Algorithm Research Of Global Aerosol Optical Depth Inversion

Accurate acquisition of atmospheric aerosol information is not only of great significance for climate change research and atmospheric environmental monitoring,but also an important parameter for quantitative remote sensing.While retrieving aerosol optical depth(AOD)from remote sensing imagery,the selection of aerosol model is one of the key factors affecting the accuracy of AOD.Existing mainstream AOD products use aerosol model information obtained from global ground observations as a priori knowledge to pre-set aerosol patterns.Due to the rapid change of aerosols in the atmosphere,the pre-setting of aerosol models often results in considerable errors and cannot respond well to sudden atmospheric changes.Therefore,a real-time and dynamic aerosol model based on multi-source data is developed in this paper.According to the variation of apparent reflectance in time series,the surface reflectance can be simulated,and the aerosol optical depth can be subsequently retrieved continuously with high accuracy globally.(1)The sensitivity of POLDER multi-angle polarization sensor to fine particles can be used to obtain the scale information of particle including radius and standard deviation;the complex refractive index of the whole atmosphere is obtained by OMI near-ultraviolet sensor which is sensitive to absorbing particles.The aerosol model is consequently constructed by setting up the parameters retrieved from POLDER and OMI at MIE scattering module of 6S radiation transfer model.(2)MODIS sensor has relatively high temporal and spatial resolution,and its spectral information has a lot of advantages for AOD retrieval.Therefore,MODIS data is chosen to obtain surface information in this paper.A time series algorithm is employed to simulate global surface reflectance.Combined with the aerosol model,the 6S model is used to solve the atmospheric transfer equation,and the accurate apparent reflectance observed by MODIS is interpolated to retrieve AOD.(3)The results of AOD obtained by the method in this paper are compared with the measured data of AERONET site,the results using continental aerosol model and MODIS aerosol products.The results show that the proposed method is more accurate than that using continental aerosol model;furthermore,the proposed method can respond to sudden atmospheric changes much better.Compared with MODIS products,the accuracy of the proposed method is similar in regions where the atmosphere is quite stable and it is much better than the MODIS AOD product in regions where the atmosphere changes dramatically.Although the proposed method is inferior to MODIS products in Europe where the environment is good and the atmosphere is stable,the proposed method has obvious advantages while comparing the effective number of AOD.