AQESHOW - Abstract

Abstract Information :

Air pollution dispersion models calculate pollution levels due to emissions from known pollution sources, such as road traffic and industry, and are widely used for air quality management. Using algorithms rooted in atmospheric science, they consider meteorological conditions, chemical reactions with other gases, local surface characteristics such as street canyons, and underlying or ‘background’ pollution levels, and can provide detailed spatial and temporal pollution information within the modelled domain. Typically, modellers use an iterative approach to model validation to improve model performance by refining model inputs to improve model agreement with available monitored data, but the increasing affordability and popularity of air quality sensor networks offers exciting opportunities to combine the benefits of models and measurements in a more systematic way. Inversion methods have been developed for this purpose, automatically adjusting emissions to improve model agreement with sensors, accounting for measurement and emissions uncertainty, leading to more accurate modelled data and providing emissions insights. In this presentation, results will be presented from a study where this technique has been applied to a network of air quality sensors hosted on telecom infrastructure in the city of Glasgow.