Modelling background air pollution exposure in urban environments: Implications for epidemiological research. (August 2018)

• Record Type: Journal Article
• Authors: Gómez-Losada, Álvaro; Pires, José Carlos M.; Pino-Mejías, Rafael
• Is Part Of: Environmental modelling & software. Volume 106(2018)
• Page Start: 13
• Page End: 21
• Publication Date: 2018-08
• Subjects: Clustering techniques -- Background pollution -- Air quality -- Time-series analysis -- Exposure -- Health risk; Environmental monitoring -- Computer programs -- Periodicals; Ecology -- Computer simulation -- Periodicals; Digital computer simulation -- Periodicals; Computer software -- Periodicals; Environmental Monitoring -- Periodicals; Computer Simulation -- Periodicals; Environnement -- Surveillance -- Logiciels -- Périodiques; Écologie -- Simulation, Méthodes de -- Périodiques; Simulation par ordinateur -- Périodiques; Logiciels -- Périodiques; Computer software; Digital computer simulation; Ecology -- Computer simulation; Environmental monitoring -- Computer programs; Periodicals; Electronic journals; 363.70015118
• DOI: 10.1016/j.envsoft.2018.02.011
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Background pollution represents the lowest levels of ambient air pollution to which the population is chronically exposed, but few studies have focused on thoroughly characterizing this regime. This study uses clustering statistical techniques as a modelling approach to characterize this pollution regime while deriving reliable information to be used as estimates of exposure in epidemiological studies. The background levels of four key pollutants in five urban areas of Andalusia (Spain) were characterized over an 11-year period (2005–2015) using four widely-known clustering methods. For each pollutant data set, the first (lowest) cluster representative of the background regime was studied using finite mixture models, agglomerative hierarchical clustering, hidden Markov models ( hmm ) and k-means. Clustering method hmm outperforms the rest of the techniques used, providing important estimates of exposures related to background pollution as its mean, acuteness and time incidence values in the ambient air for all the air pollutants and sites studied. Highlights: Four clustering techniques were compared to model background air pollution. k-means and agglomerative clustering were approximated to model-based clustering. Hidden Markov models ( hmm ) outperform the rest of the techniques used. Information derived from hmm results of interest for epidemiological research.