Climate Modelling Group
School of Earth and Ocean Sciences

Climate Research Network
Collaborative Research Agreement at the University of Victoria
on Behalf of the Canadian Institute for Climate Studies and
Environment Canada
(#11 CICS-Global Oceans)


Progress Report:
October 31, 1998

1. Principal Investigator

Andrew Weaver

2. Institution

School of Earth & Ocean Sciences
University of Victoria
PO Box 3055
Victoria, B.C., V8W 3P6

3. Research Progress

This is the final year of the CICS Global Oceans project with research progressing along three avenues. The first of these (section 3.1) will allow a smooth transition into the new AES/CICS Arctic project.

3.1 Ocean model technology and development

Integration of the Arctic Ocean with the rest of the global ocean is essential for climate modelling research. Export of Arctic water regulates deep water formation in the North Atlantic while import of Atlantic water influences the extent and thickness of Arctic ice. Small variations in fresh water supplied to the subpolar convective gyres can alter or stop convection in the North Atlantic. Arctic outflow also influences deep water properties in the Greenland, Iceland and Norwegian Seas. Global atmospheric models have shown considerable sensitivity to sea ice extent which is in turn, influenced by the ocean circulation.

Spherical coordinate systems have traditionally been used in solving the equations of motion in oceanic general circulation models. Convergence of lines of longitude to a singularity at the poles may cause the solution to become unstable. Finite difference schemes therefore require shorter time steps for areas of high latitude. To overcome these limitations, atmospheric modellers have blended different grid systems or used spectral transforms. Filtering is another technique used by both atmosphere and ocean modellers to reduce the effects of numerical instabilities. Although effective, filtering tends to corrupt the solution and may be computationally intensive.

While atmospheric models may have to cover the entire globe, ocean models have the advantage of being able to shift the "pole problem" onto land. The simplest way to shift the pole is to use a spherical grid rotation but for this to be effective, one must find sufficiently large antipodal land masses. Locations for the new poles where chosen to be in Greenland and Antarctica. This grid transformation has been carried out with a global configuration of the GFDL ocean model coupled to the UVic Climate Modelling Laboratory Atmospheric Energy Moisture Balance Model and Ice Models.

Other improvements that have been made to this coupled model include the implementation of: an elastic viscous plastic ice rheology, a simple snow model over land and sea ice, a seasonally varying planetary albedo parameterization and orographic effects. The grid has also been extended to cover the entire globe. All of these improvements have allowed us to create a state of the art coupled climate model which is especially suited for studies of how the Arctic influences the rest of the global climate. The model is currently being used to study not only today's climate and variability but is also the climate of the last glacial maximum.

There are at least 4 papers currently planned which will use this model including a climatology/Last Glacial Maximum, ice model sensitivity analysis, and two climate variability studies. On going work includes updating the manual and parallelization of the code.

3.2 Sub-grid-scale mixing and the ocean's response to greenhouse warming

E Wiebe, an MSc. student has finalized his research with submission and acceptance of his thesis and the submission of a paper for peer review. In his research an ocean general circulation model coupled to an energy-moisture balance atmosphere model was used to investigate the sensitivity of global warming experiments to the parameterisation of sub-grid scale ocean mixing. The climate sensitivity of the coupled model using three different parameterisations of sub-grid scale mixing was 3°C for a doubling of CO2 (6°C for a quadrupling of CO2). This suggests that the ocean has only a weak feedback on global mean surface air temperature although significant regional differences, notably at high latitudes, exist with different sub-grid scale parameterisations. In the experiment using the Gent and McWilliams parameterisation for mixing associated with mesoscale eddies, an enhancement of the surface response in the Southern Ocean was found. This enhancement was largely due to the existence of more realistic sea-ice in the climatological control integration and the subsequent enhanced ice albedo feedback upon warming. In agreement with earlier analyses, the Gent and McWilliams scheme decreased the global efficiency of ocean heat uptake. During the transient phase of all experiments, the North Atlantic overturning initially weakened but ultimately recovered, surpassing its former strength. This suggests that in the region around the North Atlantic the ocean acts as a negative feedback on local warming during the transient phase but a positive feedback at equilibrium. During the transient phase of the experiments with a more sophisticated and realistic parameterisation of sub-grid scale mixing, warmed Atlantic water was found to penetrate at depth into the Arctic, consistent with recent observations in the region.

3.3 Boundary layer mixing

Studies have shown that the meridional heat transport in ocean models is very sensitive to the value of the diapycnal mixing, while observations show that diapycnal mixing is at least an order of magnitude higher at lateral, bottom and seamount boundaries than in the interior of the ocean. While traditionally models have been run with basin wide constant diapycnal mixing, we are investigating the effects of various parameterizations of the diapycnal mixing on the meridional heat transport in a simplified OGCM. OGCM results are being understood through the development of parallel scaling theories. In order to clarify the physics, only thermal forcing is used in a single (temperature) tracer model, with a linearized equation of state and momentum equations. A rotated diffusion tensor is also required, as regions of steeply sloping isopycnals may create spurious diapycnal diffusivities if the mixing is parameterized in the vertical and horizontal directions. As well, a regularly spaced model is used to avoid potential numerical diffusion. The scaling relationships are being developed by using the OCGM to find the term balances at each point in the basin, in order to find realistic approximations.

Simple scaling arguments suggest that for a basin wide constant diapycnal diffusivity kv, the meridional heat transport should vary as kv1/2, and the maximum meridional stream function as kv2/3. These relationships have been examined for basin-wide and boundary-only kv, and both were found to vary as ~ kv3/5; that is, isolating the diapycnal diffusivity to the boundaries does not change the relationships, suggesting the boundary diffusion is the major contributor to the heat transport. The actual amount of heat transported is 50% larger for a boundary kv than for the same volume of diffusive ocean in the central regions, showing the boundary to have a disproportionate effect on the heat transfer. However, when the basin is narrowed the relationships with respect to kv change, so that the boundaries do not seem to be the sole contributor to the heat transport. Varying the bottom and thermocline kv has had no effect on the relationships or the amount transported. Accounting for the energetic source for baroclinic instability (i.e., potential energy in sloping isopycnals) is being done using both the Gent-McWilliams bolus transport and the Gent-McWilliams skew-diffusive flux suggested by Griffies, but as yet only the former has been run. When the term balances are calculated for the steady-state basin, although most points show a geostrophic balance many have 3 or 4 significant terms; however there are also many points which show a balance between the acceleration term (partial derivative of u with respect to t) and the pressure gradient. Furthermore the acceleration term is on the same order as the pressure gradient in about 95% of the points.

While this project is in its early phases, initial results are extremely encouraging. We plan to continue funding this research through NSERC Strategic support once CICS Global Oceans comes to an end.

Return to the Climate Modelling Group WWW page
This page is maintained by
Last updated: