Mining Public Datasets for Modeling Intra-City PM2.5 Concentrations at a Fine Spatial Resolution

Abstract

Air quality models are important for studying the impact of air pollutant on health conditions at a fine spatiotemporal scale. Existing work typically relies on area-specific, expert-selected attributes of pollution emissions (e,g., transportation) and dispersion (e.g., meteorology) for building the model for each combination of study areas, pollutant types, and spatiotemporal scales. In this paper, we present a data mining approach that utilizes publicly available OpenStreetMap (OSM) data to automatically generate an air quality model for the concentrations of fine particulate matter less than 2.5 μm in aerodynamic diameter at various temporal scales. Our experiment shows that our (domain-) expert-free model could generate accurate PM2.5 concentration predictions, which can be used to improve air quality models that traditionally rely on expert-selected input. Our approach also quantifies the impact on air quality from a variety of geographic features (i.e., how various types of geographic features such as parking lots and commercial buildings affect air quality and from what distance) representing mobile, stationary and area natural and anthropogenic air pollution sources. This approach is particularly important for enabling the construction of context-specific spatiotemporal models of air pollution, allowing investigations of the impact of air pollution exposures on sensitive populations such as children with asthma at scale.

DETECT: Deep Trajectory Clustering for Mobility-Behavior Analysis

“…The Tobler's First Law of Geography reported that "Everything is related to everything else, but near things are more related than distant things" [36]. This phenomenon forms the basis for the study of Geographical Influences, which has seen wide adoption in the literature: from POI recommendation [42], [44], [45] to air quality prediction [27]. In our work, the geographical context is constituted of categories (e.g., restaurants, apartments, hospitals) of the points of interest in the local vicinity of a stay point.…”

Statistical evaluation of spatial uncertainty in schistosomiasis mapping

“…land use and roads, derived from OpenStreeMap, to predict PM2.5 concentrations. When generating relative importance measures for the different risk factors, MAUP effects reduced when applying a random forest model that was trained with the distances between the features and the monitoring PM2.5 stations, (Lin et al, 2017). The rapid development of geoAI methods, their advantage to deal with big data, and their rapid computational time, makes them an attractive and advantageous tool to tackle limitations with modelling schistosomiasis and other diseases.…”

“…Innovations in this field aim to address real-world problems like the ones related to human health, in particular, spatial epidemiology (Baker and Nieuwenhuijsen, 2008). geoAI in spatial epidemiology looks to target issues related to inefficient computational processing and data constraints regarding coarse spatial and temporal supports, for exposure assessment (Lin et al, 2017;Vopham et al, 2018).…”

“…The use of SBD could improve the spatio-temporal resolutions of the output predictions, as well as identify significant risk factors of SCH prevalence (Vopham et al, 2018). For instance, Lin et al (Lin et al, 2017) eliminated the need of previous selection of predictors of air pollution, by letting the data decide which risk factors were significant for exposure. This was done using a data mining-based algorithm.…”

“…To further explore these issues I encourage to investigate geoAI applications for spatial epidemiology. For instance, Lin et al (Lin et al, 2017) presented a data mining approach to automatically select important geographic features for PM2.5 concentration predictions. These features were selected using a random forest classifier that quantified their influence on SCH prevalence.…”