Environmental Modeling and Traffic Simulation: A multivariate approach to monitor urban air pollutant agents.

Abstract

This work presents an interdisciplinary assessment examining air quality tracking in urban environments. This application is well suited to be approached with wireless sensor networks' paradigm in their overall variations. The proposed modeling application takes advantage of Vehicle Sensor Networks (VSN) by embedding sensor nodes in public transportation, addressing this study case with bus lines so that the mobiles spread the sampling activity through many different places visited during the route. Simultaneously, it alleviates power management restrictions, packaging dimensions (size and weight), and general maintenance issues. We perform environmental modeling based on real data considering temporal and spatial multivariate behavior on observed phenomena. We consider the city of São Paulo in our case study and parse the asserted data to create a multivariate map of samples, showing the behavior of five different air pollutants from fossil-fueled vehicles (CO, O3, PM10, NO2 and SO2) simultaneously while it also varies in time. Furthermore, the experiment considers a detailed description of roads, bus lines, vehicle itineraries, and general traffic information. The input data that has unformatted or missing information due to being sourced from real sensors is handled to create the map mentioned above. Our methodology addresses the following: 1) the mentioned environmental simulation, 2) the deployment of mobile sensor nodes and performing sensing process, 3) the implementation of network activity and delivery of collected data, 4) visualization of monitored environment based on gathered data using Voronoi Diagrams to fill blank data at non-reached areas. Finally, our VSN-based approach improved 126 times lower error and 11 times higher coverage compared to conventional monitoring with air quality stations.