• No results found

Numeric ocean models ‐A short description of 5 models

Ocean models, or general circulation models, are numerical models with a focus on the properties of oceans and their circulation. The models solve the same set of primitive equations for motion of a fluid element on a rotating sphere, and differs mainly in numerical methods for solving the equations and in coordinate system used. Most models, if not all, are written in Fortran F90. All the models can be run in parallel on many processors on

super computers. All models have opportunity to include sea ice, biology and assimilation of measurements, and those extra options will have differences between them as the ocean models do.

Ocean models play a large role in aiding our understanding of the ocean’s influence on weather and climate.

There are many ocean models in use today, and five of them are briefly described here. Some other commonly used models are listed at the end of the document. The first three models described, ROMS, NEMO and FV-COM, are open source models, with freely available source code. The last two models, SINMOD and MIKE3 are not freely available. MIKE3 is a commercial model with licences for purchase, and SINMOD is a research and development model for SINTEF.

6.3.1 ROMS

”Regional Ocean Modelling System” (ROMS, http://myroms.org) is a three-dimensional current model de-veloped by Rutgers University, University of California Los Angeles and contributors worldwide (Shchepetkin and McWilliams, 2005). Sigma coordinates in the vertical and orthogonal curvilinear coordinates on a staggered Arakawa C-grid in the horizontal is used. Both Cartesian and spherical coordinates can be used. ROMS has been widely used globally for the last∼15 years. There are lots of possibilities in the specification of the model, so the users can choose methods for horizontal and vertical mixing of momentum, turbulence, horizontal advection, lower boundary layer, stability and more.

The terrain-following sigma coordinates enables the model to have the same number of vertical layers both in shallow seas and in the deep ocean, but the trade off is that the topography can not be too steep (like in Norwegian fjords). The result is a smoothed version of the bathymetry in such areas to avoid unphysical currents along the sloping coordinate lines. The model is well suited for large scale simulations. A lot of the data in the former European MyOcean project and the associated web portal was produced with the ROMS model. It is also used in COAWST:CoupledOceanAtmosphereWaveSedimentTransport Modeling System (Warner et al., 2010).

6.3.2 NEMO

Nucleus for European Modelling of the Ocean (NEMO) is a state-of-the-art modelling framework of ocean re-lated engines that is developed as a collaboration between six European institutes (CMCC (The Euro-Mediterranean Center on Climate Change, Italy), CNRS (The National Center for Scientific Research, France), INGV (National Institute of Geophysics and Volcanology), Mercator Ocean (France), Met Office (UK) and NERC (Natural En-vironment Research Council, UK). NEMO is an ocean modelling framework which is composed of ”engines”

nested in an ”environment”. The ”engines” provide numerical solutions of ocean, sea-ice, tracers and bio-chemistry equations and their related physics. The ”environment” consists in reference configurations, pre- and post-processing tools, interface to the other components of the Earth system, user interface, computer dependent functions and documentation of the system.

NEMO allows several ocean related components of the Earth system to work together or separately (i.e ”stan-dalone mode”). It also allows a two-way nesting via the AGRIF software. It is interfaced with the other com-ponents of the Earth system (atmosphere, land surfaces, ...) via the OASIS coupler.

OPA is the physical ocean component of NEMO containing the dynamics and thermodynamics. OPA is prim-itive equation model adapted to regional and global ocean circulation problems down to kilometric scale. Pro-gnostic variables are the three-dimensional velocity field, a linear or non-linear sea surface height, the temper-ature and the salinity.

In the horizontal direction, the model uses a curvilinear orthogonal grid and in the vertical direction, a full or partial step z-coordinate, or s-coordinate, or a mixture of the two (z* vertical coordinates is also available). The distribution of variables is a three-dimensional Arakawa C-type grid. Various physical choices are available to describe ocean physics.

The range of applications includes oceanographic research, operational oceanography, seasonal forecast and (paleo) climate studies. Used by a large community of users since 2008. The global simulations available at www.marine.copernicus.eu are performed with NEMO.

6.3.3 FVCOM

FVCOM (Finite Volume Community Ocean Model) is developed at The Marine Ecosystem Dynamics Modeling Laboratory at University of Massachusetts-Dartmouth (USA). The model is an unstructured grid, Finite-Volume, primitive equation Community Ocean Model that is well suited for simulating the circulation and ecosystem dynamics from global to estuarine scales, particularly for regions characterized by irregular complex coastlines, islands, inlets, creeks, and inter-tidal zones. The horizontal grid is composed of triangles rather than squares, which ensures a geometric flexibility for a closer fitting to the coastal boundary. The vertical levels are terrain-following (sigma). FVCOM solves the governing equations on Cartesian or spherical coordinates in integral form by computing fluxes between non-overlapping horizontal triangular control volumes.

The model has been set up for small (bay) to global scale. In the global setup, a hybrid coordinate system in the vertical is used with z-coordinates closest to the surface and the bottom, and sigma-coordinates in between.

The current version of FVCOM is fully coupled ice-ocean-wave-sediment-ecosystem model system.

It handles grid cells that are wet/dry depending on water level (tides), and this makes it particularly useful where such processes are important, like some estuaries, river outlets and areas of large tidal difference. FVCOM is more costly in terms of CPU time than many other models.

6.3.4 SINMOD

The SINMOD model system is an ocean model with modules for including the ecosystem up to zooplankton (Slagstad and McClimans, 2005). The model has been in continuous development and use at SINTEF since 1987. Horizontally the model uses a structured grid, and z-coordinates in the vertical. Originally developed for the Barents Sea, SINMOD has been used mostly for the Nordic Seas/Arctic Ocean, but can easily be set up anywhere with coupling to a global model (i.e. www.marine.copernicus.eu). SINMOD uses an Arakawa C-type grid like all the other models.

SINMOD has sea ice, sedimentation, and ecosystem integrated, and functionality for including waves (SWAN model) is under development. SINMOD is often used to provide input data to other models developed at SIN-TEF, like DREAM (particle model) and OSCAR (oil spill).

The model is well suited for regional to local studies, as a global setup is not available.

6.3.5 MIKE3

MIKE 3 is a component of the MIKE by DHI software. It is a three-dimensional hydrodynamic model based on a flexible mesh approach and it has been developed for applications within oceanographic, coastal and estuar-ine environments. The model is based on the numerical solution of the 3D incompressible Reynolds Averaged Navier–Stokes equations (RANS) invoking the assumption of Boussinesq and hydrostatic pressure approxima-tion. The model is module based, so one buys the modules one need. There is a choice of grids available; single, multiple and flexible mesh. All versions can run on Windows, and the latter can also run on Linux (and hence on a High Performance Computing system).

Mike3 is a commercial model and support is available. There is a wide range of modules available.

HYCOM (Hybrid Coordinate Ocean Model), developed at University of Miami, USA.

POM (the Princeton Ocean Model), developed at Princeton University, New Jersey, USA.

MPI-OM (The Max-Planck-Institute Global Ocean/Sea-Ice Model, Hamburg, Germany).