The spatio-temporal variability of particulate pollution through modelling: new insigths from a dense network of micro-sensors in urban environment

Current high-resolution models generally present a relatively flat signal that poorly represents the spatial and temporal variability of particulate pollution at the scale of a city. This study is based on the SIRANE model, an urban air quality model that enables to simulate the dispersion of atmospheric pollutants according to the city geometry at a 10-meter spatial resolution. The model outputs provided by the ATMO Bourgogne - Franche - Comté air quality monitoring agency are actually post-processed based on a physical equation defined for the PM₁₀ (d < 10 μm) and PM_2.5 (d < 2.5 μm) pollutants and based on mass concentration levels measured by 4 micro-sensors deployed in representative sites in Dijon for the year 2021.

This study first aims at verifying if the equations established could be applied to another period, taking into account any differences in atmospheric circulation. The second objective is to test if the integration of more measurement stations enables to significantly refine the physical equations and improve the SIRANE maps correction. More generally, we would like to evaluate to what extent micro-sensors can improve the information provided by high-resolution models and when with respect to the particle season.

The work is conducted in three steps: first, apply the physical equations to 2022 and 2023 SIRANE maps and compare the outputs with reference stations; second, analyse measurements from the micro-sensors implemented in Dijon in Summer 2023; third, use this dataset to select new representative traffic, background and intermediate sites in order to refine the physical equations and quantify the added value.

The first results are encouraging as the SIRANE corrected maps based on the first 4 representative micro-stations in Dijon enable a more realistic spatial variability in the city, illustrated by a clear pollution gradient from the city centre to the suburbs (with a greater range between concentration levels and a higher number of classes), and a less flat season cycle with more realistic PM levels in Winter everywhere in the city.