Spatial urban air quality prediction model utilizing high spatiotemporal resolution direct and remote sensing, socio-demographic and urban fabric data
(2021) AGU Fall Meeting 2021 (13.December.2021)
Files
No attached file found for this publication.
Details
- Authors
- Abstract
- Recent advancements in direct sensing technologies have revealed the heterogenous spatial environmental gradients that characterize our cities. Through the study of these datasets, significant correlations between urban microclimate and air quality, and socio-economic, urban greenery, and urban fabric design parameters have also been reported. However, physics-based urban microclimate and air quality modeling approaches generally fail to include detailed urban spatial datasets and thus tend to oversimplify spatial urban environmental gradients. This research proposes a machine learning approach utilizing Support Vector Machines (SVM) and Bayesian methods for spatial prediction of PM2.5 concentration and air temperature. High spatiotemporal urban environmental datasets collected from dense municipal and citizen science urban air quality sensing networks, in combination with remote sensing datasets on land surface temperature and normalized difference vegetation index, and geospatial datasets on socio-demographic and urban fabric parameters, have been collected for Los Angeles and New York City. Through a footprint-based study, and for all direct sensing station locations, the featured variables and the buffer sizes with higher correlations have been identified to train the machine learning models. The results show that supervised machine learning approaches trained utilizing dense urban environmental sensing datasets can provide a reliable approach for urban spatial environmental predictions. The research also contributes to the understanding of the required density of urban air quality sensing networks to capture urban air pollution gradients, as well as the possibilities to utilize citizen science air quality data.
- Affiliations
Citations
Llaguno, M., & Shu, X. (2021). Spatial urban air quality prediction model utilizing high spatiotemporal resolution direct and remote sensing, socio-demographic and urban fabric data. AGU Fall Meeting 2021. https://hdl.handle.net/2078.5/228144