Category 3: Ice-Shelf/Ocean
It is increasingly apparent that the ocean is strongly influencing the ice sheet. The ocean’s ability to impact the ice sheet occurs primarily along the underside of the floating ice shelves that surround most of Antarctica and the termini of Greenland tidewater glaciers which sometimes have short floating tongues. Further, the ice shelf geometry and, thereby, its ability to provide backpressure to the grounded ice, is strongly determined by the spatial pattern of sub-ice-shelf melting.
The challenge faced by models that simulate the sub-ice-shelf ocean cavity and the interaction with the ice across the cavity’s upper surface is that beneath the ice shelf, the upper surface of the ocean is variable, a condition much different than the normal free surface of the open ocean. A number of 3D and 2D models are addressing this situation, including the following (grouped by dimensionality and listed by model name, institution and investigator):
• HYPOP and CISM (Los Alamos National Lab; Ringler/Price/Lipscomb
• ROMS (Old Dominion University; Klinck/Dinniman and ACE CRC; Galton-Fenzi/Hunter; no dynamical ice yet)
• HIM (Ocean) (Geophysical Fluids Dynamics Lab; Little; no dynamical ice yet)
• MICOM (Ocean) and CISM (University of Bergen ; Drange)
• FESOM (Ocean) (Alfred Wegener Institute; Hellmer; and Hadley Center Ice Sheet) 2D-Vertical
• Stream Function Ocean Flowline Ice (Penn State University; Walker/Dupont; includes an ice stream)
• Plume Ocean, CISM (New York University; Gladish/D.Holland; British Antarctic Survey; P.Holland; and Los Alamos National Lab; Lipscomb)
Vertical models cannot reproduce the horizontal circulations that are known to occur. A major limitation of potentially more complex models is the fact that the bathymetry beneath most ice shelves remains unmeasured. Vertical-slice models will have to parameterize the effects of “backstress” from lateral drag.
Forcing the variability of the ocean circulation and, thus, the waters entering the sub-ice-shelf cavity are the tides and the surface winds. Tides are well known and winds are a common output from GCMs and data reanalysis activities. Simulations of the ocean circulation forced by reanalysis fields reveal both seasonal and longer time-scale variability that will affect the properties of sub-ice-shelf waters and, thus, melting rate and ice sheet discharge. The role of these models is to provide better simulations of the ocean forming that drive basal ice shelf melt.
These regional models serve more as a one-way bridge between possible ocean conditions and the co-evolution of the ocean and ice shelf. These will then be translated into boundary conditions for the whole ice sheet models through knowledge gained with the ice-stream/ice-shelf models.
Set Up and Initialization
The more distant linkage of this category of models with the whole ice sheet models lessens the requirement that they spin-up with a consistent set of climate variables to the whole ice sheet models.
Bathymetry is poorly known beneath most ice shelves and throughout much of the sea ice zone surrounding ice sheets. While much of the Amundsen Sea is a welcome exception, even in this region the bathymetry in the ice shelf cavities is poorly known. Ocean conditions at depth on the continental shelf surrounding Antarctica remain vastly undersampled in both space and time. Nevertheless, appealing again to the primary purpose of this study, the deviations of hypothesized climate changes from some control run are most important and extreme scenarios remain the starting point to help define the upper bound of ice sheet contributions to sea level this century.
The control runs can use either reanalysis fields, or climatological fields, or prescribed values to derive the ocean forcing applied to the undersides of the simulated ice shelves. In subsequent simulations, what will be of most interest is the sensitivity of the coupled ice shelf/ice-sheet system underside to changes in this control run.
It has been shown that there is seasonal and interannual variation of the ocean circulation on the continental shelf when reanalysis winds are used to force the ocean. Model results suggest that these changes in circulation strongly modulate the delivery of oceanic heat to the sub-shelf cavities. What is less well known is how this affects the sub-ice-shelf melting pattern. This experiment can either be done contrasting model results from runs that force the model with temporally varying winds from a run that uses climatological forcing, or runs that amplify the temporal variations in separate runs. These runs are most suitable for the Antarctic, but can be initially run for an idealized geometry, then progressing to the Amundsen Sea before concluding with the Ross and Weddell Seas.
Warm ocean versus cold ocean
Ocean waters proximal to an ice shelf are sometimes characterized as either “warm” or “cold” in referring to the relative contribution of various water masses to the interaction with the ice shelf. “Cold” conditions refer to the dominance of High Salinity Shelf Water (HSSW), a type of cold and very saline water that forms by wintertime freezing of sea ice. “Warm” conditions refer to the dominance of warmer intermediate depth water (called Circumpolar Deep Water, or CDW in the Antarctic) that exists in the deep polar ocean. Depending on other factors, either water mass can reach the grounding line of the ice shelf, with vastly different melt rates resulting.
This “warm/cold” categorization can obscure the fact that surface waters appear to also have an influence on ice shelf state and, thus, ice sheet response. This category of model should be used to run an experiment to examine the quantitative effect of warming surface waters on ice shelf shape.
The evolution of water that reaches the grounding line also affects the pattern of melt along the ice shelf underside and this pattern, in turn, affects the resulting shape of the ice shelf and the force experienced by the ice stream or glacier discharging into the ice shelf. A range of scenarios varying the amounts of HSSW and CDW is possible, but a first-order contrast of the unlimited warm water experiment is the cold water extreme.