Remote Sensing Retrieval Of AOD/SSA Based On AHI/TROPOMI

Aerosol in the atmosphere is the key factor leading to the uncertainty of climate change and the implementation of air pollution monitoring and control.With decades of rapid industrialization and urbanization,China has become a significant high-value area of the global Aerosol Optical Depth(AOD).The currently widely used remote sensing AOD products are mainly from polar orbit satellites,with limited observation frequency,which is difficult to satisfy the dynamic monitoring needs of the management of the atmospheric environment.At the same time,compared to the total amount of aerosols,the proportion of anthropogenic aerosols is getting more and more attention: its key parameter,Single Scattering Albedo(SSA),directly affects the positive and negative levels of radiative forcing.However,due to the complex aerosol properties and the limited spectral bands of many previous satellite sensors,there are few satellite remote sensing products about SSA.Therefore,based on the geostationary satellite Himawari-8/AHI and the latest hyperspectral satellite Sentinel-5P/TROPOMI of atmospheric composition,the following research works have been carried out in this study:(1)According to the three versions of Himawari-8/AHI level-3’s hourly aerosol AOD products of AHI JAXA,cross-validation analysis is conducted with the measurements from 11 ground-based sunphotometer sites and MODIS C6 AOD products in eastern China.The results show that the V010 AOD can effectively capture the hourly variation of aerosol,with serious underestimation(RMSE>0.3).After adjustment,the underestimation has been basically solved in the latest V030 product(slope is close to 1),but its accuracy is significantly different at each site(R: 0.73-0.91),and the overall performance is not as good as MODIS product,which still needs improvement.(2)An improved time series(ITS)algorithm is applied to AOD retrieval of the AHI sensor.Based on the core strategy of K-Ratio invariance,this algorithm can make full use of the high time resolution characteristics of the geostationary satellite.According to the statistical analysis of the long-term ground-based AERONET observations,the aerosol type suitable for China is reconstructed for the ITS algorithm,then multiple AODs per hour can be retrieved.The cross-comparison results with ground-based observation data and MODIS AOD data show that the ITS derived AODs have a significant improvement than the AHI JAXA Level-2 products(R> 0.8,RMSE <0.2).(3)A fast SSA inversion algorithm based on Gradient Boost Regression Tree(GBRT)model is proposed.Based on the high sensitivity of UV spectral band to absorbing aerosols and the advantage of same bands for TROPOMI and OMI,this study takes OMI UV band products(Ultraviolet absorbing index(UVAI),SSA,AOD,etc.)as the training dataset,and constructs the prediction model through the quality control of AERONET measurements to obtain high-resolution SSA products by TROPOMI UVAI data,this algorithm has high efficiency.(4)A joint retrieval algorithm by multi-source satellite data based on a radiative transfer model is proposed.This study first uses the Optical Properties of Aerosols and Clouds(OPAC)aerosol model combined with the AERONET observation data of eastern China to predefine an aerosol model,and reconstruct the shape of the vertical aerosol profiles based on the ground-based Lidar observations in China,then a look-up table(LUT)was established to retrieve SSA.Compared with the observations from SONET and other satellite products,our results have a good consistency and high accuracy.There are 87 figures,16 tables,and 195 references in this dissertation.