Air Pollution Transport Analysis And Analysis In Northern Guangxi Based On Spatial Analysis Technology Forecast Research

In recent years,the tourism cities in the south of China with low industrialization level also had serious air pollution.Guilin City was taken as an example.The tertiary industry accounted for 56.9%of the city’s GDP,and the per capita GDP was close to that of Wuzhou and Chongzuo.Although Guilin had lower pollutant emission than these two cities,its atmospheric environment quality was obviously poorer,among which the total PM_2.5 of Quanzhou County,Pingle County,Yongfu County and urban area was on the high side as a whole.For this phenomenon,the deep research and analysis would be conducted gradually in this dissertation.In order to explore this phenomenon,it was found that the local pollution in Guilin was not enough to cause serious environmental pollution.That was to say that,the environmental pollution in Guilin also may be closely related to external pollution sources except local pollution sources.As is well-known,Guilin is located at the junction of Yongzhou City and Shaoyang City in Hunan Province,which are highly industrialized cities.In the south,Guilin is adjacent to Liuzhou City,a heavy industrial city.In the west,north and southeast,Guilin has high terrain.In the northeast,there was a long and narrow corridor formed by the mountain systems of Yuechengling and Dupangling,namely the"Xiang Gui Corridor".According to this topographic feature,it was presumed that the external environmental pollution of Guilin City may transmit along the"Xiang Gui Corridor",which caused pollution to the environment of Guilin.The environmental pollution in Guilin was closely related to the internal topography and climate factors.Based on the above preliminary analysis,firstly,the HYSPLIT model was used to study and analyze the transmission direction of Guilin’s atmospheric external environment.Secondly,the spatial interpolation method was used to determine the distribution of internal pollution sources,and the distribution characteristics of internal and external sources were analyzed in combination with the external and internal pollution sources.Finally,non-parametric analysis and multiple regression analysis were utilized to study the influence of meteorological factors on pollutant concentration,and the ANN neural network model was employed for prediction.Tracing the source and transmission direction of regional air pollution,locating and quantifying the potential source areas of air pollutants,and analyzing the transmission path and concentration distribution of pollutants in combination with topography can better provide scientific basis for the prevention and control of air pollution.The main conclusions were made as follows.1.The most serious pollution happened in winter in Guilin,followed by autumn and spring,the lightest in summer.The heaviest month was January,and the lightest was June.2.The simulation clustering of the air masses in the whole year showed that the trajectory in the northeast direction（2）accounted for 63.09%,which was the main source direction of the air masses,and the average pollution concentration of PM_2.5 and PM₁₀ was 45.94μg/m³ and 67.78μg/m³ respectively,so the direction of trajectory（2）was the main source direction of the external pollutants.3.The analysis of the four pollution processes in winter in Guilin showed that the distribution of the main trajectory lines of air masses was consistent with the"Xiang Gui Corridor".By spatial interpolation,it was found that there were weakening zones in Guanyang County,Xing’an County and Ziyuan County,which hindered the air masses from directly reaching the urban area of Guilin,so it made the concentration of PM_2.5 and PM₁₀ different in time and space.4.The spatial statistical method found that PM_2.5and PM₁₀ concentrations were spatially correlated and there was an aggregation trend.The main influencing factor was wind speed.The prediction accuracy of the concentration of PM_2.5 and PM₁₀ via the neural network model was 71.7%and 75.6%respectively.