Observations of Thermally-Driven Circulations in the Pyrenees: Detection Methods and Impact on Atmospheric Composition Measured at a Mountain Top

Atmospheric composition measured at the Pic du Midi high altitude observatory (2875m asl) in the Pyrenees is frequently affected by upward transport of boundary-layer air during anabatic circulations at different scales. The Pyrenean Platform of Atmospheric Observation (P2OA) includes two observatories located 28-km apart form each other: the Pic du Midi, and a low altitude site (580 m asl) located in the plain north of the chain. Based on a long data series (about 10 years) collected at P2OA, different methods are carried out to detect thermally-induced circulations from observations at three key locations in the plain-to-mountain circulation cell: within the altitude return flow above the plain, close to the surface in the plain, and at the mountain top. The main aims of this study are (i) to present and compare three independent detection methods, and (ii) to evaluate their impact on in situ atmospheric composition measurements at Pic du Midi.

The first method uses radar wind measurements at 3000m above the plain in order to detect the return flow of the thermally-induced plain-mountain circulation (27% of days). The second, based on surface wind data from the plain site, reveals days during which surface breeze occurs locally (27% of days). The third method, based on surface data at the mountain-top, focuses on moisture diurnal cycles in order to detect and rank days with decreasing anabatic influence (50 % of days).

The selected day ensembles obtained from the three methods overlap only partly (10 % of days), revealing that thermally-induced winds occur not always at the mountain-plain scale, but in some cases at more local scales. Nevertheless, the seasonal distributions of the three ensembles are to some extent similar. A consequence when considering the impact on atmospheric composition at Pic du Midi, is that the mean diurnal cycles of ozone, carbon monoxide and dioxide, methane, and aerosol cocentrations, obtained from the different day ensembles, show some consistency. However, considering only surface wind measurements in the plain (method 2) gives results differing most from the other two methods (1 and 3) for anabatic days, suggesting the occurrence of local breezes at low altitude over the piedmont, but with little influence in altitude. Conversely, the non-local methods 1 and 2 miss days with anabatic influence at Pic du Midi: the local method 3 thus appears better to select days with no anabatic influence.