The Elusive Arctic        

The Elusive Arctic

(by Willy Østreng)


Many criteria have been used to delineate the Arctic from the temperate zone. Neither of them has achieved common approval or acceptance.

Figure 1: Different Definitions of the Arctic

Source: AMAP (1997)

The oldest and most common definition focuses on solar radiation, defining the Arctic as the area north of the Arctic Circle (66 degrees 32 minutes North), which encircles the area of the midnight sun. This area is about 21 million square kilometres in extension – making up about 8 percent of the Earth’s surface1 . This definition excludes sub-regions like the Bering Strait, the White Sea, the southern part of Greenland and the Hudson Bay (see Figure 1). In most expert opinion, these areas are as arctic in natural conditions as most of the areas situated north of the Circle. This has made the noted Canadian researcher, Moira Dunbar to characterize the Arctic Circle – a purely astronomical concept – to be meaningless “from any other point of view2 . A more reasonable approach would in her mind be to define the Arctic by physical criteria that are domestic to the region.

In line with this thinking, multiple alternative criteria have been suggested and used: the tree line, the 10 degree C July Isotherm, the continuous permafrost, the sea ice cover, the marine boundary between cool and warm waters etc. (see Figure 1). A fairly popular and long-lived assumption has been that the limits of the tree line, which in certain areas coincides with the 10 degree July Isotherm , are more meaningful from the point of view of human activity than any of the climatic ones2 .  Although the terrestrial and marine environments vary considerably throughout this area, they nevertheless share certain Arctic characteristics that distinguish them from temperate regions, including cold temperatures, extensive snow and ice cover, large seasonal fluctuations in solar energy, continuous permafrost, polar darkness and lack of nutrients3 . Thus, in terms of natural features, this definition makes the Arctic stands out as a relatively coherent and uniform region. But also these criteria suffer from weaknesses that act counter to common perceptions of what the Arctic is and is not. For instance, the blue oceans of the North Atlantic and North Pacific are here defined as constituent parts of the Arctic along with the ice covered waters of the Central Arctic Ocean. This definition also excludes certain coastal areas of the Northeast- and Northwest Passages from the Arctic.

In short: The application of single definitional criteria does not suffice to include all geographical areas hosting polar characteristics in the concept of the Arctic, or for that matter, to exclude areas featuring the characteristics of the temperate zone. The differences in aerial extension between the various definitions amount to thousands of square kilometres. The competing delineations depend upon the perspective from which one approaches it. This is not problematic when working within the confines of single scientific disciplines; rather, the difficulty arises when one seeks a definition to suit all subject areas – the multi- and cross-sections of the sciences as well as of industries.

In the multilateral process of defining the Arctic in a way that is relevant for all areas of science, the Arctic Monitoring Assessment Program (AMAP) of the Arctic Council decided not to define the region in too precise terms. Instead the Program provides guidelines about a core geographical area to be used by all scientists irrespective of disciplinary orientation. Although the boundaries are not fixed once and for all they are to be located somewhere in the latitudinal area between 60 degree N and the Arctic Circle. This approach provides flexibility in boundary drawing without repudiating borders as a useful measure for multi- and interdisciplinary science. The flexibility achieved also shows in the modifications of borders suggested by AMAP. One such modification is that the Hudson Bay and the White Sea – which were excluded as Arctic territory by the Arctic Circle - are defined as integral parts of the Arctic. A second modification concerns the terrestrial environment of the eight Arctic states. Here the southern boundary of the region shall not be determined by any concrete natural criteria but by the individual countries in locations of their own liking between the Arctic Circle and 60 degree N. (see Figure 1)4 . In this definition, parts of the land territories of all the eight Arctic states have been defined as belonging to the Arctic. Previously, only the five littoral states to the Arctic Ocean were thought of having land territory in the region. What is more, by this definition the blue warm waters of the Norwegian Sea has become part of Arctic waters, which again deviates from common perception of what is characteristic of polar natural conditions. By this extension, the Northern Maritime Corridor becomes an integral part of Arctic waters along with the three Passages contained within the ice-covered part of the Arctic Ocean (see Figure 1). In AMAP terms, the Arctic covers an area of approximately 33, 4 million square kilometers, out of which 60 percent is defined as Arctic water5 . Here, the AMAP applies multiple scientific, political and pragmatic criteria that have been blended together to reach consensus across sectors and between the states. However, the ambiguity and flexibility of this definition causes overlaps with the Sub-arctic, which in principle should be the transition zone between the Arctic proper and the temperate zone (see Figure 1).

The Sub-Arctic is even more difficult to define than the Arctic proper. In fact no completely satisfactory definition has ever been put forward, and it has become necessary to compromise on some arbitrary climatic boundaries to sketch out what the Sub-arctic is. The complexity of the matter is shown in Figure 1 where the boundaries of the Sub-arctic at certain stretches in North America and Siberia are located north of the southern boundary of the Arctic proper. In other places the borderline is situated south of the Arctic as it should according to theory. The AMAP definition has become a ‘bastard’ between two definitional traditions and the Sub-Arctic. How then does this broad scale definition relate to recent studies on shipping in these waters?

The Shipping Industry and Broad Based Definitions

Traditionally, and based purely on scientific criteria, the boundaries of the exit areas – or the open sea borders - of the Arctic Ocean has been drawn along the Barents shelf edge from Norway to Svalbard, across the Fram Strait, down the western margin of the Canadian Archipelago and across the Bering Strait. Thus defined and including the Canadian Archipelago the total area of the Arctic Ocean is 11.5 million square kilometres6 . In AMAP terms the “marine Arctic” or “Arctic waters” has been extended far beyond the Arctic Ocean to denote an area that includes Baffin, Hudson and James Bays; the Labrador, Greenland, Iceland, Norwegian, and Bering Seas (see Figure7  1). This definition has been embraced by some of the most cited Arctic shipping assessment of the present.

The Arctic Marine Shipping Assessment (AMSA) of the Arctic Council has adopted the AMAP definition lock stock and barrel, and allowed the five coastal states to define their own Arctic waters for the purpose of data collection efforts. In this way the AMSA definition stand out as the “least common denominator” of the eight Arctic States, covering an extensive area of a variety of operational conditions. Shipping is here defined to cover all types of marine transport including tankers, bulk carriers, offshore supply vessels, passenger ships, tug/barge combinations, fishing vessels, ferries, research vessels and government and commercial icebreakers1 . The Canadian Arctic Shipping Assessment (CASA) follows suit when claiming that shipping “.. is best defined in the broadest terms, whilst recognizing that there will be differences that depend on the issues under consideration8 .” Also in the CASA study shipping is taken to include: cargo vessels, passenger vessels, military/defence vessels, navigational support vessels, oil/production vessels, support vessels and fishing vessels8 . These lists suggest that Arctic waters comprise blue waters as well. Several of these vessels – among them fishing boats - have no sea ice capability and are functionally bound to operate in open waters only. As shown in Figure 2, the most busy ‘Arctic waters’ of vessel activity in 2004 were, according to the AMSA, the Norwegian and Bering Seas.  Fishing vessels were responsible for an important part of this total.

Figure 2: Overview of all vessel activity in Arctic waters for 2004, including fishing vessels

Source: AMSA (2009), p. 73

In this definition, Arctic waters span a variety of different sea areas – cold, warm, ice free, ice covered, salty, less salty, etc. - waters of no, few and multiple polar characteristics. By implication, AMSA defines these waters along three continuums. The first continuum is, as discussed above,  definitional extending from temperate warm waters in the southern part of the Norwegian Sea via the cold but ice free waters of the Barents Sea to the permanently ice covered waters of the Central Arctic Ocean. Only the northern part of this long ocean stretch contains true Arctic characteristics. The second continuum is seasonal were operational conditions varies throughout the year according to the annual cyclical process of melting and freezing of sea ice. This continuum relates to the Arctic Ocean only (as defined by scientific criteria only, see above). The third continuum is generational and concerns the man made long-term melting and reduction of sea ice in the Arctic Ocean. This continuum spans from the continuous reduction of sea ice in the Central Arctic Ocean to the geographically expanding ice free stretches of the Northeast and Northwest Passages in summer, but not in winter. The latter continuum has moved the shipping industry to take a second look at the sailing conditions of the Arctic Ocean.

If all three continuums are put together – as the AMSA study does - the challenges encountered by ships in these waters are gradual as one sails northward from the southern part of the Norwegian Sea and into and through the Arctic Ocean. In the south the ships may encounter storms that are domestic to the temperate zone; when reaching the ice free parts of the Barents Sea some of the operational conditions of the Arctic like icing from sea spray and polar darkness in the winter will pose a challenge to shipping. Further north and east the ship will enter into areas of light sea ice conditions in addition to icing, polar darkness, fog etc. followed by more and more heavy ice conditions.

There is something basically wrong with definitions of this kind. They all comprise temperate waters that have been sailed for centuries and where the operational parameters are well known to the shipping industry. What makes the Arctic a true shipping challenge is its domestic characteristics - sea ice, cold temperatures, spraying, whitening, pitch darkness in winter etc. In this perspective, there is a need to get away from “politically motivated” definitions of Arctic waters like the ones applied by the AMSA and CASA studies and focus on the issue specific needs of the shipping industry in the Arctic Ocean proper. The AMSA borders are much too liberal to make the concept of Arctic waters a uniform geographical and operational entity. As seen from the point of view of economic planning, delineations have one function only: to encircle areas of operations that present relatively uniform challenges. As is well known, ships built for sea ice operations are performing inefficiently in blue waters.

Today, some define the Arctic as the physical and biological environment spanning across the northern reaches of Asia, Europe and North America, and the expanses of sea and ocean in between them9 . Here the Arctic is thought of as a group of concepts and attributes unique to a region far from the industrial centres of the world10 . Definitional attempts of this kind share a common feature: The region is defined borderless and portrayed as an aerial entity fading away somewhere in the south without telling where and why it happens. In principle as well as in operational terms, lack of defined borders creates on the one hand uncertainties about the extension of the region, allowing for definitions like the ones suggested by AMSA and CASA. On the other hand, it opens up for an issue specific border setting, leaving it to the mining and shipping industry to tailor made borders that make sense from their respective operational points of view. One issue specific definition that has achieved broad international acceptance in the shipping industry is launched and applied by the International Maritime Organisation (IMO).

Issue Specific Definitions Relating to Shipping in Arctic Waters

The IMO issued in December 2002 Guidelines for ships operating in Arctic Ice-covered waters (IMO Guidelines), in which it made sea ice concentrations the only definitional criteria to define Arctic waters. The IMO definition states that waters that have “..sea ice concentrations of 1/10 coverage or greater.. and which pose a structural risk to ships” are Arctic in character11 .  In this definition the three Passages, that can be delineated by fairly exact historical coordinates, and the three types of routes located in the ice-covered waters of the North are all safe within the confines of the IMO definition (see chapter 1). At the same time, connecting transportation corridors extending from the Arctic to southern markets are singled out on the basis of their own ice free operational characteristics.

Figure 3: The IMO definition of Arctic waters


Source: IMO Guidelines (2002)

The IMO definition applies just one arctic criterion – sea ice concentration – to delineate its waters of operations. In so doing, it is Arctic specific and therefore useful in regional shipping operations. In this study we take the IMO definition as our point of departure but add two more sea areas – the White Sea, which freezes over in winter and the southern part of the Barents Sea – which host some of the other natural characteristics of Arctic waters - sea ice spraying, cold temperatures, darkness in winter etc.

IMO used 60 degree N to define Arctic waters in North America (see Figure 3). The same parallel can be used to identify those parts of the Canadian and Alaskan land territories that in due time may prove to have a bearing on shipping in these waters. Thus, in functional terms we regard these land territories including Greenland to be part of the Arctic region. On the Eurasian side, the land territories north of the Tree line in Russia and Norway are part of the Arctic because these areas already have a function in Arctic shipping (see Figure 1).

Some Concluding Remarks

In short, all of the different criteria used to reach a general agreement on the delineation of the Arctic from southern zones are arbitrary and only partially satisfactory. Although the AMAP, AMSA and CASA definitions have reached a certain degree of international acceptance, the business of reaching an overall consensus on what the Arctic is or is not, is still a matter of functionality and debate. As of the present, competing definitions live side by side and are used interchangeably as if they cover identical territories. Representatives of different industries are therefore well advised to be aware of the definitional confusions applying to the region, not least in multi-sector projects like shipping.

The only definitions to be used consistently for operational purposes seems to be functional - issue specific in character - reflecting the uniqueness and pragmatic needs of the respective industries. In this study, the concept of Arctic waters comprises the areas of the IMO definition plus the White Sea and the southern part of the Barents Sea. These waters are in turn part of the Arctic region which also comprises the land areas north of the tree line in Eurasia and the 60th parallel in North America.


  •  1. AMSA (2009), Arctic Marine Shipping Assessment, Report, PAME, Arctic Council, Terragraphica, Anchorage, April 2009
  •  2. Dunbar, M. (1966), The Arctic Setting, R. Macdonald (ed.): The Arctic Frontier, University of Toronto Press, Toronto, 1966, pp. 3-25
  •  3. AMAP (1998), p. 117
  •  4. AMAP (1998), p. 7
  •  5. AMAP (1998), pp. 7 and 10
  •  6. CIA ( ), p. 26
  •  7. AMAP ( ), p.26
  •  8. CASA (2007), Canadian Arctic Shipping Assessment, By The Mainport Group Ltd. for Transport Canada, June 2007
  •  9. Polunin, N. (1951), The Real Arctic: Suggestions for its Limitation, Subdivision and Characterization, Journal of Ecology, vol. 39, 1951, pp. 308-315
  •  10. Rogers, T.G., and G. Rowley (1978), The Circumpolar North, Methuen & Co., London, 1978
  •  11. IMO Guidelines (2002). Guidelines for ships operating in Arctic ice-covered waters. IMO Doc. SC/Circ.1056 MEPC/Circ.399, 23 December 2002.

Willy Østreng, 2010, The Elusive Arctic, CHNL.©

Search Guide [toggle]

Last-Visited Pages [toggle]