基于Argo资料的西太平洋混合层和温跃层数据产品研制

基于2017版全球海洋Argo网格数据集(BOA-Argo),利用最大角度法和梯度比值法等客观分析方法计算了2004年1月—2016年12月期间,西太平洋海域(25°S~40°N,120°~180°E)的上混合层和温跃层上、下界深度,并计算了混合层温盐度以及温跃层强度等海洋环境参数,制作完成水平分辨率为1°×1°的月平均Argo数据衍生产品。将本数据产品和采用阈值法计算得到MILA GPV数据集做比较,结果显示:对于混合层的主要空间分布特征和时间序列变化特征,两者都十分吻合;将西太平洋海域温跃层上、下界深度和强度等参数与人们利用传统的温度梯度法计算结果相比较,其季节分布特征及变化趋势也大体相符。     

一种海洋混合层深度的智能识别方法研究

文章提出了一种识别混合层深度的人工智能方法。该方法在温度(密度)与压强(或深度)间建立线性模型, 并且将其系数和方差做成一组表征廓线特征的统计量。初始时为模型设定一个主观的先验分布, 在一个自海表向下移动的窗口内通过贝叶斯链式法则和最小描述长度原理学习新数据, 得到系数均值的最大后验概率估计。用F-检验识别系数发生突变的位置, 以此确定混合层的存在性及其深度。通过2017年2月太平洋海域的地转海洋学实时观测阵(Array for Real-time Geostrophic Oceanography, ARGO)数据进行测试, 并且以质量因子(Quality Index, QI)值作为判断识别混合层深度结果准确性的依据, 发现该方法相比于梯度法、阈值法、混合法、相对变化法、最大角度法和最优线性插值法在识别结果上具备更大的QI值。表明该方法能够准确识别混合层深度。     

基于观测资料的海浪与混合层深度相关性分析

从观测数据角度出发,考察海浪与上层海洋混合层深度的变化关系。采用卫星高度计和三套温度观测数据,利用改进的混合层深度提取方法,获得海洋混合层深度。简要分析了多年月平均的有效波高和混合层深度的空间分布特征及时间变化规律,并进一步分析了它们的相关性。二者直接相关性分析的结果表明,在南北半球的中纬度地区二者的相关系数较大,而赤道地区较小。滤除年周期的气候态月平均场后,计算的距平相关系数在赤道区域较小;但在太平洋东部、南部和南印度洋存在一个大值区。此外,进一步研究了有效波高和混合层深度年际距平的相关系数,其空间分布特征与二者的距平相关系数的分布特征类似。为探究混合层深度的影响因素,同时也分析了风场与混合层深度的相关系数。综合上述结果,海浪和上层海洋的混合层深度之间存在着一定的相关性,海浪过程是风输入能量向次表层海洋传播的一个重要途径。     

基于2007—2018年Argo数据分析全球混合层和障碍层时空特征

为了进一步认识上层海洋中混合层和障碍层的时空变化特征。本文基于Argo (Array for real-time geostrophic oceanography)海洋观测网2007—2018年的温盐数据,使用差值法计算了全球海洋混合层深度(Mixed layer depth, MLD)和障碍层厚度(Barrier layer thickness, BLT),讨论了二者的月均值、季节均值和年均值的空间分布特征和形成机制。研究表明,全球海洋的混合层普遍在夏季浅、在冬季深,随季节变化的特征显著。北半球混合层变化幅度较大,大西洋混合层比同纬度的太平洋深;赤道海区混合层较浅;南半球混合层呈纬向带状分布,60°S附近大洋海域存在显著的深混合层带,南极大陆与该深混合层带之间的海域混合层常年较浅。全球障碍层呈"哑铃状"分布,两半球的高纬度海区是障碍层高发区,障碍层不仅厚且持续时间长,以半年为周期变化,南大洋60°S附近海域显著的厚障碍层带随季节变化;南半球中低纬度海区长期存在障碍层,障碍层冬厚夏薄,且厚度大部分不超过40 m。     

南沙群岛海域混合层深度季节变化特征

基于南沙群岛海域综合科学考察11个航次的实测资料,研究了南沙群岛海域的混合层深度季节变化特征。研究结果表明,南沙群岛海域混合层深度存在明显的季节变化,并且与季风和海表热通量的变化密切相关。春季,风速较小且风向不稳定,海面得到的净热通量全年最大,上层水体层结稳定,混合层深度较小;夏季,南海西南季风盛行,上层为反气旋式环流,海面得到的净热通量减少,混合层呈加深的趋势;秋季,海面净热通量继续减少,混合层深度达到最大值;冬季,东北季风驱动下形成的上层气旋式环流引起深层冷水的上升,限制了混合层的加深。     

海水硝酸盐跃层深度计算方法研究

硝酸盐是海洋中浮游植物生命活动可利用的主要氮形态,其跃层深度(Z_N)会直接影响硝酸盐垂向输送、海洋初级生产力以及海洋碳循环。随着海洋观测技术的不断发展,硝酸盐剖面数据的采集呈现多样化,包括船基CTD观测和生物地球化学浮标BGC-Argo自动观测等,且垂向采样分辨率差异较大(CTD较低,BGC-Argo较高)。针对不同采样数据,亟需对硝酸盐跃层深度计算方法进行系统且定量化的对比分析研究。本文利用西北太平洋历史船测CTD数据和BGC-Argo浮标数据,采用差值法、梯度法和阈值法分别计算对应硝酸盐跃层深度。研究结果表明：就单一硝酸盐剖面,基于BGC-Argo数据,差值法计算的Z_N与目视解译的Z_N相差仅为0.2 m,阈值法次之为20.0 m,梯度法相差最大为202.8 m;基于CTD数据,差值法计算的Z_N与目视解译的Z_N相差2.0 m,阈值法相差49.0 m,梯度法相差155.0 m。相较于梯度法和阈值法,差值法计算的Z_N与目视解译的Z_N相差...     

热力学效应及斯托克斯漂流对混合层影响的研究

本文通过理想化的外部强迫以及海洋站点实测数据驱动普林斯顿海洋模式来研究海洋热力学效应和斯托克斯漂流对上混合层数值模拟的影响。在Mellor-Yamada湍流闭合方案中，经常出现夏季海表面温度偏暖和混合层深度偏浅的模拟误差。实验表明，斯托克斯漂流在冬季和夏季均能增强湍流动能，加深混合层深度。这种效应可以改善夏季的模拟结果，但与观测数据相比，将增大冬季混合层深度的模拟误差。斯托克斯漂流可以通过增强湍动能来加深混合层深度。结果表明，将斯托克斯漂流与冷皮层和暖层对上部混合层的热效应相结合，可以正确地模拟混合层深度。在夏季，海洋冷皮层和暖层通过“阻挡结构”和双温跃层结构模拟出更真实的上混合层变化。在冬季，海洋热力学效应通过增强上层海洋层结平衡了斯托克斯漂流的影响，并且由斯托克斯漂流引起的过度混合被校正。     

南海温跃层深度计算方法的比较

基于1986-2008年的中国近海及邻近海域再分析产品(CORA)气候平均海温资料,分别运用S-T法、垂向梯度法和最大曲率点3种温跃层定义计算了南海温跃层上界深度,揭示了南海温跃层季节变化特征。对3种不同定义确定的温跃层上界深度进行比较发现:采用不同定义计算南海温跃层上界深度存在差异,S-T法确定的温跃层上界深度最浅,垂向梯度法其次,最大曲率点法最深;在深水区(水深200 m)运用S-T法计算的温跃层上界深度与垂向梯度法的结果比较一致,都与实际温跃层深度符合较好;在浅水区(水深200 m),垂向梯度法和最大曲率点法可以准确判定无跃区,但对于温跃层深度计算,3种定义误差均较大。     

基于遥感数据的三维温度场参数化分析方法研究

为了满足海洋研究以及海洋调查的需求，本文基于Argo剖面和海表面温度数据开发了一个新的拟合三维温度场的算法。选取西北太平洋区域作为验证算法有效的实验海区。该水域的经纬度范围设定为:30°～40°N, 140°～155°E, 水平分辨率为0.25°。深度方向为从海表到2 000 m水深，水域划分为29层。拟合算法首先将Argo温度剖面以5个深度划分为6层，分别为混合层、夹层、温跃层、过渡层、第一深层、第二深层，然后以第一猜想值和线性回归得到的海表面温度作为初始条件重构三维温度场。重构的三维温度场的剖面与原观测剖面的均方根误差较小，相关性较好，表明该算法是合理有效的。     

飓风Isidore与海洋相互作用数值模拟研究

本文利用MM5大气模型和HYCOM海洋环流模型,模拟研究了2002年飓风Isidore过境墨西哥湾后上层海洋的响应以及墨西哥湾常态水(Gulf common water,GCW)和湾流(Loop Current,LC)对飓风Isidore的不同响应。飓风Isidore穿过GCW区后,导致海表最大降温接近5℃,混合层深度由30m加深到70m,上层海洋损失热量较多,上层海洋的结构恢复时间较长;但在LC区,由于亚热带持续不断的高温、高速水体输运,该区域水体具有较深的混合层和较大的能量,虽然在飓风过程中损失了较多的热量,但仅造成2℃的海表面降温,上层海洋的结构恢复时间较短;此外,飓风过境后,在上层海洋激发近惯性震荡,在LC区可导致1.0m/s近惯性震荡流,能量可传播至1500m以深,但在GCW区,由于较浅的上混合层、较弱的流速和较强的层化结构,近惯性震荡运动影响深度较浅。     

Maximum angle method for determining mixed layer depth from seaglider data

A new maximum angle method has been developed to determine surface mixed-layer (a general name for isothermal/constant-density layer) depth from profile data. It has three steps: (1) fitting the profile data with a first vector (pointing downward) from an upper level to a depth and a second vector (pointing downward) from that depth to a deeper level; (2) identifying the angle (varying with depth) between the two vectors; (3) after fitting and calculating angle all depths, and then selecting the depth with maximum angle as the mixed layer depth (MLD). Temperature and potential density profiles collected from two seagliders in the Gulf Stream near the Florida coast during 14 November–5 December 2007 were used to demonstrate the method’s capability. The quality index (1.0 for perfect identification of the MLD) of the maximum angle method is about 0.96. The isothermal layer depth is generally larger than the constant-density layer depth, i.e., the barrier layer occurs during the study period. Comparison with the existing difference, gradient, and curvature criteria shows the advantage of using the maximum angle method. Uncertainty in determining MLD because of varying threshold using the difference method is also presented.     

Exponential leap-forward gradient scheme for determining the isothermal layer depth from profile data

Two distinct layers usually exist in the upper ocean. The first has a near-zero vertical gradient in temperature (or density) from the surface and is called the isothermal layer (or mixed layer). Beneath that is a layer with a strong vertical gradient in temperature (or density), called the thermocline (or pycnocline). The isothermal layer depth (ILD) or mixed layer depth (MLD) for the same profile varies depending on the method used to determine it. Also, whether they are subjective or objective, existing methods of determining the ILD do not estimate the thermocline (pycnocline) gradient. Here, we propose a new exponential leap-forward gradient (ELG) method of determining the ILD that retains the strengths of subjective (simplicity) and objective (gradient change) methods and avoids their weaknesses (subjective methods are threshold-sensitive and objective methods are computationally intensive). This new method involves two steps: (1) the estimation of the thermocline gradient G _th for an individual temperature profile, and (2) the computation of the vertical gradient by averaging over gradients using exponential leap-forward steps. Such averaging can filter out noise in the profile data. Five existing methods of determining the ILD (difference, gradient, maximum curvature, maximum angle, and optimal linear fitting methods) as well as the proposed ELG method were verified using global expendable bathythermograph (XBT) temperature and conductivity–temperature–depth (CTD) datasets. Among all the methods considered, the ELG method yielded the highest skill score and the lowest Shannon information entropy (i.e., the lowest uncertainty).     

Temporal variability of winter mixed layer in the mid-to high-latitude North Pacific

Temperature and salinity data from 2001 through 2005 from Argo profiling floats have been analyzed to examine the time evolution of the mixed layer depth (MLD) and density in the late fall to early spring in mid to high latitudes of the North Pacific. To examine MLD variations on various time scales from several days to seasonal, relatively small criteria (0.03 kg m⁻³ in density and 0.2°C in temperature) are used to determine MLD. Our analysis emphasizes that maximum MLD in some regions occurs much earlier than expected. We also observe systematic differences in timing between maximum mixed layer depth and density. Specifically, in the formation regions of the Subtropical and Central Mode Waters and in the Bering Sea, where the winter mixed layer is deep, MLD reaches its maximum in late winter (February and March), as expected. In the eastern subarctic North Pacific, however, the shallow, strong, permanent halocline prevents the mixed layer from deepening after early January, resulting in a range of timings of maximum MLD between January and April. In the southern subtropics from 20° to 30°N, where the winter mixed layer is relatively shallow, MLD reaches a maximum even earlier in December–January. In each region, MLD fluctuates on short time scales as it increases from late fall through early winter. Corresponding to this short-term variation, maximum MLD almost always occurs 0 to 100 days earlier than maximum mixed layer density in all regions.     

Mixed layer formation and restratification in presence of mesoscale and submesoscale turbulence

Recent realistic high resolution modeling studies show a net increase of submesoscale activity in fall and winter when the mixed layer depth is at its maximum. This submesoscale activity increase is associated with a reduced deepening of the mixed layer. Both phenomena can be related to the development of mixed layer instabilities, which convert available potential energy into submesoscale eddy kinetic energy and contribute to a fast restratification by slumping the horizontal density gradient in the mixed layer. In the present work, the mixed layer formation and restratification were studied by uniformly cooling a fully turbulent zonal jet in a periodic channel at different resolutions, from eddy resolving (10 km) to submesoscale permitting (2 km). The effect of the submesoscale activity, highlighted by these different horizontal resolutions, was quantified in terms of mixed layer depth, restratification rate and buoyancy fluxes. Contrary to many idealized studies focusing on the restratification phase only, this study addresses a continuous event of mixed layer formation followed by its complete restratification. The robustness of the present results was established by ensemble simulations. The results show that, at higher resolution, when submesoscale starts to be resolved, the mixed layer formed during the surface cooling is significantly shallower and the total restratification is almost three times faster. Such differences between coarse and fine resolution models are consistent with the submesoscale upward buoyancy flux, which balances the convection during the formation phase and accelerates the restratification once the surface cooling is stopped. This submesoscale buoyancy flux is active even below the mixed layer. Our simulations show that mesoscale dynamics also cause restratification, but on longer time scales. Finally, the spatial distribution of the mixed layer depth is highly heterogeneous in the presence of submesoscale activity, prompting the question of whether it is possible to parameterize submesoscale effects and their effects on the marine biology as a function of a spatially-averaged mixed layer depth.     

Characteristics of the surface mixed layer depths in the northern South China Sea in spring

The nature and characteristics of the mixed layer depth (MLD) remain uncertain in the northern South China Sea. Using in situ data, we examined the quality of different MLD definitions, investigated the spatial and diurnal variation in the MLD, and examined the mechanisms of mixed layer development during March 23–31, 2014. We made distinct calculations of the MLD; of which two are (a) the depths between two different temperatures (0.2, 0.6 °C) and (b) the depths between two density differences (0.125, 0.25 kg/m³); and the fifth calculation is a depth derived from the optimal linear fitness method. We found that the optimal linear fitness MLD was the best definition for our study region ,and that it deepened from the shelf to the slope. Twenty-four-hour diurnal variation in the MLDs and mixing layers was observed when the ship was moored. Mixing layers were characterized by turbulent dissipation rates. We found that the mixed layer underwent a ‘stable-decaying–developing’ process. During the stable period, the MLD was close to that of the mixing layer, but during the decay/development periods, the MLDs were larger/smaller than those of the mixing layers. We suggest that both velocity shear and buoyancy flux were important in mixed layer development. We quantitatively examined the mechanisms of mixing in the shelf region, with air–sea net heat flux determined to be the major factor, rather than wind speed or current velocity.     

基于二次多项式拟合的超短基线声线修正方法

超短基线定位解算中的距离观测值是指换能器与水下应答器之间的直线距离,而海水声速的不均匀分布导致声波在海水中的实际传播路径为连续弯曲的曲线,需要结合实测声速剖面进行声线修正。根据声速在分层介质中的传播特性,本文提出了一种基于二次多项式拟合的声线跟踪算法,采用线性插值方法对声速剖面数据进行合理加密并按等深度进行分层,设定每层声速梯度是不断变化的,用二次多项式拟合声速,基于运动学原理建立了完整的数学解算模型。仿真结果表明,该方法修正后的水下目标分布具有明显的收敛性,且优于等梯度声线跟踪算法和等效声速剖面法,显著提高了超短基线水声定位系统的定位精度。     

南海混合层年循环特征

通过分析Levitus1994版气候平均温盐资料，得到南海混合层的时空分布特征，剖析了混合层浓度及其内部温度的季节变化规律。资料分析表明：季风通风流场调整对南海混合层的时空分布着明显的影响。这种影响的复杂性在于它不但通过海洋表层Ekman效应来影响混合层深度，而且还通过大尺度环流造成的幅散或辐合来限制或促进混合层深度的发展。研究发现混合层深度与混合层内温度存在着如下关系：夏季最大混合层的形成是28℃等温线与混合层底达到相互贴合的过程；冬季最大混合层的形成是28℃水体完全消失并且等温度线与混合层达到相交最多、相交最为垂直的过程，这时对应着冬季南海北部温跃层的通风；大于或等于28℃的水体总是位于混合层以内。     

海表Ekman流不稳定性致新型次级环流

海洋上混合层中的次级环流可通过物质和能量的垂直输运和混合过程把海洋表层的热量、动量与物质携带到次表层,对海洋上层次级环流生成机制的研究可以丰富对上层海洋的理解和认识。文中利用线性稳定性理论讨论了经典海表Ekman流的不稳定性,提出Ekman流的不稳定性可生成一种新型的次级环流。这种次级环流的空间尺度与雷诺数、Ekman流的垂向衰减速率、水平湍黏性系数与垂向湍黏性系数比值等密切相关,尺度范围从数十米到数千米。数十米量级的次级环流其垂向结构以及次级环流流轴与主流场偏角都与Langmuir环流的特征极为相似,是Langmuir环流形成机制的一种新解释。千米量级的次级环流能够解释黄海浒苔的条带分布。此外,所得次级环流的流轴与主流之间的偏角与科氏力有显著关系,北半球次级环流流轴偏向主流左侧,南半球反之。     

Prediction of short-term distributions of load extremes of offshore wind turbines

This paper proposes a new methodology to select an optimal threshold level to be used in the peak over threshold (POT) method for the prediction of short-term distributions of load extremes of offshore wind turbines. Such an optimal threshold level is found based on the estimation of the variance-to-mean ratio for the occurrence of peak values, which characterizes the Poisson assumption. A generalized Pareto distribution is then fitted to the extracted peaks over the optimal threshold level and the distribution parameters are estimated by the method of the maximum spacing estimation. This methodology is applied to estimate the short-term distributions of load extremes of the blade bending moment and the tower base bending moment at the mudline of a monopile-supported 5MW offshore wind turbine as an example. The accuracy of the POT method using the optimal threshold level is shown to be better, in terms of the distribution fitting, than that of the POT methods using empirical threshold levels. The comparisons among the short-term extreme response values predicted by using the POT method with the optimal threshold levels and with the empirical threshold levels and by using direct simulation results further substantiate the validity of the proposed new methodology.