Spatial-Temporal Characteristics Of PM_2.5 And Its Influencing Factors In Hunan Province

In recent years,with the continuous improvement of people’s living standards,more attention has been paid to environmental issues,especially the issue of air pollution has become the focus of daily attention.As the key target of air pollution control,PM_2.5 poses a huge threat to the human body and the environment.While Hunan’s economy is developing rapidly,how to control and control PM_2.5 is a major problem.Based on the air pollution data collect from 2014.6 to 2018.5.Using the mathematical statistics method to analyze the variation characteristics of PM_2.5 concentration in Hunan province under different time series.Analysis of spatial distribution of PM_2.5 in Hunan province by spatial geostatistical analysis.The autocorrelation analysis method was used to analyze the spatial aggregation characteristics of PM_2.5 in Hunan province.Finally,the relationship between PM_2.5 and other atmospheric pollutants,meteorological factors and socio-economic factors in Hunan Province was explored by using grey correlation model and deep neural network model.The main results are as follows:（1）The average annual concentration of PM_2.5 in Hunan province continued to decline.As for seasonal variation,the concentration of PM_2.5 showed a order of winter>spring≈autum-n>summer,showing wave-like fluctuations.The peak appeared in the winter every year,and the valley appeared in the summer every year.As for monthly variation,the PM_2.5concentration in June-August was the lowest in the whole year,while the highest in December-January.The PM_2.5 concentration in September and October was greatly affected by the burning straw;For one day,the PM_2.5 concentration reached the peak at 11:00 during the day and 22:00 at night.（2）The pollution of PM2.5 in Hunan province generally showed a trend of decreasing from east to west and from south to south.The trend of PM_2.5 pollution in spring and autumn was similar,from west to east and from central to south.The northwest direction and the southeast direction decreased sideways;the PM_2.5 pollution in winter showed a patchy distribution;statistics showed that Changsha,Zhuzhou,Xiangtan and Hengyang had more pollution days,and Xiangxi prefecture had the least number of pollution days.Winter of each year,the Chang-Zhu-Tan area were seriously polluted,and the pollution of PM_2.5 in Xiangxi,Zhangjiajie and Chenzhou was light.（3）From 2015 to 2017,the global Moran’s I index is basically positive,indicating that there is spatial dependence of PM_2.5 pollution in Hunan province,that is,urban areas with higher or lower PM_2.5 concentration tend to be adjacent.The local autocorrelation analysis shows that there was spatial aggregation characteristics of PM_2.5 in Hunan province,the H-H clustering feature is Chang-Zhu-Tan area while the L-L cluster is Xiangxi.（4）PM_2.5 has a close relationship with PM₁₀.The average value of PM_2.5/PM₁₀ in Hunan province is 0.62,the ratio of PM_2.5/PM₁₀ is 0.52-0.76,Changsha has the highest ratio.The PM_2.5/PM₁₀ is rising from early morning to 08:00.The PM_2.5 content reached the highest level at 08:00,and the PM_2.5/PM₁₀ decreased during the day to night;There is a significant positive correlation between PM_2.5 and SO₂,NO₂ and CO,indicating that the increase of SO₂,NO₂ and CO concentration will aggravate PM_2.5 pollution.There is a significant negative correlation between PM_2.5 and O₃,indicating that the increase in O₃ concentration has the effect of lowering PM_2.5.There is a significant positive correlation between PM_2.5 and air pressure.PM2.5 concentration increases with increasing pressure;PM_2.5 has a significant negative correlation with wind speed.Wind speed and wind direction determine the dilution rate and transmission direction of atmospheric pollutants;PM_2.5 has a significant negative correlation with temperature.The increase of temperature is beneficial to reduce the concentration of PM_2.5 in the atmosphere;PM_2.5 has a significant negative correlation with relative humidity.The relative humidity becomes larger and the water vapor content in the air increases,which can effectively reduce the PM_2.5 concentration.PM_2.5 has a negative correlation with precipitation and is an effective way to reduce PM_2.5 concentration.Based on the neural network model analysis,it is concluded that the increase of PM₁₀,SO₂,NO₂ and CO concentration will aggravate PM_2.5 pollution.The increase of wind speed,temperature and relative humidity will reduce PM_2.5 pollution,and the similar influencing factors will be“joint”Expand the impact on PM_2.5.（5）The degree of influence of socio-economic factors on PM_2.5 concentration is:urbanization level>green coverage rate of built-up area>industrial structure>car ownership>population size>urban population density>economic development level>energy consumption>whole society electricity consumption quantity>smoke dust emissions.The influence of socio-economic factors on PM_2.5 concentration has spatial differences in different cities.