| Due to the important role of aerosol in environmental science,climate change and human health,its research has received wide attention at home and abroad.Due to the rapid economic development and accelerated urbanization in China in recent years,the air pollution is increasingly serious,and aerosol is an important part of air pollution.Satellite remote sensing technology has advance in time,space and cost,therefore it has become an important means to observe atmospheric aerosols.Aerosol optical thickness(AOD)is an important parameter in the study of the key factors of climate change and the evaluation of air quality.Polar orbit satellite Suomi-NPP launching at the end of 2011 and geostationary satellite Himawari-8 launching at the end 2014,are a new generation of earth observation satellites.They provide an complementary observation way with high time and space resolution which can meet aerosol monitoring tasks and research needs in China.This study analyzed the existing problems in the international AOD inversion products,and studied the AOD retrieval algorithm from three main aspects: cloud mask,aerosol model and surface reflection.This work develop an AOD retrieval algorithm of VIIRS and AHI,improve the inversion precision,product coverage and spatial resolution,and eventually form the operational inversion algorithm.The results of the inversion were verified by using the ground-based AOD data of AERONET and CARE-CHINA observation network,and were cross-validated with MODIS products.The results showed that the AOD product of VIIRS and AHI was good.The specific contents of the paper include the following aspects:[1] Analyze the current aerosol inversion products and point out several key problems that affect the accuracy and effectiveness of AOD inversion.The confusion of cloud/haze separation and the recognition error of pheremeral water body resulted in the lack of partial AOD results;Low accuracy of AOD products under heavy pollution condition;The influence of aerosol type selection error on inversion results;The influence of the accuracy of the misjudgment of the aerosol type.[2] As for cloud identification method for research,based on the reflectivity threshold and special variation characteristics,we develop a method which is more suitable for cloud/haze distinguish cloud recognition algorithm,is applied to VIIRS and AHI load respectively.At the same time,based on the band characteristics,the haze and dust are identified,and the prior information is provided for the subsequent inversion.[3] Take Beijing AERONET site as an example to analyze the optical characteristics of atmospheric aerosol.According to the conclusion,the information of aerosol type only divided by region and season is not sufficient to characterize its change.Therefore,based on the analysis of the data of AERONET in the condition of heavy pollution in China,two types of aerosols suitable for China's pollution condition are obtained.[4] The treatment method of surface reflectivity relation is considered for two loads respectively.AHI-AOD algorithm uses the method of band conversion to transform the mature surface reflectance relation of MODIS into the relationship applicable to AHI.VIIRS-AOD algorithm based on the assumption of the surface reflectance is constant over a period of time,with two years of apparent reflectance data for atmospheric correction,creating a griding database of the surface reflectance ratio.[5] Based on three improvements of cloud recognition,aerosol model and surface reflectance,we develop an aerosol optical thickness retrieval algorithm module for VIIRS and AHI,realizing the daily business operation.The results show that the improved AOD product has better accuracy and coverage. |
PDF Full Text Request