by Judith Curry
I spent the 1990’s conducting research on the climate dynamics of the Arctic Ocean, and then moved onto other things circa 2002. My interest in the Arctic has recently been reinvigorated by the increasing societal implications of reduced sea ice extent in terms of security issues, the prospect of a northern sea route, implications for resource exploration and extraction, and adaptation issues for coastal villages along the Arctic Ocean coast.
This multi-part series will begin with an overview of what we know about the climate dynamics of the Arctic Ocean sea ice.
The definitive paper on the history of Arctic sea ice is this recent paper by Polyak et al.
History of sea ice in the Arctic
Leonid Polyak, Richard B. Alley, John T. Andrews, Julie Brigham-Grette, Thomas M. Cronin, Dennis A. Darby, Arthur S. Dyke, Joan J. Fitzpatrick, Svend Funder, Marika Holland, Anne E. Jennings, Gifford H. Miller, Matt O’Regan, James Savelle, Mark Serreze, Kristen St. John, James W.C. White, Eric Wolff
Quaternary Sciene Reviews 29 (2010) 1757–1778
Abstract. Arctic sea-ice extent and volume are declining rapidly. Several studies project that the Arctic Ocean may become seasonally ice-free by the year 2040 or even earlier. Putting this into perspective requires information on the history of Arctic sea-ice conditions through the geologic past. This information can be provided by proxy records from the Arctic Ocean floor and from the surrounding coasts. Although existing records are far from complete, they indicate that sea ice became a feature of the Arctic by 47 Ma, following a pronounced decline in atmospheric pCO2 after the Paleocene–Eocene Thermal Optimum, and consistently covered at least part of the Arctic Ocean for no less than the last 13–14 million years. Ice was apparently most wide- spread during the last 2–3 million years, in accordance with Earth’s overall cooler climate. Nevertheless, episodes of considerably reduced sea ice or even seasonally ice-free conditions occurred during warmer periods linked to orbital variations. The last low-ice event related to orbital forcing (high insolation) was in the early Holocene, after which the northern high latitudes cooled overall, with some superimposed shorter- term (multidecadal to millennial-scale) and lower-magnitude variability. The current reduction in Arctic ice cover started in the late 19th century, consistent with the rapidly warming climate, and became very pronounced over the last three decades. This ice loss appears to be unmatched over at least the last few thousand years and unexplainable by any of the known natural variabilities.
Full paper available online [link]
Additional background material is available in the CCSP Synthesis and Assessment Report on Arctic Climate Change
Millennial to decadal variability
The current interglacial is of particular interest in terms of understanding the recent sea ice variability and decline, particularly in terms of millennial to decadal varaibility. The relevant text from Polyak et al. is reproduced here (note, for readability I have removed the reference citations):
4.3.2. Suborbital variability
Owing to relatively high sedimentation rates at continental margins, paleoceanographic environments and ice drift patterns can be constructed on suborbital (millennial to decadal) scales from some sedimentary records. These high-resolution records reveal a considerable complexity in the system including forcings operating on suborbital scales (such as solar activity, volcanic eruptions, and atmospheric and oceanic circulation), changes in seasonality, and links with lower latitudes. Thus, the periodic, centennial-scale influx of large numbers of iron-oxide grains from the Siberian shelves to the northern Alaska margin has been linked to a reduced Beaufort Gyre and a shift in the Trans-Polar Drift toward North America. Under present conditions such a shift occurs during a positive phase of the Arctic Oscillation. The finding that similar changes may have occurred on century to millennial-scales argues for the existence of longer-term atmospheric variability in the Arctic than the decadal Arctic Oscillation observed during the last century.
Variations in the volumes of IRD in the subarctic North Atlantic indicate several cooling and warming intervals during Neoglacial time (late Holocene), similar to the so-called ‘‘Little Ice Age’’ and ‘‘Medieval Warm Period’’ cycles of greater and lesser areas of sea ice known from the last millennium. Southward polar-water excursions have been reconstructed at several sites in this region as multi-century to multidecadal-scale variations superimposed on the longer-term trends. Millennial variations in ice conditions have also been suggested for the Barents Sea reflecting changes in atmospheric and oceanic interactions between the North Atlantic and the Arctic.
In the Arctic centennial climate variability during the cooling of the last 2 kyr was more subdued than in the Northern hemisphere as a whole, although major temperature anomalies like the warmings around 1000 and 500 A.D. can be discerned. A final peak of bowhead bones appears to have culminated shortly prior to 1000 A.D. in the Beaufort Sea and somewhat later in the eastern CAA [Canadian Arctic Archipelago], suggesting the possibility of temporarily open channels. This inference is consistent with the IP25 record, which indicates a relatively decreased spring ice occurrence between ca 1.2 and 0.8 ka (800– 1200 A.D.). At the end of this time the bowhead-hunting Thule Inuit (Eskimos) expanded eastward out of the Bering Sea region and ultimately spread to Greenland and Labrador. The subsequent decline of bowhead abundances in the CAA is consistent with the abandonment of the high Arctic of Canada and Greenland by bowhead hunters, while Thule living in more southern Arctic regions increasingly focused on alternate food resources. The warming event around 1500 A.D. is identified by climatic simulations in the Atlantic sector of the Arctic and is explained by the internal variability of atmospheric circulation. The subsequent cooling culminated in the ‘‘Little Ice Age’’, between ca 1600 and 1850 AD, when ice conditions in the high- Arctic remained especially prohibitive for navigation.
Historical observations in the Nordic Seas since mid-18th century indicate multidecadal oscillations in ice extent super- imposed on an overall trend of retreating ice margin. Similar oscillations, although with somewhat variable frequencies, are inferred for a longer, 800-yr period from an ice- core/tree-ring proxy record. These multidecadal changes are probably related to variability in the North Atlantic thermohaline circulation, but their mechanism is not well understood and may involve a combination of internal variability in the circulation with external factors such as solar and aerosol forcings. No record of similar oscillations prior to the 20th century is known from other parts of the Arctic, although most of the paleo-data series existing to date lack sufficient detail.
Historical sea ice record
With regards to the historical record of sea ice extent, again from Polyak et al:
The composite historical record of Arctic ice margins shows a general retreat of seasonal ice since about 1900, and accelerated retreat of both seasonal and annual ice during the last five decades (Fig. 2a). The most reliable observations are from 1979 onwards, corresponding to the modern satellite era. Patterns of ice-margin retreat may differ between different periods and regions of the Arctic, but the overall retreat trend is clearly larger than decadal-scale variability, consistent with observations and modeling of the 20th-century ice concentrations and water temperatures. The severity of present ice loss can be highlighted by the breakup of ice shelves at the northern coast of Ellesmere Island, which have been stable until recently for at least several thousand years based on geological data. On the basis of satellite records, negative trends in sea-ice extent encompass all months, with the strongest trend in September. As assessed by the U.S. National Snow and Ice Data Center, the September trend over the period 1979–2009 is 11% per decade. Conditions in 2007 serve as an exclamation point on this ice loss. The average September ice extent in 2007 of 4.28 million km2 was the lowest in the satellite record and 23% lower than the previous September 2005 record low of 5.56 million km2. On the basis of an extended sea-ice record, it appears that the September 2007 ice extent is only half of that estimated for the period 1950–1970 based on the Hadley Center sea ice and sea-surface temperature data set (HadlSST)). While the ice extent rebounded slightly in September 2008 and 2009, these months rank second and third lowest in the satellite record, respectively.
Many factors may have contributed to this ice loss, such as general Arctic warming, extended summer melt season, and effects of the changing phase of large-scale atmospheric patterns such as the Northern Annular Mode and the Dipole Anomaly. These atmospheric forcings have flushed some thicker multi-year ice out of the Arctic and left thinner first-year ice that is more easily melted out in summer, changed ocean heat transport, and increased recent spring cloud cover that augments the long- wave radiation flux to the surface. Strong evidence for a thinning ice cover comes from an ice-tracking algorithm applied to satellite and buoy data, which suggests that the amount of the oldest and thickest ice within the multi-year pack has declined significantly. The area of the Arctic Ocean covered by predominantly older ice (5 or more years old) decreased by 56% between 1982 and 2007. Within the central Arctic Ocean, the coverage of old ice has declined by 88%, and ice that is at least 9 years old (ice that tends to be sequestered in the Beaufort Gyre) has essentially disappeared. Examination of the distribution of ice of various thicknesses suggests that this loss of older ice translates to a decrease in mean thickness for the Arctic from 2.6 m in March 1987–2.0 m in 2007.
The satellite data record: 1979-present
The most detailed time series representation of sea ice extent is obtained from the UIUC cryosphere web site [link]. While these data are most often interpreted in the context of a linear trend, it is instructive to interpret the record in the context of a (qualitative) change point analysis, defined by changes in trend, mean value, amplitude of the annual cycle, and interannual variability.
- 1979-1988: little trend, consistent interannual variability in the amplitude of the annual cycle.
- 1989-1996: small negative trend (more prominent in the summertime minima), large interannual variability.
- 1997- 2003: lower values relative to the period prior to 1996, with the most noticeable decrease in ice extent being the wintertime maximum; small amplitude and fairly regular annual cycle.
- 2003-2007: marked decrease in wintertime maxima; strong negative trend in both winter max and summer min; continued small amplitude of the annual cycle. A steady decline in wintertime maxima from 2003 to 2007 seems to have led the decline in summertime minima, with a marked decline beginning in 2005 that culminated in the major anomaly of summer 2007.
- 2007-present: return to a large amplitude annual cycle (with an increase in the wintertime max), but with a an overall shift to lower summertime values. The winter 2011 values look anomalously low, possibly with a pattern resembling 2006.
Impact of teleconnection and flow regimes
To what extent can we relate these change points to known climate shifts or changes in teleconnection regimes?
- 1989: shift to strong positive phase of the Arctic Oscillation. This resulted in the transport of multiyear ice out of the Arctic Ocean, setting the stage for reduced ice extent (Rigor et al. 2004)
- 1995/1996: shift to warm phase of the Atlantic Multidecadal Oscillation (AMO); shift to near neutral or negative phase of the Arctic Oscillation
- 2003: vicinity of shift to cool phase of the PDO
The period of strongly positive Arctic Oscillation during 1989-1995 and the loss of multi-year ice set the stage for the large decline in sea ice extent in the past decade.
Possible impacts of the AMO and PDO can be inferred from looking at the longer historical time series of sea ice extent (Fig 2a in Polyak et al.):
- PDO: The 1976 shift to the warm phase of the PDO is evident by a jump to lower wintertime maxima. This would be consistent with the increase in wintertime maxima with the circa 2003 shift to the cool phase of the PDO.
- AMO: The broad cycle of the AMO has some apparent correlation with summertime minima, with high values circa 1880 and 1940’s-1950’s, characterized by the warm phase of the AMO.
There are also several regional studies that examine the impact of teleconnection regimes on regional sea ice variability:
Mahoney, A. R., R. G. Barry, V. Smolyanitsky, and F. Fetterer (2008), Observed sea ice extent in the Russian Arctic, 1933–2006, J. Geophys. Res., 113, C11005, doi:10.1029/2008JC004830. [Link]
We present a time series of sea ice extent in the Russian Arctic based on observational sea ice charts compiled by the Arctic and Antarctic Research Institute (AARI). These charts are perhaps the oldest operational sea ice data in existence and show that sea ice extent in the Russian Arctic has generally decreased since the beginning of the chart series in 1933. This retreat has not been continuous, however. For the Russian Arctic as a whole in summer, there have been two periods of retreat separated by a partial recovery between the mid-1950s and mid-1980s. The AARI charts, combined with air temperature records, suggest that the retreat in recent decades is pan-Arctic and year-round in some regions, whereas the early twentieth century retreat was only observed in summer in the Russian Arctic. The AARI ice charts indicate that a significant transition occurred in the Russian Arctic in the mid-1980s, when its sea ice cover began to retreat along with that of the rest of the Arctic. Summertime sea ice extents derived from the AARI data set agree with those derived from passive microwave, including the Hadley Centre’s global sea ice coverage and sea surface temperature (HadISST) data set. The HadISST results do not indicate the 1980s transition or the partial recovery that took place before it. The AARI charts therefore add significantly to our understanding of the variability of Arctic sea ice over the last 8 decades, and we recommend their inclusion in future historical data sets of Arctic sea ice.
Vinje, Torgny, 2001: Anomalies and Trends of Sea-Ice Extent and Atmospheric Circulation in the Nordic Seas during the Period 1864–1998. J. Climate, 14, 255–267. [Link]
Abstract. The extent of ice in the Nordic Seas measured in April has decreased by 33% over the past 135 yr. Retrospective comparison indicates that the recent decrease in the ice extent is within the range of variability observed since the eighteenth century. Temporal, monotonically reduced extreme events occur with intervals of 12–14 yr, suggesting that series longer than 30 yr should be considered to obtain statistical significance regarding temporal changes. Otherwise, decadal temperature variation is also found in the northbound warmer ocean currents. The temperature in the upper layers of these currents seems moreover to have increased by the order of 1°C since the cooling during the Little Ice Age. This temperature increase accounts for most of the ice extent reduction since 1860. A strong negative correlation is found between the larger North Atlantic oscillation (NAO) winter index and the Nordic Seas April ice extent, and a corresponding positive correlation is observed for the Newfoundland–Labrador Sea. It is not until the warming of the Arctic, 1905–30, that the NAO winter index shows repeated positive values over a number of sequential years, corresponding to repeated northward fluxes of warmer air over the Nordic Seas during the winter. An analog repetition of southward fluxes of colder air during wintertime occurs during the cooling period in the 1960s. Concurrently, the temperature in the ocean surface layers was lower than normal during the warming event and higher than normal during the cooling event. Northward atmospheric winter fluxes are observed after the enhanced global warming after 1970, and, for the first time over the period considered, a positive correlation is observed between atmospheric and oceanic reducing effects on the ice extent. The enhanced global warming over the past two decades seems also to be manifest in an intensified winter circulation at higher latitudes, rather than a contemporary change in the Arctic Ocean surface temperature.
Liu, J., J. A. Curry, and Y. Hu (2004), Recent Arctic Sea Ice Variability: Connections to the Arctic Oscillation and the ENSO, Geophys. Res. Lett., 31, L09211, doi:10.1029/ 2004GL019858. [Link]
Abstract. Trends in the satellite-derived Arctic sea ice concentrations (1978 – 2002) show pronounced decreases in the Barents/Kara Seas, between the Chukchi and Beaufort Seas, the central Sea of Okhotsk and a portion of the Hudson/Baffin Bay by 2 – 8% per decade, exceeding the 95% confidence level. Qualitatively speaking, positive phases of the Arctic Oscillation (AO) and El Nin ̃o- Southern Oscillation (ENSO) produce similar ice changes in the western Arctic, but opposite ice changes in the eastern Arctic. The manner in which the ice changes are related to the AO and ENSO are demonstrated. Over the last 24 years, the magnitude of the ice changes associated with the positive AO trend and the negative ENSO trend is much smaller than the regional ice trends. Thus, more local or less understood large scale processes should be investigated for explanations.
Role of global warming?
There are many indicators that natural variability has a strong influence on the variability of sea ice extent on decadal to millennial timescales. IMO, the strongest argument for sea ice decline over the last decade for being unusual and at least in part attributable to global warming is this (from Polyakov et al.): The severity of present ice loss can be highlighted by the breakup of ice shelves at the northern coast of Ellesmere Island, which have been stable until recently for at least several thousand years based on geological data.
The overall decline of sea ice with global warming has been predicted by climate model simulations since the past several decades. However, the observed decline does not follow in a simple way the increase in CO2 or variability in global or local Arctic surface air temperatures.
If natural variability is dominant, the sea ice extent could increase if the AO stays predominanly negative, the PDO stays cool, and the AMO switches to the cool phase (a scenario that might occur sometime in the next 2-3 decades).
A complex interplay between natural internal variability and CO2 forcing is the most like explanation. Further research is needed particularly on role of natural internal variability in influencing sea ice thickness and extent.