【作者】 贺双颜；

【导师】 贺明霞； J(?)rgen Fischer；

【作者基本信息】 中国海洋大学 ， 海洋信息探测与处理， 2011， 博士

【摘要】 东中国海以具有混浊沿岸水著称，属典型的二类水体。由于黄河、长江河水的入海，工业城市的沿海布局，使东中国海富集浮游植物、悬移质和黄色物质；蒙古和中国北部沙漠地区黄沙和陆源煤烟随大气传输入侵，使东中国海和东亚海域的大气气溶胶呈现强吸收性。因此，国外提供的海色卫星数据产品诸如ME RIs、MODIs在东中国海沿岸水域具有很大误差；中国的在轨海色传感器FY₃A／MERSI和HY₁B／cOcTs实际上缺乏可靠的数据产品。本文针对欧空局Envisat／MERIs数据，它是目前相对最佳的在轨海色传感器，利用1998—2010年东中国海海洋光学和大气光学现场测量数据集以及Mie散射和大气一海洋辐射传递模拟计算软件，对东中国海的大气气溶胶模型和海水后向散射模型进行了较深入的研究，它们是解决东中国海海色反演算法的关键问题。主要研究结果如下：1．ME RIs、MODIs大气校正算法在东中国海的评估。评估结果表明：经EsA大气校正算法获得的遥感反射比Rrs（？）在东中国海的印证平均相对误差达71％，最大相对误差高达300％。单次散射反照率ssA—Angstrom指数Ⅱ分布图被用于检验气溶胶模型的粒径分布和折射率特征，上述二者的气溶胶模型均不能涵盖东中国海的大气气溶胶特征。EsA大气校正算法在东中国海的性能总体上优于NAsA SeaI)As 6．1大气校正算法。后者的Angstrom指数Ⅱ在东中国海的性能优于前者。2．东中国海大气气溶胶光学性质。根据各类气溶胶光学厚度（？）和Angstrom指数Ⅱ的数值范围，利用AERONET和sKYNET气溶胶观测网数据库19个东中国海岸站以及中国海洋大学海洋遥感研究所海洋光学数据库c)RSI02DB 6个航次620个站位的数据，首次给出东中国海大气气溶胶（?）分类分布图，以及三种气溶胶类型的粒径分布和单次散射反照率光谱。东中国海大气气溶胶以海盐（cM）、煤烟（Bu）和黄沙（DD）共同作用的混合型为主导；煤烟的作用大于黄沙；黄沙的作用与区域和季节有关；几乎没有海盐作用为主的区域和季节。根据现场数据统计结果，给出东中国海大气气溶胶光学性质统计参数。3．东中国海大气气溶胶模型及气溶胶反射率查找表。利用气溶胶粒径分布和折射率现场数据以及气溶胶粒径双峰对数正态分布，首次给出东中国海大气气溶胶模型，包括粒径分布模型参数表和折射率参数表。利用Mie：散射计算软件，由该气溶胶模型计算气溶胶光学性质并与现场测量数据比较，二者吻合；用另一组现场测量数据作比较印证。利用德国柏林自由大学研发的MOMO大气海洋辐射传递计算软件，进一步给出适用于东中国海的大气气溶胶反射率查找表。4．东中国海海水后向散射模型利用中国海洋大学的oRSIo2DB数据库东中国海6个航次575站位的数据，选择【30xarfdl]et a1．（2009）的海中悬浮粒子散射系数b。（?）光谱模型，拟合获得适用于东中国海的b。（?）模型参数r计算公式和平均值。b。《¨模型计算值与现场测量值的相对误差分别为4．76％和5．22％，均方根误差RMStj为0．1117和0．1199。基于Morel的悬浮粒子后向散射系数（?）光谱模型，拟合给出适用于东中国海的悬浮粒子后向散射系数bu。(”光谱幂指数的计算公式和平均值。bb。(”模型计算值与现场测量值的相对误差分别为5．22％和5．24％，RMsE均为0．0016。在此基础上，给出东中国海（?）的光谱模型。进一步利用同步测量的悬浮粒子浓度sPM数据，给出sPM与bp（532）、bbp（532）、bbp（532）／bp（532）的经验关系式。更多还原

【Abstract】 East China Seas are famous for containing turbid coastal waters and are typical case II waters. East China Seas are abundant of phytoplankton, suspended particulate matter and colored dissolved organic matter due to the Yellow River and Yangtze River plumes and a coastal industrial city layout. The atmospheric aerosols have strong absorption properties resulting from the invasion of sand and land-based sources of soot transporting from Mongolia and Desert regions in northern China atmospheric. Therefore, the ocean color data products provided with overseas sensors, e.g., MERIS, MODIS, have big errors for coastal waters in East China Seas. The Chinese in-orbit ocean color sensors, i.e., FY-3A/MERSI and HY-1b/COCTS are actually lack of reliable data products.This paper is focused on Envisat/ MERIS data, which is relatively best in-orbit ocean color sensor at present. Using data sets of ocean and atmospheric field measurements during 1998-2010 in East China Seas and atmosphere - ocean radiative transfer modeling, further study on aerosol model and seawater backscattering model for East China Seas were made. The models are the key to solve the ocean color retrieval algorithm for East China Seas. The main results are listed as follows:1.Evaluation of atmospheric correction algorithm for MERIS and MODIS in East China Seas.The evaluation results suggest that the average validation error of Rrs( ) from ESA atmospheric correction algorithm in East China Seas can reach 71% and the maximum of the validation error is as high as 300%. The distributions of single scattering albedo SSA-?ngstr?m exponent were used to examine the characteristics of aerosol models, aerosol size distribution and index of refraction. The ESA and NASA aerosol models cannot represent the aerosol characteristics over East China Seas. The performance of ESA atmospheric correction algorithm in East China Seas is generally better than the NASA SeaDAS 6.1 atmospheric correction algorithm.2.Optical properties of aerosols over East China SeasAccording to the range of aerosol optical depth a and ?ngstr?m exponent for various types of aerosols, the category distribution of a– over East China Seas, and the size distribution and single scattering albedo for three aerosol types were first provided by use of data from 19 ground-based stations in East China Seas of AERONET and SKYNET aerosol-monitoring network and 620 stations in 6 cruises of Ocean Remote Sensing Institute Ocean Optics DataBase ORSIO2DB, Ocean University of China. The aerosols over East China Seas are dominated by the mixture of sea salt (CM), soot (BU) and sand (DD). The effect of soot is greater than sand, and the effect of sand is associated with areas and seasons. Little areas and seasons are dominated by sea salt aerosols. Based on the statistics of field data, the statistical parameters for aerosol optical properties over East China Seas were given.3.Aerosol models over East China Seas and aerosol reflectance look-up table.According to the field measurements of aerosol size distribution and index of refraction, as well as bimodal lognormal particles size distribution, aerosol models over East China Seas were provided for the first time, including the tables of aerosol size distribution and index of refraction. Based on the established aerosol models, the aerosol optical properties computed with Mie theory were compared with field measurements. They agreed well. Another group of field measurements were used for validation. The aerosol reflectance look-up table over East China Seas was established using atmosphere-ocean radiative transfer modeling MOMO developed by Free University Berlin, Germany.4.Seawater backscattering models for East China SeasThe scattering coefficient spectral model for suspended particulate matter bp( )published in Doxaran et al. (2009) was selected as a reference. The expression and average value of parameter for East China Seas were obtained by fit. The relative errors between the computed value of bp( ) from the model and the in-situ value are 4.76% and 5.22%, and the RMSE are 0.1117 and 0.1199. Based on Morel’s backscattering coefficient spectral model for suspended particulate matter bbp( ), the expression and average value of the power exponent in the bbp( ) model were derived by fit. The relative errors between the modeled and in-situ values of bbp( ) are 5.22% and 5.24% , and the RMSE is both 0.0016 . On this basis, the spectral model of bbp( )/ bp( ) for East China Seas was established. Using simultaneous measurements of concentration of suspended particulate matter SPM, the empirical relationship between SPM and bp(532), bbp(532), bbp(532) / bp(532) were further established respectively.更多还原