Characterization of outdoor-indoor air pollution at a city-wide scale in Barcelona using a high-resolution urban air quality model

Urban air quality presents a significant global environmental challenge, with 77% of the urban population exposed to NO2 concentrations exceeding the 2021 guidelines established by the World Health Organization. Moreover, recent analyses show that people typically spend approximately 90% of their lives indoors. In this context, understanding indoor and outdoor air quality levels is key to better assess population exposure. Despite there is extensive research on modelling and monitoring air quality, studies modelling outdoor and indoor pollution simultaneously remain limited. In this study, we address this research gap by combining NO2 outdoor levels from the air quality model CALIOPE-Urban with indoor-outdoor parametric relations.

Bias-corrected NO2 concentration levels at the street scale (20 m x 20 m) are used. This information is based on a kriging data-fusion method that merges monitoring stations and the dispersion model CALIOPE-Urban. Additionally, it integrates a machine learning-driven microscale-Land Use Regression model, which is calibrated using observational data gathered from short, intensive passive dosimeter campaigns.

Two different parametric relations for deriving indoor-outdoor air quality levels are assessed. On the one hand, we use infiltration rates, which quantifies the fraction of outdoor pollutants  entering into buildings. On the other hand, we apply indoor-outdoor ratios, which besides infiltration also include indoor sources. Both parameters are obtained from available data in the literature. They depend on socio-demographic data (e.g., building use) and  seasonality (e.g., summer, winter, and spring-autumn) which can affect ventilation patterns.

To obtain unprecedented city-wide maps containing both indoor and outdoor NO2 concentrations, bias-corrected NO2 concentration levels and the parametric relations are combined. We present results for the 2019 annual mean of NO2 in the domain of Barcelona. Results are validated using existing indoor and outdoor measurements from various experimental campaigns available in the city. Furthermore, a sensitivity analysis is carried out to quantify  uncertainties associated with parametric relations and their effects upon final results.

This comprehensive approach allows us to refine our understanding of both indoor and outdoor air quality dynamics in Barcelona. Such findings hold promise in improving population exposure in urban environments, revealing both indoor and outdoor NO2 hotspots of pollution exposure.