【作者】 林绍迎；

【导师】 高会旺；

【作者基本信息】 中国海洋大学 ， 环境科学， 2007， 博士

【摘要】 黄海是西太平洋上一个半封闭的陆架浅海,陆源影响严重;在黄海环流系统的作用下,悬浮物质分布具有显著的区域性与季节性特征,绝大部分海域属典型的二类水体。本文首先利用SOLAS第一航次在整个黄海的现场观测资料,分析了研究海区光合有效辐射PAR的衰减特征及PAR在水体中衰减的影响因素;然后在观测资料的基础上,结合黄海海区生物光学模型的历史数据,利用水体中的光传输模式AOMC,进一步模拟了黄海不同海区光在水体中的衰减特征,探讨了影响光传输的主要衰减因子及其与光学特性量的关系,具体结论如下:（1）黄海海区生物光学特征:①黄海海域的浊度与叶绿素a浓度分布具有明显的区域性差异,浊度在黄海北部成山头附近的近岸水域为高值区,黄海中部与南部离岸水域为低值区;叶绿素a浓度[Chl-a]在整个海域较低,基本上随深度分布均匀,黄海南部水体中的[Chl-a]要高于北黄海,近岸水体高于离岸水体。②浊度反映水体中总悬浮颗粒物的含量,与叶绿素a浓度的相关不明显,表明非藻类的悬浮粒子是引起浊度增大的主要贡献来源。③在真光层深度以上,光合有效辐射PAR随深度以及积分浊度IT的衰减呈指数函数关系,在大多数站位的相关性显著（R2≥0.98）;线性回归与指数回归法分别计算的KT（PAR）和KO（PAR）两者一致性较好,除了在离岸深水区的Gp站,相对差△=12.3%,其他站位的△均低于10%,表明PAR随深度衰减的指数率模式更适用于近岸浅水区。④PAR衰减系数K_d（PAR）存在分层结构,分别为近表面2m处的极大值层与20-30m处的渐进层。⑤不同站位影响光衰减的主要因子不同,具有一定的区域性特征。K_d（PAR）在大多数站位与[Chl-a]无明显相关性,随浊度的增加而增大,呈一定的正相关性,说明这些水体中的浮游植物不是影响光传输的主要衰减因子,主要光衰减因子为非藻类的悬浮粒子;而在某些近岸站,K_d（PAR）与浊度、叶绿素a浓度的相关均不明显,影响光传输的因子复杂。（2）利用水体光传输模式AOMC,在进行准确性必要验证的基础上,对黄海海区的光学衰减特征进行了模拟,研究结果表明:①在多数站位,AOMC模式模拟计算的平均PAR衰减系数K_dM（PAR）与实测数据计算的平均PAR衰减系数K_dO（PAR）具有很高的一致性,相对差控制在20%。②K_dM（PAR）随深度的变化规律与观测结果K_dO（PAR）相似,出现分层结构;近表层（1-4m）,模拟值与观测值差别较大,在4m深度以下到真光层深度处,多数站位的K_dM（PAR）值与K_dO（PAR）值接近,差别很小,这是因为在表层4m以上PAR的观测值受船体阴影以及表面波浪的影响较大,从而在计算K_dO（PAR）时产生了一定了误差,造成模拟值与观测值差别较大,在4m以下,PAR衰减系数K_d（PAR）受环境光场的影响较小,受水体中成分的影响较大,因而模拟值与观测值接近。③指数率模型反演PAR值与实测PAR值在各深度上比较一致,相对差最大不超过45%,表明指数律模型中使用的K_dM（PAR）准确性较高。④光从大气中传输进入悬浮粒子含量高的黄海水体中时,下行辐照度的光谱性质发生改变,随水深增加,辐照度光谱峰值从大气中的蓝光区右移到了水体中的绿光区。⑤模式的敏感性分析表明,非藻类的悬浮粒子是影响黄海海区光传输的主要因素,非藻类粒子对海水中光的衰减作用相当于其他物质作用的总和,在PAR衰减系数中占53.3%,其次是浮游植物。⑥非藻类的悬浮粒子对光谱的影响主要集中在波长小于600nm的短波区间,与其他光传输影响因子相比,非藻类粒子的物质种类与浓度变化更容易影响光的传输特征,从而导致水体中的光量与光场性质的改变,对海洋生态系统产生深远影响。⑦黄海的光学衰减特征具有显著的区域性差异,在同一深度处,北黄海的光谱辐照度值低于南黄海,近岸水域低于离岸水域,PAR衰减系数的平均值K_d（PAR）,在北黄海最大（0.298m-1）,依次是近岸水域(0.276 m^-1)、离岸水域(0.201 m^-1),南黄海最小(0.175 m^-1),在一定程度上反映了影响光传输主要因子的分布特征。⑧目前生态模式中使用的光学模型将K_d（PAR）简单表示为纯水与浮游植物的衰减系数和,该光学模型在非藻类粒子是光衰减主要因子的黄海中不适用,本文利用AOMC模式并结合实测资料,建立了浊度与K_d（PAR）的简单参数化表达式,为光学模型在海洋生态模式中的应用提供了参考。更多还原

【Abstract】 The Yellow Sea is a semi-enclosed shallow sea on the continental shelf of the West Pacific. Large quantities of terrigenous substances are deposited into the sea by several rivers, such as Huai, Han, and Yalv rivers etc., and accordingly it has come to be seen as a sea area containing the most suspended materials in the world. Under the dynamic process of circulation system, the distribution of the suspended particulate matters has distinct regional and seasonal characteristic. As a result, optical properties of the Yellow Sea are relatively complex and the majority of it belongs to typically Case II waters.The optical properties, including planar PAR （Photosynthetically Available Radiation,400-700nm）, scalar PAR, turbidity and concentrations of chlorophyll a （[Chl-a]）, were measured across the Yellow Sea during a cruise of the China SOLAS from 19 to 27 March 2005. Based on these data, this paper first analyses the vertical attenuation characteristic of PAR and the factors that affect light attenuation in the water. Then an aquatic optical radiative transfer model （AOMC） is used to simulate the light attenuation characteristic in different sea areas of the Yellow Sea. Relationship between the main light attenuation factors and the apparent optical properties is studied. Main conclusions are as follows:（1） Bio-optical properties of the Yellow Sea:①The distribution of turbidity and the concentration of Chlorophyll a [Chl-a] has distinct regional difference in the sea area. For turbidity, the high value area appeared at inshore stations near Cheng Shantou which located in the northern Yellow Sea and the low value area are those offshore stations located in the middle and southern Yellow Sea. [Chl-a] is lower over the study sea areas and distributes uniformly along the water column. [Chl-a] is higher in the south Yellow Sea than in the north Yellow Sea and so is it in the near shore water than the offshore water.②Turbidity reflects the content of total suspended substances and it has no obvious relationship with [Chl-a] which indicate that non-algal suspended particles contribute most to the turbidity.③Above the euphotic depth, PAR decreases exponentially with both depth and integrated turbidity with high correlation coefficients （R2≥0.98） at most stations. The diffuse attenuation coefficient of PAR, K_d （PAR） is calculated by two kinds of regressive methods, denoted as KT（PAR） and KO（PAR）. There is good agreement between KT（PAR） and KO（PAR） at all stations with the relative error△lower than 10% except offshore station Gp station which has a relative high△=12.3%. The result shows that the exponential law is more applicable to inshore shallow sea waters.④Two K_d （PAR） layers are found at the offshore stations, one layer has a maximum value appeared in the surface layer and the second layer, ranging from 20 to 33m, K_d （PAR） approaches an asymptotic.⑤The main factors causing light attenuation are different between stations which show some regional characteristic. At most stations K_d （PAR） has no coherent relationship with [Chl-a] but increase with turbidity which implied that not the phytoplankton but the non-algal particles are the primary light attenuation factors. But at some nearshore stations there is no obvious relationship between K_d （PAR） and turbidity and [Chl-a], and accordingly the controlling factors are complex in inshore waters.（2） It is difficult to study the light attenuation characteristic over the whole sea area with full observed optics data during different season. But the problem can be largely solved by numerical models without limit of time and place. AOMC model developed on the basis of Monte Carlo method. After the necessary validation of exact degree, it is used to study the light attenuation characteristic of the Yellow Sea. The results are showed below:①at most stations, the depth averaged PAR attenuation coefficient computed by AOMC K_dM（PAR） has a high agreement with that calculated by the observed data. The relative errors between K_dM（PAR） and K_dO（PAR） are within the range of 20%.②The vertical distribution of K_dM（PAR） with depth has come to be the layer structure which is similar to the shape of K_dO（PAR）. In the near surface layer （1-4m）, the modeled values have large differences with the measured value. From 4m down to the euphotic depth the differences between modeled value and measured value are relative small. PAR is influenced by ship shadow and surface waves when measured above 4m which causes the calculation error of K_dO（PAR） and results in the large difference. The environmental factors have little influence on K_d （PAR） below 4m, so the modeled value approaches to the measured value.③The inversion value of PAR derived by an exponential model has small difference with the measured PAR and the lager relative error is less than 45% which indicates that the K_dM（PAR） used in the exponential model has a high degree of exact.④When light propagates from the atmosphere to the ocean which contained a high quantity of suspended particles, the spectrum property of the downwelling irradiance is changed. The spectrum peak value shifts rightward from the blue region to the red region.⑤Sensitive study shows that the non-algal particles are the main factor influencing light propagation in the Yellow Sea and phytoplankton is the secondary factor. Light attenuation by the particles accounts for about 53.3% of average K_d（PAR） and phytoplankton accounts for 27.4%.⑥The non-algal particles have great effect on the spectrum less than 600nm. Compared with other factors, changes in the composition and concentrations of the non-algal particles can more easily alter the light attenuation characteristics, and accordingly have a profound effect on the marine ecosystem.⑦Light attenuation characteristic of the Yellow Sea has distinct regional differences. At the same depth the irradiance was lower in the north Yellow Sea than in the south Yellow Sea, lower in the inshore water than in the offshore water.⑧At present, the optical model used in the marine ecosystem model expresses K_d（PAR） as the sum of light attenuation by water Kw（PAR） and phytoplankton Kph（PAR） which is no longer applicable to the Yellow Sea. On the basis of the AOMC model simulation results and the observed data in the Yellow sea, an simple relationship between turbidity and K_d（PAR） has been derived which can serve as a referenced parameterized optical model for the marine ecosystem model.更多还原