Fine simulation of PM2.5 combined with NPP-VIIRS night light remote sensing and mobile monitoring data

Human activity plays a crucial role in influencing PM2.5 concentration and can be assessed through nighttime light remote sensing. Therefore, it is important to investigate whether the nighttime light brightness can enhance the accuracy of PM2.5 simulation in different stages. Utilizing PM2.5 mobile monitoring data, this study introduces nighttime lighting brightness as an additional factor in the PM2.5 simulation model across various time periods. It compares the differences in simulation accuracy, explores the impact of nocturnal human activities on PM2.5 concentrations at different periods of the following day, and analyzes the spatial and temporal pollution pattern of PM2.5 in urban functional areas. The results show that (1) the incorporation of nighttime lighting brightness effectively enhances the model's accuracy (R2), with an improvement ranging from 0.04 to 0.12 for different periods ranges. (2) the model's accuracy improves more prominently during 8:00–12:00 on the following day, and less so during 12:00–18:00, as the PM2.5 from human activities during the night experiences a strong aggregation effect in the morning of the next day, with the effect on PM2.5 concentration declining after diffusion until the afternoon. (3) PM2.5 is primarily concentrated in urban functional areas including construction sites, roads, and industrial areas during each period. But in the period of 8:00–12:00, there is a significant level of PM2.5 pollution observed in commercial and residential areas, due to the human activities that occurred the previous night. [Display omitted] •Introduce nighttime lighting brightness as an independent variable in the PM2.5 model for more accurate simulation.•To explore the influence of nocturnal human activities on PM2.5 in different periods of the next day.•Analyze the fine-scale spatial and temporal pattern of PM2.5 pollution in the city.•To explore the degree of PM2.5 pollution in different Urban functional areas.