The Arve river valley airshed in the French Alps experiences particularly severe air pollution during wintertime stable atmospheric conditions associated with persistent cold-air pools. PM 10 data recorded in the region indicate that the urbanised area of the central basin-shape section of the valley is generally the most polluted, with a harmful impact on the health of inhabitants. In the present work, we examine the air pollution transport potential of the Arve river valley airshed using results from high-resolution numerical simulations of a cold-air pool documented as part of the Passy-2015 field campaign. Passive tracers were used to model PM 10 with emissions provided by a detailed inventory developed by the local air-quality agency. The observed differential in PM 10 levels between valley sections was well captured by the numerical model and could not be explained solely by the differential in emissions. The stagnation, recirculation and ventilation potential of the airshed was evaluated spatially and temporally using integral quantities. The analysis indicated that the central basin-shape section of the valley is poorly ventilated and hence air pollution there would originate mostly from local emission sources. This stagnation zone appears to be almost decoupled from the rest of the airshed. The airshed was decomposed in separate valley sections so as to quantify the fate of the pollutants emitted within each section. Air pollution apportioned according to the contribution of emissions from the different valley sections shows that indeed the central basin-shape section is dominated by local sources. The situation was found more complex in the valley sections further downstream, where the contribution from the sum of the non-local sources can be as large as that from local sources. This study allows to identify the origin of the strong pollution in the Arve river valley, through the link between the local topography, emission sources and pollutant transport. Image 1 • The magnitude of PM 10 concentration predicted by the numerical model compares well with air quality station measurements. • In the central part of the valley, PM 10 predicted by the numerical model is trapped in the first 100 m above the ground. • The basin-like atmospheric dynamics around Passy and high local emissions account for persistent strong pollution there. • Under stable atmospheric conditions over complex terrain, local sources proved to be key in observed high pollution levels. [ABSTRACT FROM AUTHOR]