Econometric Analysis Of Influencing Factors And Prediction Of Air Quality Based On Spatial Statistics

Air quality has a direct impact on people's daily life.In recent years,air quality has seriously declined,haze weather occurs from time to time.Travel difficulties and disease outbreaks caused by severe air pollution in winter and spring have attracted great attention of the country.The analysis of spatial distribution characteristics and influencing factors of air quality can provide reference for formulating control measures.At present,real-time monitoring data are the main means of air quality broadcasting,and there is no unified air quality prediction system in the whole country.Taking Xi'an and its surrounding cities as an example,the spatial distribution characteristics of air quality index and major pollutants are analyzed by using spatial statistics method,and positive spatial correlation is found.Because the traditional regression model neglects the spatial correlation of the research object,the spatial econometric model considering the temporal and spatial effects is introduced.In order to comprehensively analyze the influencing factors of air quality and introduce as many variables as possible,but in order to make the results more explanatory,the principal component analysis method is used to reduce the dimension of various influencing factors and then as influencing indicators to be introduced into the econometric model.This paper takes PM10 and PM2.5 as explanatory variables,public transport,urban development level,environmental protection,air pollution and fuel combustion as explanatory variables,and establishes a spatial econometric model with spatial inverse distance weight matrix as spatial weighting matrix.In the spatial measurement model of PM2.5,the error of space-time Durbin model with double fixed effects is the smallest,and the impact of social and economic development,public transport and environmental protection on PM2.5 is negative.This shows that with the development of economy,it will help to reduce environmental pollution in a certain period of time;the development of public transport and environmental protection will help to alleviate air pollution;fuel combustion and atmospheric pollution and so on.PM2.5 is positively correlated,so we should strengthen the control of sewage discharge and develop new energy sources.For PM10,the goodness of fit of spatial Durbin model without fixed effect is higher than other models,and the public transport and environmental indicators are negatively correlated with PM10,which shows that a perfect public transport system can help reduce the use of private cars,thereby reducing air pollution;and the increase of environmental protection is also helpful to reduce air pollution.By studying the existing air quality prediction methods,we find their limitations.On the premise that the air quality index has spatial correlation,taking Xi'an as an example,considering both time and spatial effects,we combine the traditional Kriging interpolation prediction method with time series model to propose a two-stage space-time interpolation prediction method,and the prediction results are compared with the general time series model.The results are compared with the results of space-time Kriging prediction.The results show that the prediction error of two-stage spatial interpolation is smaller than that of spatio-temporal Kriging.The conditions and limitations of each prediction method are further analyzed.