| The rapid economic development and the advancement of urbanization have caused the concentration of fine particulate matter PM2.5 in the atmosphere to continue to increase,causing serious air pollution and adversely affecting people’s health,life and production.The research on PM2.5 pollution and prevention is an important task in the current environmental protection and ecological civilization construction.The Yangtze River Delta region is the most economically developed region in China,and many environmental pollution problems have arisen in the process of economic development.In this context,conducting research on the estimation of PM2.5 concentration in the Yangtze River Delta region and analyzing the temporal and spatial characteristics of PM2.5 concentration are of great significance for the further control of air pollution and the protection of the ecological environment.Based on the monthly PM2.5 monitoring data,night light data,weather data,elevation data,and population density data from 2015 to 2019,this thesis selected the Yangtze River Delta region as the research area and simulated the monthly PM2.5concentration from 2015 to 2019 by establishing multiple linear regression models,BP neural network models and Random forest model.By comparing the simulation effects of different models,the model with the highest relative accuracy was selected to estimate the PM2.5 concentration.Based on the estimated results,the regular pattern and characteristics of PM2.5 concentration changes in the Yangtze River Delta were analyzed from two scales of time and space.The results of the study are as follows:（1）In the simulation of the monthly PM2.5 concentration,the random forest model has the highest simulation accuracy（R2=0.6604）and the best effect.（2）From the time scale,the monthly maximum value of PM2.5 data was114.06ug/m3,and the minimum value was 11.80ug/m3.The maximum quarterly value was 88.11ug/m3,and the minimum value was 13.68ug/m3.The maximum value of the annual average PM2.5 data was 67.34ug/m3,and the minimum value was 35.16ug/m3.In addition,the concentration of PM2.5 showed a downward trend on three scales:monthly,quarterly,and annual.From a spatial scale,the PM2.5 concentration in the Yangtze River Delta region generally presented a spatial pattern of"high in the north and low in the south",and the low PM2.5 concentration area was gradually expanding and spreading northward.Areas with high PM2.5 concentrations were concentrated in the areas on both sides of the Yangtze River and were also shrinking.（3）From 2015 to 2019,the degree of dispersion of PM2.5 concentration in the Yangtze River Delta region decreased,gradually shifting to a few value intervals and converging to a low value interval.For the cities at all levels,the PM2.5 concentration showed a downward trend.In addition,the PM2.5 concentration in each city showed similar seasonal variation characteristics,that is,the PM2.5 concentration was lowest in summer and highest in winter.Compared with prefecture-level cities,Wuxi,Taizhou and Zhenjiang had the highest contribution to PM2.5 concentration in the Yangtze River Delta.（4）The PM2.5 concentration in the Yangtze River Delta had a significant spatial autocorrelation relationship.The high-high-type agglomeration area and the low-low-type agglomeration area are in"north-south opposition",and the area of the former is shrinking,while the latter continues to spread northward.The above analysis shows that from 2015 to 2019,the PM2.5 concentration in the Yangtze River Delta had been continuously reducing,and the air quality had been continuously improved.This shows that the air pollution control in this area has achieved good results. |
PDF Full Text Request