Data science approach for urban monitoring network assessment with application in Buenos Aires

Abstract

Low air quality is a public health problem with severe global repercussions. This work presents a data science approach for obtaining relevant information from an urban monitoring network assessment. The long-term concentration of carbon monoxide (CO), nitrogen dioxide (NO2), and particulate matter (PM10) were examined at three stations (Córdoba, Centenario, and La Boca) in Buenos Aires for the period 2009-2017. The methodology allowed the evaluation of the monitoring network, the analysis of extreme events, and local meteorology´s influence on the recorded levels. The description of the hourly, daily, and seasonal behaviour of contaminants and the determination of concentration trends using the Mann-Kendall trend test were also included. The analyzed pollutants have reached high maximum values, which poses a risk to the exposed population´s health. All stations showed a significant reduction in NO2 concentrations, but only two stations had a reduction in PM10. We observed significant increases in CO levels. These results provide evidence of air quality degradation and provide critical information for regional government policy development.