Model projections of greenhouse-gas-induced climate warming indicate enhanced sensitivity at high latitudes owing largely to positive feedbacks involving sea-ice and snow (e.g. Mitchell et al., 1990; Kattenberg et al., 1996; Rind et al., 1995). A worry is that this sensitivity may be an artifact of the crude treatment of sea-ice and snow processes in such models. For example, some coupled models treat sea-ice as a motionless thermodynamic slab, ignoring ice deformation and advection, while others use a simple "ice drift" approach.
In the course of my CMIP1 subproject, I compared the (control) model results and produced tables and figures for both ice and snow for use in the IPCC TAR (Chapter 8). The results indicated no clear connection between errors in simulated ice or snow extent (i.e. difference between model and observed climatology) and the level of sophistication of the model parameterizations of ice or snow. There was some tendency for models employing flux adjustment to have smaller errors, but this was not universally true. Although I believe these intercomparison results are a valuable contribution to the IPCC TAR, the rather limited analysis possible using the CMIP1 results precludes a stand-alone publication at this point.
In this proposal, I wish to expand on my CMIP1 study by investigating the relationship between fidelity of the control run simulation of ice and snow, and the changes, both in ice and snow and also in other climatic quantities (notably surface air temperature), which occur as a result of greenhouse-gas forcing.
The goals of this proposal are twofold:
The change in sea-ice extent and thickness, snow extent and depth, surface air temperature and other climatic quantities will be computed as the difference between the first and last 20-year averages from the CMIP2 transient model runs. There is interest in these ice and snow difference plots for potential use in the IPCC TAR (U. Cubasch, pers. comm.) if they can be done in time. These differences will be stratified according to the type of cryospheric parameterization employed (e.g. ice dynamics vs. motionless ice) and according to the error (model minus observed) in sea-ice and snow extent determined in the previous CMIP1 subproject. Where suitable observations are not available (e.g. for sea-ice thickness), the model results will be stratified according to model means for these quantities (e.g. models which produce 'thin' sea ice vs. those which produce 'thick' sea ice relative to the ensemble model mean).
The approach will then be to seach the results for common features and relationships. For example, do models which include ice dynamics exhibit less high-latitude amplification in climate warming than those which include motionless ice? -- a result expected from earlier ice-only model experiments (Hibler, 1984; Flato, 1996). Do models with unusually deep snow in their control climates exhibit greater warming over land than those with unusually shallow snow (a result which might be inferred from the albedo feedback arguments of Fyfe and Flato [1998]). As a final example, the CMIP1 results indicated a particularly large range in simulated southern hemisphere ice extent -- is the magnitude of Antarctic warming correlated with ice extent in the control climate?
Answers to these and other questions will useful in assessing the role of cryospheric processes in climate change, and will help guide future improvements in the representation of cryospheric processes in global climate models by illustrating aspects which have a potentially large impact on climate sensitivity. Results from this and the previous CMIP1 sub-project will be combined for publication.
We will use the basic CMIP model output: atmospheric surface fields (especially snow and ice depth); ocean annual means and zonal cross-sections; and surface air temperature time-series. In addition we will make use of climatological data available at our institution (e.g. ice and snow extent, surface air temperature and pressure, ocean temperature and salinity).
Flato, G.M., 1996. The role of dynamics in warming sensitivity of Arctic sea ice models. Proc. Workshop on Polar Processes in Global Climate, Am. Meteorol. Soc., 13-15 November, Cancun, Mexico, 113-114.
Fyfe, J.C. and G.M. Flato, Enhanced climate change and its detection over the Rocky Mountains. J. Clim., 12:230-243, 1998.
Hibler, W.D. III, 1984. The role of sea ice dynamics in modeling CO2 increases. Climate Processes and Climate Sensitivity, Geophysical Monograph 29, Am. Geopys. Union, 238-253.
Kattenberg, A. plus 8 others, Climate Models - Projections of Future Climate, Chapter 6 in Climate Change 1995, Cambridge University Press, 572pp., 1996.
Mitchell, J.F.B., S. Manabe, V. Meleshko, and T. Tokioka, Equilibrium Climate Change - and its Implications for the Future. Chapter 5 in Climate Change: The IPCC Scientific Assessment, Cambridge University Press, 365pp., 1990.
Rind, D., R. Healy, C. Parkinson, D. Martinson, The role of sea ice in 2xCO2 climate model sensitivity. Part I: The total influence of sea ice thickness and extent. J. Climate, 8:449-463, 1995.
Flato, G.M. and W.D. Hibler III, Modeling pack ice as a cavitating fluid. J. Phys. Oceanogr., 22:626-651, 1992.
Flato, G.M. and R.D. Brown, Variability and climate sensitivity of landfast Arctic sea ice. J. Geophys. Res., 101:25,767-25,777, 1996.
Flato, G.M. and D. Ramsden, Sensitivity of an atmospheric general circulation model to the parameterization of leads in sea-ice. Annals of Glaciol., 25, 96-101, 1997.
Lemke, P. W.D. Hibler III, G. Flato, M. Harder, and M. Kreyscher, On the improvement of sea ice models for climate simulations: the sea ice model intercomparison project. Annals of Glaciol., 25, 183-187, 1997.
Flato, G.M., The thickness variable in sea-ice models. Atmos.-Ocean, 36(1): 29-36, 1998.
Flato, G.M., G.J. Boer, W.G. Lee, N.A. McFarlane, D. Ramsden, M.C. Reader, and A.J. Weaver, The Canadian Centre for Climate Modeling and Analysis Global Coupled Model and its Climate. Climate Dynamics, in press, 2000.