高程异常模型的已知点框架约束

在确保GPS/水准点成果可靠的前提下,用其对重力法计算所得的高程异常模型进行框架约束,以削弱重力系统与高程和坐标系统间的系统差,实现系统转换,提高最终模型的精度。通过对不同离散数据网格化方法的比较,给出相对计算快捷、精度高的框架约束方法。     

基于GIS的琼州海峡潮流空间插值模型比较研究

为了客观准确地反映海洋潮流空间分布特征,文中选取琼州海峡表层大潮流速数据,运用泛克里金插值法中的球状、指数、高斯和有理二次方程式4种不同半变异模型进行拟合插值,采用交叉验证和点验证法,通过对比4种半变异模型4个指标精度,研究结果表明在琼州海峡区域,有理二次方程式模型对表层大潮流速拟合效果最优,该研究为描述琼州海峡潮流空间分布和建立精确插值模型提供依据。     

最优插值法用于天气雷达测定区域降水量

本文通过数据模拟和实测资料分析，表明用最优插值法测量区域降水量的测量精度和变分法相当，而最优插值法对雨时计的分布和密度要求较低，计算处理也简单。     

几种卫星测高海洋重力场模型精度比较分析

以中国南海部分海域作为试验区,依据测区船载重力测量结果分别对S&S V28.1、S&S V29.1、DTU15和DTU17卫星测高海洋重力场模型开展了精度评估与分析工作。首先利用EIGEN6C4超高阶位系数模型计算得到的参考重力异常,使用多项式拟合技术对船载重力异常进行了长波误差改正;然后借鉴3σ准则进一步控制船载重力测量结果质量;最后以RMS作为精度主要评价指标,完成了对4种海洋重力场模型精度评估。结果表明,4种海洋重力场模型检核精度整体相当,其中S&S V29.1检核精度略高于S&S V28.1,4种重力异常模型在试验海域精度大致在1.5~6mGal之间。试验结论对于后续重力场模型应用具有一定参考价值。     

空间内插法在航道测量水深检测中的应用

针对目前航道测量水深检测中主要依赖人工判读比对的主观随意性问题,探讨空间插值法在水深检测中应用的可行性。选取地形条件各异的测点样本数据,用反距离加权插值法、克里格插值法、样条函数插值法、自然邻点插值法对其插值分析。通过实验比较发现,需针对具体的地形特征选择最优空间插值法,对河床地形起伏大、水深测点密度较低的区域,空间插值误差的比对超限比例会超出规范允许的限值,故较难进行实际应用;反之,空间插值法具备应用可行性。     

基于DTU重力数据检查海洋重力测量成果的质量

在海上复杂的环境下进行重力测量,重力仪及技术人员的操作难免会出现各种各样的问题,导致测量的成果精度降低,甚至不符合规范要求。目前检查海洋重力测量成果质量的主要方法是内符合法,主要采用主测线、检测线交叉点的差值,或重复测线相邻点的差值来评估测量成果的质量。介绍了内符合法存在的弊端,比较了DTU重力数据与重力仪实测数据的精度,叙述了基于DTU重力数据检查海洋重力测量成果较大系统差的可行性,以模型数据为外符合数据源来检查实测成果的质量,作为内符合法的补充,是十分必要的。     

高精度测高重力场反演南海海底地形

随着测高技术的不断发展,测高海洋重力场的精度不断提高,由其反演计算的海底地形精度也相应提高。利用Sandwell 2014年最新发布主要由Jason-1末期大地测量任务和Cryosat-2观测资料反演的测高重力场V23.1,采用重力地质法(GGM)反演了中国南海海域海底地形模型,并对结果进行了精度评价。该过程中首先直接计算一系列密度差下的反演结果,并通过船测数据检核选定优化的密度差范围,然后利用向下延拓的方法确定了最优密度差异常数为7g/cm3。与船测水深数据相比,反演得到的GGM模型与检核点船测水深数据差值的标准差达到了±70.32m;此外,还计算了由测高重力异常V15.2反演的海底模型,比较这两个模型发现:测高重力场短波部分的改善对海底地形反演精度的提升作用有限,为得到更高精度的海底地形模型,需引入短波更为敏感的重力梯度资料。     

DGS AT1M-3海洋重力仪的应用及精度评估

为了评估DGS AT1M-3海洋重力仪的精度,首先通过分割的有效重力测线进行重力异常质量的内符合精度评价,得出该海洋重力仪测量精度符合海洋调查规范要求。然后与Sandwell v23测高卫星重力异常进行比对分析,对重力异常质量进行外符合精度评价,可看出DGS测量异常与Sandwell卫星测高异常在整体变化趋势上基本一致。最后以西南印度洋断桥热液区为例,简要阐述DGS所测重力数据的有效性。通过评估,可知该海洋重力仪在动态环境下的工作性能较好,测量精度较高,为今后的海洋矿产资源勘探提供新的测量工具。     

一种经验相似方法在沿海海岛阵风预报中的应用

近地面阵风的精确估算对航海航空临界危险天气确定和防范建筑物和农作物损害等具有重要意义。利用2006—2012年的NCEP/NCAR再分析数据、舟山群岛68个自动站的小时极大风速数据,通过定义大气环流形势相似指数和地面风场相似指数,建立一种阵风经验统计降尺度映射预报模型。利用模型对2012年7——12月一天2次(08时和20时)进行试报,并对比评估其与权重插值法的预报,结果表明,经验预报方法除在ScAn(均值评估)评分与权重插值法接近外,其他评分均占优,尤其ScHi(极值击中率评估)和ScVar(方差评估)评分优势明显,ScDisCen(大风区位置重心)的评分也提高了约10%。说明模型可以在充分发挥了经验统计降尺度技术的优点基础上,结合模式精细化预报,较好的将经验数据映射到预报空间,使预报数据空间分布与地形一致,弥补单纯动力模式的不足。     

测量系统差对大地水准面精度影响分析

从边值理论出发,推导了顾及Stokes核函数改化的观测值系统误差传播公式,基于超高阶重力场模型设计了仿真试验,以定量分析顾及核函数改化的重力异常系统差对大地水准面计算精度影响。仿真试验结果表明,重力异常系统差对大地水准面的影响表现为系统性特征,顾及核函数改化的Stokes积分公式能显著改进大地水准面计算精度,即使重力异常系统误差达3 mGal,大地水准面精度也优于1 cm,这些成果将为大地水准面精化工程的方案优化和质量评估提供有益参考。     

Mapping Submerged Aquatic Vegetation with GIS in the Caloosahatchee Estuary: Evaluation of Different Interpolation Methods

This article evaluates different spatial interpolation methods for mapping submerged aquatic vegetation (SAV) in the Caloosahatchee Estuary, Florida. Data used for interpolation were collected by the Submersed Aquatic Vegetation Early Warning System (SAVEWS). The system consists of hydro-acoustic equipment, which operates from a slow-moving boat and records bottom depth, seagrass height, and seagrass density. This information is coupled with geographic location coordinates from a Global Positioning System (GPS) and stored together in digital files, representing SAV status at points along transect lines. Adequate spatial interpolation is needed to present the SAV information, including density, height, and water depth, as spatially continuous data for mapping and for comparison between seasons and years. Interpolation methods examined in this study include ordinary kriging with five different semivariance models combined with a variable number of neighboring points, the inverse distance weighted (IDW) method with different parameters, and the triangulated irregular network (TIN) method with linear and quintic options. Interpolation results were compared with survey data at selected calibration transects to examine the suitability of different interpolation methods. Suitability was quantified by the determination coefficient (R2) and the root-mean-square error (RMSE) between interpolated and observed values. The most suitable interpolation method was identified as the one yielding the highest R2 value and/or the lowest RMSE value. For different geographic conditions, seasons, and SAV parameters, different interpolation methods were recommended. This study identified that kriging was more suitable than the IDW or TIN method for spatial interpolation of all SAV parameters measured. It also suggested that transect data with irregular spatial distribution patterns such as SAV parameters are sensitive to interpolation methods. An inappropriate interpolation method such as TIN can lead to erroneous spatial representation of the SAV status. With a functional geographic system and adequate computing power, the evaluation and selection of interpolation methods can be automated and quantitative, leading to a more efficient and accurate decision.     

Mapping Submerged Aquatic Vegetation with GIS in the Caloosahatchee Estuary: Evaluation of Different Interpolation Methods

This article evaluates different spatial interpolation methods for mapping submerged aquatic vegetation (SAV) in the Caloosahatchee Estuary, Florida. Data used for interpolation were collected by the Submersed Aquatic Vegetation Early Warning System (SAVEWS). The system consists of hydro-acoustic equipment, which operates from a slow-moving boat and records bottom depth, seagrass height, and seagrass density. This information is coupled with geographic location coordinates from a Global Positioning System (GPS) and stored together in digital files, representing SAV status at points along transect lines. Adequate spatial interpolation is needed to present the SAV information, including density, height, and water depth, as spatially continuous data for mapping and for comparison between seasons and years. Interpolation methods examined in this study include ordinary kriging with five different semivariance models combined with a variable number of neighboring points, the inverse distance weighted (IDW) method with different parameters, and the triangulated irregular network (TIN) method with linear and quintic options. Interpolation results were compared with survey data at selected calibration transects to examine the suitability of different interpolation methods. Suitability was quantified by the determination coefficient (R2) and the root-mean-square error (RMSE) between interpolated and observed values. The most suitable interpolation method was identified as the one yielding the highest R2 value and/or the lowest RMSE value. For different geographic conditions, seasons, and SAV parameters, different interpolation methods were recommended. This study identified that kriging was more suitable than the IDW or TIN method for spatial interpolation of all SAV parameters measured. It also suggested that transect data with irregular spatial distribution patterns such as SAV parameters are sensitive to interpolation methods. An inappropriate interpolation method such as TIN can lead to erroneous spatial representation of the SAV status. With a functional geographic system and adequate computing power, the evaluation and selection of interpolation methods can be automated and quantitative, leading to a more efficient and accurate decision.     

Pragmatic aspects of uncertainty propagation: A conceptual review

When quantifying the uncertainty of the response of a computationally costly oceanographic or meteorological model stemming from the uncertainty of its inputs, practicality demands getting the most information using the fewest simulations. It is widely recognized that, by interpolating the results of a small number of simulations, results of additional simulations can be inexpensively approximated to provide a useful estimate of the variability of the response. Even so, as computing the simulations to be interpolated remains the biggest expense, the choice of these simulations deserves attention. When making this choice, two requirement should be considered: (i) the nature of the interpolation and (ii) the available information about input uncertainty. Examples comparing polynomial interpolation and Gaussian process interpolation are presented for three different views of input uncertainty.     

有限元方法在台湾以东海域黑潮流速计算中的应用

1956-1960年间由Turner[13],Clough[16]等引入有限元概念、方法以来,首先在结构分析领域内取得极大进展.与此同时,它也应用于固体力学的许多领域[2,4,7].在七十年代,有限元方法较广泛地应用于流体力学的许多领域[3,11,12].最近几年来,有限元方法也应用于物理海洋学中的一些问题中,例如湖泊与海洋环流问题[14,15,19],潮汐与风暴潮问题[17]等.本文从流体力学原理出发,应用有限元理论建立了黑潮的三维流场有限元方程组.采用了海流理论中常见的物理模式[1,5,18],考虑了压力梯度、柯氏力、重力、水平与垂直方向上的湍流摩擦效应,而不考虑惯性力的影响.本文对所应用的有限元几何形状——三角棱柱进行了插值误差分析,证明了三角形棱面不能出现较大的钝角,也证明了完备性的问题.并将动力计算和有限元二种方法计算的流速进行了比较.     

中国沿岸海域深度基准框架点布设研究

根据已有深度基准面模型,研究了深度基准面在不同方向的非线性变化程度,并且从深度测量的极限误差出发,推算得到深度基准面内插的极限精度要求,二者结合确定深度基准面线性内插的有效范围。实验结果证实,潮波传播特征差异会导致深度基准面线性内插有效范围的差异。最后,结合深度基准面线性内插的有效范围,在渤海湾部分沿岸海域设计了框架点的布设方案。     

GPS网络RTK内插算法分析与比较

在GPS网络RTK中,利用不同的内插算法,生成的流动站误差改正数不同,它们对定位结果的影响也不同。通过算例分析表明,选择合理的内插算法,不仅可以生成与流动站真实误差较为接近的误差改正数,还可以提高整周模糊度的解算成功率和导航定位的精度。     

基于BP神经网络的高频地波雷达海流空间插值

高频地波雷达是海洋环境监测的重要手段，当前已经实现对海流的业务化观测，但是外部因素常引起海流空间探测的不连续性。为解决此问题，尽量保障区域数据的完整性和准确性，本文将BP神经网络技术与空间插值相结合，建立了海流的BP神经网络插值模型，并进行了针对实测数据的缺失插值仿真，通过与反距离权重法和线性插值法插值结果的对比，分析该模型在区域海流大面积缺失、流速整体较大和流速整体较小3个方面的性能。结果表明，BP神经网络插值模型的海流预测效果明显优于其他两种方法，且在流场数据大范围缺失下也取得了良好的效果。     

黄海渔业资源密度空间插值方法的比较研究

空间插值是渔业生态学研究,特别是海洋生态系统模型构建中的常用工具。为找到适合黄海渔业资源密度的插值方法,本文对四种常用的空间插值方法进行了比较研究。四种插值方法分别为反距离加权（Inverse distance weighted,IDW）、全局多项式（global polynomial interpolation,GPI）、局部多项式（local polynomial interpolation,LPI）和普通克里格（ordinary kriging,OK）。采用交叉验证分析（cross-validation diagnostic）确定不同插值方法的准确度。同时,结合渔业资源密度空间分布的可视化表达,比较不同插值方法结果的优劣。结果表明,原始的生物学数据具有典型的非正态分布特征,对数转化方法能让其显著地接近或符合正态分布。四个调查阶段中,指数模型在2014年8月和10月为最佳半方差函数,2015年1月和5月数据具有明显的纯块金效应。配对样本T检验表明,预测数据和实测数据之间没有显著差异（P>0.05）。交叉验证结果表明,OK在2014年10月插值效果最好,IDW在剩余的三个航次表现最佳,GPI和LPI表现效果不佳。OK在空间分布的可视化表达方面效果最好,既不像IDW一样有较多的“牛眼”效应,也不像GPI和LPI一样具有过多的平滑效果。然而,纯块金效应的存在有时会影响OK的应用。综上所述,我们推荐操作简单且处理快速的IDW作为黄海渔业资源密度常规的插值方法。     

Evolving a Bayesian network model with information flow for time series interpolation of multiple ocean variables

Based on Bayesian network (BN) and information flow (IF), a new machine learning-based model named IFBN is put forward to interpolate missing time series of multiple ocean variables. An improved BN structural learning algorithm with IF is designed to mine causal relationships among ocean variables to build network structure. Nondirectional inference mechanism of BN is applied to achieve the synchronous interpolation of multiple missing time series. With the IFBN, all ocean variables are placed in a causal network visually, making full use of information about related variables to fill missing data. More importantly, the synchronous interpolation of multiple variables can avoid model retraining when interpolative objects change. Interpolation experiments show that IFBN has even better interpolation accuracy, effectiveness and stability than existing methods.     

Modeling hurricane waves and storm surge using integrally-coupled,scalable computations

The unstructured-mesh SWAN spectral wave model and the ADCIRC shallow-water circulation model have been integrated into a tightly-coupled SWAN + ADCIRC model. The model components are applied to an identical, unstructured mesh; share parallel computing infrastructure; and run sequentially in time. Wind speeds, water levels, currents and radiation stress gradients are vertex-based, and therefore can be passed through memory or cache to each model component. Parallel simulations based on domain decomposition utilize identical sub-meshes, and the communication is highly localized. Inter-model communication is intra-core, while intra-model communication is inter-core but is local and efficient because it is solely on adjacent sub-mesh edges. The resulting integrated SWAN + ADCIRC system is highly scalable and allows for localized increases in resolution without the complexity or cost of nested meshes or global interpolation between heterogeneous meshes. Hurricane waves and storm surge are validated for Hurricanes Katrina and Rita, demonstrating the importance of inclusion of the wave-circulation interactions, and efficient performance is demonstrated to 3062 computational cores.