Since the reform and opening up,China’s industrialization and urbanization have developed rapidly,and the problem of air quality pollution has become increasingly serious,which has seriously affected public health and living environment.In recent years,Nanchang has successively built an ecological civilization demonstration area and carried out the battle of pollution prevention and control.The air quality has been greatly improved,but PM2.5remains high as the primary pollutant.Therefore,this thesis will start from the air quality indicators,integrate meteorological and economic factors,and study the PM2.5 in Nanchang from the aspects of temporal and spatial variation characteristics,source analysis,model prediction and feature monitoring.Firstly,the time variation law of PM2.5 in Nanchang from 2014 to 2020 is analyzed by year,quarter,month and day.It is found that the average annual concentration is lower than the national first-class standard;The seasonal pattern is obvious,which is winter > autumn >spring > summer;The monthly variation of concentration shows a "U" distribution with high on both sides and low in the middle;Except that the diurnal variation characteristics in summer are not obvious,it is generally of double peak and double valley type.Then,using Kriging spatial interpolation method,it is found that the spatial distribution of PM2.5 in the main urban area increases from northwest to southeast,and the high-value areas are concentrated in the old urban area,Honggutan new area and Xinjian area;The analysis techniques of backward trajectory clustering,potential source contribution factor analysis(PSCF)and weighted concentration trajectory(CWT)are used to identify the PM2.5diffusion path,potential source area and its corresponding daily average concentration contribution level of PM2.5 in four seasons were determined.The results show that only the airflow trajectory from the south accounts for the largest proportion in summer,and the trajectory with the largest proportion in other seasons comes from the airflow in the north or northeast;The analysis results of PSCF and CWT are basically the same.The high-value areas are mainly concentrated in and around Jiangxi province,and the regional contribution to the north of Nanchang is the main.Finally,functional data analysis is used to explore the main influencing factors of PM2.5in Nanchang,and predict and monitor them.Traditional multiple regression and functional regression analysis show that air pollutants,meteorological conditions and regional economic development have a significant influence on the change in PM2.5 concentration.The functional regression model with interactive items constructed by using nine single variables and six groups of interactive items selected by variable selection can significantly improve the prediction effect,reveal the interactive effects among influencing factors,and then comprehensively reflect the internal mechanism of PM2.5 change.In order to test whether a series of measures taken by Nanchang in recent years to improve air pollution have a significant impact on PM2.5,the policy implementation is introduced into the model as a label variable.The research shows that the reverse effect on PM2.5 reaches the maximum after building a demonstration area and carrying out the battle of pollution prevention and control for a period of time.Especially in 2018,due to the simultaneous implementation of the two measures,the average annual concentration of PM2.5 reached a record low.To monitor whether PM2.5 has abnormal fluctuations over a given period of time,the functional regression control chart(FRCC)is further used to analyze its stability,monitor and feed back the out of control points in the process.The results show that the time when PM2.5 is out of control or unstable is mainly concentrated in winter and autumn.By objectively grasping the pollution characteristics and improvement of PM2.5 in Nanchang,using an appropriate functional regression model,this thesis explores the main influencing factors of PM2.5 in Nanchang,reveals the dynamic regularity between variables from a continuous perspective,and further predicts and monitors the PM2.5 concentration level,so as to help citizens better grasp the air pollution situation,reasonably arrange travel,and help relevant departments carry out air pollution control,promote the sustainable development of urban economic construction. |