[ home / oceanography ]
Note: These are no longer available as stated in their announcement about closing down the project.
A synthesis of enhancements and additions to the GFDL MOM package by Australian researchers. This is a series of modifications to various MOM versions as follows:
"ADCIRC is a system of computer programs for solving time dependent, free surface circulation and transport problems in two and three dimensions. These programs utilize the finite element method in space allowing the use of highly flexible, unstructured grids. Typical ADCIRC applications have included: (i) modeling tides and wind driven circulation, (ii) analysis of hurricane storm surge and flooding, (iii) dredging feasibility and material disposal studies, (iv) larval transport studies, (v) near shore marine operations."
A three-dimensional hydrodynamic multi-purpose model for coastal and shelf seas, which is coupled to biological, resuspension and contaminant models, and resolves mesoscale to seasonal scale processes.
"The Bergen Ocean Model is a numerical simulation tool developed by the Institute of Marine Research , Norway and the Hydrodynamics group at Dept. of Mathematics , University of Bergen, with contributions from Dept. of Informatics and Parallab also at the University of Bergen.[http://www.mi.uib.no/BOM/]
The code development was initiated in 1995, with the main motivation to incorporate modern numerical techniques into ocean modeling, and this process is continuously ongoing. The model is implemented in Fortran 90 with source code freely available.
The mathematical basis for the model is the velocity field, pressure, density, salinity and temperature governed by momentum, continuity and conservation equations. The discretization applies finite differences on a staggered grid with a vertical sigma-coordinate representation."
A regional climate system modeling system consisting of
The CLIO model comprises a global, free-surface Ocean General Circulation Model (OGCM) coupled to a comprehensive sea ice model. The OGCM is a primitive equation model adopting the usual set of assumptions, i.e., the hydrostatic equilibrium and the Boussinesq approximation. The sea ice model has a representation of both thermodynamic and dynamic processes. A relatively sophisticated parameterization of vertical mixing, a parameterization of the effect of meso-scale eddies as well as a parameterization of dense water flow down topographic features are included in the latest version of the model. A 3-layer model, which takes into account sensible and latent heat storage in the snow-ice system, simulates the changes of snow and ice thickness in response to surface and bottom heat fluxes. The variation of ice compactness due to thermal processes is a function of the energy balance of the surface layer in the region occupied by leads. For calculating ice dynamics, sea ice is considered to behave as a viscous-plastic continuum. At the ice-ocean interface, the sensible heat flux is proportional to the temperature difference between the surface layer and its freezing point and to the friction. The ice-ocean stress is taken to be a quadratic function of the relative velocity between ice and the uppermost level of the ocean. Considering salt and freshwater exchanges between ice and ocean, brine is released to the ocean when ice is formed, while freshwater is transferred to the ocean when sea ice or snow melts.
"ELCIRC is an unstructured-grid model designed for the effective simulation of 3D baroclinic circulation across river-to-ocean scales. It uses a finite-volume/finite-difference Eulerian-Lagrangian algorithm to solve the shallow water equations, written to realistically address a wide range of physical processes and of atmospheric, ocean and river forcings. The numerical algorithm is low-order, but volume conservative, stable and computationally efficient. It also naturally incorporates wetting and drying of tidal flats. While originally developed to meet specific modeling challenges for the Columbia River, ELCIRC has been extensively tested against standard ocean/coastal benchmarks, and is starting to be applied to estuaries and continental shelves around the world."
"ECOM-si is a 3D ocean circulation model developed principally by Alan Blumberg of HydroQual. It is similar to the Princeton Ocean Model (POM) described in Blumberg and Mellor (1987), but incorporates an implicit scheme developed by Vincenzo Casulli for solving the gravity wave so that the need for separate barotropic and baroclinic time steps is eliminated. ECOM-si is not generally available, but if you are looking for a freely available 3D circulation model for conducting research in the coastal model, I suggest checking out either POM or the very nicely documented and maintained Rutgers/UCLA ROMS model. Note that POM and ROMS are not available for commercial use. More information about modeling applications with ECOM-si can be found on the HydroQual web site."
The numerical code used in FLAME is based on GFDL's Modular Ocean Model MOM 2.1 (Pacanowski 1995), but it has experienced numerous modifications, most notably the addition of parallelization.
One of the main considerations in FLAME has always been to keep the code operational for use on different architectures. At the moment, DEC Alpha workstations, Cray vector and parallel machines and NEC SX5 architectures are supported. Work is in progress to use the FLAME code on other types of machines as well.
FMS is built on top of the MPP modules, a modular parallel computing infrastructure written in Fortran 90. This infrastructure includes:
It is also capable of a high degree of flexibility - the model has been designed with the intention of performing experiments using highly idealised configurations, so it can accept new continental geometry, land surface orography and ocean bottom topography easily; the resolution of the ocean (and to a lesser extent, the atmosphere) is also configurable.
Its principal components are ocean and atmosphere General Circulation Models (GCMs) originally designed for solo operation. The ocean model is MOMA (Webb,93) a derivative of the well-known Modular Ocean Model, a z-coordinate primitive equation model based on the GFDL code. The atmosphere is adapted from IGCM3, a spectral GCM developed at Reading University (Hoskins,75 Forster,00). They are coupled together using OASIS (Terray,00), a flexible coupler that allows the models to pass data between each other whilst running.
Although still beta-level, the model is currently being used by 3 groups (two here at the Southampton Oceanography Centre and one at Liverpool University) and the code is being made available as a tool for the modelling community at large. (11/2/02)
The Atmosphere-Ocean Model is a computer program that simulates the Earth's climate in three dimensions on a gridded domain. The Model requires two kinds of input, specified parameters and prognostic variables, and generates two kinds of output, climate diagnostics and prognostic variables. The specified input parameters include physical constants, the Earth's orbital parameters, the Earth's atmospheric constituents, the Earth's topography, the Earth's surface distribution of ocean, glacial ice, or vegetation, and many others. The time varying prognostic variables include fluid mass, horizontal velocity, heat, water vapor, salt, and subsurface mass and energy fields.
Note: These are all now password protected. If you want to obtain them you need to ask them for access.
Various circulation models developed for the Gulf of Maine and Georges Bank.
A finite element formulation of the non-conservative form of the vertically integrated advection/diffusion/reaction (ADR) equation that tracks any number of different depth-averaged transport variables. This is usually used in conjunction with QUODDY simulations.
A graphical Matlab interface to the C language 2-D quality finite element grid generator Triangle. BatTri performs the mesh editing, bathymetry incorporation and interpolation, provides the grid generation and refinement properties, prepares the Triangle input file, and visualizes and saves the created grid.
A 3-D diagnostic model for continental shelf circulation studies. It solves the linearized shallow water equations forced by tidal or other barotropic boundary conditions, wind or a density gradient using linear finite elements. Solutions are obtained in the frequency domain. A user's manual is available in both PostScript and HTML formats. MATLAB routines to process FUNDY and QUODDY simulations are available at OPNML. See also DROG3D.
A turbulent boundary layer model for the linearized shallow water equations. This is a time-stepping point model which uses linear finite elements to determine the vertical structure of the horizontal components of velocity and density under specified surface forcing. Both a quadratic closure scheme and the level 2.5 closure scheme of Mellor and Yamada are used in this code. The documentation for this is contained within a 45 page report and user's manual available in PostScript format.
A package for modeling 3-D shallow water hydrodynamics with implicit linear triangular finite elements.
In the mixed layer, grid points are placed vertically so that a smooth transition of each layer interface from an isopycnic to a constant-depth surface occurs where the interface outcrops into the mixed layer. HYCOM therefore behaves like a conventional sigma model in very shallow and/or unstratified oceanic regions, like a z-level coordinate model in the mixed layer or other unstratified regions, and like an isopycnic-coordinate model in stratified regions. In doing so, the model combines the advantages of the different types of coordinates in optimally simulating coastal and open-ocean circulation features. The present procedure of driving high-resolution coastal models (which invariably use fixed vertical grids) with output from a basin-scale isopycnic model can be streamlined, since HYCOM will be able to provide the required near-shore data at fixed depth intervals.
"The original Gent-Cane model (Gent and Cane, 1989) was developed for modeling the equatorial pacific on a stretched longitude/latitude A-grid. Fourth order approximations to horizontal spatial derivatives combined with a fourth order (in time) Lorenz cycle provided a high order discretization advantageous for high resolution runs. The original vertical structure was simply a mixed layer at the surface and sigma layers down to a bottom layer of no motion. The surface mixed layer could be either a true layer with no mass exchange with lower layers, or could be kept at a fixed thickness. Restoration of SST at the surface provided the surface heat flux and Shapiro filters, reducing the order near land boundaries, provided a minimal amount of horizontal diffusion. The vertical diffusion of momentum and heat was explicit.
Most of the horizontal features have been kept in the present version. The 4th order time and space (horizontal) discretization has been maintained, the Shapiro filters (reduced, conservative, narrow passage modifications,...) are still used, and most of the old options remain. The only major exception is that the reduced gravity setup (assuming no motion at depth) is not currently supported in this version (but easily could be, if the need arises ...).
Descriptions of most of the features of the current model can be found in
Keith Rodger's documentation. The major changes from the original Gent/Cane
model are the vastly improved I/O handling, a barotropic solver, a new ocean
mixed layer parameterization and an atmospheric mixed layer (AML).
Naomi Naik's barotropic solver allows the model to be run with a constant bathymetry assumption rather than the reduced gravity assumption. Vertical mixing has been generalized to use a combination of convective adjustment with a Kraus-Turner parameterization of wind mixing and Richardson number dependent mixing , the Large et al. K-profile parameterisation (KPP) , or Dake Chen 's ocean hybrid vertical mixing scheme . As either an alternative or in addition to the Shapiro filter, Isopycnal diffusion has been added, with an option to make it purely isopycnal or horizontal. The Gent-McWilliams eddy parameterization is also available, with variable coefficient as suggested by Visbeck et al. Density is computed from both temperature and salinity by a computationally efficient high order approximation to the UNESCO formula, and passive tracers can be implemented by specifying initial conditions in a user supplied data file. Richard Seager's atmospheric mixed layer (AML) computes the surface fluxes rather than restoration to SST and SSS, with data files prescribing wind, air temperature and humidity over land, clouds, solar radiation and precipitation. Thermodynamic ice has been added by Bob Newton and Martin Visbeck as an integral part of the AML. Ice advection has been coded, but has not been tested."
Related (more or less) software includes:
Documentation - A user's manual and a theory guide in French and English, and in PostScript and HTML.
"We have developed a message-passing version of MICOM called MP-MICOM that uses SHMEM on the Cray T3D/T3E and MPI on other machines. We also have a shared-memory, multi-threaded version (SC-MICOM) for SMP clusters that uses MPI to communicate between machines and direct shared memory within each machine."
"The MITgcm (MIT General Circulation Model) is a numerical model for studying the ocean and atmosphere. It is capable of simulating these fluids at a wide range of scales and can resolve many different processes. It has a non-hydrostatic capability (Marshall et al., JGR 1997a & b) and uses the finite volume method to accurately represent the bottom boundary position (Adcroft et al., MWR 1998)."
"The NLOM has become the world's first scalable portable ocean model. It will run efficiently and interchangeably on massively parallel computers (distributed memory - CM5, CRAY T3D/T3E, IBM SP2, SGI Power Challenge Array, Convex Exemplar), multi-processor shared memory computers (CRAY YMP/C90/T90, SGI Power Challenge, Convex Exemplar), or scalar computers (single processor workstations). To obtain a copy of the NLOM and related software, e-mail email@example.com."
"The NCAR CSM Ocean Model (NCOM) is based on the GFDL 's Modular Ocean Model (MOM) 1.1 global oceanic general circulation model with substantial modifications to include improved mesoscale tracer transport, boundary layer mixing, and surface forcing."
PSTSWM is written in Fortran 77 with VMS extensions and a small number of C preprocessor directives. Message passing is implemented using MPI, PICL, PVM, and/or native message passing libraries, with the choice being made at compile time. Additionally, all message passing is encapsulated in three high level routines for broadcast, global minimum and global maximum, and in two classes of low level routines representing variants and/or stages of the swap operation and the send/receive operation. Porting the code to another message passing system requires either porting the PICL, MPI, or PVM libraries and/or implementing the (few) communication routines in PSTSWM using native message passing primitives."
The model depths are based on the DBDB5 data set, with sill depths checked against original surveys.
The model was started from the Levitus annual mean temperature and salinityfields. The surface forcing uses ECMWF monthly mean winds and relaxation to the Levitus seasonal surface temperature and salinity fields.
The model is run on the UK Research Councils' multi-processor Cray-T3D
operated by the University of Edinburgh. The initial model run for
12 model years will be completed by September 1996. The results are
being used to understand the heat flows and movements of different
water types in the ocean. They are also being used to help analyse the
data from WOCE, the World Ocean Circulation Experiment.
The OCCAM ftp directory contains the full set of programs and files produced by the OCCAM project which have been made available to the academic community."
"The regular partitioning of grid based finite difference models for distribution onto parallel processors leads to a characteristic nearest neighbour boundary exchange communications pattern. In general the data structures to be exchanged are not contiguous in memory.
OCCOMM is a low-level communications kernel benchmark which determines the performance of various message passing techniques applied to contiguous, single-strided and double-strided data structures, commonly found in ocean, and other regular grid based, models."
"OPA (acronym for "Ocean PArallelise") is the Ocean General Circulation Model (OGCM) developed by the ECUME team at the Laboratoire d'Oceanographie DYnamiquexi et de Climatologie (LODYC). It is a primitive equation model applied to both regional and global ocean circulation. It is intended to be a flexible tool for studying the ocean and its interactions with the others components of the earth climate system (atmosphere, sea-ice, chemical tracers, ...).
Prognostic variables are the three-dimensional velocity field and the thermohaline variables. They are distributed on a three-dimensional Arakawa-C-type grid using prescribed z- or s-levels. Various physical choices are available to describe ocean physics, including a 1.5 turbulent closure for the vertical mixing, geopotential or isopycnal mixing, eddy induced velocity parameterization, simple bottom boundary layer representation, etc.
OPA is interfaced with several sea-ice models, a passive and biogeochemical tracer model and, via the OASIS coupler, with several Atmosphere General Circulation Models. OPA also has its adjoint and tangent linear models.
OPA can be run on many different computers, including shared and distributed memory multi-processor computers (Cray C98, T3D, T3E, Origin 2000, NEC SX4 and SX5, VPP, ...)."
"The Princeton Ocean Model directory. POM is a sigma coordinate, free surface, primitive equation ocean model, which includes a turbulence sub-model. It was developed in the late 1970's by Blumberg and Mellor, with subsequent contributions from other people. The model has been used for modeling of estuaries, coastal regions and open oceans."
The next generation version of this is called TOMS.
"Poseidon Ocean General Circulation Model (OGCM) is designed with generalized horizontal and vertical coordinates. While the formalism ensures conservation of mass, momentum, heat, salt and other tracers under any grid choice, the specific selection of the grid definition provides additional benefits. In the vertical, we adopt a three-tiered layering system comprised of a bulk turbulent mixed layer, a deep ocean treated with isopycnal layering, and a buffer zone between the two in which a sigma-like layering is employed. All of the three regions can be sub-divided into many (or zero) layers.
The isopycnal region is treated in a quasi-isopycnal fashion, in which layers do not vanish at outcrops, but retain a thin minimum thickness at all grid points. This is similar to the "massless" layer treatment of Bleck and Boudra (J. Phys. Ocean., 1981) , in contrast to the disappearing layer treatment of Oberhuber (J. Phys. Ocean., 1993). The isopycnal layering provides for far better control of diapycnal mixing and obviates the need for expensive tensorial treatment of diffusive processes in the ocean. The treatment of horizontal mixing within the model is implemented with high order Shapiro filtering which is reduced to a flux form. The mixed layer is treated with a bulk turbulence scheme derived from Kraus-Turner (Tellus, 1967) and Niiler-Kraus (The Sea, 1977) and as implemented in time-integral fashion in Schopf and Cane (J. Phys. Ocean., 1983)."
"POSUM (Parallel OSU Model) is an ocean circulation model. Vertically, it is stack of layers of constant orthobaric density. The horizontal discretization is a finite difference approximation to the primitive equations in spherical polar coordinates. POSUM is a non-Boussinesq, free-surface model, employing a stable time-stepping algorithm that uses operator splitting of the baroclinic and barotropic modes (Higdon and de Szoeke 1997). The two modes are stepped forward over the same time interval. The baroclinic mode is solved using explicit methods, and the barotropic mode is solved using an ADI (alternating direction implicit) method (Bates 1984), which is stable and very fast. POSUM is a ``massless layer'' model, meaning that all orthobaric density layers are present at every horizontal grid point, although some may become extremely thin (micrometers or less). To ensure positive definiteness, the layer-thickness equations are solved using a second-order MPDATA scheme (Smolarkiewicz 1984, Smolarkiewicz and Clark 1986, Smolarkiewicz and Margolin, 1993). The MPDATA methods have been carefully reformulated to be compatible with the barotropic-baroclinic split to guarantee mass conservation within each layer (in the absence of in the absence of reversible and irreversible diapycnal fluxes) and total water column mass conservation during the baroclinic integration (Springer and de Szoeke 1999). The MPDATA scheme is also used to advect salinity, an active tracer, as well as any passive tracers.
The model makes allowance for diapycnal mass exchange that occurs in the ocean as a result of microstructure turbulence. This is parameterized in POSUM as diapycnic diffusion. In orthopycnic coordinates, this is represented by a nonlinear diffusion equation which must be solved to predict layer thickness. We have developed a relatively simple, efficient two step scheme for solving the one-dimensional diffusion equation in isopycnic coordinates. In the first step a nonlinear mapping guarantees positivity of the layer thickness and allows the difference equations to be rewritten as a system of linear difference equations. At the surface the diffusivity coefficients of layers contained within the upper several meters are attenuated almost to zero in an implicit way so that their positiveness is guaranteed. The second, explicit step of the algorithm ensures that bottom pressure is not altered by the diffusion process. An important elaboration of this algorithm makes it capable of handling large disparities (orders of magnitude) of diffusivity coefficients between neighboring layers. The difficulty that this presents is that a large-diffusivity layer (an entraining layer, for example) grows at the expense of adjacent low-diffusivity layers, reducing the latter to zero (or worse, negative) thicknesses. This difficulty is overcome by letting the diffusivity coefficient of a layer about to be annihilated grow implicitly to the value of the entraining layer. Then the thickness of the latter layer tends to a small positive limit, and constitutes a front (jump) ahead of the entraining layer. Multiple layers can be thus entrained into the base of the growing layer.
An important feature of any ocean circulation model is its surface mixed layer, which we have implemented in a novel way (de Szoeke and Springer 1999b). We represent the mixed layer by a near-surface vertical interval of high diffusivity (typically several orders of magnitude larger than the weak interior diffusivities); the diffusion algorithm outlined above is then employed to calculate the thicknesses of density layers. At a given location and time, this procedure tends to select a single isopycnal layer (i.e., surface density) to occupy the entire vertical interval of high diffusivity. This is typically separated from the underlying low-diffusivity water column by a multiple density step consisting of several thin layers. At the surface, several infinitesimally thin layers (literally micrometers thick!) may be maintained, depending on the range of density layers pre-specified in the model set-up. (The surface density of the ocean is determined by buoyancy fluxes in the buoyancy budget; superfluously light density layers in the model are automatically shrunk to near zero thicknesses.) The vertical extent of the high-diffusivity region may be determined by macroscopic indicators of surface conditions. At a given instant, the "mixed layer depth" is determined by the KT prescription and the density-layer configuration of the previous time-step. Within this depth, diffusivity is set to a large value.
Pacific Ocean simulations have been conducted using 10 vertical layers and including realistic coastlines and bottom topography (contained in the bottom-most layer). Isopycnal outcropping defines the large scale subtropical and subpolar gyres in the northern hemisphere resulting in a large scale general circulation which agrees well with observations. Recently, the model has been running in a 1/30 x 1/30 global domain with 10-20 vertical layers. The figure above shows the model's sea surface height field (SSH) after 12,000 days of integration.
POSUM is written in Fortran 90. At the moment it uses a set of compiler directives to run on a 33 Gflops 256 node Thinking Machines CM-500e supercomputer located in COAS' Environmental Computing Center. A subset of the code has been ported to High Performance Fortran (HPF) for evaluation on other platforms including an IBM SP2 and a Cray T3E."
"QTCMs are models of intermediate complexity suitable for the modeling of tropical climate and its variability. It occupies a niche among climate models between complex general circulation models and simple models.
The primitive-equation-based dynamical framework is constructed using analytical solutions from the quasi-equilibrium convective parameterization as the first basis function in a Galerkin representation of vertical structure. A uniqueness of the QTCM is its balanced treatment of dynamics and physical parameterizations. It includes a linearized longwave radiation scheme, simple cloud prediction and shortwave radiation schemes, and the Simple-Land (SLand) land model."
The model equations are solved separately for total momentum and vertically integrated momentum and then coupled. The total momentum and tracer equations are time discretized using a third order Adams-Bashforth scheme; the vertical viscosity/diffusion terms are treated implicitly using a Crank-Nicolson scheme. The free-surface and vertically integrated momentum equations are time discretized using a trapezoidal Leapfrog scheme. Horizontal and vertical derivatives are evaluated using finite differences on a staggered horizontal C-grid and a staggered vertical grid.
An array processor version of MOM, with a reduced set of options, has been adapted by Webb (1996) and subsequently named MOMA (Modular Ocean Model - Array processor version). Although not a parallel code as such, MOMA is arranged into columns, so that the model arrays can be vectorised in the vertical and decomposed in the horizontal. In contrast to Bryan's rigid-lid approximation, which requires a stream function formulation for the barotropic (depth-averaged) mode, MOMA adopts a free surface formulation (Killworth et al., 1991). This is considered better suited to parallelism, since the finite difference representation only requires nearest neighbour information and does not require line integrals to be calculated around islands. It is also felt that the free surface formulation will allow future assimilation of altimeter data to be implemented more readily.
At the Southampton Oceanography Centre (SOC) the MOMA code has been used as the base model for the UK Ocean Circulation and Climate Advanced Modelling (OCCAM) programme, the aim of which has been to develop a high resolution model of the World Ocean. As the first truly global, eddy-resolving, ocean model, the OCCAM code is parallelised and optimised specifically for the Cray-T3D (Gwilliam, et al., 1995), the UK national MPP supercomputer at Edinburgh University.
As a spin-off from the OCCAM project, a PVM version of the MOMA code was developed at SOC (Webb et al., 1997). At that time, similar work was undertaken at the University of East Anglia (UEA), where techniques for the parallelisation and optimisation of ocean models were investigated (Beare and Stevens, 1997). As a colloborative venture the two approaches have been combined and the MOMA code has been developed into a general purpose, portable, parallel ocean circulation model. It is this code that has become known as The Southampton - East Anglia (SEA) model.
CICE has several interacting components: a thermodynamic model that computes local growth rates of snow and ice due to vertical conductive fluxes, snowfall and local temperatures; a model of ice dynamics, which predicts the velocity field of the ice pack based on a model of the material strength of the ice; a transport model that describes advection of the areal concentration, ice volumes and other state variables; and a ridging parameterization that transfers ice among thickness categories based on energetic balances and rates of strain. Additional routines prepare and execute data exchanges with an external "flux coupler," which then passes the data to other climate model components such as POP.
An extensible, user-configurable model system for nearshore wave, circulation and sediment processes.
REF/DIF is a phase-resolving parabolic refraction-diffraction model for ocean surface wave propagation.
The SHORECIRC model system consists of two parts:
WAVEWATCH III solves the spectral action density balance equation for wavenumber-direction spectra. The implicit assumption of these equations is that the medium (depth and current) as well as the wave field vary on time and space scales that are much larger than the corresponding scales of a single wave. Furthermore, the physics included in the model do not cover conditions where the waves are severely depth-limited. This implies that the model can generally by applied on spatial scales (grid increments) larger than 1 to 10 km, and outside the surf zone.
[ home / oceanography ]
Last modified or updated: May 4, 2004
Dept. of Oceanography
Texas A&M University