Over the past decades a variety of ocean general circulation models have been developed, distinguishing themselves principally in their choice of discretization for the vertical dimension. The first type of ocean model to be developed was the z-level model (Bryan, 1969) in which the vertical is discretized based on constant geopotential surfaces. Next an isopycnic-coordinate model was presented (Bleck and Boudra, 1981) in which the vertical is divided into layers of constant density. The motivation for this layered approach is the belief that, below the surface mixed layer, the flow in the ocean occurs primarily along constant density surfaces and not along constant geopotentials. Subsequently, sigma-coordinate models (Blumberg and Mellor, 1987) were developed in which the vertical coordinate is bathymetry-following. The motivation being the desire not to lose vertical resolution over shallow, continental shelf regions as may occur with z-level or isopycnic-coordinate models.
Each approach to vertical discretization has strengths and weaknesses, and the Arctic with its broad continental shelves, steep bathymetry, strong pycnocline, and deep basins, provides an ideal test-bed for intercomparing the performance of all these model classes.
In our intercomparison project, z-coordinate models dominate and are represented by the Naval Postgraduate School (NPS, USA), Alfred-Wegener Institute (AWI, Germany), Institute of Ocean Sciences (IOS, Canada), and University of Washington (UW, USA) models. Practically, all of these models are based on the same original code (Bryan, 1969) but are, nonetheless, sufficiently distinct in their detailed treatment of physical processes as to warrant intercomparison. As an example, the Institute of Ocean Sciences model replaces the traditional Newtonian formulation of viscous damping with an eddy-topography rectification, a parameterization known as the Neptune effect. A variant of the most-widely used isopycnic model (Bleck and Boudra, 1981) is represented in our study by the New York University (NYU, USA) model. We also employ two versions of a sigma-coordinate model (Blumberg and Mellor, 1987) represented by the Goddard Space Flight Center (GSFC, USA) and International Arctic Research Center (IARC, USA) models.
The basic configuration for all participating models is to have their ocean model coupled to a sea-ice model and to be driven by specified atmospheric forcing fields.
| AOMIP Model ID | AWI | GSFC | IARC | ICMMG | IOS | LLN |
| Home Institute | Alfred Wegener Institute | Goddard Space Flight Center | International Arctic Research Center | Institute of Computational Mathemetics and Mathematical Geophysics | Institute of Ocean Sciences | Louvain La Neuve |
| Ocean Model Pedigree | MOM | POM | POM | Finite Elements | MOM | OPA |
| Coupled Sea-Ice Model | Yes | Yes | Yes | Yes | Yes | Yes |
| AOMIP Model ID | NPS | NYU | RAS | UW |
| Home Institute | Naval Postgraduate School | New York University | Russian Academy of Sciences | University of Washington |
| Ocean Model Pedigree | MOM | MICOM | Finite Element | MOM |
| Coupled Sea-Ice Model | Yes | Yes | Yes | Yes |
There are a large number of parameters and parameterizations in use in the various AOMIP models. We present below a summary of these values for each of the AOMIP model components: ocean, sea-ice, atmosphere, river-runoff, and coupler. Model descriptions are also found on the PlanetWater website.
| Ocean | Sea-Ice | Atmosphere | River-Runoff | Coupler |
| Grid | Grid | All | All | All |
| Bathymetry | Thermodynamics | |||
| Surface Fluxes | Dynamics | |||
| Lateral Fluxes | ||||
| Surface Mixed Layer | ||||
| Bottom Boundary Layer | ||||
| Equation-of-State | ||||
| Momentum | ||||
| Tracers | ||||
| Convection | ||||
| Tides |
|
. © David Holland. All Rights Reserved. |
|
If you would like further information concerning any of the above topics please send email . |