Icebreaking Environments and Challenges to the U.S. Fleet
Distribution and Characteristics of Sea Ice and Icebergs in the Polar Regions
Sea ice is a defining characteristic of the polar oceans and throughout the course of the year occupies between 15 million and 25 million square kilometers or up to 7 percent of the world’s ocean surface. Polar sea ice undergoes tremendous seasonal changes every year. During the winter, the extent of the Arctic ice pack grows to the size of the United States; in the summer, more than half of the ice disappears. On the other side of the globe, ice covers nearly 98 percent of the Antarctic continent and averages 1 mile thick. Depending on the season, the sea ice may cover roughly 2 million square kilometers in the Antarctic summer (February) and increase to 20 million square kilometers during the austral winter (September). The total Antarctic ice volume is so large that it accounts for 90 percent of the world’s ice and 70 percent of its fresh water.
Antarctic Sea-Ice Characteristics
In the Southern Ocean, the vast sea-ice apron that rings the Antarctic continent attains its maximum width of around 2,000 km in the Ross and Weddell Sea sectors of the Southern Ocean (Figure 7.1). In austral summer, most of the remaining sea ice is found in the western Weddell and Amundsen Seas, with some multiyear ice also found in the Ross Sea (Comiso, 2003). First-year Antarctic drift ice typically does not grow to more than 0.5 to 1 meter thickness (Haas, 2003) because substantial heat from the ocean is transferred to the base of the sea ice (heat flux), which keeps this ice relatively thin. Substantially thicker ice may develop due to deformation processes such as rafting and ridging and multiyear ice growth. Another key aspect of sea ice in the Southern Ocean is snow accumulation, which can build up to mean depths of well above 0.5 meter in some areas (Massom et al., 2001). Sea ice is free to drift (driven by winds) in a large-scale pattern around the Antarctic continent because there are no bays or basins to restrict its movement. In general, the sea ice in the Weddell and Ross (slightly lesser extent) Seas typically exhibits a clockwise motion.
Although the Southern Ocean ice cover changes dramatically from season to season, change in ice extent between years is small, typically on the order of 2 percent of the mean maximum ice extent. The northernmost position of the ice edge in September is controlled largely by the location of the polar front separating cold Antarctic surface water from warmer sub-Antarctic water masses. The position of the front, and hence to some degree the maximum ice extent, is determined primarily by the surface wind field and the submarine topography. Large-scale coupled atmospheric circulation patterns such as the El Niño-Southern Oscillation are also impacting regional ice distribution. Scrutiny of the past three decades of satellite-derived records of global sea-ice distribution so far yields conflicting results as to whether the Antarctic sea-ice extent has marginally decreased or increased during this time period, depending on the methodology employed in extracting ice information from the satellite instrument record (Cavalieri et al., 2003; Comiso, 2003). The geological record (Gersonde et al., 2005) indicates that sea-ice conditions in the Antarctic appear to have followed a fairly consistent, stable pattern during interglacial periods. Nevertheless, there is some indication that regional warming in the Antarctic Peninsula region has led to significant reductions in perennial ice extent during the austral summer in the Bellingshausen, Amundsen, and Weddell Seas (Comiso, 2003).
In the northern hemisphere, maximum ice extent reaches up to 16 million square kilometers in March, when the entire
Arctic Ocean and its marginal seas as well as parts of the sub-Arctic seas are covered by ice. In summer, up to 6 million square kilometers of perennial ice remains in the interior Arctic (Comiso, 2003; Figure 7.1). Such multiyear ice can grow to between 3 and 5 meters in thickness while it resides in the Arctic, with deformed ice building up to more than 30 meter thickness in the form of pressure ridges (Wadhams, 1998). Sea-ice circulation in the Arctic is dominated by the clockwise Beaufort Gyre extending over much of the western and west-central Arctic (with mean ice age on the order of three to five years), and the Transpolar Drift, transporting ice from the Siberian shelves across the Pole into the Greenland Sea (with mean age of less than three years).
Interannual variability in ice extent in the Arctic is more pronounced than in the Southern Ocean, with anomalies in some years approaching 10 percent of the mean maximum extent. While maximum ice extent at the end of winter (March) has remained fairly stable since the establishment of a consistent satellite record of ice conditions (1978), summer ice extent (i.e., the perennial ice area, corresponding roughly to the extent of multiyear ice) has been reduced significantly. Thus, from 1979 through 2000, Arctic sea-ice extent has been shrinking by about 2.2 percent per decade, driven mostly by reductions during the ice melt season (Comiso, 2003). The rate of decline of summer minimum ice extent amounted to almost 8 percent per decade from 1979 to 2005 (NSIDC, 2006). At the same time, submarine sonar data collected in the central and western Arctic indicate that the Arctic ice pack thinned by approximately 40 percent from the 1950s to the 1990s. This thinning is attributed mostly to changes in atmospheric circulation and radiative forcing (Francis et al., 2005; Rothrock and Zhang, 2005). Interpretation of ice variability is hampered to some extent by incomplete instrumental records, in particular for ice thickness, and a comparatively short satellite-derived ice extent record. This renders separation of inherent decadal- and centennial-scale variability from any trends due to anthropogenic warming, which is predicted to be greatest in the Arctic (Holland and Bitz, 2003; ACIA, 2005), difficult. These challenges notwithstanding, the past several years have been nothing short of remarkable: Since 2000, four out of the five Arctic ice seasons have exhibited consecutive record summer ice minima (Stroeve et al., 2005). From the available record it appears that perennial ice extent is as low as it has been in the past few centuries. Moreover, most recent indications are that winter ice extent is now also starting to retreat at a faster rate, possibly as a result of the oceanic warming associated with a thinner, less extensive ice cover (Meier et al., 2005). These observations of a shrinking, thinning Arctic sea-ice cover are in line with climate model predictions of enhanced high-latitude warming (ACIA, 2005), which in turn is driven in significant part by ice-albedo feedback (Holland and Bitz, 2003). It has been argued that the Arctic climate system has reached a “tipping point” and is now on a trajectory to a different, stable state, characterized by a greatly reduced or absent summer ice cover (Lindsay and Zhang, 2005; Overpeck et al., 2005) and—by inference— significantly thinner, less extensive winter ice.
Floe-Scale Sea-Ice Characteristics Relevant to Icebreaking
From the perspective of sea-ice trafficability, the ice cover of the polar seas is far from homogeneous, and an assessment of the hazards posed by sea ice to navigation, coastal infrastructure, or other human activities requires a more detailed examination of ice characteristics at the scale of individual ice floes (kilometers to tens of meters). Of particular importance are the distribution of openings in the ice (leads or polynyas), the thickness and extent of level ice, and the morphology and thickness of pressure ridges or hummocks formed as a result of ice deformation (see Figure 7.2 for explanation of terms and schematic representation of ice evolution).
The thickness of level ice, growing through accretion at the base of the ice sheet (or also, as is often the case in the Antarctic, through flooding of submerged ice at the top) depends on the amount of heat that can be extracted from the ocean water and transferred to the cold atmosphere. In the marginal seas of the Arctic Ocean, where the insulating effects of a snow cover are typically low and where the amount of heat transferred from the ocean to the base of the ice has historically been small, as much as 1.5 to 2 meters of level ice can grow within a single season. Repeated cycles of summer melt and winter accretion can increase this value to at most 3-4 meters in ice that is between a few years and a decade old. In recent years, due to reduced residence time of ice in the Arctic Ocean, diminished winter growth and increased summer melt, maximum thicknesses of level multiyear ice are typically less than 3 meters throughout much of the Arctic ( Perovich et al., 2003; Haas, 2004). In the Antarctic, high ocean-to-ice heat fluxes and the insulating effects of the snow cover (typically not compensated for by snow-ice formation) result in level ice thicknesses that are mostly well below 1 meter and in many areas around 0.5 meter (Wadhams et al., 1987; Worby et al., 1998). An important exception is the narrow, stationary landfast ice belt that is attached to the coastline or the floating ice shelves.
With the exception of landfast ice that is firmly attached to a coast, sea ice is in near-constant motion, with velocities typically on the order of 5 to 10 km per day. On shorter (i.e., operationally relevant) time scales of days to weeks, this motion is mostly a result of wind forcing and—to a lesser extent—transfer of momentum from the ocean through currents, with tidal currents of particular importance. The response of the ice cover to such forcing depends in part on its variable thickness and roughness. Furthermore, wind or ocean forcing itself may be divergent or convergent on scales of tens to hundreds of kilometers. Hence, openings develop
at regular spatial and temporal intervals in areas of divergent ice motion (Figure 7.2), and deformation features such as rafting or ridging (Figure 7.2) appear in areas of convergent ice motion or along coasts with onshore ice drift. The geography and stronger wind regime of the Southern Ocean compared to the enclosed Arctic “Mediterranean” seas typically results in more openings and fewer deformation features in the Antarctic compared to the Arctic, with the exception of some of the sub-Arctic seas, such as the Bering Sea. The shallow Arctic shelves and complex coastal morphology, in particular those areas comprising the Northern Sea Route along the Siberian coastline and the Northwest Passage through the Canadian Archipelago, foster substantial interaction between drifting and stationary ice or coastlines, resulting in strongly deformed ice bodies. Russian ice navigators have coined the term “ice massifs” for such highly deformed areas—often tens of thousands of square kilometers in extent—that contain little to no open water and exhibit ridges that are more than 20 meters thick and capable of grounding on the seafloor. Often such highly deformed areas lack undeformed, level ice altogether, which can make navigation through these regions challenging or impossible even for the most powerful icebreakers (see also “Icebreaker Technology” in Chapter 6).
Icebreaker design depends on where you want to go, when you want to go, and what you want to do when you get there. The answers to these questions may well produce operational scenarios in very different ice conditions in the Arctic and the Antarctic, but the differences in Arctic and Antarctic ice do not directly explain or drive icebreaker design.
Small-Scale Properties of Sea Ice Relevant to Icebreaking
As explained in detail in the discussion of icebreaker technology in Chapter 6, modern icebreaker design aims to achieve the following, to some extent conflicting, goals: (1) increase the local bending moment resulting in failure of the ice sheet flexing under the vessel’s weight when in continuous icebreaking mode; (2) effectively fragment or displace broken ice to the sides of the channel while minimizing the amount of ice submerged to the depth of the vessel’s keel; (3) reduce friction exerted on the vessel’s hull by the ice sliding past the ship; and (4) increase the fragmentation and displacement efficiency when in ramming mode. In addition to the distribution of open water and ridges and the thickness of the ice cover discussed in the previous section, two additional, small-scale properties of the ice cover are of prime importance in this context.
First and foremost of these is the strength (typically referred to as the yield stress under which an ice sheet fails, i.e., loses mechanical integrity) of the ice cover and, to a lesser extent, a variable referred to as fracture toughness, which often scales with ice strength, that determines the resistance offered to the breaking action of an icebreaker. Ice strength depends greatly on the porosity or volume fraction of brine- and gas-filled void space of the sea ice. Sea ice retains a portion of the salt contained within seawater in the form of brine inclusions, which can occupy more than 10 percent by volume of warm first-year sea ice. In conjunction with the temperature, the amount of salt present within a volume of ice determines the bulk porosity and hence ice strength. All other factors being equal, Arctic multiyear ice has substantially lower salinity (approaching 0 in the uppermost decimeters) than first-year ice because low-salinity meltwater generated at the surface of the ice (see melt pools evident in Figure 7.2) flushes out saline brine during summer melt. As a result, the strength of Arctic multiyear ice is considerably higher than that of first-year ice. In the Antarctic, the multiyear drift ice found in the Weddell and Ross Seas typically does not exhibit reduced salinity due to the lack of surface melt (Eicken, 2003). Hence, at the decimeter scale, Antarctic multiyear ice is not substantially stronger, though still somewhat thicker, than first-year ice. As explained in more detail below, however, the situation is somewhat different in McMurdo Sound.
The second factor that needs to be considered is friction exerted on a vessel’s hull during passage through the ice. Here, it is mostly the presence and, to a lesser extent, the thickness of a snow cover that determines the magnitude of surface friction. Bare ice, as commonly found in the Arctic during summer months, but also in some locations such as McMurdo Sound, where little snow accumulation and strong winds prevent buildup of a substantial snow cover, has significantly lower friction coefficients.
Icebreakers and Breaking Ice
As a general rule, icebreakers usually do not seek to break ice per se. Whether the task at hand is escorting a less capable vessel, proceeding to a science station, or simply transiting, the icebreaker crew is almost always searching for the fastest and most economic route. This usually means avoiding ice to the degree possible, finding leads and areas of lesser ice concentration. The best route is rarely the most direct.
However, vast areas of consolidated sea ice, water depth, or other navigational obstacles may limit or preclude ice avoidance techniques. This is especially true in certain areas of the Northwest Passage and skirting the multiyear pack in the vicinity of Point Barrow, Alaska, where the shore lead may be too shallow. It is also true in establishing a fixed channel through the fast ice in southern McMurdo Sound to resupply the major U.S. Antarctic base there. In these cases the only options are to break the ice or wait for better conditions.
In areas of large sea-ice fields the process of icebreaking essentially consists of making a passage by breaking the solid ice under the weight of the icebreaker and pushing away the broken ice fragments. If there are areas of open water nearby, ice can usually be pushed into these unoccupied spaces. In high concentrations, usually represented in tenths of surface
ice coverage or in a sheet of level fast ice, the ship must force broken pieces of ice under or onto the adjacent ice cover, or under the ship itself, in order to proceed. Ice may be broken into pieces of such small size that the broken track is filled with a slurry of brash ice. Obviously, thinner ice can be broken more easily than thicker ice, and much thicker ice can be broken in lower concentrations than in higher ones (MacDonald, 1969).
The icebreaking problem must be considered in terms of both ice properties and ship design. Ice resists the ship progress. Resistance depends on flexural strength, friction, buoyancy, and wind- or current-induced lateral pressure from the surrounding ice fields (Booz Allen Hamilton, 2005). The strength of ice depends on the thickness and salinity, which are a function of its age as well as temperature. The frictional resistance of ice is most significantly affected by the presence, amount, and condition of snow cover and the condition of the hull surface. Several inches of snow will notably slow the progress of an icebreaker, increasing friction between the surface and the ship’s hull and absorbing energy from the ship’s momentum. Thicker ice exerts more upward buoyant pressure that an icebreaker must overcome, and the amount of lateral pressure in an ice field hinders the displacement of broken ice as described above and also increases the frictional force by increasing the normal force on the hull.
The ability to break ice efficiently and effectively depends directly on the icebreaker’s hull form characteristics, propulsion power, and type and arrangement of propellers. The bows of icebreaking ships are generally inclined aft toward and below the waterline, as an inclined wedge, so that as the ship moves forward, ice is broken from above and forced downward and to the sides. If the ice is sufficiently thick, the sloped bow will slide on the ice until the weight of the ship exceeds the ice’s flexural strength and breaking occurs. Ice may be broken continuously, or if the ship is stopped, continued progress would require backing down several ship lengths and ramming at high power levels.
Icebreaker hull forms have evolved continually since the nineteenth century. Early wooden ships featured, in addition to thick planking and strong framing, rounded hulls intended to counteract the crushing forces of ice under pressure. With the relatively higher power levels available in modern ships (e.g., 10,000 shaft horsepower (SHP) in the World War II-era Wind class ships versus 60,000 SHP in the Polar class), the hull form has become increasingly important for its efficiency in displacing ice to the sides, under the ship, and away from the propellers.
In addition to the sloped bow and rounded hull form, older icebreaker design characteristics have generally favored a relatively wide beam (to provide a broad track for ships to follow), a curving rather than a parallel middle body (to facilitate maneuverability in ice), a single rudder on the centerline (for maximum protection), poor sea-keeping qualities in open water (a regrettable result of the rounded hull and lack of bilge keels or other appendages to dampen roll), and higher propulsion power than conventional ships (MacDonald, 1969).
Although there has never been a single, standard icebreaker design, recent years have seen much greater diversity in designs. This variation is due to advances in engineering knowledge (especially enhanced by accurate model testing), as well as vessel specialization beyond the predominant escort role of the mid-twentieth century. Today, icebreakers tend to be bigger (reflecting increased mission demands), have greater power (enabled by propulsion advances), and feature more flat hull surfaces (producing effective performance with cheaper construction costs). There appears to be a trend toward azimuthing thrusters in icebreaking vessels, which can provide full power in all directions, over conventional shafts and propellers. Design innovation continues, exemplified by a radical conceptual design for an asymmetric, “dual-acting” icebreaker that would present different aspects of the hull in the direction of motion for icebreaking and open-water navigation.1
Linking the ice environment and the icebreaker is a third factor: skill in operating the ship. At a fundamental level, icebreaking is unnatural for mariners because it requires purposefully hitting objects with a ship. Active navigation of a ship in ice involves continual decision making: searching for obstacles and paths of least resistance in the ice field, assessing environmental conditions, changing headings, adjusting power levels, and so forth. Knowledge of the ship’s capabilities is essential. Much has been written about icebreaking technique, but following “rules of thumb” from an older text are still valid today and gives a good sense of the judgment required in icebreaking (MacDonald, 1969):
Steer courses that will take advantage of ice weaknesses—The shortest way in the ice is almost always the longest way around.
Do not hit ice at high speed.
Never hit a large piece of ice if you can proceed around it; if you must hit it, strike it head on.
Protect propellers and rudder at all times.
Refrain from using all engine power just because power is available. There is danger in the philosophy that difficulties can be overcome by weight and power alone.
Do not hesitate to apply full power when necessary.
Keep moving. If progress is unsatisfactory, either change your tactics, seek better ice conditions, or lie-to to await a change for the better.
POTENTIAL CHALLENGES TO THE U.S. FLEET
Production, Distribution, and Drift of Icebergs in the Ross Sea, Antarctica, and Their Potential Impact on Sea-Ice Conditions
In spring 2000, several of the largest icebergs ever witnessed calved from the Ross and Ronne-Filchner ice shelves. These icebergs were named B15, A43, and A44 by the U.S. National Ice Center, and together they represent approximately 5,000 km3, or about 2.5 times the annual accumulation of ice on the entire Antarctic ice sheet. Despite the fact that the titanic size of these icebergs garnered a great deal of public attention, their creation was not glaciologically unusual or unexpected. The initial width of these icebergs, approximately 40 km, represents about 50 years of the forward flow of the ice shelves—floating glacial ice that is hundreds of meters thick—from which these icebergs calved; thus, their sudden appearance after 50 years of slow northward advance of the ice shelves represents the normal maintenance of Antarctica’s glacial ice coverage. In a steady state the various ice shelves, including the Ross Ice Shelf, must undergo calving of these behemoth icebergs about once every 50 years, simply because the Antarctic ice sheet is either in steady state, or very close to steady state. Aside from a small piece of the Ross Ice Shelf’s front located near 180 degrees longitude, the entire front of the ice shelf has calved back since B15 was released; this means that the next calving, all other effects being equal, is not expected until 2050 or so.
Following the release of B15 from the eastern half of the Ross Ice Shelf’s calving margin, the iceberg broke into several small pieces. Unlike the sibling pieces, B15A failed to take a course of drift that would eject it from the Ross Sea in a matter of a few months. Instead, B15A crashed into the ice front near Ross Island, spawning a smaller iceberg, C16, which quickly ran aground in Lewis Bay and finally settled itself into a “holding pattern” just north of Cape Crozier on the eastern end of Ross Island. B15A remained adrift in this holding pattern for the next four years (until November 2004, when it began to move away) constantly gyrating on the ocean’s diurnal tide.
The reason for B15A’s apparent attraction to the area just north of Ross Island is still a subject of research and debate. It is possible, for example, that the prevailing southerly winds in the region of McMurdo are blocked by the high topography of the island’s tall volcanic cones (Mt. Erebus and Mt. Terror), and this contributes to the protection of icebergs from the effects of fierce winds. Other factors contributing to the iceberg’s failure to flush from the area may include the inverse barometer effect (there is a persistent atmospheric low in the lee of Mt. Erebus and Mt. Terror), and general localized convergence of the ocean currents generated on the western side of the Ross Sea Polynya discussed below.
By April 2006, both C16 and B15A had left the area of Ross Island; the conditions that triggered this move are unclear. B15J continues to hover near Ross Island; however, this smaller, round-shaped iceberg tends to have less of an impact on sea-ice conditions because of its size and inability to move into shallower waters west of its current position (e.g., to run aground on the pinning point that held C16 for so long in a position that could “blockade” sea ice west of Cape Bird).
Ice Conditions in the Ross Sea and McMurdo Sound, Antarctica
With annual resupply of the U.S. Antarctic Program’s major bases, McMurdo and South Pole, currently dependent on ship access into McMurdo Sound, ice conditions in this region are a major factor in recommending a sound strategy in the context of this report. Large-scale ice conditions in the Ross Sea are characterized by the Ross Sea Gyre that transports water and ice in a clockwise fashion through the south-western Ross Sea off McMurdo Sound. The prevailing strong, offshore winds coming down the Ross Ice Shelf result in the development of the Ross Sea Polynya, a vast expanse of open water and thin ice maintained by wind-driven advection of ice to the north (Figure 7.1). This results in a sea-ice thickness gradient such that the thinnest ice in the Ross Sea is found in the very north along the marginal ice zone and the very south where young thin ice emerges from the polynya region (Jeffries et al., 2001). As a consequence of this distribution of thin ice, the ice cover in the Ross Sea typically recedes from both the northern and the southern edges during summer. In most years, however, some ice survives summer melt. This ice is typically confined to the eastern Ross Sea (see Figure 7.1) and occupies less than one-tenth of the total ice-covered area.
The seasonal ice retreat typically does not start until mid-November, with the summer minimum ice extent in the Ross Sea reached during mid-February. Thus, much of the icebreaking associated with resupply efforts (which, of course, are constrained by more than ice conditions, see Chapter 10) takes place one to three months before the climatological seasonal ice minimum in the Ross Sea sector. (This seasonal sea ice is relatively thin and does not pose an undue icebreaking burden.)
In the 1980s and 1990s, the Ross Sea sector of the Antarctic experienced an increase in maximum ice extent of about 9 percent per decade. At the same time, the ice cover of the neighboring Amundsen and Bellingshausen Seas declined by 10 percent per decade, suggesting that at least part of this increase is explained by advection of ice from the west. Despite the increases in ice extent, a freshening of the Ross Sea also points toward reduced ice production and hence overall thinner ice (Jacobs et al., 2002).
McMurdo Sound itself is characterized by a complex ice regime that depends strongly on the interplay of calving
McMurdo Ice Management
In the Antarctic, polar icebreakers are tasked with opening a shipping channel for the resupply of the U.S. Antarctic Program’s McMurdo Station. The difficulty of this task, particularly in challenging ice years, and the use of icebreakers primarily as ice management tools call for an integrated ice management and resupply strategy, ensuring the success of the mission while minimizing the effort expended in high-cost, high-risk icebreaking activities. Key components of such an approach include the following.
In conjunction with changes in the resupply schedule, integrated ice management may result in significant savings and streamlining of operations in the long run. Examples of such a comprehensive approach based on optimization and coordination of all components of a mission include trafficability studies in Alaska’s coastal seas (using the Polar class vessels) or the European Union’s Arctic Demonstration and Exploratory Voyage, focusing on the implementation of a near year-round transportation system in Siberian waters.
of icebergs from and ocean circulation underneath the neighboring Ross Ice Shelf as well as the local surface climate (Box 7.1). Typically, a seasonal, landfast ice cover develops in the sound, extending out to about 20 to 60 km from McMurdo Station at the end of the ice growth season (Brunt et al., 2006; Figure 7.3). As a result of low air temperatures, lack of substantial snow cover and the contribution of underwater ice forming from water emerging from underneath the Ross Ice Shelf and accumulating or growing at the base of the ice sheet, the sea-ice cover in McMurdo Sound is in general the thickest level ice in the Antarctic, growing to more than 2 meters thick during a regular ice season. In a typical ice year, ice retreat in the Ross Sea, heating of McMurdo surface waters, and the action of swell penetrating from the open Ross Sea foster the break-up and removal of much if not all of the landfast ice by February (Crocker, 1988). How-
ever, on occasion (roughly once every decade) the landfast ice would not break or clear out far into the sound for one or two seasons.
The most notable effect of the icebergs was the creation of a natural “breakwater” that extended from Cape Bird, the normal place where McMurdo Sound opens to the Ross Sea, to Franklin Island, approximately 100 km to the north. This breakwater, acting in concert with the transverse breakwater presented by the Drygalski Ice Tongue to the north, created a mediterranean encompassing McMurdo Sound and a 200 km stretch of the Victoria Land Coast that reduced sea-ice mobility and favored the appearance of large areas of multiyear fast ice (Brunt et al., 2006). While causes remain tentative, it appears that the obstruction presented by the icebergs reduced penetration of swell and warm summer water into the sound to the extent that much of the ice in the interior sound could not clear out by the end of the summer melt
season. This led to the accumulation of ice between 4 and 6 meter thickness in large areas south of Cape Evans.
At the same time, a combination of factors resulted in surface melting of the ice cover, likely increasing the strength of this ice due to the associated desalination of the underlying ice layers. The only relatively thin ice found in the inner sound (south of Cape Evans to Cape Armitage) is now confined to the icebreaker channel maintained over the past summer seasons (discernible in the synthetic aperture radar image shown in Figure 7.3). Because of the thickness of the surrounding level ice, it may become increasingly difficult to manage (i.e., displace) the mix of ice fragments and new ice forming each year as this channel is rebroken by icebreakers. As of summer 2006, the iceberg blockade had substantially reduced: B15A left the Ross Sea in October 2005; C16 is currently approaching Coleman Island; and B15J has historically remained east of Lewis Bay, close to Cape Crozier, where it has not had a substantial impact on sea ice. Also, collisions with B15A and C16 in 2005 and 2006, respectively, have shortened the Drygalski Ice Tongue by approximately 20 km. Nevertheless, there is a possibility that the level ice in McMurdo Sound that was fostered by the icebergs over the past several years has now reached such substantial thickness that only a major “centennial” storm is capable of breaking it up and clearing it from the sound. In summary, with several independent factors required to act upon the ice cover for decay and removal, it is at this point quite difficult to predict when the ice situation will improve again from the perspective of resupply operations. If the ice does not clear out on its own, then heavy ice conditions (level ice thicker than 4 meters and high concentrations of very thick ice fragments in a partially cleared channel) in subsequent years, starting with the 2006-2007 resupply season, may present a substantial challenge and could potentially thwart efforts to maintain an open channel to McMurdo pier.
At the same time, however, the factors that led to the development of the difficult sea-ice situation have now abated (with the departure of B15A and C16), and this means that once the sea ice does clear from McMurdo Sound, renewal of difficult sea-ice conditions would be unlikely until another period of unlucky iceberg circumstances (which may not develop again for 50 years, depending on the frequency of calving of the Ross Ice Shelf).
Ice Conditions in the Western Arctic
Ice-covered U.S. waters in the Arctic and sub-Arctic include the Bering Sea as well as parts of the Chukchi and Beaufort Seas, the latter typically referred to as the western Arctic (Figure 7.1). While the Bering Sea has not experienced substantial reductions in winter maximum ice extent (Comiso, 2003), the onset of spring ice retreat has occurred progressively earlier since the late 1970s (Stabeno and Overland, 2001). The Chukchi and Beaufort Seas have seen some of the most substantial changes in ice conditions anywhere in the Arctic. Thus, the greatest reduction in summer ice extent has been observed in the northern Chukchi Sea (Comiso, 2002; Overpeck et al., 2005). At the same time, changes in ice circulation, diminished ice growth, and enhanced summer melt have greatly reduced the amount of thick, multiyear ice in the region (Tucker et al., 2001; Perovich et al., 2003). Whereas the ice edge remained within <50 km of the coastline in waters off northern Alaska during most years in the 1970s and 1980s, it now typically retreats by more than 200 km to the north by the end of summer. However, due to the rapid response of a loose ice cover to shifting winds, and possibly aided by increases in summer storm intensity, the ice conditions overall have also become less predictable, with significant impacts on both wildlife and human activities. Thus, Native hunters and coastal residents report significant impacts on their traditional lifestyle by the changing ice regime (Huntington, 2000; Krupnik and Jolly, 2002). Among other factors, the amount of local rescue operations (in northern Alaska typically supported by the North Slope Borough’s Search and Rescue Operations Center) due to hazardous ice or open water conditions has increased substantially in the past decade or two (George et al., 2004).
Regardless of the recent changes, the western Arctic remains one of the areas with the most diverse, complex ice conditions in the northern hemisphere. This is due to the fact that the clockwise Beaufort Gyre is still advecting thick multiyear ice into coastal regions (and even through the Bering Strait in the winter of 2005-2006) while the coastal wind regime still fosters growth and export of ice in coastal polynyas and leads in the Chukchi Sea. Interactions between drifting ice and the coastline result in some of the largest ridges produced anywhere in the coastal Arctic. The changes observed so far in the ice regime have not resulted in the complete loss of any of these ice types, but rather have increased spatial and temporal variability, arguably rendering prediction and hazard mitigation more difficult. This situation is exemplified by difficulties encountered by a number of vessels in the Chukchi and Beaufort Seas during the 2006 summer. Thus, a Russian icebreaker carrying tourists into the Chukchi Sea encountered thick multiyear ice that impeded the ship progress and significantly altered cruise plans. More important, offshore oil and gas exploration activities that are resurgent as a result of past and planned federal lease sales in the Chukchi and Beaufort Seas were significantly affected, with a larger number of vessels confined to coastal stretches of the Beaufort Sea for most of the summer.