A Geological Perspective on Arctic Sea Ice Extent (AKA PIP25: “Miracle on Ice”)

Guest review article by David Middleton

We’re often bombarded with headlines line this:

Arctic sea ice continues its downward spiral

At 4.6 million square kilometres in coverage, this year’s sea ice minimum is the sixth lowest on record.

The problem with headlines like this is the fact that the “record” only dates back to 1979. How meaningful could “the sixth lowest on record” be, when the record isn’t even forty years long?

NSIDC does feature an estimate of Arctic sea ice extent that goes back to 1953 based on the “Hadley data set”…

Arctic sea ice before satellites, NSIDC

Based on this analysis, Arctic sea ice extent was anomalously high from 1953 to 1978 and has shrunk to anomalously low since 1979.  This is not very meaningful from a geological perspective.  In a previous post, I cited McKay et al., 2008 as an example of a geological perspective on Arctic sea ice extent.

The Arctic was probably ice-free during summer for most of the Holocene up until about 1,000 years ago.  McKay et al., 2008 demonstrated that the modern Arctic sea ice cover is anomalously high and the Arctic summer sea surface temperature is anomalously low relative to the rest of the Holocene.


“Modern sea-ice cover in the study area, expressed here as the number of months/year with >50% coverage, averages 10.6 ±1.2 months/year… Present day SST and SSS in August are 1.1 ± 2.4 8C and 28.5 ±1.3, respectively… In the Holocene record of core HLY0501-05, sea-ice cover has ranged between 5.5 and 9 months/year, summer SSS has varied between 22 and 30, and summer SST has ranged from 3 to 7.5 8C (Fig. 7). (McKay et al., 2008)

Over most of the Holocene, >50% sea ice coverage occurred from 5.5 to 9 months each year.  During the “Anthropocene”, >50% sea ice coverage has ranged from 9 to 12 months each year.


Yes… I know there are only 12 months in a year.

Steve Mosher noted in the comments that, “The Chukchi Sea IS NOT THE ARCTIC !!!! It is part of the arctic.”

This is very true… However, sediment cores are pretty-well limited to where they have been drilled.   In another recent thread, tty brought Stein et al., 2017 to my attention.  After reading through it, I decided that there was enough material there to take a crack at a general characterization of Holocene sea ice conditions relative to modern times.

PIP25: “Miracle on Ice”

In a pioneering study by Belt et al. (2007), the ability to (semi-)quantitatively reconstruct paleo-sea ice distributions has been significantly improved by a biomarker approach based on determination of a highly branched isoprenoid (HBI) with 25 carbons (C25 HBI monoene = IP25). This biomarker is only biosynthesized by specific diatoms living within the Arctic sea ice (Brown et al., 2014) and appears to be a specific, sensitive and stable proxy for Arctic sea ice in sedimentary sections representing Late Miocene to Recent times (Stein et al., 2012, 2016; Belt and M€uller, 2013; Stein and Fahl, 2013; Knies et al., 2014). The presence of IP25 in the studied sediments is direct evidence for the presence of sea ice.


For more semi-quantitative estimates of present and past sea ice coverage, M€uller et al. (2011) combined the sea-ice proxy IP25 and phytoplankton biomarkers in a phytoplankton- IP25 index, the so-called ‘PIP25 index’:

PIP25 = [IP25]/([IP25] + ([phytoplankton marker] x c))

with c is the mean IP25 concentration/mean phytoplankton biomarker concentration for a specific data set or core.


Stein et al., 2017

The image below is of a Late Quaternary sea ice reconstruction for the western Yermak Plateau from Belt, et al., 2016

Pages from Belt_2013-2

Figure 6 from Belt et al., 2016. P8IP25 values for core PS2837-5 (western Yermak Plateau; after Stein et al., 2012) calculated from IP25 and brassicasterol data. Light to dark blue shadings indicate different degrees of sea ice cover (right). Dashed lines depict permanently ice covered intervals causing a lack of IP25 and brassicasterol, which hinders a calculation of the PBIP25 index. Red ellipse refers to the Intra-Allerød or Younger Dryas cooling, which is not properly displayed by the PBIP25 index. (Adapted from Stein et al., 2012).

Generally speaking, the PIP25 index correlates to sea ice extent as follows:

  • >0.7 = Perennial (year-round) ice cover
  • 0.5-0.7 = Seasonal ice cover
  • 0.1-0.3 = Reduced ice cover
  • <0.1 = Ice-free year-round

Most of the Arctic has had seasonal sea ice cover over the past 30 years, with a relatively small area of perennial sea ice.

Perennial sea ice is roughly equivalent to the recent minima.  Seasonal sea ice is more or less the difference between the maxima and minima.

So… How does this compare to the rest of the Holocene?

A geological perspective on Arctic sea ice extent

Kinnard et al., 2008 estimated Arctic sea ice extent minima and maxima back to about 1880 from a variety of methods.  A comparison of this with a PIP25 reconstruction of the North Icelandic Shelf sea ice extent suggests that modern sea ice extent is somewhere between that of the Little Ice Age (LIA) and Medieval Warm Period (MWP), probably closer to the MWP.


Figure 2a from Polyak et al., 2010 (originally from Kinnard et al., 2008) “Maximal (winter) and minimal (summer) Arctics ea-ice extent time series,1870–2003 (from Kinnard et al.,2008). Smooth lines are robust spline functions that highlight low-frequency changes.” Figure 4 from Belt et al., 2013: “Relative abundances of IP25 found in the core MD99-2275 for the period 800-1950 AD plotted against historical records of Icelandic sea ice interpreted from Ogilvie (1992) and Ogilvie and Jónsson (2001) (bottom scales) and diatom-based reconstructed sea surface temperature (Jiang et al., 2005) (Reproduced from Massé et al., 2008).”

Stein et al., 2017 compared sea ice extent reconstructions using PIP25 indices from four cores around the current area of perennial sea ice.


Figure 2 from Stein et al., 2017. Core locations relative to 2015 Arctic sea ice maximum and minimum extent.

A comparison of the Fram Strait core to the North Icelandic Shelf indicates that the sea ice extent of the MWP was anomalously higher than the vast majority of the Holocene.  Prior to the onset of Neoglaciation, Arctic sea ice extent appears to have ranged from nearly ice-free to the low end of seasonal coverage.


Figure 4 from Belt et al., 2013 and figure 10 from Stein et al., 2017 (cropped).

Andy May’s Arctic climate reconstruction exhibits the same general pattern as the four PIP25 sea ice extent reconstructions:


Figures 10 and 6 from Stein et al., 2017 and Andy May’s Arctic climate reconstruction.

Prior to the onset of Neoglaciation, Arctic sea ice ranged from nearly ice-free to reduced.  Since the onset of Neoglaciation, it has ranged from seasonal to perennial.

I plotted Andy May’s excellent Arctic climate reconstruction as an overlay of the Chukchi Sea core from Stein et al, 2017 to demonstrate the relatively decent correlation between Arctic temperatures and sea ice conditions.  Note that all four core locations exhibit perennial sea ice during the Late LIA… And only during the Late LIA.


Figure 10 from Stein et al, 2017 (orange stippled curve indicates insolation) and Andy May’s Arctic climate reconstruction.

While it is clear that Arctic sea ice extent has diminished since 1979, it’s also important to note that this reduction must be viewed in the context of the overall Holocene.  Holocene Arctic sea ice extent reached its Holocene maximum between 1600 and 1850 AD.

Figure 4 from Belt et al., 2013

So any reductions are from the maximum values of this geologic epoch.  The LIA was clearly the coldest phase of the Holocene and modern sea ice conditions are much closer to the LIA than they are to the pre-Neoglaciation period.


Alley, R.B. 2000.  The Younger Dryas cold interval as viewed from central Greenland.
Quaternary Science Reviews 19:213-226.

Belt S.T., Müller J.  The Arctic sea ice biomarker IP25: A review of current understanding, recommendations for future research and applications in palaeo sea ice reconstructions. (2013)  Quaternary Science Reviews,  79 , pp. 9-25.


Hoff U, Rasmussen TL, Stein R, Ezat MM, Fahl K. Sea ice and millennial-scale climate variability in the Nordic seas 90 kyr ago to present. Nature Communications. 2016;7:12247. doi:10.1038/ncomms12247.


Kinnard, C., Zdanowicz,C.M., Koerner,R .,Fisher,D.A., 2008. A changing Arctic seasonal ice zone–observations from 1870–2003 and possible oceanographic consequences. 35, L02507.


May, Andy. A Holocene Temperature Reconstruction Part 4: The global reconstruction. Watts Up With That. 2017.


Polyak, Leonid, 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.  History of sea ice in the Arctic, Quaternary Science Reviews, Volume 29, Issues 15–16, 2010, Pages 1757-1778, ISSN 0277-3791, https://doi.org/10.1016/j.quascirev.2010.02.010.


Stein, R. , Fahl, K. , Schade, I. , Manerung, A. , Wassmuth, S. , Niessen, F. and Nam, S. (2017), Holocene variability in sea ice cover, primary production, and Pacific‐Water inflow and climate change in the Chukchi and East Siberian Seas (Arctic Ocean). J. Quaternary Sci., 32: 362-379. doi:10.1002/jqs.2929



34 thoughts on “A Geological Perspective on Arctic Sea Ice Extent (AKA PIP25: “Miracle on Ice”)

  1. Not just a geologic but an historical perspective also reveals that nothing the least bit out of the ordinary is happening with Arctic sea ice extent.

    Maps of 20th century Arctic sea ice extent show that between the wars and during WWII, the Arctic was as navigable as it is now. The Northern Sea Route along the coast of Siberia was used by both sides in the war. Then the ice cycle swung up again, and the USSR needed nuclear-powered ice breakers to keep the route open in summers during the Cold War, until the PDO switch of 1977.

    Since its record low in satellite observations in 2012, no new record has been recorded. From 1979 to 2012, a new low struck at least every five years. Now six years have passed without a lower low record. The trend in Arctic sea ice summer minimum extent since 2012 has thus been up, and flat since 2007.

    In any case, Antarctic summer minimum grew from 1979 to 2014, while the Arctic was falling. Hence, well-mixed CO2 can’t have caused the NH sea ice decline during the same period.

  2. Same goes for the paleontolgical evidence. From 1990, but conditions in the High Arctic are still more severe there now than during the early Holocene:

    An Early Holocene Bowhead Whale (Buluena mysticetus) in Nansen Sound, Canadian Arctic Archipelago


    At about 10,O00 B.P. northern Axel Heiberg and Ellesmere islands underwent a climatic amelioration that caused the demise of the last glaciation. Generally, by8000B.P. accelerated retreat left extensive coastal areas ice free. The occurrence of an early Holocene (7475 +/-220 B.P.) bowhead whale (Balaena mysticetus L.) skeleton several hundred kilometres north of its present range concurs with other biological and glaciological evidence to indicate that the early Holocene climate in the High Arctic was less severe than at present.

  3. How meaningful could “the sixth lowest on record”…..

    ….could be interpreted as the ice is increasing….it was lower in the past 5 times

  4. Is it possible to compile a list of active ice breakers deployed in the Arctic and the number of hours each is used annually? What would a chart of those total hours look like juxtaposed against the chart of Arctic Sea Ice (contrived) extent?

    Are we’re actively breaking up Arctic Sea Ice while also allowing wind and water currents to impact Sea Ice extent values?

    • I have often wondered about this. However, I think that when ice breakers are being used (Summer), the ice isn’t a solid thick sheet. If I am correct, Ice Breakers are more “clearing a path through the chunks of floating ice”, than actually breaking the ice.

  5. David, here’s more historical-geological evidence re. the ice-free status of far-northern Greenland (Hans Tausen ice cap) pre the ‘record’.

    (1)Textures, fabrics and meltlayer stratigraphy in the Hans Tausen ice core, North Greenland – indications of late Holocene ice cap generation? in: (U.C. Hammer, ed.) The Hans Tausen Ice Cap Glaciology and Glacial Geology , Meddelelser om Grønland, Geoscience, 39 , pp. 97-114 .

    A thin section study of crystal structure has been carried out on a 345 m long ice core drilled to bedrock on Hans Tausen Iskappe, 1995. In addition, a meltlayer stratigraphy was set up showing how the fraction of meltlayer-ice in the core increases with depth. Main characteristics of crystal structure are increasing mean crystal size from top to bottom in the core and development of a weak single maximum c-axis fabric. The rate of ice crystal growth in the well dated upper half of the core is much lower than expected from studies of the normal grain growth regime in other polar ice cores. Probably the grain boundary movements are impeded by impurities, which are present in relatively high concentrations in the Hans Tausen ice. Assuming the applicability of the calculated growth rate throughout the core, a late Holocene origin of the oldest ice is suggested by the size of the crystals close to the bedrock. Presented data furthermore implies that bottom ice temperatures were never near the melting point and it is concluded that there was no ice cap on the Hans Tausen plateau earlier in Holocene.

    (2)On the Holocene evolution of Hans Tausen Iskappe (Greenland)
    Zekollari, Harry; Huybrechts, Philippe
    EGU General Assembly 2015, held 12-17 April, 2015 in Vienna, Austria. id.1540

    Hans Tausen Iskappe (Greenland), situated at 82.5°N, 27.5°W, is world’s northernmost ice cap. During several field campaigns in the 70s and 90s, its ice thickness was measured, mass balance and meteorological measurements occurred, and a 345 m deep ice core was drilled. From this ice core it is known that the ice cap (largely) disappeared during the Holocene Thermal Maximum. The present-day ice cap started building up some 3500-4000 years ago in a wetter and warmer climate than at present.

    (3) In the seventies two firn cores were drilled a bit south of the Central Dome. One of the main discoveries was that the ice cap did not survive the higher temperatures of the Climatic Optimum 6000-8000 years ago. Different studies of the ice core show that the age estimate of the ice cap lies in the range 3500-4000 years Madsen and Thorsteinsson [2001]; Clausen et al. [2001]; Hammer et al. [2001].

    The Hans Tausen Iskappe is not the only ice cap not to survive the warming of the climate following the last ice age and to regrow in the second half of the Holocene. Koerner and Fisher [2002] found from climate proxy data from ice cores that this is the case for many smaller ice caps throughout the Arctic.

    The present Hans Tausen Iskappe is only 3500-4000 years old as indicated by different studies of the ice core drilled to bedrock at the central dome. With the onset of the Holocene (our present warm period) around 10000 years ago [Lowe and Walker, 1997] temperatures started to rise and deglaciation began culminating during the Climatic Optimum 6000-8000 years ago. The Hans
    Tausen Plateau became ice free prior to 8100 BP while the deglaciation in the outer parts of the Independence Fjord (See Fig A.3 for location) occurred some thousand years before. During the deglaciation even the highest located parts of the ice cap were below the equilibrium line and experienced surface melt. The decay occurred by a lowering of the surface and not by horizontal
    recession. The warmer climate meant that the fjords were seasonally ice free from the period 6700 − 6300 years BP documented by finds of driftwood at Kap Bopa and the eastern end of Nordpasset. In Jørgen Brønlund Fjord finds extend this period to 6600 − 2550 years BP.

    During the last ice age Hans Tausen Iskappe is thought to have been confluent with the Greenland Ice Sheet. Studies in Nordpasset show evidence that the northern part of Hans Tausen Iskappe was part of an ice sheet that covered the adjacent area during the same period. [Landvik et al., 2001; Bennike, 1987] At present the ice cap lies ~20 km north of the Greenland Ice Sheet. Studies of sedimentation rates from a lake near Jørgen Brønlund Fjord show a richer vegetation than presently in the period 5000-3300 BP probably caused by an increase in precipitation due to the seasonally open fjords [Landvik et al., 2001]. This interval fits well with the age of the ice cap at the drill site, 3500 − 4000 years assessed by studies of stratigraphy, ice crystals and annual layer thickness by Madsen and Thorsteinsson [2001] and Clausen et al.[2001]. The build-up of the ice cap probably began as a response to the increased precipitation rate. The sedimentation rate dropped at 3500 years BP and 2100 years BP marking the transition to colder summers. During the last century up until 1978 the glacier margins seem to be stable or show a slight retreat of 10-100 m from aerial photographs [Weidick, 2001]. Recession is generally related to north and western facing glaciers. The oldest moraines in the area are ~100 years old, and therefore the culmination of re-advance of local glaciers following the Climatic Optimum is believed to be at the end of the Little Ice Age (1900 AD) Weidick [2001].

      • Quote above re: Hans Tausen Iskappe (Greenland): “The present-day ice cap started building up some 3500-4000 years ago in a wetter and warmer climate than at present.”

        The “Neoglacial” in some schemes distinguishes a very cold period at 35000-4000 which the above evidence supports.

        I have published evidence that the period 4000-2500 or thereabouts was very cold and windy in the Bering Sea, with extensive seasonal ice pushed further south in spring than modern records (Crockford and Frederick 2007, in The Holocene).

        What’s interesting is that the LIA did not have the same impact, suggesting the LIA cold was not as cold in the western Arctic as the Neoglacial cold (although it may have been colder elsewhere).

  6. In 1967 the Russians sent three cargo ships through the NSR to Japan in 1967 to illustrate it was open for international shipping- this offer was shelved for political reasons- the Suez crisis. The offer was made formal in the 1980s.

    “According to the Intergovernmental Panel On Climate Change (1990 IPCC report) – NOAA has sea ice satellite data going back to the early 1970s. You can see why the New York Times cherry picked the 1979 start date, because it was the high point. By starting at the top of the mountain, all directions go down. They lied to their readers about the 1979 start of satellite records, and it is clear why they did that”

    “This 1985 DOE climate change report had Arctic sea ice data back to 1925, which showed little ice from the 1930s to the 1950s”

    More at –


        • Wait! You’re both right!

          On 5 June 1967, at the beginning of the Six Day War, Egypt closed the Suez Canal. The closure was sudden and unexpected. Fifteen cargo ships known as “The Yellow Fleet”‘ were trapped inside during the closure. At the end of the war, the Egyptian and Israeli armies were stationed on either side of the canal and the prospects for reopening were very uncertain. The canal remained closed until the end of a second conflict, the 1973 Yom Kippur War, and subsequent peace negotiations, eight years later.

          So the Canal was closed from 1967 to 1975.

  7. There is a kind of proxy which possibly might add to the resolution of sea ice extent back in time past 1953. That is the log books of seal catchers. The seals give birth to their pups on “young ice”. At heavy ice conditions the more pups were born further south. No need to elaborate more on that. The info has not been compiled. The records of Canadian, Norwegian and Russian sealers should be given a try for compilation and analyses. The books may be hard to get hold of and stored at random. The notes were well kept family secrets of vital competitive value. Think they may add to the analyses of both ice extent and the interplay of PDO, AMO and NAO as they cover a time period where other data sets hold useful information. Not sure how much annual variations in sea currents may add to noise (variations) though. Hope this don’t spur a crusade by the greens to burn these for them awful notes in many a way.

    • Speaking of proxies, this biomarker in the article looks very robust and distinctive.
      It’s unusually good, at first glance, for a biological marker.

      Of course, the problem with biological marker s is that they look good in relation to one parameter but then something else comes along to kill them (like a new invasive organism) and the proxy collapses.

      But that’s speculation. It looks good.

  8. On WUWT several years ago there was a Danish study that examined the well developed beaches on the presently ice-locked north [coast] of Greenland. Driftwood on the [beaches] was dated at 6-8000ybp and it was stated that formation of the beaches would have needed a considerable [expanse] of open sea to have been created.

  9. During the whole Cold War the Arctic was under surveillance by the NATO forces… plus the WW2 operations, with convoys bringing supplies to the USSR and the Luftwaffe attacking them, the Allies hunting the Nazi battleship Tirpitz, British Commandos raids on the coal mines occupied by the Nazis and so on…
    Also some clandestine Nazi meteo bases in Greenland and in the far northern Norway (Spitzbergen); the last Nazi crew surrendered in september, 1945….
    There should be a lot of data, in military archives….

  10. Another thing i read recently

    It’s in the Definition
    It’s all in the definition of “Arctic.” The NSIDC definition is essentially “the northern hemisphere,” while the IARC definition is much more restrictive. Therefore, for example, the NSIDC includes in its definition of the Arctic such areas as the Bay of Fundy and the St. Lawrence River estuary, areas that are as far south as latitude 43 N, and other areas well south of the Polar Circle along the North American and Asian continents.


  11. There is a typo in the definition of PIP25, it should read:
    PIP 25 = [IP25] ∕ ([IP25] + ([phytoplankton marker] × c))
    Although IP25 is mentioned as a HBI (highly branched isoprenoid), the molecule has only one side-chain (as an isoprenoid). As the molecule has only one double bond, I would call it a highly saturated sesterterpene.
    The steroids can be considered as HBIs (of which the linear triterpene squalene is the precursor).

    Ice-free conditions in the Artic Ocean are not always related to high temperature. As Stein et al. mentions in the article ‘Arctic Ocean sea ice cover during the penultimate glacial and the last interglacial’:

    During late Marine Isotope Stage 6, polynya-type conditions occurred off the major ice sheets along the northern Barents and East Siberian continental margins, contradicting a giant Marine Isotope Stage 6 ice shelf that covered the entire Arctic Ocean.

  12. “Steve Mosher noted in the comments that, “The Chukchi Sea IS NOT THE ARCTIC !!!! It is part of the arctic.”

    This is very true… However, sediment cores are pretty-well limited to where they have been drilled.”

    well duh.

    My point was that you knew this or should have known this and you implied that the study covered the whole arctic.

    you cherry picked. found something you liked and ran with it.

    Let me ask you simple question:

    1. do you think ice is declining since 1979
    2. if so, list all the possible causes

    • Mosh,

      Your presumption is wrong… But I should have better clarified my citation of McKay, as I have done every previous time I cited it.

      1. It has declined since 1979… That’s bleeding obvious.
      2. Besides GHG-forcing, the same processes that drove previous Holocene decadal-centennial scale fluctuations.

      Furthermore, thank you for making me improve my “game”… Not sure if you or Nick Stokes has been more helpful in “upping my game.”

  13. “Stein et al., 2017 compared sea ice extent reconstructions using PIP25 indices from four cores around the current area of perennial sea ice.”

    almost too funny

    Imagine if we measured sea ice today by only looking at photos of those 4 locations and
    argued that sea ice was declining because of measures at 4 locations.

    You’d all laugh. and yell shit like, “you cant estimate the extent from 4 locations.”

Leave a Reply

Your email address will not be published. Required fields are marked *