This paper describes the simulation of the flow of a viscous incompressible Newtonian liquid with a free surface. The Navier–Stokes equations are formulated using a streamline upwind Petrov–Galerkin scheme, and solved on a Q-tree-based finite element mesh that adapts to the moving free surface of the liquid. Special attention is given to fitting the mesh correctly to the free surface and solid wall boundaries. Fully non-linear free surface boundary conditions are implemented. Test cases include sloshing free surface motions in a rectangular tank and progressive waves over submerged cylinders. 相似文献
It is important to estimate hard-to-observe parameters in the ocean interior from easy-to-observe parameters. This study therefore
demostrates a reconstruction of observed temperature and salinity profiles of the sea east of Japan (30°≈40°N, 140°≈150°E).
The reconstruction was done by estimating suboptimal state from several values of the observed profiles and/or sea surface
dynamic height (SDH) calculated from the profiles. The estimation used a variational method with vertical coupled temperature-salinity
empirical orthogonal function (EOF) modes. Profiles of temperature and salinity in the subtropical region are effectively
reconstructed from in situ temperature profile data, or sea surface temperature (SST) and SDH. For example, the analyzed temperature field from SST
and SDH has an accuracy to within 1°C in the subtropical region. Salinity in the sea north of Kuroshio, however, is difficult
to estimate because of its complex variability which is less correlated with temperature than in the subtropical region. Sea
surface salinity is useful to estimate the subsurface structure. We also show the possibility that the estimation is improved
by considering nonlinearity in the equation calculating SDH from temperature and salinity analysis values in order to examine
the misfit between analysis and observation. Analysis using TOPEX/POSEIDON altimetry data instead of SDH was also performed.
This revised version was published online in July 2006 with corrections to the Cover Date. 相似文献
The turbulent motions responsible for ocean mixing occur on scales much smaller than those resolved in numerical simulations
of oceanic flows. Great progress has been made in understanding the sources of energy for mixing, the mechanisms, and the
rates. On the other hand, we still do not have adequate answers to first order questions such as the extent to which the thermohaline
circulation of the ocean, and hence the earth's climate, is sensitive to the present mixing rates in the ocean interior. Internal
waves, generated by either wind or flow over topography, appear to be the principle cause of mixing. Mean and eddy flows over
topography generate internal lee waves, while tidal flows over topography generate internal tides. The relative importance
of these different internal wave sources is unknown. There are also great uncertainties about the spatial and temporal variation
of mixing. Calculations of internal tide generation are becoming increasingly robust, but we do not know enough about the
subsequent behavior of internal tides and their eventual breakdown into turbulence. It does seem, however, that most internal
tide energy flux is radiated away from generation sites as low modes that propagate over basin scales. The mechanisms of wave-wave
interaction and topographic scattering both act to transfer wave energy from low modes to smaller dissipative scales.
This revised version was published online in August 2006 with corrections to the Cover Date. 相似文献
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. 相似文献
A fluorescent sand-tracer experiment was performed at Comporta Beach (Portugal) with the aim of acquiring longshore sediment transport data on a reflective beach, the optimization of field and laboratory tracer procedures and the improvement of the conceptual model used to support tracer data interpretation.
The field experiment was performed on a mesotidal reflective beach face in low energetic conditions (significant wave height between 0.4 and 0.5 m). Two different colour tracers (orange and blue) were injected at low tide and sampled in the two subsequent low tides using a high resolution 3D grid extending 450 m alongshore and 30 m cross-shore. Marked sand was detected using an automatic digital image processing system developed in the scope of the present experiment.
Results for the two colour tracers show a remarkable coherence, with high recovery rates attesting data validity. Sand tracer displayed a high advection velocity, but with distinct vertical distribution patterns in the two tides: in the first tide there was a clear decrease in tracer advection velocity with depth while in the second tide, the tracer exhibited an almost uniform vertical velocity distribution. This differing behaviour suggests that, in the first tide, the tracer had not reached equilibrium within the transport system, pointing to a considerable time lag between injection and complete mixing. This issue has important implications for the interpretation of tracer data, indicating that short term tracer experiments tend to overestimate transport rates. In this work, therefore, longshore estimates were based on tracer results obtained during the second tide.
The estimated total longshore transport rate at Comporta Beach was 2 × 10− 3 m3/s, more than four times larger than predicted using standard empirical longshore formulas. This discrepancy, which results from the unusually large active moving layer observed during the experiment, confirms the idea that most common longshore transport equations under-estimate total sediment transport in plunging/surging waves. 相似文献
A new algorithm is proposed, called the stream function method (SFM) for producing vector current maps from radial data measured
by dual-site high frequency surface wave radar (HFSWR). In SFM, a scalar stream function is constructed under some oceanographic
assumptions. The function describes the two-dimensional (2-D) ocean surface water motion and is used to obtain the distribution
of vector currents. The performance of SFM is evaluated using simulated radial data, which demonstrates that SFM has advantages
over typical vectorial combination methods (VCM) both in error acceptance and robustness, and excels another method based
on least-squares fitting (LSF) in recovering the complicated current models. Furthermore, SFM is capable of providing the
total currents based on radials from single-site radar. We also test the assumptions of horizontal non-divergence in the simulation.
The new algorithm is applied to the field experiment data of Wuhan University’s ocean state measuring and analyzing radar
(OSMAR), collected in the coastal East China Sea during April 11–17, 2004. Quantitative comparisons are given between radar
results by three current algorithms and in-situ current meter measurements. Preliminary analysis of the vertical current shear is given based on the current meter measurements. 相似文献