Weather regimes and the related atmospheric composition at a Pyrenean observatory characterized by hierarchical clustering of a 5-year data set

At high-altitude stations worldwide, atmospheric composition measurements aim to represent the free troposphere and intercontinental scale. The high-altitude environment favours local and regional air mass transport, impacting the sampled air composition. Processes like mixing, source-receptor pathways, and chemistry rely on local and regional weather patterns, necessitating station-specific characterization. The Pic du Midi (PDM) is a mountaintop observatory at 2850 m above sea level in the Pyrenees. The PDM and the Centre de Recherches Atmospheriques (CRA) in the foothills form the Pyrenean Platform for the Observation of the Atmosphere (P2OA). This study aimed to identify recurring weather patterns at P2OA and relate them to the PDM's atmospheric composition. We combined 5 years of data from PDM and CRA, including 23 meteorological variables (temperature, humidity, cloud cover, and wind at different altitudes). We used hierarchical clustering to classify the data set into six clusters. Three of the clusters represented common weather conditions (fair, mixed, disturbed weather), one highlighted winter north-westerly windstorms, and the last two denoted south foehn conditions. Additional diagnostic tools allowed us to study specific phenomena such as foehns and thermally driven circulations and to affirm our understanding of the clusters. We then analysed the PDM's atmospheric composition statistics for each cluster. Notably, radon measurements indicated a regional background dominance in the lower troposphere, overshadowing diurnal thermal effects. Cluster differences emerged for the anomalies in CO, CO 2 , CH 4 , O 3 , and aerosol concentrations, and we propose interpretations in relation to chemical sources and sinks.