Spatial Correlation Analysis And Prediction Research Of PM_2.5 In Urumqi

As a kind of air fine particle pollutants,PM_2.5 affects human health,atmospheric environment and ecosystems seriously,so it has been widely concerned and studied by the society.Urumqi is the famous international dry port city as the country's throat to Central Asia and West Asia and the window opening to the west.Analyze and predict the historical data of PM_2.5 concentration in Urumqi to provide a scientific basis for its air pollution control,and then provide a theoretical basis for the formulation of environmental protection control measures in the core area of the“Belt and Road”.This article does the following statistical modeling research on the monitored PM_2.5 data:(1)The first part of this article analyzes the current status of PM_2.5 in Urumqi,this paper analyzes the spatial autocorrelation of PM_2.5 concentration data of 7 monitoring points in Urumqi.It is concluded that the spatial correlation of PM_2.5 pollution in monitoring sites in Urumqi is relatively small.In other words,the level of PM_2.5 pollution among the seven districts of Urumqi may have a small mutual influence,but the pollution in each district is also relatively serious.The pollution situation in each district of Urumqi is not optimistic.(2)The second part of this article analyzes the influencing factors of PM_2.5.5 in Urumqi,to study the relationship between PM_2.5 and other monitoring indicators in AQI.This article first analyzes the relationship between PM_2.5 concentration and PM₁₀,O₃,SO₂,NO₂,CO through linear regression model,linear parameter quantile regression model,single index quantile regression model.Using statistical software R language to estimate the specific model parameters,it is concluded that PM₁₀,O₃ and SO₂ coefficients do not change very much at different quantiles,but they have a great impact on PM_2.5.The estimated value of the CO coefficient is relatively large,indicating that the higher the CO concentration,the greater the impact on PM_2.5.NO₂ has a negative effect on PM_2.5,but its influence coefficient is not very large.From the results,we can see that the quantile regression method can reflect the whole picture of the data more than the linear regression.Secondly,the influence of meteorological factors on PM_2.5 concentration was studied.The results show that PM_2.5.5 concentration in Urumqi is negatively correlated with ground temperature,sunshine duration,air temperature and wind speed,and positively correlated with precipitation,air pressure,humidity and sunshine duration.Finally,it analyzes the impact of heating and non-heating periods on the concentration of PM_2.5 in Urumqi,and provides PM_2.5.5 early warning for citizens of Urumqi.(3)When predicting the PM_2.5 concentration in Urumqi at the end of this paper,the PM₁₀,CO,O₃,SO₂,NO₂ concentration data is used as the input variable sequence,and established ARIMAX model.The parameters of the ARIMAX model affecting the variable sequence are estimated.It is concluded that the ARIMAX model passes the significance test.The PM_2.5concentration error calculated based on the prediction value of the ARIMAX model is small,indicating that the prediction effect is good,and relevant suggestions are made based on the prediction result and influencing factors.According to the above research,in order to reduce the PM_2.5 concentration in Urumqi,it is recommended that you buy new energy vehicles to reduce exhaust emissions.On the one hand,improve the technology for producing high-quality coal and reduce the combustion of low-quality coal that contains a large amount of polluting gas;on the other hand,promote the use of low-emission energy such as natural gas.