Bias-adjustment method for street-scale air quality models

Abstract

Air quality (AQ) is a growing concern, especially in urban areas where high-density populated regions are exposed to frequent exceedances of regulated pollutants. To take action in reducing citizen exposure to pollution, a reliable assessment of the pollutants’ ambient concentrations across the city is required. Street-scale AQ models are designed to capture the typical spatial variability that pollutants exhibit in the urban morphology. Such urban models are generally nested to regional AQ models and use the information of traffic emissions, together with meteorological conditions, and a geometric description of the building’s layout, to provide an estimation of the dispersion of target pollutants at the street scale. However, results of urban AQ models are subjected to uncertainties, mainly due to the multiscale behavior of the phenomenon and to the challenges of characterizing the wind flow within street-canyons, which encompasses multiple emission sources and the downscaling of meteorological variables. To minimize these uncertainties, we present a data-fusion method that combines the model results, obtained using the CALIOPE-Urban [1] model, with publicly available observations from the official monitoring network in Catalonia (XVPCA). This method is derived to preserve the spatial variability of the urban model. As a test case, we then present annual bias-corrected results of the NO2 levels across the city of Barcelona for the year 2019. Results correspond to the legislated annual mean and the 19th daily maximum value of the year.