Box-Jenkins stochastic models for studying air pollutants in a Latin American megacity

Abstract

The objective of this paper is to show the development of Box-Jenkins stochastic models to study the behavior of air pollutants concentrations in the megacity of Bogotá, Colombia. Information was collected from 10 stations in the city’s air quality monitoring network over a ten-year period. The temporal relationship between air pollutants, their spatial variation, and the occurrence of extreme pollution episodes was studied using Box-Jenkins models. The results showed that the moving average term of the models developed was the main indicator of spatial distribution for the daily pollutant concentrations. In the case of atmospheric particulate matter < 10 μm, the following spatial distribution was identified in the megacity: northwestern, center-southwest, and southeast. For atmospheric particulate matter < 2.5 μm: north, center, and southwest. For ozone: northwest, center, and south. Maximum hourly particulate matter concentrations were observed between 6-10 a.m., and between 11 a.m. - 4 p.m. for ozone. Monthly, the highest particulate matter concentrations were observed in February (14.1%), January (13.5%), and March (12.2%). In the context of atmospheric physics, this study was relevant for the following findings: The usefulness of Box-Jenkins models in simulating the temporal behavior of air pollutants, and for their adequate performance in detecting urban spatial trends.