Decision support tool to improve the spatial distribution of air quality monitoring sites

Abstract

The air quality is an increasingly concerning topic. Efficient air quality monitoring system is highly required in urban areas due to the existing different types of air pollution sources. Consequently, air pollution profiles of the monitored region should be regularly evaluated to infer the efficiency of the monitoring system. This study applied principal components analysis (PCA) to air quality data from Porto (Portugal), aiming to characterize the spatial distribution of NO2,O-3 and PM10, concentration profiles. In the analysed period (2006 and 2013), some of the selected monitoring sites have been closed and others were moved to other locations. PCA allowed the evaluation of these changes in geographical distribution of the monitoring sites and the definition of proposals to improve the efficiency of air quality monitoring system. The performed analysis concluded that these last changes were correctly performed, but the reduction of the monitoring sites for NO2 and O-3 is still possible, being optimized the current number for PM10. In addition, the application of computational model HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory) allowed analysing the air pollution transport over long distances in high pollutant concentration episodes, having found that the variation of NO2 and O-3 concentrations may be strongly influenced by air masses from the north of Spain. For PM10, there are two different air mass trajectories: the transport of air masses from the Galicia region and from North Africa.