Distribution patterns and influencing factors of population exposure risk to particulate matters based on cell phone signaling data

Chuyi Zhang,Yuanman Hu,Matthew D. Adams,Rencang Bu,Zaiping Xiong,Miao Liu,Yan Du,Binglun Li,Chunlin Li
DOI: https://doi.org/10.1016/j.scs.2022.104346
IF: 11.7
2022-12-10
Sustainable Cities and Society
Abstract:In this study, spatial-temporal characteristics of particular matter (PM) exposure risk in Shenyang were analyzed with landscape patterns using data from land use, cell phone signaling, and PM mobile monitoring. Pollution surfaces were established with geographically weighted regression models and impact factors analysis was implemented by boosted regression tree models. The results showed that weekdays and weekends had different spatial distributions of PM, and the exposure risk was lower on weekends. High exposure risks of PM 10 were concentrated in the first ring zone (76.53 people·m −2 ·μg·m −3 ) and residential-commercial land (292.34 people·m −2 ·μg·m −3 ). Exposure risks of PM 2.5 were most affected by residential-commercial land and fourth-class (relative contribution: 59.69 and 8.88, respectively). However, the exposure risks of PM 10 were more influenced by first-class roads (relative contribution: 2.01). The results indicated that independent modeling analysis of different types of PM and periods contribute to more detailed studies of spatial-temporal variation of PM. For human activity studies, cell phone signaling data can effectively distinguish spatial-temporal distribution characteristics of the population on weekdays and weekends. Multi-source big data combined with mobile monitoring and model simulations were used to make population exposure risk studies more accessible, real-time, and cost-effective for sustainable urban planning and development.
energy & fuels,construction & building technology,green & sustainable science & technology
What problem does this paper attempt to address?