relative humidity in four seasons and temperature in summer and negatively correlated with wind speed in spring. The contribution of green landscape composite factors to PM2.5 concentration is higher than that of single factors. The research results can provide references for Beijing’s urban planning and design.