Development Of Hourly AOD Dataset Based On Geostationary Satellites And Fusion Of Multi-source AOD Datasets

Aerosols can affect global climate in both direct and indirect ways and can also cause a series of environmental problems such as the increase of fine particles in the atmosphere and the decline of atmospheric visibility.In addition,aerosols have a negative impact on the accuracy of satellite sensor calibration and quantitative remote sensing.Aerosol optical depth(AOD)is the parameter used to estimate the quantity of aerosols,which is the vertical integration of the aerosol extinction coefficient from the ground to the top of the atmosphere.To accurately assess the effects of aerosols on climate,atmospheric environment,and quantitative remote sensing,we must first obtain an accurate aerosol distribution.However,aerosol distribution has a strongly spatiotemporal difference due to the short life cycle of aerosols and the spatial variability of aerosol emissions.Therefore,it is very important to accurately monitor the global distribution of AOD at high frequencies.The aim of this study is to obtain more complete and accurate aerosol datasets by developing global hourly AOD data sets based on geostationary satellites and fused AOD data sets,which can provide more complete and accurate data support for climate change,environmental monitoring,atmospheric correction and other research.In the work of developing global hourly AOD data sets based on geostationary satellites,we developed two aerosol retrieval algorithms,including the two-channel algorithm developed for SEVIRI and the optimal estimation algorithm developed for ABI and AHI.Then,we obtain hourly global aerosol optical depth dataset by integrating AOD datasets retrieved from four geostationary weather satellites(GOES-16,MSG-1,MSG-4 and Himawari-8).The integrated geostationary satellite AOD datasets from April 2018 to August 2018 are validated using AERONET data.The validation results are follows: the Mean Absolute Error(MAE),Mean Bias Error(MBE),Relative Mean Bias(RMB),and Root Mean Square Error(RMSE)are 0.07,0.01,1.08 and 0.11,respectively.The ratio of the error of satellite retrieval within ±(0.05+0.2AODAERONET)is 0.69.As a representative of polar orbit satellites,the spatial coverage and accuracy of MODIS C61 AOD product released by NASA are also analysed.The analysis results show that the integrated AOD dataset has similar accuracy to that of the MODIS AOD dataset and has higher temporal resolution and spatial coverage than the MODIS AOD dataset.In the work of developing fused AOD data sets,we proposed a fusion algorithm that can fully consider the systematic error and uncertainty of AOD data sets at the pixel scale.The fusion algorithm consists of three parts: the first part is to remove the systematic errors;the second part is to calculate the uncertainty and fuse multi-source AOD datasets based on the corresponding uncertainty data sets;and the third part is to mask outliers with a threshold of 0.12.The algorithm has been successfully applied to the fusion of three AOD data sets of ATSR-2/ATSR dual-view aerosol retrieval algorithm(ADV),the Oxford-RAL Retrieval of Aerosol and Cloud algorithm(ORAC)and the Swansea algorithm(SU).The spatial coverage of fused AOD dataset masked with a threshold of 0.12 is 148%,13% and 181% higher than ADV,ORAC and SU respectively.The fused AOD dataset in 2009 is validated using AERONET data.The validation results are follows: the MBE,MAE,RMB,and RMSE of the fused data after the mask operation with a threshold of 0.12 in 2009 are 0,0.05,0.97,and 0.07 respectively,which are superior to the three original datasets.Therefore,we can conclude that our fusion method can effectively improve the spatial coverage and the precision of the data.