Analysis Of PM₁ Concentration Characteristics In China And Its Correlation With Satellite Aerosol Optical Depth

PM₁ is an important factor causing air pollution,and its pollution has caused great harm to environmental quality and human health.At present,domestic research on PM₁ pollution concentration is mostly concentrated in a short time and small sample area,and it is impossible to macroscopically understand the current status of the temporal and spatial migration and diffusion of atmospheric pollution,and the correlation between PM₁ concentration and satellite AOD data on a fine national scale There are few analyses of the temporal and spatial variation laws.In addition,there is still a problem of low accuracy in the PM₁ concentration estimation method.To this end,this paper uses the PM₁ mass concentration data of all ground air quality monitoring stations in the country from 2014 to 2017 as the data source,and explores its distribution characteristics in time and space based on time series statistical methods and spatial system clustering methods,revealing its The law of time and space evolution.Combined with the MODIS remote sensing satellite AOD data,the spatial and temporal changes of the correlation between the two are explored on a fine scale across the country.In addition,using the revised"dry"extinction coefficient,meteorological data,geographic coordinate data and ground PM_2.5 concentration data,the method of modeling and estimating PM₁ concentration based on linear and random forest algorithms was explored.The main research conclusions are as follows:（1）From 2014 to 2017,the annual average PM₁ concentration in China decreased year by year.The PM₁ concentration in the four seasons showed a"high in winter and low in summer".Type change characteristics,PM₁ concentration is a high value point on Monday and Friday during the week,and a low value point on Sunday.Based on the spatial system clustering method,the average annual PM₁ concentration value in the country is divided into seven categories.The highest annual average PM₁concentration in central China is 54.592μg·m^-3,and the lowest annual average PM₁concentration in New Qinghai-Tibet is 11.375μg·m^-3."High west and low,north high and south low"spatial distribution pattern.（2）The PM₁-AOD relationship is positively correlated as a whole.The highest correlation coefficient is 0.55 in central China,and the lowest correlation coefficient is 0.36 in central and southern China.Seasonal PM₁-AOD correlation coefficient changes are affected by humidity,and its correlation coefficient is negatively correlated with atmospheric humidity.However,the PM₁/AOD ratio and the slope or intercept of the regression equation vary in different regions in different seasons:the minimum value in summer and the maximum value in winter.（3）The PM₁-AOD relationship differs from the correlation results of individual monitoring stations on a regional scale.When the correlation coefficient on the regional scale is higher than the average of the correlation coefficients of each monitoring station,it indicates that the PM₁ in this region is consistent with the source of AOD.The regional correlation coefficient between the New Qinghai-Tibet region and the Neganning region is significantly higher than the average correlation coefficient of its monitoring stations,indicating that the urban pollution in the region is similar.（4）Based on the"dry"extinction coefficient and meteorological and geographic data obtained from the vertical correction and humidity correction of AOD data,a random forest algorithm is used to build a PM₁ and AOD relationship model.The modeling accuracy R² is 0.96,the RMSE is 3.13,and the verification accuracy R² is0.63 and RMSE is 7.07.