Abstract
Different functions of various POIs result in different concentrations of PM2.5 at their locations and a different number of people gathering near them. This study divided the Baoding into 500 m × 500 m grids and studied several representative moments. The frequency of 16 kinds of POIs in each grid was calculated, and the kriging interpolation model was used to simulate PM2.5 concentration. The average pollutant concentration at each kind of POI was calculated. With mobile signal data, we count the number of people in each grid at each moment in real time. Then we used stepwise and principal component regression (PCR) to identify the relationship between the population distribution and various kinds of POIs. We found that the types of POIs with the highest and lowest pollutant PM2.5 concentrations in the study area did not change much most of the time. The relationship between each kind of POI and population distribution is also varied. Combining these two POI characteristics, we analyzed the causes and significance of these features, considering the functions of the various kinds of POIs and the main groups of people attracted. This study also provides suggestions on environmental exposure and urban planning according to the results obtained.