Analysis And Reasearch Of PM_2.5 Based On Time Series Method And Stepwise

With the rapid development of the economy,pollutants released into the environment are gradually increasing,affecting human health and survival.In recent years,frequent PM_2.5pollution incidents have gradually attracted the attention of human beings,and people have paid more attention to PM_2.5.5 and environmental pollution incidents.They have also become the research objects of many scholars,from PM_2.5.5 chemical components,changing characteristics,and pollution sources.This study used the air quality monitoring data and meteorological data of 51monitoring stations in Hubei Province in 2016,using time series analysis method and stepwise regression method to model the PM_2.5.5 concentration data of each city in Hubei Province,and applied the time series method.Based on the PM_2.5.5 data,the model fitting effect was analyzed,and the stepwise regression method was used to analyze the relationship between meteorological elements and PM_2.5.The main conclusions of this paper are as follows:?1?The monthly,daily and seasonal variations of PM_2.5.5 concentrations in all cities of Hubei Province are basically the same.The peak concentration of PM_2.5.5 is concentrated in January and December,and PM_2.5.5 concentrations in June,July and August.During the low period,the annual PM_2.5.5 concentration gradually decreased from January to July,and the PM_2.5concentration reached the lowest in the whole year.The PM_2.5.5 concentration increased in August and September,and decreased again in October.The concentration in November and December rose again;the daily variation of PM_2.5.5 concentration in each city was basically the same,and the PM_2.5.5 concentration changed gently,showing a distinct bimodal distribution.In the early hours of the morning,the PM_2.5.5 concentration decreased slowly,and the PM_2.5concentration gradually increased at around 6:00.The first peak appeared at 12:00 during the day,and then the PM_2.5.5 concentration slowly decreased again,at 17:00 in the evening.There was a trough area around 18:00.The concentration of PM_2.5.5 increased gradually at night,and the second small peak appeared at 23:00.The seasonal variation of PM_2.5.5 concentration in each city was basically the same,showing winter>spring>fall>summer,and individual cities have a higher autumn concentration than spring.?2?The fitting effect of PM_2.5.5 concentration model in each city is basically the same.In the high value period of PM_2.5.5 concentration,the model fitting effect is not ideal.In the low value period,the model fitting result is smaller than the actual value,which is ideal;during the period of gently changing concentration,the model fitting effect is good,and the degree of variation is large,the model fitting effect is poor;the fitting effect of PM_2.5.5 concentration model in each city is about the same,and the period of PM_2.5.5 concentration is gently changing.The fitting effect is good,indicating that the time series model is suitable for data analysis with gentle changes;the relative error of time series method that is used to predicts PM_2.5concentration will increase with the increase of predicted days,and the weather factor has a great influence on the prediction result of PM_2.5.5 concentration,further reducing the prediction accuracy.?3?The influence of meteorological elements in different cities on the concentration of PM_2.5.5 is different.The fitting effect of individual cities is not ideal.Generally speaking,the wind speed has the greatest influence on PM_2.5.5 concentration,followed by temperature,air pressure,humidity and sunshine.