Research And Implementation Of 3D Visualization System For Satellite Remote Sensing Of PM2.5 In Haze

With the rapid development of global industrialized economy and the acceleration of urbanization in recent years,air pollution has become increasingly serious.The air pollution control led by PM2.5 is extremely urgent.The remote sensing method is used to invert the near-surface PM2.5 mass concentration and obtain its three-dimensional spatial and temporal characteristics,so as to explore the three-dimensional spatial and temporal transport law of haze aerosol and provide theoretical and technical support for the government to conduct joint regional governance and pollution control of haze.This paper takes zhengzhou city of henan province as the research area to carry out relevant research,and mainly obtains the following research results.1.Study on the contribution rate of AOD or PM2.5 by relative humidity and vertical elevationBased on the CALIPSO satellite lidar level 2 aerosol data products from December 2013to November 2014 in Zhengzhou City,the inversion models of near-surface PM2.5 and AOD are established by introducing aerosol humidity correction and vertical elevation correction methods,and the optimal inversion models was selected.Based on this model,the contribution rate of relative humidity and vertical elevation to PM2.5 or AOD is quantitatively studied by cumulant slope rate comparison method for the first time.Based on the winter period,without considering influence factors such as temperature,precipitation,wind speed,wind direction during the inversion PM2.5,relative contribution rate of relative humidity of PM2.5 in the spring and autumn period were 31.60%and 48.40%respectively,the vertical elevation of AOD relative contribution rate in spring and autumn period were 72.28%and 40.23%respectively,the relative humidity of AOD relative contribution rate in spring and autumn period were 24.59%and 26.23%respectively,the research results provide a theoretical basis for further improving the inversion accuracy of PM2.5.2.A spatio-temporal adaptive PM2.5 inversion model is proposed for the first timeBased on the geographically weighted regression?GWR?model,a four dimensional geographically and temporally weighted regression?4D-GTWR?model was proposed by introducing DEM and time factors,which solved the problem of four-dimensional spatio-temporal non-stationarity in PM2.5 inversion.Based on the PM2.5 near-ground observation data of Zhengzhou City,Henan Province from December 2017 to February 2018,combined with meteorological factor data and MYD04₃K AOD products of MODIS remote sensing data,the inversion of PM2.5 was studied by using GWR model,time weighted regression?TWR?model,geographically and temporally weighted regression?GTWR?model and 4D-GTWR model.The results showed that the 4D-GTWR model R²?0.9496?showed better goodness of fit than GWR model R²?0.7761?,TWR model R²?0.7763?and GTWR model R²?0.8811?.3.Based on component GIS,a 3d visualization system for satellite remote sensing of PM2.5in haze was developedThe three-dimensional kriging spatial interpolation algorithm was studied,and a unified nested model was established to realize the three-dimensional spatial distribution expression of PM2.5 in each layer.Based on the Arcgis Engine secondary development platform,the PM2.5three-dimensional visualization system was constructed to realize the three-dimensional visualization of PM2.5 in haze and obtain the PM2.5 mass concentration in the three-dimensional space.The innovative points and features of this paper are:A spatio-temporal adaptive PM2.5 inversion model,the four dimensional geographically and temporally weighted regression?4D-GTWR?model,was proposed for the first time to improve the inversion accuracy and realize the three-dimensional spatial distribution expression and visualization of PM2.5 in haze.