Research On PM_2.5 Concentration Inversion And Spatio-temporal Variation In The Yangtze River Economic Belt

In recent years,due to the rapid pace of industrialization and urbanization,there has been an increase in the frequency of haze events in China’s major cities.Such events are primarily caused by the accumulation of particulate matter 2.5(PM_2.5),which are small in diameter and mass,and can persist in the atmosphere for extended periods,and can even be transported over long distances.Because of its harmful impact on human health,it is crucial to monitor the spatio-temporal variations in PM_2.5 concentration.Since 2013,the number of air monitoring stations in China has increased significantly,but the majority of these stations are concentrated in urban areas,with fewer being distributed in suburban and rural areas.This lack of distribution can result in some errors in studies.Therefore,it is necessary to adopt more effective methods for monitoring PM_2.5 concentrationsBy integrating PM_2.5 monitoring data,aerosol optical depth（AOD）data,ECMWF Reanalysis v5（ERA-5）meteorological data,and related auxiliary data,a random forest regression（RF）model was used to fit and generate PM_2.5 concentration raster data for each month from 2015 to 2020.The spatio-temporal evolution of PM_2.5 concentration in the Yangtze River Economic Zone was investigated by integrating the kernel density curve,statistical analysis,Theil-Sen Median slope estimation and Mann-Kendall significance test,as well as global and local Moran’s I.The findings are presented below:（1）An improved linear regression data integration method was utilized to effectively integrate Terra and Aqua AOD data.Subsequently,missing values were filled in using ordinary Kriging interpolation,leading to monthly Multi-Angle Implementation of Atmospheric Correction AOD（MAIAC AOD）data from 2015 to 2020.（2）A random forest regression model was developed using PM_2.5monitoring data,MAIAC AOD,ERA-5 meteorological data,and relevant auxiliary variables.The optimal training and prediction performance of the model was achieved with n＿estimators of 400,max＿depth of 20,and max＿features of 0.5.The inversion result provided a dataset of PM_2.5 concentration in the study area at a 1 km resolution on a monthly basis from 2015 to 2020.（3）The downstream areas of the study area had the highest PM_2.5 concentrations,while the upstream areas had the lowest concentration.Furthermore,there was an overall decreasing trend of PM_2.5 concentration in the study area during the research period.Regarding temporal variation,the concentration exhibit a U-shaped temporal trend throughout the year,with initial decrease and subsequent increase in mean values of concentration across different months.Moreover,in different seasons,the concentration showed a pattern of being high in winter,low in summer,and moderate in spring and autumn.（4）The study reveals distinct spatio-temporal variations and agglomeration features of PM_2.5 concentrations at monthly,seasonal,and annual scales,all exhibiting a predominant downward trend.Moreover,PM_2.5 concentration agglomeration types,including high-high,low-low,and low-high clusters,were identified at different temporal scales.Low-low agglomeration regions were predominantly located in the southwestern part of the study area,while high-high agglomeration regions were primarily situated in two urban clusters in the northeastern region.This study provides a comprehensive analysis of the spatio-temporal trends in PM_2.5 concentration in the Yangtze River Economic Belt,which can improve our understanding of the evolution of PM_2.5 pollution in the region.The identification of high-pollution areas can inform the development of effective policies for joint prevention and control of PM_2.5 pollution in the region.