Identification and characterization of foehn events in Beijing and their impact on air-pollution episodes

Abstract. This study proposes a method for identifying foehn events in Beijing using automatic weather station (AWS) data, considering upper-air wind direction, topography, meteorological changes, and foehn propagation. Analysis of AWS data from 2015 to 2020 revealed an annual average of 56.5 foehn days, with these days occurring most frequently in winter and least frequently in summer. High-frequency foehn areas exhibit a band-like distribution from the northwestern mountainous region to the southeastern plains, while low-frequency areas are primarily concentrated in the northeastern plains. The horizontal extent of the foehn influence is maximal in spring and minimal in summer. Foehn-induced hourly temperature increases can exceed 11 °C, peaking from night to early morning. Approximately 67 % of pollution episodes are accompanied by foehn events, with foehn duration negatively correlated to pollution episode duration. 60.4 % of foehn events coincide with decreasing concentrations of particulate matter of 2.5 μm diameter (PM2.5), while 39.6 % show increases. Rapid PM2.5 concentration increases (> 50 μg m^-3/h) primarily correspond to weak foehn events (temperature increase < 2 °C). Foehn winds influence pollution through direct and indirect effects. The direct effect, associated with strong northwesterly pressure gradients, can ;rapidly decrease pollutant concentrations. The indirect effect, linked to weak pressure gradients, alters the boundary-layer structure, causing slight decreases followed by rapid increases in pollutant concentrations. This foehn identification method, applicable to long-term historical surface observations, facilitates in-depth exploration of the relationships between foehn events and high-impact weather phenomena.