Spark平台下的海表面温度并行预测算法

面对海量的海表面温度数据,如何使用大数据处理平台和新的处理技术来实时处理、分析并预测海表面温度数据,是一个亟待解决的问题。本文基于现阶段的时间序列方法和专家意见,首先,将类比合成方法引入到海表面温度预测应用中;其次,基于Spark平台提出了一种改进的快速DTW算法SparkDTW;最后,为了充分利用通过时间序列挖掘得到的信息,将SparkDTW与SVM相结合,提出了SparkDTW+SVM混合模型,为海表面温度预测的应用研究提供了较好的理论基础和技术支持。实验结果表明,SparkDTW算法预测精度优于SVM,提高了海表面温度预测效率,验证了将类比合成方法应用在海表面温度预测的可行性;SparkDTW+SVM在精度方面要优于SparkDTW和SVM,表明SVM模型能充分利用时间序列挖掘的信息,验证了SparkDTW+SVM在海表面温度预测的有效性。     

TOPEX/Poseidon and Jason Equatorial Sea Surface Slope Anomaly in the Atlantic in 2002: Comparison with Wind and Current Measurements at 23W

A time series of velocity profile in the upper 150 m of the equatorial Atlantic was gathered at 23W in 2002 within the PIRATA program. It constitutes the first time series of near surface current measurements simultaneous with altimetric data in the equatorial Atlantic. The surface slope anomaly along the equator is computed from satellite altimetry, and, as a proxy for the pressure gradient along the equator, compared with the wind and near surface current data. In a first step, a time series of the surface slope anomaly along the equator in the Atlantic is computed from the 10-year-long TOPEX/Poseidon sea level anomalies. A sensitivity study establishes the robustness of the calculation. Apart from a 15 cm bias, the equatorial sea surface slope anomalies estimated either from TOPEX/Poseidon or from Jason over the 6-month overlap (Feb.–Aug. 2002) do not reveal drastic differences. We produce two sea surface slope anomaly composite time series for 2002 (one with T/P data, the other with Jason data during the commissioning phase) and compare them to the wind and velocity data at 23W. As expected, the near surface velocity and depth of the upper limit of the equatorial undercurrent (EUC) are extremely well correlated with the surface pressure gradient anomaly. 10-year-long time series of altimetry-derived zonal sea surface slope anomaly and ECMWF ERA40 wind stress are also well correlated. They exhibit similar spectral content and similar anomalous years. This is a first step towards a full analysis of the EUC dynamics using altimetry, PIRATA data (near surface current and subsurface thermohaline structure) and model. These initial comparisons reinforce the utility of Jason measurements for continuing the 10-year and highly accurate TOPEX/Poseidon time series for study of equatorial signals.     

TOPEX/Poseidon and Jason Equatorial Sea Surface Slope Anomaly in the Atlantic in 2002: Comparison with Wind and Current Measurements at 23W

A time series of velocity profile in the upper 150 m of the equatorial Atlantic was gathered at 23W in 2002 within the PIRATA program. It constitutes the first time series of near surface current measurements simultaneous with altimetric data in the equatorial Atlantic. The surface slope anomaly along the equator is computed from satellite altimetry, and, as a proxy for the pressure gradient along the equator, compared with the wind and near surface current data. In a first step, a time series of the surface slope anomaly along the equator in the Atlantic is computed from the 10-year-long TOPEX/Poseidon sea level anomalies. A sensitivity study establishes the robustness of the calculation. Apart from a 15 cm bias, the equatorial sea surface slope anomalies estimated either from TOPEX/Poseidon or from Jason over the 6-month overlap (Feb.-Aug. 2002) do not reveal drastic differences. We produce two sea surface slope anomaly composite time series for 2002 (one with T/P data, the other with Jason data during the commissioning phase) and compare them to the wind and velocity data at 23W. As expected, the near surface velocity and depth of the upper limit of the equatorial undercurrent (EUC) are extremely well correlated with the surface pressure gradient anomaly. 10-year-long time series of altimetry-derived zonal sea surface slope anomaly and ECMWF ERA40 wind stress are also well correlated. They exhibit similar spectral content and similar anomalous years. This is a first step towards a full analysis of the EUC dynamics using altimetry, PIRATA data (near surface current and subsurface thermohaline structure) and model. These initial comparisons reinforce the utility of Jason measurements for continuing the 10-year and highly accurate TOPEX/Poseidon time series for study of equatorial signals.     

Structure function characteristics for 2 metre temperature and relative humidity in different horizontal resolutions

The isotropic correlations of forecast errors in the HIRLAM system are investigated for different horizontal grid sizes in order to achieve an improved representation of the structure functions for high-resolution surface analysis. The investigation is performed for 2 metre temperature and relative humidity and makes use of operational forecasts from DMI-HIRLAM at the Danish Meteorological Institute (DMI), which can support the background for a surface analysis in three different horizontal resolutions. Two different well-known methods for determining isotropic forecast error correlations are applied. The first method compares forecasts to observations (the Observation Method), while the second makes use of two different forecasts valid for the same time (the NMC Method). The latter method is also used to investigate isotropy as well as the influence of land–sea contrast and orography. A comparison of the two mentioned methods reveals a good correspondence between them, and the investigation of monthly changes shows some seasonal tendencies. The isotropy assumption is shown to be acceptable to a first approximation, despite a slight dependency on the predominant flow. The results further suggest a decrease in the background error correlation scales when going to higher horizontal resolution in the forecast model.     

Validation of interannual simulations from the 1/8° global Navy Coastal Ocean Model (NCOM)

A 1/8° global version of the Navy Coastal Ocean Model (NCOM) is used for simulation of upper-ocean quantities on interannual time scales. The model spans the global ocean from 80°S to a complete Arctic cap, and includes 19 terrain-following σ- and 21 fixed z-levels. The global NCOM assimilates three-dimensional (3D) temperature and salinity fields produced by the Modular Ocean Data Assimilation System (MODAS) which generates synthetic temperature and salinity profiles based on ocean surface observations. Model-data intercomparisons are performed to measure the effectiveness of NCOM in predicting upper-ocean quantities such as sea surface temperature (SST), sea surface salinity (SSS) and mixed layer depth (MLD). Subsurface temperature and salinity are evaluated as well. An extensive set of buoy observations is used for this validation. Where possible, the model validation is performed between year-long time series obtained from the model and time series from the buoys. The statistical analyses include the calculation of dimensionless skill scores (SS), which are positive if statistical skill is shown and equal to one for perfect SST simulations. Model SST comparisons with year-long SST time series from all 83 buoys give a median SS value of 0.82. Model subsurface temperature comparisons with the year-long subsurface temperature time series from 24 buoys showed that the model is able to predict temperatures down to 500 m reasonably well, with positive SS values ranging from 0.18 to 0.97. Intercomparisons of MLD reveal that the model MLD is usually shallower than the buoy MLD by an average of about 15 m. Annual mean SSS and subsurface salinity biases between the model and buoy values are small. A comparison of SST between NCOM and a satellite-based Pathfinder data set demonstrates that the model has a root-mean-square (RMS) SST difference of 0.61 °C over the global ocean. Spatial variations of kinetic energy fields from NCOM show agree with historical observations. Based on these results, it is concluded that the global NCOM presented in this paper is able to predict upper-ocean quantities with reasonable accuracy for both coastal and open ocean locations.     

台风海面风场动力诊断模式的建立与模拟计算

本文总结和归纳了已有的台风海面风场模式,按照风场模式物理背景的不同进行了分类．在此基础上,本文选择并建立了一种新的台风海面风场动力诊断模式．首先,利用台风影响范围内某条具有代表性的闭合等压线的拟合方程表示出台风海面气压场,并利用改进的气压场模式和修正的梯度风方程求得台风系统风场,同时还利用宫崎正卫的热带气旋合成风假设建立移行台风风场．然后将两者作权重订正后进行迭加,即得到台风模型风场．该模式考虑了包括台风气压场的非对称性、边界层摩擦效用、气压梯度的切向变化及台风中心移动的影响等多种因素．经过对0519号“龙王”台风的模拟,结果表明本文所建立的台风风场模式可以比较准确的模拟出非圆对称的台风海面气压场和海面风场,较为真实地反映实际台风风场的特征．     

GPS浮标数据反演海浪谱的理论仿真与试验验证

GPS浮标作为一种新型的海洋测量设备,近年来在海面高度现场测量和星载高度计定标方面取得了重要应用。通过仿真试验对反演海浪谱的方法和流程进行研究,旨在探索从GPS浮标测量的海面高度序列中提取海浪谱的方法。首先,使用Longuest模型生成了海浪波面位移时间序列,并通过Pierson-Moscowitz风浪谱对波面位移的统计特性进行约束,其随机性由相位引入。结合典型潮汐和GPS浮标仪器噪声的仿真时间序列,合成了仿真时间长度1h的1Hz(每秒1次)随机海面高度序列。然后,利用自相关函数法,进行高通滤波和数据压缩,得到了仿真的海浪谱。该仿真结果和理论海浪谱非常接近,可满足海浪谱反演的需求。最后,通过山东石岛外海的GPS浮标现场试验,验证了本文提出的反演方法的适用性。本文的研究解决了GPS浮标反演海浪谱的关键问题,丰富了海浪谱反演的手段,拓展了GPS浮标的应用领域,有望为未来我国的星载波谱仪定标服务。     

Neural network for wave forecasting among multi-stations

Unlike in the open sea, the use of wind information for forecasting waves may encounter more ambiguous uncertainties in the coastal or harbor area due to the influence of complicated geometric configurations. Thus this paper attempts to forecast the waves based on learning the characteristics of observed waves, rather than the use of the wind information. This is reported in this paper by the application of the artificial neural network (ANN), in which the back-propagation algorithm is employed in the learning process for obtaining the desired results. This model evaluated the interconnection weights among multi-stations based on the previous short-term data, from which a time series of waves at a station can be generated for forecasting or data supplement based on using the neighbor stations data. Field data are used for testing the applicability of the ANN model. The results show that the ANN model performs well for both wave forecasting and data supplement when using a short-term observed wave data.     

基于支持向量回归的海面高度统计预报方法研究

本文提出一种基于支持向量回归的统计预报方法,通过经验正交分解对原始数据矩阵进行时空分解,提取出空间模态和时间系数。由于海面高度变化具有非线性、大惯性的特点,对时间系数进行小波分析,能有效滤除其中的高频信号,得到表征海面高度变化的低频信号。利用支持向量回归方法对小波分解后的低频信号构建预报模型。最后,进行小波重构,还原时间序列长度,实现未来7天的海面高度预报。通过黑潮附近海域的海面高度预报结果验证,该预报方法的预报效果优于整合滑动平均自回归预报方法。本文通过机器学习的算法实现了海面高度的预报,为海洋预报方法提供了新的思路。     

CO2 exchange at air-sea interface in the Huanghai Sea

INTRODUCTIONTheroleoftheoceaniscrucialintheoverallcycleofCOZ,withitsspecialpumpingmechanismssuchassolubilitypumpingattheair-seainterfacewithcarbonatechemistry,biologicalpumpinginsurfacewatersandalsointhewatercolumn,anddynamicpumpingassociatedwithoceancirculation(BroeckerandPeng,1982).Inordertounderstandthesevariouspumpingprocessesintheocean,muchresearchhasbeencarriedoutonaglobalscaleasapartofeffortstounderstandtheglobalgeochemicalcycleofCOZ.TheHuanghaiSea,atypicalmid-latitudeepicontine…     

An inverse problem methodology for multiple parameter estimation in bend stiffeners

The flexible riser top connection is a critical region for lifetime assessment due to large tension/curvature variations and modeling uncertainties. The bend stiffener polyurethane mechanical response not only presents a nonlinear loading rate and temperature dependency but is also subjected to weather ageing during operation, which may affect its mechanical behavior over time. The top tension, employed for riser local cross-section stress calculation, is usually obtained from global dynamic analyses performed under selected environmental conditions, if direct measurement is not available. As a consequence, both the bend stiffener effect on the curvature distribution and the top tension time series present inherent uncertainties for riser lifetime (re)assessment. In the present work, a proposed monitoring approach composed by gyrometers installed along flexible riser/bend stiffener top connection system length combined with an inverse problem methodology is numerically investigated to estimate the following parameters: (i) polyurethane hyperelastic response and (ii) effective top tension. The top connection system is modeled using a large deflection beam bending model and the parameters are estimated using a damped least-square minimization approach with the Levenberg–Marquardt algorithm. For the preliminary feasibility investigation, the gyrometer experimental data is numerically estimated through Monte Carlo simulations. A case study is carried out to investigate the influence that the number of sensors, sensors arrangement, loading conditions and top connection model have on the inverse parameters estimation. The results indicate that the proposed monitoring approach and inverse parameter estimation methodology may effectively reduce flexible riser lifetime calculation uncertainties.     

Numerical hindcasting of sea surface pCO2 at Weathership Station Papa

A one-dimensional model of temperature, salinity, nutrients, oxygen, carbon, and argon chemistry is used to hindcast the annual cycle of sea surface pCO₂ at weathership Station Papa in the subarctic Pacific (50°N, 145°W), based on recent biological and chemical measurements made in conjunction with the SUPER program. Heat fluxes are calculated from the meteorological time series data from the Canadian weathership program. The Price, Weller and Pinkle (1986) model is used for predicting mixed layer dynamics. The rate of new production in the model is based on sediment trap data (Welschmeyer, personal communication) and a comparison of model oxygen and argon concentrations with the recent data of Emerson, Quay, Stump, Wilbur and Knox (1991). The balances of nutrients and oxygen in the permanent halocline require isopycnal ventilation on a time frame of approximately 10 years; this estimate is consistent with estimate of Van Scoy, Fine and Ostlund (1991) based in tritium data from Geosecs and Long Lines programs. The model results are compared with the 5 year time series data presented by Wong and Chang (1990). The model appears to capture the mean sea surface pCO₂ and the magnitude and timing of the seasonal cycle. The data, howeber, contain much greater high frequency variation than the model results, which we believe is caused by patchiness in the horizontal distribution of NO₃. The model pCO₂ sensitivity to the chemistry of the upwelling water and the rate of biological new production are presented.Although the model simulation of pCO₂ is satisfactory, this study illustrates the limitations of modelling the chemistry of the upper ocean in one dimension. The slow currents and horizontally homogeneous sea surface in the subarctic Pacific make Papa one of the best available candidates for modelling in 1-D. In spite of this, a 1-D formulation is inadequate to address the chemistry of the halocline (a crucial lower boundary condition to the mixed layer) and does not constrain the transport of the nutrients by wind-driven currents in the mixed layer. We conclude that further progress in modelling the upper ocean nutrient and carbon cycles will requires simulation in three dimensions.     

基于相空间重构与模糊神经网络的海温垂直分布预测模型

在海温预报中引入混沌理论,将相空间重构与模糊神经网络相结合,提出了海温垂直建模预测模型。通过相空间重构,把海温时间序列拓展为多维序列,而多维序列包含着各态历经的信息,从而挖掘出了丰富的海温变化空间的信息,有利于模糊神经网络的训练。利用建立好的模糊神经网络模型,较好地对海温的垂直结构进行了建模、训练和预测。实际的预测结果表明,该模型预报精度较高,超前1~5个月的预测值的相对误差均控制在10%以内,预测结果可以为业务工作提供一定的参考与借鉴。     

A hybrid efficient method to downscale wave climate to coastal areas

Long-term time series of sea state parameters are required in different coastal engineering applications. In order to obtain wave data at shallow water and due to the scarcity of instrumental data, ocean wave reanalysis databases ought to be downscaled to increase the spatial resolution and simulate the wave transformation process. In this paper, a hybrid downscaling methodology to transfer wave climate to coastal areas has been developed combining a numerical wave model (dynamical downscaling) with mathematical tools (statistical downscaling). A maximum dissimilarity selection algorithm (MDA) is applied in order to obtain a representative subset of sea states in deep water areas. The reduced number of selected cases spans the marine climate variability, guaranteeing that all possible sea states are represented and capturing even the extreme events. These sea states are propagated using a state-of-the-art wave propagation model. The time series of the propagated sea state parameters at a particular location are reconstructed using a non-linear interpolation technique based on radial basis functions (RBFs), providing excellent results in a high dimensional space with scattered data as occurs in the cases selected with MDA. The numerical validation of the results confirms the ability of the developed methodology to reconstruct sea state time series in shallow water at a particular location and to estimate different spatial wave climate parameters with a considerable reduction in the computational effort.     

Question of the pitch-roll buoy response to ocean waves as a simple harmonic oscillator?

The assumption that the East and North deck slopes of a pitch–roll buoy respond to East and North sea slopes as simple harmonic oscillators is routinely made by the National Data Buoy Center (NDBC) and others producing directional wave data. Although directional wave data derived with this assumption usually appear to be of good quality, the validity of the assumption has not previously been more directly demonstrated. In this paper, a method is proposed to judge the validity of the assumption for any set of time series records of buoy angular motion. The proposed method is applied to 200 record sets taken by an NDBC buoy located at ocean station 46024, and to five record sets taken by another NDBC buoy at 46051. For the 46024 data, it was demonstrated that the simple harmonic oscillator assumption was near perfectly valid. For the smaller 46051 data set, the simple harmonic oscillator assumption was shown to be slightly less valid.     

Experiments in reconstructing twentieth-century sea levels

One approach to reconstructing historical sea level from the relatively sparse tide-gauge network is to employ Empirical Orthogonal Functions (EOFs) as interpolatory spatial basis functions. The EOFs are determined from independent global data, generally sea-surface heights from either satellite altimetry or a numerical ocean model. The problem is revisited here for sea level since 1900. A new approach to handling the tide-gauge datum problem by direct solution offers possible advantages over the method of integrating sea-level differences, with the potential of eventually adjusting datums into the global terrestrial reference frame. The resulting time series of global mean sea levels appears fairly insensitive to the adopted set of EOFs. In contrast, charts of regional sea level anomalies and trends are very sensitive to the adopted set of EOFs, especially for the sparser network of gauges in the early 20th century. The reconstructions appear especially suspect before 1950 in the tropical Pacific. While this limits some applications of the sea-level reconstructions, the sensitivity does appear adequately captured by formal uncertainties. All our solutions show regional trends over the past five decades to be fairly uniform throughout the global ocean, in contrast to trends observed over the shorter altimeter era. Consistent with several previous estimates, the global sea-level rise since 1900 is 1.70 ± 0.26 mm yr⁻¹. The global trend since 1995 exceeds 3 mm yr⁻¹ which is consistent with altimeter measurements, but this large trend was possibly also reached between 1935 and 1950.     

常规海面风预报质量检验的实现

常规海面风预报内容是“文字加字符”形式,实况数据是间隔均匀的数据序列,所以常规海面风预报无法像数值预报一样直接应用实况数据进行预报质量的检验评估。本文在详细调研全国海洋预报机构发布的海面风预报产品基础上,针对不同的常规预报内容,制定了一套客观、合理、标准化的处理方法,成功实现了对常规海面风预报的检验,并应用东海区实况数据对常规海面风预报进行了检验测试,通过对存在的问题进行研讨与论证,完善了检验评估规则,提升了其实用性、合理性和适用性。