Ma, Y; Ma, B; Jiao, H; Zhang, Y; Xin, J; Yu, Z
In recent years, near-surface ozone (O-3) concentrations have been increasing, which aggravates O-3 pollution. Due to the environmental threat it poses to human health, O-3 pollution has become a hot topic among researchers. In this paper, we used a generalized additive model (GAM) to evaluate the complex nonlinear relationships between O-3 concentration and the factors influencing O-3 concentration from 2013 to 2017 in Lanzhou in Western China. We considered factors such as long-term trend, seasonality at a quarterly interval, and the weekend effect. The results showed that O-3 concentration in Lanzhou was affected by many factors, and these influencing factors were correlated with each other. In the single-factor model, the adjusted coefficient of determination (R-2) was 0.594, and the total deviance explained of O-3 concentration was 62.3%. We detected significantly nonlinear relationships between O-3 concentration and the influencing factors like air temperature, sunshine hours, wind speed, relative humidity, and the concentrations of NO2 and PM2.5. In particular, air temperature was the main driving factor for O-3 concentration, which explained 24.2% of the variance (F = 350.84). In addition, we built a double-factor model to investigate the interactive influences of these influencing factors on O-3 concentration variation. In the fitted double-factor model, R-2 was 0.636 and the total deviance explained was 68.1%, both of which were higher (i.e., better performance) than that in the single-factor model. Among these studied influencing factors, the interaction between air temperature and air pollutants showed the greatest influence on O-3 concentration variation. The results of this study can be used to assist local environmental authorities to take proactive measures for O-3 pollution control in Lanzhou, Gansu.