Sea Ice and Issues of Scale
Lyn McNutt
University of Alaska
Fairbanks, Alaska
Sea Ice and Issues of Scale
Lyn McNutt
University of Alaska
Fairbanks, Alaska
Sea ice plays an important role in the climate system as a result of the immense area it covers, its seasonal changes and heat flux effects, and its brightness, which leads to albedo feedbacks to climate. Sea ice is also significant for economic and habitat reasons, and can be important for modeling in ways that go beyond the physical system. McNutt discussed what sea ice is, how it grows, why and how we model it, problems with how it is modeled, and scaling and hierarchy considerations. She says that it is important, when modeling sea ice, to be very careful about jumping too many levels in scale without taking into account emergent properties and aggregate behaviors that come into play. She also stressed the need for more conscious validation of models and the need for more data collection and more accurate monitoring.
How Sea Ice Forms
Ice formation begins when platelets form as water reaches the freezing point; these platelets move with the wind and aggregate into pancake ice (1-3 meters). These pancakes then aggregate to form ice floes (large blocks) which can support snow. At this point, the ice floe starts to grow from the bottom and extrudes salt out. Then, as melting at the surface takes place, water puddles form which absorb incoming radiation because they are dark, and so they expand in size, becoming self-generating. There is constant brine flushing and constant change in albedo and heat flux from ocean to atmosphere based on the thickness of the ice. Sea ice is very dynamic and understanding these dynamics is important to our understanding of the effects of ice on climate.
Another type of ice is "fast ice" (see Figure 1.19) which grows out from the shore where platelets line up, and anchors itself to the sea bottom near the coast where it interacts with the sea ice pack. The length and extent of fast ice depends on how much the sea ice compresses into the shore. This fast ice is of interest because it behaves like a geological entity but moves so fast (up to 200 mm per second) that it can completely alter regional climate patterns.
Roles
of Sea Ice
It is believed that sea ice in the Arctic Ocean plays a significant role in controlling deep water formation and thus the conveyor belt of oceanic circulation. The interplay between the Bering Sea and the Arctic Ocean leads to warm water formation that effects oceanic circulation, but general circulation models (GCMs) don't account for this interplay across the shelf. In the important fisheries of the Bering Sea region, ice has significant controls on fish habitat. Ice also allows for nutrient transport and creates whale habitat through openings in the ice. Contaminants in Russia River runoff are transported around the Arctic by sea ice, as is air pollution, acid deposition, and particulate matter. In all of these ways, ice acts as a conveyor and the provider of key feedbacks to climate through atmospheric and oceanic circulation.
Figure 1.19 (left)
View of the Beaufort Sea from the edge of the fast ice, four miles off Point Barrow, Alaska (photo by Carl Byers).
An Arctic Climate Regime Shift
Air temperature and precipitation have risen in the Arctic since the mid-1970s and there has been a marked decline in sea ice extent (see Figures 2.6 and 2.7). There have also been serious changes in Bering Sea species distribution associated with this change in climate; these species changes lag the regime shift by about 3 years. Increased concentrations of greenhouse gases globally as well as Alusian Low and El Niño effects may be implicated in this regime shift which has effected the climatology of the entire Arctic. For example, the melt back of ice creates pools of bottom water and the relative heat of that water determines the longevity and growth rate for pollock. As a result, the pollock fishery in the Bering Sea has been extremely productive since the 1970s. This is an example of how climate effects involving sea ice move all the way up the food chain.
Sea Ice in the Physical Realm
In the physical domain, ice extent is important, as is its concentration, thickness, age, surface albedo, surface temperature, and roughness (both top and bottom). Ice extent changes seasonally and annually. Marginal ice zones are those that have ice in winter but no ice in summer. These seasonal ice zones (such as the Beaufort Sea) are very dynamic, have large annual variations, and have the most interplay with pack ice (and thus play a role in oceanic circulation patterns). In addition, the climatic regime shift mentioned above has caused additional changes in ice extent.
With regard to ice concentration, we must consider how much open water exists, where the leads are, and how they are oriented. Ice thickness and age are very important as well, and it is clear that the ice in the Arctic is thinning. The ice extent in summer has declined while in winter it hasn't changed much, but both summer and winter ice is getting thinner, and this changes the heat flux in the Arctic. The surface albedo of the ice depends on the extent of melt ponds making these a key feature. Roughness, both at the top and bottom surfaces of the ice, is also important.
Ice acts as a conveyor and the provider of key feedbacks to climate through atmospheric and oceanic circulation.
The Role of Sea Ice in Habitat
Polynyas are openings in the ice that occur when winds blow the ice away from the shore. Polynyas are highly productive areas biologically. The ice edge is also very important from a habitat standpoint because fish and whales follow the ice edge. Shelf mixing is also significant because that is how nutrients and pollutants are transported around the Arctic. For example, after an oil spill, oil is sucked into the ice, rests in a layer there, and forms oil ponds in the melt water pools which are then distributed around the Arctic through shelf mixing.
The Economic Role of Sea Ice
Sea ice has a large effect on navigation because in the Arctic re-supply is accomplished almost entirely by ship. Economically valuable fisheries and whaling are extremely connected to sea ice because, as mentioned above, fish and whales follow the ice edge. In addition, indigenous people rely on subsistence, and sea ice has large effects on their ability to subsist on harvests from their surroundings. Could climate change eliminate this entire way of life by reducing sea ice to the point that subsistence is no longer possible? Sea ice also effects development of resources such as oil and natural gas. Finally, ice in the Arctic has become a collector of contaminants that did not originate in the region. Through the global distillation effect, pollutants from industrialized areas in the lower latitudes have concentrated in the Arctic.
Sea Ice in Global Climate System Modeling
As explained above, sea ice is a very important and interactive climate parameter. Despite this fact, it has not been modeled effectively in general circulation models (GCMs). It has not been modeled responsively to the role it plays, but rather simply as a deforming slab, in most cases. There is, however, a simple and effective way to model sea ice, which was suggested by Zubov in 1943 and is widely used in operational sea ice forecasting. To determine the movement of sea ice, Zubov said, take 3 to 4 percent of the wind speed, 20 degrees to the right of the wind. Despite the fact that this is a very good rule of thumb for short-term modeling of sea ice, McNutt says, it has been ignored in GCMs (perhaps because it seems inelegant, she suggests).
How has sea ice been treated in global climate models? Sea ice has been modeled as an elastic, plastic, and/or viscous material in terms of its response to stress/strain. Elastic materials bounce back. If a material is viscous, its density increases but it doesn't come back if the pressure on it is released. Plastic materials break under sufficient pressure. Sea ice has been modeled mainly as a slab, and either as an elastic plastic material, a viscous plastic material or an elastic viscous material.
Sea ice has not been modeled effectively in GCMs.
The earliest attempts to account for sea ice in GCMs involved simply treating it as a swamp - a thick layer that doesn't do anything. In such a model, the sea surface temperature (SST) is derived from the surface energy balance. The next level of sophistication involves a simple mixed layer or slab ocean and accounts for a layer of sea ice, but one that is deformed with elastic, plastic or viscous properties, which is not a correct representation and does not account for interplay. Even the current best treatments are not realistic and do not properly include the interaction of sea ice with climate. Initially, the problem of sea ice treatment in GCMs was a resolution problem, but now it's a process problem, McNutt says.
Issues of Scale in Modeling Sea Ice
As discussed and summarized by session chair Danny Harvey, two distinct scaling issues arise in the treatment of sea ice:
(a) how the dynamic behavior of sea ice changes with scale, and
(b) determination of the scale at which atmospheric forcing of sea ice motion is most directly applicable.
With regard to the first issue, an aggregate of ice floes behaves in ways that are quite different from the behavior of individual ice floes; in particular, ice behaves like a granular medium at the 0.1-1 km scale, while at a regional scale it behaves like a viscous fluid.
With regard to the second issue, the local velocity of sea ice cannot be directly related to the local atmospheric shear stress. Rather, the valid linkage is between regional atmospheric forcing and regional sea ice deformation. However, atmospheric forcing varies much more rapidly in time than the sea ice response, so the history of atmospheric forcing must also be taken into account.
Validation in Modeling
Sea ice parameters in GCMs are calculated internally making validation impossible, McNutt says. The fact that large scale information used in models is derived from existing models, rather than from actual mesoscale data, makes the models internally consistent, but not consistent with reality. Even for smaller scale models (5 to 20 km), parameters are derived from large scale models or generated internally from the models. McNutt believes that large scale sea ice behavior is probably being modeling reasonably well but that problems arise when aggregate behavior from the floe scale is applied to the mesoscale.
Initially, the problem of sea ice treatment in GCMs was a resolution problem, but now it's a process problem.
Forward to Next Section//Back to Table of Contents//AGCI Homepage//Comments: agcimail@agci.org