Ocean Modelling最新文献

Data assimilation (DA) combines observational data and the dynamical ocean model to forecast the ocean state in a matter that is not possible from either observations or models by themselves. However, the incorporation of data-derived corrections into the model introduces the potential to disrupt the dynamical balance of the model state, leading to initialization shocks. These shocks arise as the model undergoes a process of adjustment to restore the perturbed dynamic balance, involving the generation of spurious near-inertial motions. Notably, the US Navy’s global ocean forecast system strives to mitigate these issues through the implementation of the Incremental Analysis Update (IAU) method, distributing the DA corrections to the model state across a specified time window (3 h in the operational system). Our study shows that, despite the implementation of this 3 h IAU period, the initialization shocks still persist in the model analysis fields. We find that during DA updates, spurious internal waves are generated that are in a broad near-inertial frequency range and propagate long distances from their generation sites in the form of low-mode near-inertial waves. The depth-integrated, time-mean near-inertial kinetic energy in a simulation with DA is 68% higher than in a corresponding forward simulation (free-run, without DA) of the simulation with the same surface wind forcing. The presence of these spurious near-inertial waves disrupts the ocean model energetics, and minimizing them is crucial for using the assimilative model simulations to study small scale/high-frequency ocean dynamics. We also examine a possible way to minimize the spurious internal waves by extending the IAU period in numerical experiments using regional model simulations. We demonstrate that the generation of spurious near-inertial waves can be minimized if we insert increments of smaller magnitude at each time step during the update, which can be achieved by extending the IAU period. Our findings indicate that in simulations with a 24 h IAU period, the variance of near-inertial kinetic energy between the assimilative and forward simulations is reduced to approximately 1%.

The static energy encodes all possible information about the thermodynamics and potential energy (and all related forces) of stratified geophysical fluids. In this paper, we develop a systematic methodology, called static energy asymptotics, that exploits this property for constructing energetically and thermodynamically consistent sound-proof approximations of the equations of motion. By approximating the static energy to various orders of accuracy, two main families of approximations are (re-)derived and discussed: the pseudo-incompressible (PI) approximation and the anelastic (AN) approximation. For all approximations, the background and available potential energies (in Lorenz sense) can be constructed to match their exact counterparts as closely as feasible and to be expressible in terms of the exact (as opposed to ad-hoc) thermodynamic potentials. For hydrostatic motions, the AN approximation (of which the Boussinesq approximation is a special case) has the same structure as that of legacy Seawater Boussinesq primitive equations. The energetics of such models could therefore be made transparently traceable to that of the full Navier–Stokes equations at little to no additional cost, thus allowing them to take full advantage of the Gibbs Sea Water (GSW) library developed as part of the new thermodynamic standard for seawater TEOS-10.

Tides play a crucial role in regulating the dispersal and dynamics of a river plume. However, the impact of tides on the dynamics and transport of freshwater in a medium-scale river plume, particularly with multiple outlets, is still not well understood. Using the Hanjiang River Plume in the northern South China Sea as an example, we analyze the momentum and volume of the plume based on salinity space. We also investigate the effects of tidal advection and tidal mixing. Tidal advection propels plume water from the bulge downstream, resulting in a plume type with intermediate surface-advected and bottom-advected characteristics. Tidal mixing causes the plume to come into contact with the seafloor, leading to bottom-advected plumes. Tidal advection leads to the accumulation of plume water in high salinity space, while tidal mixing mitigates this effect, as the plume water near the estuary with relatively low salinity is effectively mixed. In the absence of tidal forcing, vertical shear is the main contributor to the total freshwater flux. However, when tidal effects are taken into account, the contribution of vertical shear to the total freshwater flux decreases and becomes comparable to the advection term. The downstream buoyant flow is primarily controlled by geostrophic balance. The barotropic current carries freshwater downstream, overpowering the upstream transport by the baroclinic current, resulting in a net downstream freshwater transport. Tidal advection enhances this downstream freshwater transport, while tidal mixing has the opposite effect.

The linear equations of motion are solved to obtain dispersion diagrams with Rayleigh friction ($-\gamma u$) and Laplacian friction ($\nu {\partial }_{xx}u$), the latter being solved numerically. Laplacian friction is more efficient at eliminating the short-wavelength Rossby waves, whereas Rayleigh friction is more effective at dissipating long-wavelength Rossby waves. For Rayleigh friction, short-wavelength Rossby waves do not exist at periods longer than the damping time scale ($1/\gamma$); for Laplacian friction, they do not exist for wavenumbers less than $\sqrt{3}/\left(2\sqrt[3]{\nu }\right)$. For both damping forms, the imaginary wavenumber ($k_i$) no longer separates the upper branch (gravity waves) from the lower branch (Rossby waves) of the real wavenumber ($k_r$), and the waves are damped at all frequencies. The two solutions of $k_r$ do not overlap, and for Rayleigh damping, they meet at $\sqrt{0.5-{\gamma }^2}$, which roughly corresponds to $\sim$13.7 days for very low friction. For very high Rayleigh damping, which is an unrealistic scenario, only gravity waves exist in the dispersion diagram. The consequence of adding friction, even if negligible, is that the discontinuity evident in the inviscid solution at the critical latitude no longer holds, or the critical latitude ceases to exist.

The prediction of ocean temperature using sea surface data is crucial for studying ocean-related events and climate change. However, current temperature predictions mainly focus on surface data and rarely consider the temporal relationship of ocean temperature. In this study, we propose a novel deep-learning model to predict ocean temperature for the next two months, which fully considers both temporal and spatial features. The input consists of satellite remote sensing data from the past month, including weekly sea surface temperature, salinity, height, and velocity. The model consists of four modules: convolutional, attention, residual, and transposed convolutional. We compare the model estimation with reanalysis data and conduct temporal, spatial, and vertical distribution analyses. The results demonstrate that the model can accurately predict ocean temperature at different lead time, depths, and locations. Finally, we compare the predicted temperature with actual sea observations to ensure the model's good performance in practical applications. This study shows the tremendous potential of the proposed model in predicting 4-D ocean temperature, providing powerful data support for ocean-related events and climate change research.

Storm surges can give rise to extreme floods in coastal areas. The Norwegian Meteorological Institute (MET Norway) produces 120 h regional operational storm surge forecasts along the coast of Norway based on the Regional Ocean Modeling System (ROMS), using a model setup called Nordic4-SS. Despite advances in the development of models and computational capabilities, forecast errors remain large enough to impact response measures and issued alerts, in particular, during the strongest storm events. Reducing these errors will positively impact the efficiency of the warning systems while minimizing efforts and resources. Here, we investigate how forecasts can be improved with residual learning, i.e., training data-driven models to predict the residuals in forecasts from Nordic4-SS. A simple error mapping technique and a more sophisticated Neural Network (NN) method are tested. The simple error mapping technique provides a reduction in the Root Mean Square Error (RMSE) of less than 4%. Using the NN residual correction method, the RMSE in the Oslo Fjord is reduced by 36% for lead times of one hour, 9% for 24 h, and 5% for 60 h. Therefore, the residual NN method is a promising direction for correcting storm surge forecasts, especially on short timescales. Moreover, it is well adapted to being deployed operationally, as (i) the correction is applied on top of the existing model and requires no changes to it, (ii) all predictors used for NN inference are already available operationally, (iii) prediction by the NNs is very fast, typically a few seconds per station, and (iv) the NN correction can be provided to a human expert who may inspect it, compare it with the model output, and see how much correction is brought by the NN, allowing to capitalize on human expertise as a quality validation of the NN output. While no changes to the hydrodynamic model are necessary to calibrate the neural networks, they are specific to a given model and must be recalibrated when the numerical models are updated.

Near real-time surface current measurements from shore-based high-frequency (HF) radars have increasingly proved to be an essential observation for ocean data assimilation (DA) into operational forecasting systems. For the first time in Brazil, a high-resolution operational system was developed assimilating HF ocean currents data. The system comprises a well known ocean model, the Regional Ocean Modeling System (ROMS), applied to the Southeastern Brazilian shelf and oceanic regions. The ROMS Restricted B-preconditioned Lanczos 4D-variational DA method is employed using real-time coastal radar, remote sensing, and in situ observations, and the DA solution is used as initial fields to produce hourly forecasts for the next two days. The performance of the system in providing accurate forecasts by using this source of initial condition (IC) was evaluated in an experiment in which multiple sources of IC were used. In situ and remote sensing data were used to assess the quality of predictions obtained in the forecasting experiments. The results indicate that the employed DA technique significantly reduced the misfit between model and assimilated observations, leading to improved forecast results. By using this IC, the system was capable to provide forecasts with errors reduced by up to 85%, 14%, and 12%, respectively for sea surface temperature, velocities, and heights, compared to forecasts based on global models. The system was also able to accurately predict the positioning and intensity of the Brazil Current flow and its spatiotemporal variability along the studied region.

The Southwestern Atlantic Continental Shelf (SWACS) is a large oceanic region with remarkably barotropic dynamics. Several scientific studies have described how processes, such as tide or surface winds, affect the variability of the sea surface height and currents. However the tidal dynamics has not received attention for at least the last 15 years, in spite of their importance for both local and global dynamics. Since the last works, the amount of available observations and numerical models (physics, resolution, numerics, etc.) have all greatly improved. In this context, data assimilation (DA) becomes an relevant tool to merge both the observations and the model solutions, producing a better representation of the regional processes. Particularly, DA provides, in addition, an objective methodology to calibrate model parameters. Thus, the aim of this work is to perform, for the first time for this outstanding region, the calibration of the numerical model bottom friction coefficient ($c_D$) by means of DA; then, the opportunity of a better simulation is seized to update the description of tidal dynamics. The spatial variability of the derived $c_D$ is consistent with the bathymetry, with a mean value of $2.0\times 10^{-3}$ along the coast and $2.5\times 10^{-3}$ nearby the shelf-break. Results show that the incorporation of a spatially varying $c_D$ improves the representation of the tidal amplitude and phase compared to the case when it is considered homogeneous, and drives in a single model to results of a better quality than previous nested models with much larger resolution. The optimal representation of the regional tide with a single model allowed us to provide a deeper, improved and novel description of the tidal dynamics. It was found that the energy enters the domain not only from the south but also from the north, being the flux to the north two orders of magnitude larger; those two fluxes produce an cyclonic circulation consistent with the behaviour of the SWACS as a semidiurnal tidal resonant canal theoretically proposed by Webb (1976). This explains why the energy flux is, by far, domained by the potential energy and the large amplitudes of the tide. Finally, a remaining and weaker branch exits along the coast; it enters the Río de la Plata Estuary from its southwesternmost tip and travels upstream along the Argentinean coast, reaching the upper est