Impact of Ocean Mean Dynamic Topography on Satellite Data Assimilation

The response of an eddy-permitting ocean model to changes imposed by the use of different mean dynamic topographies (MDT) is analyzed in a multivariate assimilation context, allowing the evaluation of this impact, not only on the surface circulation, but also on the interior ocean representation. The assimilation scheme is a reduced-order sequential Kalman filter (SEEK). In a first set of experiments, high resolution sea surface temperature, along-track sea surface height and sea surface salinity from climatology are assimilated into a 1/3° resolution North and Tropical Atlantic version of the HYCOM model. In a second experiment, in situ profile data are assimilated in addition to the surface measurements.

The first set of experiments illustrates that important differences in the representation of the horizontal model circulation pattern are related to differences in the MDT used. The objective of assimilation is to improve the representation of the 3D ocean state. However, the imperfect representation of the mean dynamic topography appears to be an important limiting factor with regard to the degree of realism obtained in the simulated flow.

Vertical temperature and salinity profiles are key observations to drive a general circulation ocean model toward a more realistic state. The second set of experiments shows that assimilating them in addition to sea surface measurements is a far from trivial exercise. A specific difficulty is due to inconsistencies between the dynamic topography diagnosed from in situ observations and that diagnosed from sea surface height. These two fields obtained from different data sources do not contain exactly the same information. In order to overcome this difficulty, a strategy is proposed and validated.     

多源数据推求的西太平洋区域海面动力地形比较分析

简述利用空间大地测量观测数据和海洋水文数据推求海面动力地形的方法。基于EGM96重力场模型和卫星重力恢复的重力场模型GL04C,联合卫星测高平均海面高模型分别推算西太平洋海域的平均海面动力地形,并与根据海洋水文数据推算之结果进行比较分析。结果表明:卫星重力场模型GL04C更好地表现了海面地形的细节特征。卫星重力和卫星测高的联合应用将成为确定海面动力地形的有效途径之一。     

卫星重力梯度测量在海洋科学中的应用分析

简述了卫星重力梯度测量技术的基本原理和GOCE数据特点;基于三个不同的重力场模型,采用不同阶次,联合卫星测高平均海面高模型分别推算出全球海面地形,并对结果作了比较分析;探讨了卫星重力梯度测量技术在海洋科学各相关领域的具体应用前景,指出卫星重力梯度测量技术的发展将为海洋科学发展带来巨大的变化。     

基于POM模式与blending同化法建立中国近海潮汐模型

利用POM海洋数值模式建立了中国近海（2°N-41°N,99°E～132°E）分辨率为5′×5′的潮汐模型,模式采用blending同化法同化了由10年TOPEX／Poseidon测高数据反演的潮汐参数与沿岸52个验潮站观测。精度分析表明建立的潮汐模型的8分潮RSS为12．5cm。     

Velocity Statistics Inferred from the TOPEX/Poseidon-Jason-1 Tandem Mission Data

Using a parallel-track approach to estimate geostrophic surface velocities, an estimate of the statistics of ocean geostrophic surface currents and momentum stresses is provided on a 10 km along-track resolution from the first 49 repeat cycles (16 months) of the Jason-TOPEX/Poseidon tandem altimetric sea surface height (SSH) data. Results are compared with estimates obtained in a traditional way from along-track SSH data at crossover points and with in situ, Acoustic Doppler Current Profiler (ADCP) measurements obtained on board the VOS Oleander along a nominal path connecting Bermuda with the U.S. mainland. Agreements with the Oleander data are reasonable when simultaneous (in space and time) sampling is available. However, amplitudes of parallel-track geostrophic velocity variances are about 25% lower as compared to Oleander measurements which represent geostrophic and ageostrophic flow components. Estimates of velocity variances show clear signs of an anisotropic eddy field in the vicinity of all major current systems. At the same time estimates of Reynolds stresses and eddy momentum fluxes show a convergence of eddy momentum in all those regions, suggesting a forcing of the mean flow by the eddy field there.     

Velocity Statistics Inferred from the TOPEX/Poseidon-Jason-1 Tandem Mission Data

Using a parallel-track approach to estimate geostrophic surface velocities, an estimate of the statistics of ocean geostrophic surface currents and momentum stresses is provided on a 10 km along-track resolution from the first 49 repeat cycles (16 months) of the Jason-TOPEX/Poseidon tandem altimetric sea surface height (SSH) data. Results are compared with estimates obtained in a traditional way from along-track SSH data at crossover points and with in situ, Acoustic Doppler Current Profiler (ADCP) measurements obtained on board the VOS Oleander along a nominal path connecting Bermuda with the U.S. mainland. Agreements with the Oleander data are reasonable when simultaneous (in space and time) sampling is available. However, amplitudes of parallel-track geostrophic velocity variances are about 25% lower as compared to Oleander measurements which represent geostrophic and ageostrophic flow components. Estimates of velocity variances show clear signs of an anisotropic eddy field in the vicinity of all major current systems. At the same time estimates of Reynolds stresses and eddy momentum fluxes show a convergence of eddy momentum in all those regions, suggesting a forcing of the mean flow by the eddy field there.     

The Jason-1 Mission Special Issue: Jason-1 Calibration/Validation

On December 7, 2001, the Jason-1 satellite was successfully launched by a Boeing Delta II rocket from the Vandenberg site in California, USA. Its main mission was to maintain the high accuracy altimeter measurements, provided since 1992 by TOPEX/Poseidon (T/P), ensuring continuity in observing and monitoring the ocean for intraseasonal to interannual changes, mean sea level, tides, and so forth. Despite four times less mass and power, the Jason-1 system has been designed to have the same performances as T/P, measuring sea surface topography at the centimeter level. This new Centre National d'Etudes Spatiales/National Aeronautics and Space Administration (CNES/NASA) mission also provides near real-time data for sea state and ocean forecast. The first 10 months of the Jason mission were dedicated to the verification of the system performance and cross-calibration with T/P measurements. A complete CALVAL plan was conducted by the Science and Project Teams of the mission based on in situ and regional experiments, global statistical approaches, and multisatellite comparisons, taking advantage of the T/P-Jason overlap during the first months of the mission. CALVAL and first science results showed that the Jason-1 performances were compliant with prelaunch specifications. This was a needed preamble before starting the routine phase of the mission in July 2003 with generation and distribution of validated geophysical data records to the whole user community.     

国际全球海洋环流预报系统的现状与展望

回顾近年来国外主要海洋预报业务机构全球海洋环流预报技术的发展现状,特别是最近5～10 a来在全球海洋动力数值预报模式和多源资料同化技术方面的主要进展。结合未来相关学科发展的趋势,围绕观测体系建设、数据共享机制、大数据分析等新技术应用等方面提出国内相关研究工作的发展建议。     

A New Methodology for Incorporating Tide Gauge Data in Sea Surface Topography Models

As part of the Vertical Offshore Reference Frames (VORF) project sponsored by the U. K. Hydrographic Office, a new model for Sea Surface Topography (SST) around the British Isles has been developed. For offshore areas (greater than 30 km from the coast), this model is largely derived from satellite altimetry. However, its accuracy and level of detail have been enhanced in coastal areas by the inclusion of not only the 60 PSMSL tide gauges with long-term records around the coasts of the United Kingdom and Ireland but also some 385 gauges established at different epochs and for different observation spans by the U. K. Admiralty. All tide gauge data were brought into a common reference frame by a combination of datum models and direct GPS observations, but a more significant challenge was to bring all short-term sea level observations to an unbiased value at a common epoch. This was achieved through developing a spatial-temporal correlation model for the variations in mean sea level around the British Isles, which in turn meant that gauges with long-term observation spans could be used as control points to improve the accuracy of Admiralty gauges. It is demonstrated that the latter can contribute point observations of mean sea level (MSL) with a precision of 0.078 m. A combination of least squares collocation and interpolation was developed to merge the coastal point and offshore gridded data sets, with particular algorithms having to be developed for different configurations of coastal topology. The resulting model of sea surface topography is shown to present a smooth transition from inshore coastal areas to offshore zones. Further benefits of the techniques developed include an enhanced methodology for detecting datum discontinuities at permanent tide gauges.     

利用验潮站数据进行三种潮汐模型的精度分析

采用全球分布的565个验潮站水位资料对NAO.99b,CSR4.0和TPXO7.2三种潮汐模型进行精度评估。结果表明:在全球海洋范围内,NAO.99b模型精度最高;在黄海海域,TPXO7.2模型的精度最高;在东海和南海海域,则是NAO.99b模型最优;在深海海域,三种模型精度差异不大;在浅海海域,采用同化方法的潮汐模型比采用经验方法的潮汐模型更有优势。     

卫星遥感海表温度资料和高度计资料的变分同化

利用国家气候中心正在发展的第二代全球海洋资料同化系统(BCC_GODAS2.0),针对多变量同化的协调性问题,发展了一种基于三维变分框架(3DVAR)下的高度计和海表温度(SST)相互约束的同化方法。该方法使海面高度和SST资料在同一个动力约束关系下进行同化。在一般方法中,海面高度和SST观测项是代价函数中2个独立的观测项,海面高度项引入动力高度计算公式,海表温度项用统计关系进行垂向投影。在代价函数的实际求解的计算过程中,虽然其总体积分效应受海面高度观测的约束,但整个水柱中各层温盐分析变量的调整是无序的。针对这个问题,文章提出一种新的同化方案。该方案将SST的观测项并入海面高度观测项中,海面高度的一部分,确切说是上层海洋部分,由SST决定,因此至少在SST的统计关系能影响到深度的上层海洋,在代价函数的求解过程中,温盐的调整是受较强的统计关系约束的,而这种统计关系的有效性已经在很多SST的同化试验中被其他学者广泛应用并证明。利用该方法,对1993—1997年的AVHRR卫星遥感海表温度资料进行变分同化试验,用TAO、OISST和SODA数据集进行检验证明,通过对卫星遥感资料的同化能够有效改进对海洋温度和盐度的估...     

海洋数据同化与数据融合技术应用综述

简述了不同数据同化和数据融合方法在海洋环境监测与预测方面的应用、国内外相关业务单位的海洋分析和预报系统的现状,以及海洋数据同化将来的业务化应用的发展趋势。四维变分和集合卡尔曼滤波正在成为国际上海洋环境分析与预报的主要应用方向,海-气耦合数据同化以及海冰数据同化是目前数据同化方法研究的热点。     

Simultaneous Ocean Wave Measurements by the Jason and Topex Satellites, with Buoy and Model Comparisons

The verification phase of the Jason-1 satellite altimeter mission presents a unique opportunity for comparing near-simultaneous, independent satellite measurements. Here we examine simultaneous significant wave height measurements by the Jason-1 and TOPEX/Poseidon altimeters. These data are also compared with in situ measurements from deep-ocean buoys and with predicted wave heights from the Wave Watch III operational model. The rms difference between Jason and TOPEX wave heights is 28 cm, and this can be lowered by half through improved outlier editing and filtering of high-frequency noise. Noise is slightly larger in the Jason dataset, exceeding TOPEX by about 7 cm rms at frequencies above 0.05 Hz, which is the frequency at which the coherence between TOPEX and Jason measurements drops to zero. Jason wave heights are more prone to outliers, especially during periods of moderate to high backscatter. Buoy comparisons confirm previous reports that TOPEX wave heights are roughly 5% smaller than buoy measurements for waves between 2 and 5 m; Jason heights in general are 3% smaller than TOPEX. Spurious dips in the TOPEX density function for 3- and 6-m waves, a problem that has existed since the beginning of the mission, can be solved by waveform retracking.