Errors in dynamical fields inferred from oceanographic cruise data: Part II. The impact of the lack of synopticity

Diagnostic studies of ocean dynamics based on the analysis of oceanographic cruise data are usually quite sensitive to observation errors, to the station distribution and to the synopticity of the sampling. The first two sources have been evaluated in the Part I of this work. Here we evaluate synopticity errors for different sampling strategies applied to simulated unstable baroclinic waves. As suggested in previous studies, downstream and upstream cross-front samplings produce larger errors than along-front samplings. In our particular case study, the along-front sampling results in fractional errors (rms error divided by the standard deviation of the field) of about 15% for dynamic height and more than 50% for relative vorticity and vertical velocity. These values are significantly higher than those obtained in Part I for typical observation errors and sampling limitations (about 6% for dynamic height and between 15 and 30% for geostrophic vorticity and vertical velocity).We also propose and test two methods aimed at reducing the impact of the lack of synopticity. The first one corrects the observations using the quasi-geostrophic tendency equation. The second method combines the relocation of stations (based on a system velocity) and the correction of observations (through the estimation of a growth rate). For the fields simulated in this work, the second method gives better results than the first, being able to eliminate practically all synopticity errors in the case of the along-front sampling. In practice, the error reduction is likely to be less effective, since actual fields cannot be expected to have a system velocity as homogeneous as for the single-mode waves simulated in this work.     

The use of simulated drifters to estimate vorticity

A primitive equation ocean model is used to generate trajectories of simulated clusters of drifters in the California Current (CC) region. These trajectories allow us to evaluate a least squares (LS) method of estimating vorticity and vertical velocity along a cluster's path. Two clusters provide examples of successful and less successful estimates of vorticity and vertical velocity. Our analysis quantifies the dependence of estimate quality on several parameters that can be used as error predictors in the LS estimate of vorticity: cluster separation, number of drifters in a cluster, and cluster shape. A combination of cluster separation and ellipticity shows the most promise as an indicator of quality for the vorticity estimate.     

Variability of the shelf water off the northeast Spanish coast

During the seasonal stratification of 1983 and 1984, intensive sampling of the shelf region of the northeast Spanish coast was carried out. This unique dataset is used to define the most salient features of the shelf water dynamics. We show that region has a strong spatial and temporal variability associated with the southward spreading of continental waters from the Rhone river, in the northern gulf of Lyons. Southward spreading begins in early spring and was particularly obvious during June 1983 when strong horizontal density gradients were observed. Spreading almost completely disappears in August when observed upper layer salinities are uniform. We also show that significant shelf/slope water exchange takes place associated with the abrupt topographic canyons found in the region. In particular we observed in June 1983 a small positive geostrophic vorticity region at the Palamós canyon while a larger anticyclonic eddy was observed over the shelf south of this canyon. The biological implications of these two features have been demonstrated. We conclude indicating that in this region where no dominant winds exist, the shelf water dynamics is dominated by the southward spreading of low salinity continental waters and by the intrusion of high salinity open ocean water through submarine canyons.     

Assimilation of buoy and satellite data in wave forecasts with integral control variables

The quality of numerical wave forecasts can be improved significantly by assimilating wave observations prior to the forecast. In the present study a technique for such assimilation is developed that exploits (a) the efficiency of a limited number of integral control variables, and (b) the effectiveness of variational (model-consistent) assimilation. The formal procedure is independent of the type of control variables and of the wave model (moreover, no adjoint wave model is required). In the present study, integral control variables are chosen to represent large-scale errors in the driving wind fields and uncertainties in the wave model. The assimilation technique is validated with observations of the ERS-1 satellite altimeter and two waverider buoys in two consecutive storms in the Norwegian Sea. The assimilation of the observations reduced the errors in the forecasted significant wave height at the buoy locations typically from 25% to 12%. For low-frequency waves the effect of the assimilation is similarly significant at one buoy location but marginal at the other buoy location.     

Measurements of a small scale eddy at a tidal inlet using an unmanned automated boat

An unmanned automated boat equipped with an acoustic Doppler current profiler was used in field surveys at a tidal inlet, the Southwest Pass of Vermillion Bay, Louisiana on Sept 6 and Oct 6, 2007. During the first survey, under calm weather conditions, a small scale eddy with a diameter of 300 m was discovered with strong upwelling and downwelling zones. A detailed analysis of this small eddy shows that the eddy's velocity field is relatively uniform in the vertical and the eddy is formed by a flow convergence, tidal velocity shear induced relative vorticity, and the interaction between the horizontal flows and bathymetry. The major upwelling area is where an uphill flow occurs while the major downwelling area is where a downhill flow occurs. The vorticity of this eddy is on the order of 0.013 s^? 1, which is two orders-of-magnitude larger than the planetary vorticity, and one-order-of magnitude larger than that in a typical tidal inlet without eddies. The Coriolis effect is thus insignificant and the generation of the eddy cannot be affected by the earth rotation. The maximum upwelling and downwelling velocities exceed 0.3 m/s. This high vertical velocity in a tidal inlet does not appear to have been reported before. The second survey, conducted under a thunder storm condition, did not reveal a similar eddy at the same location during roughly the same tidal phase. Though the measurements of 3-D flow structure under a thunder storm condition in a tidal channel does not appear to have been reported before, the second survey is of important value in providing support of the mechanism of the eddy formation during the first survey: the wind tends to produce downwind flow in shallow water than in deep water, producing a velocity shear counterproductive to the formation of the eddy. Therefore, the second survey under a thunder storm condition did not show an eddy. A scaling analysis of the non-hydrostatic flow shows that the uphill and downhill flows introduce a non-hydrostatic flow component proportional to the square of the bottom slope which leads to the conclusion that the non-hydrostatic flow component affects less than 10% of the vertical momentum balance.     

对《港口工程地基规范》关于垂直固结度计算的修正意见

通过计算证明,现行港口工程地基规范中关于垂直固结度的公式有局限性,特别是当固结度在30%以内时,误差较大,对此提出了几点建议。     

Recovery of North-East Atlantic temperature fields from profiling floats: Determination of the optimal float number from sampling and instrumental error analysis

Argo is an international project that is deploying an array of temperature and salinity profiling floats over the global ocean. Here we use the error formulation derived from Optimal Statistical Interpolation to estimate statistical errors associated with the recovery of the temperature field in the North-East Atlantic ocean. Results indicate that with the present distribution of floats (119 in the considered domain), scales of wavelength larger than 500 km can be recovered with a relative uncertainty (rms error relative to the standard deviation of the field) of about 7% at 50 m, 8% at 200 m and 10% at 1000 m. This corresponds to mean absolute errors of 0.111 °C at 50 m, 0.104 °C at 200 m and 0.073 °C at 1000 m.The splitting of total errors into instrumental and sampling contributions reveals that, in the present scenario, errors are more due to the small number of floats than to instrumental errors, especially at upper levels. For scales larger than 500 km this will hold true until 200–250 floats are deployed (less than 200 for deep levels). In such a simulated scenario, the number of observations and the technology become approximately equally limiting factors for the accuracy of the temperature field mapping, with total relative errors of less than 2% at upper levels and about 3% at 1000 m.     

用于船舶电站控制器的交流采样算法研究

对交流采样算法的分析表明,离散傅立叶算法不能算出频率,在非整周期同步采样时误差较大。非同步算法中的加窗插值算法通过加窗插值减小频谱泄漏,不需整周期同步采样,可直接计算频率,对船舶电站采样具有很好的适应性。该算法多针对谐波分析,当用于船舶电站,只计算基波参数时,需选择合理采样方案。通过对算法的分析与推导,选择合理采样方案,在满足精度要求的同时使计算量适中,并应用于船舶电站控制器,很好地满足采样要求。     

垂直及斜出水流场的二维及三维TR-PIV试验

本文对钝头回转体垂直及斜出水流场采用TR-(Time-Resolved)PIV技术进行测量,并对斜出水流场进行3D-Stereo PIV(三维体视 PIV)测量.文中介绍了测试技术及测量结果,揭示了出水过程中流动结构及其演变,展示了TR-PIV技术对具有瞬态历程特征的出水流场研究的适应性.     

考虑算法的实船试航船速测量不确定度分析

实船试航中影响船速测量结果的因素很多,对其进行测量不确定度分析时应考虑的不确定度来源也有很多。采用“测量不确定度表示指南”规定的方法,将船速算法作为不确定度源,对其测量结果进行了不确定度分析。不确定度分析以某油船为例,分别考虑了实际航速平均法和逐点法两种算法,并对两种算法得到的结果作了讨论。     

An inverse model for calculation of global volume transport from wind and hydrographic data

The P-vector inverse method has been successfully used to invert the absolute velocity from hydrographic data for the extra-equatorial hemispheres, but not for the equatorial region since it is based on the geostrophic balance. A smooth interpolation scheme across the equator is developed in this study to bring together the two already known solutions (P-vectors) for the extra-equatorial hemispheres. This model contains four major components: (a) the P-vector inverse model to obtain the solutions for the two extra-equatorial hemispheres, (b) the objective method to determine the Ψ-values at individual islands, (c) the Poisson equation-solver to obtain the Π-values over the equatorial region from the volume transport vorticity equation, and (d) the Poisson equation-solver to obtain the Ψ and depth-integrated velocity field (U, V) over the globe from the Poisson Ψ-equation. The Poisson equation-solver is similar to the box model developed by Wunsch. Thus, this method combines the strength from both box and P-vector models. The calculated depth-integrated velocity and Ψ-field agree well with earlier studies.     

On the parameterization of interleaving and turbulent mixing using CTD data from the Azores Frontal Zone

CTD-data obtained in the Azores Frontal Zone using a towed undulating vehicle are analyzed to study the relationship between characteristics of intrusions and mean parameters of the thermohaline field. A self-similar dependence between intrusion intensity and hydrological parameters is obtained. The most well-founded interpretation of the empirical dependence is as follows: (a) the main source supporting intrusive layering is the salt finger convection; (b) the abrupt decrease of intrusion intensity with the reduction of geostrophic Richardson number obtained from the analysis is explained by the beginning of turbulence when salt fingers do not work any longer, so the “driving force” for intrusive motion disappears. These results are consistent with the conclusions of the paper [Kuzmina N.P., Rodionov V.B., 1992. About the influence of baroclinicity upon generation of the thermohaline intrusions in the oceanic frontal zones. Izvestiya Akad. Nauk SSSR, Atmosperic and Oceanic Physics 28 (10–11), 1077–1086]. These conclusions imply that there are three main mechanisms of intrusive layering at oceanic fronts, namely the 2D baroclinic instability of geostrophic flow, the vertical shear instability and the thermohaline instability where the driving source of intrusive motion is double diffusive convection. The baroclinic and thermohaline instabilities can generate intrusions of large vertical scale, while vertical shear instability usually gives rise to thin turbulent layers. Turbulence in these thin layers can prevent salt finger convection and thus destroy the energy source of the intrusive motion conditioned by thermoclinicity. Therefore, the baroclinicity plays two parts in the processes of the intrusive layering: (1) it prevents double-diffusion interleaving by means of turbulence, and (2) it generates intrusions due to the 2D baroclinic instability of geostrophic current. Using features of thermohaline interleaving as a specific tracer of turbulent mixing, we have estimated turbulent mixing coefficient as k_tRi^−0.8 (Ri>1), where Ri is the geostrophic Richardson number. Application of the proposed approach to other frontal zones is discussed.     

台风波高重现值的泊松最大熵分布估计

以日本中城湾观测站实测台风过程出现的最大波高序列为例,提出了泊松-最大熵复合极值分布,并将其用于计算台风海浪的重现值波高.采用年极值法与过阈法对台风波高进行长期统计分析.计算结果表明:观测时段相同,阈值的选取对重现值的估计影响不大.而阈值相同,抽样时段的长度对重现值估计的影响较大;若样本时段较短,但极值波高的最大值包含于其中,则估计得到的重现值偏大,该结果对于结构设计是偏安全的.     

基于MRF方法和滑移网格的螺旋桨水动力性能研究

本文基于计算流体力学(CFD)方法,对多重参考系模型(MRF)及滑移网格模型(SM)在计算螺旋桨水动力性能时的差异进行了探讨。将以上两种模型应用到4381螺旋桨的水动力性能计算中,首先将计算得到的推力系数及转矩系数与试验数据进行了对比,考察了两种计算模型对螺旋桨的敞水性能的预测情况,并进一步对两种模型计算得到的螺旋桨盘面的速度场、桨叶的压力分布、桨后涡量云图等进行了对比分析。计算结果表明,滑移网格模型相较于多重参考系模型,对螺旋桨的推力系数的模拟结果误差更小,扭矩系数方面,两种模型的模拟结果相差不大;对于进速系数较大时,两种模型模拟得到的压力分布及速度分布较为相似,但对于高负荷情况,滑移网格模型可以更好地捕捉桨叶的压力分布及桨盘面处的速度分布情况;进速系数较小时,多重参考系模型可以模拟出涡结构的发散现象,而滑移网格模型可以更好的在高进速系数情况下捕捉到梢涡结构。     

码头前系泊船舶撞击速度研究

采用时域数值模型,对横浪作用下码头前系泊船舶的运动响应进行了研究,分析了波高、周期、水深、船舶尺度及码头型式等因素对撞击速度的影响.结果表明:当波浪周期小于般舶自振周期时,撞击速度随波高的增加线性增长;波浪周期越大,撞击速度越大;船舶装载度(吃水)越小,撞击速度越大;墩式码头(无限开敞水域)中系泊船舶的撞击速度大于岸壁...     

单个透空四面体水槽试验

针对透水框架附近的流场分布规律的相关问题,通过采用高精度河流动力学水槽试验,利用基于变频技术的DCMS控制系统,进行单个透空四面体模型试验。试验在透空四面体周围布置972个测点,每个测点采用旋桨式光电流速仪测量。通过与无框架流场的对比,分析横断面流速沿垂直水流方向和沿高度方向的变化规律,得到透空四面体周围流速分布规律和具体减速、增速部位、平均减速率,揭示了透空四面体对水流扰动作用。     

变态河工模型垂线流速分布不相似问题的初步研究

变态河工模型垂线流速分布不完全相似的问题早为人所知 ,但其偏差有多大至今未有定量研究 ,对此进行了初步分析 ,指出变态模型表层流速偏大而底层流速偏小 ,即流速垂向分布梯度增大     

Numerical investigation of turbulence near a sheared air–water interface. Part 2: Interaction of turbulent shear flow with surface waves

Large-eddy simulations (LES) of the interactions between turbulent shear flow and surface waves are presented. The formation mechanism of Langmuir circulation and its contribution to the vertical momentum transport are investigated in detail. The effect of surface waves is modeled in two ways in the LES runs. One model includes only the phase-averaged effect of the waves as an added source term in the momentum equation, and the other model includes the full effect of the waves by use of the fully nonlinear conditions of the air-water interface. Langmuir circulations are clearly indicated in both cases, indicating that the phase-averaged effect is essential for the formation of this circulation. It is shown that the formation of Langmuir circulations enhances the vertical transport of momentum. As a result, the mean velocity gradient and the streamwise component of the turbulence intensity are decreased, while the spanwise and interface normal components are increased. Examination of the turbulence energy budget equations shows that production is due to the interaction between the vorticity and the Lagrangian drift as the phase-averaged effect of the wave becomes the dominant source of turbulence. Received: August 17, 2000 / Accepted: December 22, 2000     

Modelling of hydrodynamic circulation in Benoa Bay, Bali

A simulation of water level, velocity, salinity, and temperature in the Bay of Benoa has been carried out using a three-dimensional hydrodynamic Estuarine and Coastal Ocean Model incorporating a main characteristic of southward transport of the Indonesian throughflow at the offshore area of the bay. In other respects, two types of boundary conditions have been tested: (1) specifying elevation at all boundaries; and (2) implementing a combination of elevation and velocity at the boundaries. Performance of the model results has been quantified in terms of mean absolute errors, root-mean square errors, and correlation coefficients based on the availability of water level and current data. The general agreement between simulated and observed values of water elevation and currents is encouraging. Errors in computed water levels are less than 5 % of the local tidal range, and correlations between the data and model exceed 0.95. Meanwhile, errors and correlations for simulated currents are less than 22 % and are about 0.75, respectively.     

双惯导改善导航性能的组合数据研究

在传统的水平阻尼网络的基础上,针对双惯导系统的特性设计了新的阻尼方法并分析了各种误差与这种阻尼网络的关系。通过仿真分析,比较2套阻尼系统的误差结果,可以看出双惯导外水平阻尼能够显著降低外速度误差的影响,减少阻尼时间,抑制经度误差。