Arctic Ice Shelves and Ice Islands

Ice shelves are relatively thick ice masses that are afloat but attached to coastal land rather than adrift. They form by the seaward extension of glaciers or ice sheets or by build up of multiyear landfast sea ice. They thicken further by surface accumulation of snow and superimposed ice and by accretion of ice from water beneath. Composite ice shelves are composed of sea ice and glacier ice. Glacier tongues are floating ice margins that are narrow relative to their length. Ice shelves comprise 55% or 18,000 km of the Antarctic coast. ‘Classical’ Antarctic ice shelves are fed from glaciers or ice streams and are dynamically part of the parent ice sheet; the largest, the Ross and Ronne, are 10⁵ km² and hundreds of metres thick. Where they ground on isolated bedrock peaks, ‘ice rises’ are formed. Arctic ice shelves are restricted to several archipelagos fringing the Arctic Ocean and to a few Greenland fjords. The Ward Hunt Ice Shelf is the largest at about 400 km². Arctic and Antarctic ice shelves have expanded and contracted during the Holocene. The Ellesmere Ice Shelf developed about 5500 years ago in response to Holocene cooling. In the warmer Twentieth century, calving events have broken this continuous ice-shelf into several remnants. Floating glacier tongues of the Greenland Ice Sheet have also broken up recently. The entire Arctic Ocean may have been covered by a huge ice shelf during the coldest Late Cenozoic glacial periods. Large, often tabular icebergs calve from ice shelves. Ice islands are a form of tabular iceberg in the Arctic Ocean which have a characteristic undulating surface. Icebergs drift mainly under the influence of currents and Arctic Ocean ice islands have been used occasionally as research stations.

The Ellesmere ice shelves are located on the northernmost edge of Canada, the northwest coast of Ellesmere Island, facing the perennial pack ice of the Arctic Ocean. They are the only ice shelves in Canada and the most extensive in the entire Arctic. The special nature of the Ellesmere ice shelves extends to the essential role that sea ice has played in their initiation, maintenance and replacement. This chapter describes ice physical features, types and properties that inform the knowledge and understanding of the origin and development of the ice shelves. There are three main sections. The first is a brief history of the scientific investigation of the ice shelves. The second section describes features at the ice surface, including the characteristics and origin of the rolling topography, geological and biological materials, fractures and channels, and ice rises. The third section describes what is found below the ice surface, including the thickness and types of ice. These include marine ice (ancient and modern), formed by freezing of seawater and brackish water, and meteoric ice formed by surface accumulation of superimposed ice and lake ice (both originating as snow), and bottom accretion of freshwater.

Despite the presence of about 4000 km of marine-terminating glaciers and ice caps in the Eurasian Arctic, there are few floating ice shelves. Neither are there extensive areas of multi-year shorefast sea ice which might thicken into composite ice shelves themselves. The archipelagos of Severnaya Zemlya and Franz Josef Land contain some ice shelves in addition to grounded tidewater ice fronts. The largest Eurasian Arctic ice shelf was the Matusevich Ice Shelf, Severnaya Zemlya, at about 240 km² with a drainage basin of about 1100 km²; this ice shelf largely broke up in 2012. In Franz Josef Land, a number of ice caps have smooth and very low surface gradient seaward margins, covering over 300 km² or 2% of the total area of the ice caps in the archipelago. These low-gradient areas are located mainly in relatively protected embayments and produce large tabular icebergs of up to several kilometres in length. Whether individual areas are floating in hydrostatic equilibrium or are simply close to buoyancy, they provide the major modern source of tabular icebergs to the Barents Sea. Svalbard has about 860 km of coastal ice cliffs, but almost none of the ice margin appears to be afloat. There may be short periods, during the active phase of the surge cycle, where marine margins become afloat. Neither is there evidence that the margins of the marine-terminating glaciers on Novaya Zemlya are floating. Twenty-five to fifty percent of the bed of the three largest ice caps in the Eurasian Arctic lies below sea level. Thus, in a warming Arctic, the ice margin would eventually retreat onto land, curtailing mass loss by iceberg production and providing a break on rapid ice-cap disintegration through calving.

This chapter focuses on a review of the glaciers on north and northeast Greenland that terminate in fiords with long glacier tongues and floating, ice-shelf-like margins. There is some debate as to whether these glacier tongues can be classified as a traditional ice shelf, so the relevant literature and physical properties are reviewed. There exists a difference between: (1) Floating glaciers in northern Greenland (>77°N) which experience bottom melting as their dominant ablation mechanism and calve relatively thin, but large (km-sized) tabular icebergs (‘ice islands’), and (2) Grounded glaciers further south (<77°N), where iceberg calving provides the dominant ablation mechanism. The relatively smaller iceberg discharge in northern Greenland is closely related to the occurrence of extended floating glacier sections, allowing bottom melting estimated at up to 10 m year⁻¹ for locations such as Petermann Glacier. A case study is described of the physical characteristics and historical changes of Nioghalvfjerdsfjorden Glacier, NE Greenland, based on field and remote sensing studies.

The ice shelves along the northern coast of Ellesmere Island have been in a state of decline since at least the early twentieth century. Available data derived from explorers’ journals, aerial photographs and satellite imagery have been compiled into a single geospatial database of ice shelf and glacier ice tongue extent over 13 observation periods between 1906 and 2015. During this time there was a loss of 8,061 km² (94%) in ice shelf area. The vast majority of this loss occurred via episodic calving, in particular during the first six decades of the twentieth century. More recently, between 1998 and 2015, 515 km² of shelf ice calved. Some ice shelves also thinned in situ, transitioning to thinner and weaker ice types that can no longer be considered ice shelf, although the timing of this shift is difficult to constrain with the methods used here. Some ice shelves composed partly of ice tongues (glacier or composite ice shelves) also disintegrated to the point where the ice tongues were isolated, representing a loss of ice shelf extent. Our digitization methods are typically repeatable to within 3%, and generally agree with past determinations of extent. The break-up of these massive features is an ongoing phenomenon, and it is hoped that the comprehensive dataset presented here will provide a basis for comparison of future changes in this region.

This chapter chronicles the mass balance and thickness of the Ward Hunt Ice Shelf and Ward Hunt Ice Rise, Ellesmere Island, Canada (1952–2008). The surface mass balance of the ice shelf and ice rise followed the mass balance changes of other monitored Canadian Arctic glaciers, but their overall mass losses have been comparatively low due to their proximity to the Arctic Ocean. Nevertheless, thinning of the Ward Hunt Ice Shelf via surface and basal melting has reduced its structural integrity as evidenced by the large-scale breakup of its eastern section between 2008 and 2011. In this context basal melting is considerably more significant in terms of overall ice shelf thickness and stability than the associated surface mass losses. The Ward Hunt Ice Shelf cannot readily reform again unless climatic conditions deteriorate for a prolonged period of time. The other remaining Canadian Arctic ice shelves appear to be in a similar situation as open water conditions on the Arctic Ocean become more prevalent and the dynamic stresses on the ice shelves related to wind, wave, and tidal action increase. This will eventually leave the Ward Hunt Ice Rise as one of the last remnants of Peary’s ‘Glacial Fringe’ along the northern coast of Ellesmere Island, although its long-term survival is also threatened by current and predicted climatic change. A comprehensive measurement and modeling program is urgently needed to document and better understand the behaviour of these unique ice masses.

The modern Ellesmere ice shelves constitute the oldest landfast sea ice in the Northern Hemisphere, originally described by late nineteenth century sledging expeditions. At that time, the ‘Ellesmere Ice Shelf’ formed a contiguous, coastal apron that extended hundreds of kilometres in length and covered up to 9000 km². By the mid-twentieth century the Ellesmere Ice Shelf was reduced to several large remnants and half its original area. Early efforts to document the age of ice shelf inception used conventional radiocarbon dating of internal, aeolian layers and marine organisms incorporated by basal accretion, all providing problematic results. The most widely cited age (3000 ¹⁴C yr before present (BP)) was provided by a radiocarbon date on driftwood stranded behind the Ward Hunt Ice Shelf.

Here we constrain the age of ice shelf inception based on 69 accelerator mass spectrometry radiocarbon dates on driftwood collected from raised marine shorelines inland of the Ward Hunt Ice Shelf, including adjacent fiords deglaciated ~9500 cal yr BP. Driftwood entered Disraeli Fiord, behind the Ward Hunt Ice Shelf, as well as Phillips Inlet to the east, continuously from 9500 to 5500 cal yr BP, after which it abruptly terminated until present. This termination is interpreted to record blockage of the coast by establishment of multiyear landfast sea ice, rather than the lack of driftwood availability due to changing ocean currents, because driftwood delivery continued farther east at several locations through the same interval. This driftwood delivery was favoured by the seasonal expansion of the Lincoln Sea polynya that extends from northern Nares Strait to Clements Markham Inlet.

The radiocarbon chronology of entrapped driftwood behind the former Ellesmere Ice Shelf provides a proxy for the severity of Arctic Ocean pack ice with which it co-varies. This record extends back at least five millennia, far beyond the four decades of satellite surveillance, demonstrating that modern sea ice reduction extends back to the early twentieth century. The dramatic acceleration of seasonal pack ice reduction in the Arctic Ocean during the early twenty-first century heralds the imminent demise of the remaining ice shelf remnants across northern Ellesmere Island, considered an unprecedented event on the scale of millennia. Fiords adjacent to the former Ellesmere Ice Shelf—historically occupied by mulit-year landfast sea ice—have also become seasonally ice-free within the last decade.

Information about the long-term responses of Arctic ice shelves to changes in climate is critical for understanding the significance of their recent declines. However, the history of these systems is poorly understood, and few methods exist to reconstruct these missing datasets. Paleoenvironmental studies of sediment cores can provide valuable information about past ice shelf fluctuations and provide the context in which to evaluate recent changes. This chapter discusses the methods available for the reconstruction of ice shelf dynamics from sediment cores. It reviews the proxy indicators that provide insights into different ice shelf characteristics and their effects on sedimentation and water column properties, and gives examples where these have been applied in studies of ice shelves from both polar regions.

Arctic ice shelves are microbial ecosystems with a rich biodiversity. Until recently, polar ice shelves were seen as mostly abiotic glaciological features, however they are oases for life, with snow, meltwater pools and sediments providing cryohabitats for microbiota. The biological communities are composed of diverse forms of microscopic life, including cyanobacteria, heterotrophic bacteria, viruses, algae, other protists and microfauna, and occupy a variety of habitats: supraglacial meltwater lakes, englacial microhabitats within the ice and snow and planktonic environments in ice-dammed, epishelf lakes. These habitats are defined by seasonal light availability, cold temperatures and nutrient poor conditions. In the supraglacial pools, production is dominated by benthic microbial mat assemblages that have diverse stress adaptation systems and that use internal nutrient recycling and scavenging strategies. Despite short growth periods and perennial low temperatures, biomass accumulations are considerable, with a striking diversity of light-harvesting, UV-protection and other accessory pigments. The chemical characteristics such as conductivity and origin of salts are defined by the underlying ice types, and microbial mat studies from adjacent habitats show a high resilience to solute concentration during freeze-up. The structural integrity of these cryoecosystems is dependent on ice, and they are therefore vulnerable to climate change. Many of these unique Arctic ecosystems have been lost by ice shelf collapse over the last two decades, and they are now on the brink of complete extinction.

A review of historical literature and remote sensing imagery indicates that the ice shelves of northern Ellesmere Island have undergone losses during the 1930s/1940s to 1960s, and particularly since the start of the twenty-first century. These losses have occurred due to a variety of different mechanisms, some of which have resulted in long-term reductions in ice shelf thickness and stability (e.g., warming air temperatures, warming ocean temperatures, negative surface and basal mass balance, reductions in glacier inputs), while others have been more important in defining the exact time at which a pre-weakened ice shelf has undergone calving (e.g., presence of open water at ice shelf terminus, loss of adjacent multiyear landfast sea ice, reductions in nearby epishelf lake and fiord ice cover). While no single mechanism can be isolated, it is clear that they have all contributed to the marked recent losses of Arctic ice shelves, and that the outlook for the future survival of these features is poor.

Ice islands are large tabular icebergs produced from the calving of Arctic ice shelves. The loss of ~8000 km² of ice shelves from the northern coast of Ellesmere Island over the past century has resulted in the production of numerous ice islands, with the first detected in the 1940s. Once calved, these ice islands take one of three routes: (1) they drift west and remain in the Arctic Ocean, where they can circulate for up to several decades; (2) they drift west and enter the interior islands of the Canadian Arctic Archipelago, where they disintegrate relatively rapidly; (3) in rare cases they drift east after calving and enter Nares Strait and then drift south along the east coast of Canada, reaching as far south as Labrador. Historically, the drift paths and disintegration patterns of ice islands were of military interest as they provided mobile platforms for the measurement of oceanographic and atmospheric properties, and they could potentially act as staging posts for Soviet or US access to the opposite side of the Arctic Ocean. Today, interest in ice islands primarily arises from the risks that can pose to shipping and offshore oil exploration, and their indication of the effects of climate change.

For most of the twentieth century, multiyear landfast sea ice (MLSI) existed in semi-permanent plugs across Nansen Sound and Sverdrup Channel in the northern Queen Elizabeth Islands (QEI), Canada. Between 1961 and 2004, these ice plugs only experienced simultaneous break-ups in 1962 and 1998. However, break-ups of both ice plugs have occurred in 9 out of the 12 years since 2005, indicating that these features are not reforming. The history of these plugs is reviewed using Canadian Ice Service ice charts, satellite imagery and a literature review. The weather systems associated with plug break-up events are related to the synoptic patterns defined by Alt (Atmos-Ocean 3:181–199, 1979). Most break-ups occur during Type III synoptic conditions, when a low centers over the Asian side of the Arctic Ocean and a warm pressure ridge develops over the QEI, creating warm temperatures, clear skies, and frequent wind reversals. Ice plug break-ups are also associated with reductions in sea ice concentration along the northwest coast of Ellesmere Island. The removal of these MLSI plugs in recent years aligns with ice shelf losses and reductions in age and thickness of sea ice in the Canadian Arctic Archipelago, with implications for ice import and export through these channels and the response of Arctic sea ice to a changing climate.

The Arctic Ocean, because it is ice-covered, facilitates research difficult to conduct in the open seas. The polar pack and, upon discovery, ice islands offer natural (if problematic) platforms. When first detected, ‘floating islands’ aroused immediate interest by the U.S. Air Force then the U.S. Navy. Large, deep-draft masses, they resist sea ice pressure and breakup; as platforms, ice islands confer long-term occupation. Superpower rivalry propelled Cold War science. The circumpolar North had become a theater of operations, an exposed flank. Field programs multiplied on both sides of the Central Arctic, as did the number of ice-based outposts for research. Air-deployed drifting stations are a Soviet logistic invention; from 1954 to 1991, the USSR sustained a continuous presence on drifting ice islands. As U.S. Air Force concern for the Arctic eased, the under-ice capabilities of nuclear submarines intensified and programs to understand (and exploit) the Polar Basin were developed: oceanography, geophysics, underwater acoustics, sea-ice physics, meteorology, marine biology. The ballistic-missile submarine introduced anti-submarine warfare into Arctic waters, further stimulating research pertaining to the floating ice cover and Arctic Ocean acoustics.

Aviation helped open the Arctic to systematic observation. To date, Russia has deployed five drifting stations onto ice islands, the United States two, and Canada one. These were tiny communities, with deep isolation at close quarters. On sea ice or ice island, the potential for emergency haunted every camp manager. Success depended on the cooperation of researchers with logistics-support staff. In the post-Cold War era there are different challenges in the Arctic: climate change, some now argue, poses a threat to national security.

A summary of Russian discoveries of Arctic ice islands – peculiar tabular icebergs – is presented, complete with a chronological account of drifting stations installed on ice islands. Of 40 ‘North Pole’ drifting stations established from 1937 through 2013, six were set up on five ice islands: North Pole-6, 18/19 (same ice island), 22, 23, and 24. These ice islands served as reliable long-term research platforms as evidenced by the extensive bibliography of scientific publications based on observations made from manned ice island stations. Studies were conducted of structure and morphology of ice islands; under-ice biota; deep Arctic Ocean benthos; meteorology and climate; and oceanography. Biological collections from these ice islands are still being analyzed.

When considering man-made structures for offshore Arctic regions where ice islands may transit, the probability of encountering different sizes of ice islands and ice island fragments needs to be estimated. This can be difficult to determine in near-shore regions where the occurrence of ice islands is infrequent, yet if there is an occurrence and the ice island grounds in shallow water, it could break into a number of smaller ice island fragments. Probabilistic design methods can be used to determine if consideration of ice island impact in structural design criteria is required and to choose appropriate levels of ice strengthening. This chapter includes a brief description of the design issues involved, available approaches and areas where additional information would be useful.