Local modeling of $\text{PM}_{2.5}$ concentration in Beijing-Tianjin-Hebei area: A multi-scale geographically weighted regression analysis

There are various environmental and economic factors influencing the concentration of $\text{PM}_{2.5}$ , which is the focus of many fields. Although many studies have proven that the influential factors of $\text{PM}_{2.5}$ exist spatial heterogeneity, few have focused on the effects of multi-scale in spatial relationships. This study applies a multi-scale geographically weighted model (MGWR) to explore the spatial relationships between the influential factors and $\text{PM}_{2.5}$ at different operational scale in Beijing-Tianjin-Hebei area, China. The results show that variables such as NDVI, aspect, pressure, percentage of green space, urban function mixing degree, and relative humidity are local-scale variables affecting $\text{PM}_{2.5}$ . Those variables have significant spatially heterogeneous effects on $\text{PM}_{2.5}$ . The variables of slope, temperature, wind speed, population density, road density, and percentage of construction are global scale variables. Among the global scale variables, slope, wind speed, and population density has negative effect on $\text{PM}_{2.5}$ . road density, percentage of construction, temperature and has positive effect on $\text{PM}_{2.5}$ . Our work can deeply understanding the influential mechanisms of $\text{PM}_{2.5}$ , which can help to develop the relevant air pollution control recommendations and policies.