Abstract:Regional meteorological conditions and air quality may be closely related to global climate changes. Global climate changes were categorized as El Nio and La Nia. Based on the large data mining technologies of k-nearest neighbor (KNN), historical meteorological data from 1951 to 2017 and air quality data from 2013 to 2017 of Chengdu were both used for trend analysis and its correlation with global climate changes. The K-nearest neighbor (KNN) was used to evaluate the contribution of both meteorological conditions and emissions reduction to the air quality improvement. Results showed that global climate changes have significant impacts on both regional meteorological conditions and air quality. Both El Nio and La Nia have caused regional changes of meteorological conditions such as air temperature, precipitation, wind speed, and sunshine hours. It has facilitated the diffusion and dispersion of atmospheric pollutants. As a result, the air quality has been improved locally except for tropospheric zone. Compared to years with typical meteorological conditions, higher ozone concentrations were observed under El Nio or La Nia due to a variety of reasons. For the year of 2015, A KNN model was developed using the observed meteorology factors and heave pollution days. It is found that for the reduced heave pollution days in 2015(El Nio year), 73% can be attributed to the emission reduction, and 27% was caused by the improvement of meteorological conditions. In 2016(La Nia year), 42% was caused by the improvement of meteorological conditions. Thus, policy-making regarding air quality improvement and management should take into account both meteorological conditions and emissions sources.