The impact of mountain-valley wind circulation on the typical examples of pollution was analyzed through the selected pollution process, combining with the hourly PM2.5 concentrations and meteorological data in Haidian, Shangdianzi and Lishuiqiao in Autumn and Winter from 2013 to 2015, and also the data of Tower of atmospheric, wind profile of Haidian and automatic meteorological stations in the same period. The analysis showed that the average wind speed of valley wind was greater than that of the mountain wind, and they both would be "broken" during the conversion time in the mountain-valley wind days. In contrast with the mountain wind, the average duration of valley wind in autumn was longer than that in winter, and the start time of valley wind in autumn was earlier than the same wind in winter; influenced by the topography of Beijing area, the direction boundary of the transformation between mountain-valley wind was northeast-southwest. The frontier of mountain wind in autumn could fall down to the South Second Ring Road, and it could be pressed to the South Third Ring Road in winter; the average thickness of valley wind was greater than the mountain wind. Whether the moment was in autumn or winter, in the south, the average time when the PM2.5 concentration began to rise, was earlier than in the north in a day; the time when concentration of pollutants began to rise in the fall was earlier than in the winter, but the time when the concentration began to decline showed the opposite trend. The transition zone of different PM2.5 concentration in Beijing in autumn or winter located in South Second Ring Road (South Third Ring Road), and it would move to south over time. Duration autumn and winter seasons, this phenomenon lasted about 4 and 2 hours, respectively. Furthermore, the positive and negative feedback effects may exist between pollutant concentrations and mountain-valley wind.
Keywords: PM2.5; air pollution; mountain-valley wind circulation; positive and negative feedback; thickness.