Main | Sessions | Programme (download talks) | Poster Abstracts | Committee | WGOMD

The following sessions are planned for the two-day workshop.

Day 1: Basics

Overview of equations and methods (0800-0930): Alistair Adcroft

Vertical coordinates (1000-1200): Robert Hallberg

Non-rectangular structured meshes and unstructured meshes (1300-1500): Todd Ringler, Matthew Piggott, Eric Deleersnijder

Parameterization of physical process (1530-1800): Richard Greatbatch and Martin Schmidt

Day 2: Applications

Coastal/Regional modelling (0800-1030): Eric Blayo, Jarle Berntsen

Basin/Global modelling: (A) mesoscale eddy non-permitting (1100-1230) and (B) mesoscale eddy permitting (1330-1500): Claus Böning, Anne Marie Treguier and Stephen Griffies

Ocean processes and inverse methods (1530-1700): Detlef Stammer

Recommendations to WGOMD and LOM (1700-1730): Stephen Griffies

---

1. Fundamental equations and methods

Coordinator: Alistair Adcroft, Princeton University

Session time: 90 minutes

This session is aimed at establishing a foundation for fundamental issues involved with building an ocean model.

The following points will be addressed in this session:

• Physical assumptions and resulting mathematical equations discretized in an ocean model
  • Boussinesq versus nonBoussinesq
  • Hydrostatic versus nonhydrostatic
• Finite differences, finite volume, and finite element methods
• Horizontal and vertical meshes
• Solution algorithms for Eulerian and Lagrangian vertical coordinates, including time stepping for fast and slow modes
• Advective form momentum versus vector invariant form velocity
• Calculation of pressure forces
• Advective tracer transport
• Fundamental principles that an ideal ocean model should respect (e.g., conservation, realizability, etc.)

2. Vertical coordinates

Coordinator: Robert Hallberg, NOAA/GFDL

Session time: 120 minutes

This session addresses perhaps the most important decision to be made regarding the design and use of an ocean model: what is the vertical coordinate? There are numerous approaches being considered in the community, with each having their advantages and disadvantages both in theory and in practice.

Amongst other topics, this session will address the following points:

• Advantages and disadvantages of various vertical coordinates, with discussion of issues related to
  • bottom flows and interactions with the solid earth boundary
  • interior flows
  • surface boundaries and interactions with atmosphere and ice
  • properties of an ideal vertical coordinate
• Spurious diapycnal mixing and the importance of accuracy in transport/remapping operators
• Treatment of the nonlinear equation of state: is the solution posed for isopycnal models sufficient for generalized vertical coordinate models?
• Can physical parameterizations (e.g., mixed layer schemes, mesoscale eddy parameterizations) be treated in a unified manner across all vertical coordinate choices?
• Are there remaining scientific reasons precluding the use of terrain following models for global climate?

3. Non-rectangular structured meshes and unstructured meshes

Coordinators: Todd Ringler, LANL; Matthew Piggott, Imperial College; Eric Deleersnijder, Université Catholique de Louvain

Session time: 120 minutes

All global ocean models in use for the latest IPCC report on climate change use the structured Arakawa grids first developed in the 1960s. Most major modelling centres will continue using these sorts of structured grids for many years. Nonetheless, there are good reasons to believe that in the next decade or two, there will be more non-Arakawa gridded models in use. In particular, various international efforts are making headway into structured icosahedral meshes, and other efforts are focused on unstructured fixed or adaptaive meshes. This session aims to educate the broader community of model developers on the problems and prospects of these approaches for high end ocean modelling.

Topics to be discussed include the following:

• Formulating the ocean equations non-rectangular structured meshes and finite element meshes
• Developing solution algorithms
• Effects of staggering variables within a cell
• Reasons why finite elements have taken so long to move from engineering coastal applications to large-scale applications
• Vertical physical processes: similarities and distinctions from rectangular meshes
• State of art in existing approaches for structured and unstructured non-rectangular approaches
• Prospects and problems with these approachs for use large-scale modelling

4. Physical process parameterizations

Coordinators: Richard Greatbatch (Dalhousie University) and Martin Schmidt (Baltic Sea Research Institute)

Session time: 150 minutes

The issues of physical parameterization for large-scale basin and global models have often been distinct from those of coastal and regional models. This is largely due to the distinct space and time scales of the simulations. Nonetheless, as the regional models extend their boundaries outward in space-time, and the global models enhance resolution to capture regional effects, the issues of physical parameterizations are increasingly overlapping. In this session, we bring together various issues of physical parameterizations that will remain central in the next generation of ocean models.

Topics discussed in this session include the following:

• Mesocale and submesoscale parameterizations
• Diapycnal mixing from internal wave breaking
• Surface and solid earth boundary layer turbulence
• Stresses from internal wave drag
• Rectified tidal effects
• Double diffusion
• Generalized turbulence closure modelling, and prospects for a unified suite of closure methods

5. Coastal and regional modelling

Coordinator: Eric Blayo (Laboratoire Jean Kuntzmann, Université Joseph Fourier), Jarle Berntsen (University of Bergen)

Session time: 150 minutes

Coastal/regional efforts have generally been posed on relatively fine meshes using open boundary conditions. Nesting methods are also becoming more common in these models, whereby ultra-fine meshes next to the coasts are nested with coarser meshes further off-shore. Some of this technology, especially two-way nesting, could be very useful in the larger scale global models. This session aims to update participants on the state of the art in open boundary condition modelling and nested modelling.

Topics to be discussed include the following:

• Open boundary conditions: theory and practice
• One way and two way nesting: algorithms and methods for ensuring transparent behaviour at boundaries, and conservation of mass and tracer
• Nesting refined resolution coastal or regional models within global models: technical and practical issues
• At what scale and for which applications do we need an explicit representation of nonhydrostatic dynamics?

6. Basin and global modelling

Coordinator: Claus Böning, IFM-GEOMAR; Anne Marie Treguier, IFREMER; Stephen Griffies, GFDL

Session time: 180 minutes

Basin/global scale models have long been used for studies of ocean climate dynamics. In the last 10 years, the computer power and model numerics have improved to the point that we can explicitly permit a representation of the World Ocean's mesoscale eddy field in decade long simulations. It is anticipated that in the next few years, more groups will embark on fully coupled climate simulations with these mesoscale eddy permitting oceans. Even so, coarse resolution models will remain a key component in many applications, particularly millenial scale paleo-climate studies.

This session is split into two parts, with the first focusing on the needs of coarse global models run for centennial to millenial time scales, with the second session focused on the eddy permitting models.

Topics to be explored in this session include the following:

• Can we establish the minimal needs of a coarse resolution for global ocean climate model for long term milennial studies? What is essential to represent explicitly (e.g., tropical currents for ENSO), and what can be reasonably parameterized (e.g., mesoscale eddies)?
• Can we establish useful methods for running ocean-ice models, short of running models fully coupled to realistic atmospheric models?
• What are the key numerical features of global models required to provide a robust simulation? For example, tracer advection properties; boundary layer processes; feasibility of incorporating realistic ocean biogeochemistry models; representation or parameterization of marginal seas and overflows; how can we parameterize flow through underresolved straits or capture small-scale topographic effects in the momentum equation (e.g. headland eddy drag)?
• What parameterizations are required for eddy permitting simulations (e.g., friction, adiabatic closures in the tracer equation)? How sensitive are simulations to the choices?
• What resolution is required to capture the bulk of the mesoscale eddy spectrum?
• What new numerical issues arise when admitting mesoscale eddies in global climate models?
• What about terrain following models for basin and global applications at either the coarse or fine resolutions?

7. Ocean processes and inverse methods

Coordinator: Detlef Stammer, Hamburg

Session time: 90 minutes

Ocean data assimilation (ODA) is a fundamental aspect of ocean forecasting systems and state estimation. Central to ODA systems is the posing of an inverse problem, whereby various sophisticated methods aim to solve the inverse problem in a robust and unbiased manner, given the best of the forward model and our understanding of errors in the observational data. Combined forward/inverse systems can provide a great deal of understanding of ocean processes and help to develop mechanistic descriptions of observations.

The purpose of this session is to flesh out those aspects of inverse methods that are directly applicable to understanding fundamental physical processes in the ocean. The emphasis is thus not on operational applications, but on using these methods as a research tool.

Topics in this session include the following:

• Inverse methods for parameter estimation
• Flow regimes most appropriate for inverse methods (e.g., with or without mesoscale eddies?)
• What can inverse methods tell us about fundamental dynamical balances in the ocean, and biases in the models?
• Applications of linear sensitivity/stability maps

8. Recommendations to WGOMD and LOM

Coordinator: Stephen Griffies, GFDL

Session time: 30 minutes

This is a general wrapup session. Here, we attempt to bring together some of the various recommendations made throughout the workshop for use by the WGOMD and LOM. Most specifically, we aim to address the question: