Empirical Analysis Of Factors Affecting Air Quality In Fuzhou City

Since the Ministry of Environmental Protection announced the PM_2.5 concentration data of various cities in 2013,the problem of haze and air pollution has received more and more attention,and a large amount of air quality data has been accumulated.There are many kinds of pollutants in the air,and the main concern is PM_2.5.Meteorology is an important factor affecting air quality.The impact of meteorology on air quality is multifaceted.Traditional statistical methods are limited due to the dramatic increase of data.By using data mining methods,data features can be discovered from large data sets,providing a basis for further research.Taking the data of Fuzhou City from 2014 to 2018 as an example,this paper attempts to find out the law in data from data-driven data mining,and then studies the influence of meteorological factors on air quality from the two aspects.Firstly,using the association rule analysis to mine and screen the strong correlation rules between meteorological conditions and pollutant concentration.These rules can reflect the correlation between the air quality of meteorological conditions in Fuzhou City,and can be used to find the main meteorological factors affecting air quality.Second,the overall relationship explores the use of canonical correlation analysis to analyze the overall ability of meteorological interpretation of changes in the overall concentration of pollutants.Finally,the focus is on PM_2.5 concentration,fitting PM_2.5 concentration and meteorological conditions,and other pollutant,and comparing the effects of parametric regression and nonparametric regression.In a conclusion.There is a strong correlation between higher temperature,lower PM_2.5concentration and lower wind speed in most association rules,indicating that temperature has a significant impact on air quality.Canonical correlation analysis found that meteorological conditions as a whole can account for 30%of the variance in the overall variation of air pollutant concentrations.Can explain 56.8%of NO₂ change,28.9%of CO,7.8%of O₃,15.7%of PM₁₀,47.7%of PM_2.5.The nonparametric regression effect is better,that is,nonparametric regression can better explain the complex relationship between variables.