Analysis Of The Air Quality Index In Beijing-Tianjin-Hebei-Shanxi Region Based On Functional Data

Since the 18th National Congress of the Communist Party of China,General Secretary Xi Jinping has repeatedly emphasized that green water and green mountains are golden mountains and silver mountains.With the rapid development of my country's economy and the acceleration of industrialization,the phenomenon of air pollution has become more and more serious,which has attracted great attention from the people.It is urgent to improve the air pollution problem.Due to geographical and economic factors,the Beijing-Tianjin-Hebei-Shanxi area has been affected by haze weather all year round,which has caused a great impact on people's lives.This article is based on the weather data of 24 cities in the Beijing-Tianjin-Hebei-Shanxi region from 2018 to 2020,using functional data analysis methods combined with generalized quantile regression and vector autoregressive models with exogenous variables to determine the air quality index(AQI)The influencing factors are studied,and the four variables of daily average temperature,daily average wind speed,average daily precipitation,and PM2.5 content are analyzed on the AQI.The results of the two models explain the impact from different dimensions.The research results are complementary to each other.The follow-up study of the air quality index and the improvement of air quality in my country provide new ideas.This paper firstly analyzes the characteristics of the air quality index in the Beijing-Tianjin-Hebei-Shanxi area in terms of time and space,and obtains the regional air quality distribution characteristics;secondly,the function of the air quality index is constructed based on the AQI daily data of the Beijing-Tianjin-Hebei-Shanxi area in 2019 Type regression model,using the Fourier expansion method to fit the function variables,and using the harmonic acceleration operator to punish,select the best smoothing parameter according to the GCV value,and then perform the parameter test on the regression model.The model passes the significance test.R~2 is 0.9895,indicating that the model has a good interpretation.In this way,the influence of the four covariates on the AQI at different times of the year is obtained.According to the regression analysis results,the following conclusions are obtained:temperature can be said to be the"adjuster"of AQI,and the temperature changes in the four seasons of each year are used to adjust AQI in a timely manner,so that it will not develop towards the extremes of too good or too bad;The influence of wind speed on AQI is roughly divided by season,and the degree of positive influence and negative influence is about the same;the influence of precipitation on AQI is the strongest in winter,the first two months of the year are negative,and the end of the year is positive.As one of AQI's accounting items,PM2.5 has always had a positive impact on AQI.Finally,using Beijing's AQI time data in 2018 and 2019,based on the seven-quantile level,a progressive generalized quantile regression,using the least asymmetric weighted square(LAWS)criterion combined with a punitive B-spline curve to get generalized quantile function of AQI.Perform functional principal component analysis on this generalized quantile.The analysis results show that the first three principal components have accumulated to explain 97%of the variation in the response variable.Therefore,the first three principal components are extracted and the meaning of each principal component is analyzed.Explanation.This paper aims to find out how the four variables of daily average temperature,daily average wind speed,daily average precipitation,and PM2.5 content affect AQI.Therefore,a vector autoregressive model(VARX)with exogenous variables is established,and the above four variables Join the model as covariates,and the three principal component scores obtained above are used as response variables.In this way,the regression results of the three principal component scores on the four covariates at the seven-quantile level are obtained,that is,four covariates are obtained.The influence of the variable on the AQI varies throughout the distribution of the AQI.Conclusions as below:1.The first principal component is dominated by temperature,and the regression coefficient is positive,indicating that as the temperature increases,the AQI value also increases.And the coefficient has a tendency to increase with the increase of the quantile level;2.The second principal component is dominated by wind speed.After the 95%quantile,the regression coefficient turns from positive to negative,indicating that the AQI is followed by As the wind speed increases,the score of the second principal component will decrease,but before this,the situation is the opposite;3.The third principal component is also dominated by wind speed,and the regression coefficient is negative,that is,as the wind speed gradually increases,the first The three principal component functions decrease accordingly,and the degree of influence is relatively stable.