Dynamic effect analysis of meteorological conditions on air pollution: A case study from Beijing

Abstract

Air quality directly relates to human health and economic and social sustainable development. This study collected the meteorological data of Beijing from November 1, 2013 to October 31, 2017, employed vector autoregression (VAR) model, Granger causality test, impulse response function and variance decomposition to explore the dynamic effects of average humidity, extreme wind speed, sunshine duration, average wind speed and rainfall capacity on air quality index (AQI). The results indicated that the air pollution in Beijing was mainly a self-aggregation and self-diffusion process, the self-cumulative effect accounted for around 88.9318% during 5 periods, once the diffusion conditions of air pollution worsen, air pollution would be formed within 3 days. Meteorological conditions, especially extreme wind speed, sunshine duration and average humidity affected the concentration and spatial-temporal distribution of air pollutant. Extreme wind speed as atmospheric dynamic factor rather than average wind speed was the most important meteorological element influencing the AQI change in Beijing, which caused more atmospheric motion and turbulence, improving the diffusion and dilution ability of air pollutant, whose self-cumulative influence was around 7.5270% during 5 periods. Sunshine duration as atmospheric thermal factor was the secondary important meteorological element affecting AQI change in Beijing for it was associated with the formation of temperature stratification and inversion, the self-cumulative effect accounted for around 2.1402% during 4 periods. This study deepens the insights about the formation and diffusion mechanism of air pollution in Beijing, introduces nontraditional methods to review traditional issue and draw valuable conclusions. Other natural or human action factor should be further analyzed in the future research.

Keywords: Air quality index; Granger causality; Impulse response function; Meteorological condition; VAR; Variance decomposition.