Arctic Ocean may have been covered by an ice shelf nearly double the size of Greenland’s ice sheet

An ice shelf over a kilometre thick once formed over the Arctic Ocean, a new study by researchers from the University of Sheffield has revealed.

The study is the first of its kind and strongly supports the theory that a floating ice shelf might have formed in the Arctic Ocean during some of the Earth’s Ice Ages.

An ice shelf over a kilometre thick once formed over the Arctic Ocean, a new study by researchers from the University of Sheffield has revealed.

The research, led by Dr Ed Gasson, from the University of Sheffield’s Ice and Climate research group published today (17 April 2017) in Nature Communications, used a numerical ice sheet/shelf computer model to show that an ice shelf in the Arctic Basin could only have become a kilometre thick if it covered the entire basin. This model also provided an estimate of the ice shelf’s size – up to 1.7 times the size of Greenland’s current ice sheet.

The study also found that the Arctic ice shelf might have increased the volume of ice sheets on land surrounding the Arctic Ocean. Its presence may also explain discrepancies in sea level reconstructions.

Dr Ed Gasson from the University of Sheffield’s Department of Geography said: “The study is important because it opens up further research into what role this previously missing piece played in Earth’s glacial periods. Although these periods of Earth’s history have been extensively studied already there are many things that are not fully explained. What impact an Arctic ice shelf had on the climate system is a big unanswered question.”

The results are consistent with a recent discovery of ice scours – a narrow ditch on a seabed caused by the movement of pack ice – on the Lomonosov Ridge, which runs across the middle of the Arctic Basin. This indicated that a floating ice shelf might have formed in the Arctic Ocean during some of the Earth’s Ice Ages.

The hypothesis of an Arctic ice shelf was first put forward in the 1970s, but there was limited evidence supporting it. When ice reaches the seafloor it causes erosion. Although ice scour marks have been found on the Arctic seafloor previously, it was argued whether the keels of icebergs or an ice shelf caused the scours. However, a recent research cruise to the Arctic discovered new traces that had the tell-tale signs of a former ice shelf.

The ice shelf most likely formed during the penultimate glacial period around 140,000 years ago. There is no evidence yet that an ice shelf also formed during the Last Glacial Maximum around 20,000 years ago, when much of North America and northern Europe was covered by ice sheets. The study suggests that this could be due to differences in the Earth’s orbit or because the Eurasian ice sheet did not expand as fully towards the Arctic Basin. One challenge with detecting floating ice shelves if that they can leave no traces unless they ground on the seafloor, leaving open the possibility that a thinner ice shelf also formed during the Last Glacial Maximum.

###

Read the full paper in Nature Communications

Advertisements

  Subscribe  
newest oldest most voted
Notify of
rbabcock

.. used a numerical ice sheet/shelf computer model

Ok. I hate to disparage all computer models (ask Boeing or Airbus on new aircraft design), but who wrote the code and how is it verified? I suppose you can calculate energy in/out requirements, etc., but really isn’t this just an exercise in computer science and code writing? It doesn’t really prove anything because at present, it isn’t provable.

Was going to make the same comment. I read the peer review file, very much a modellers convention, no mention anywhere of model validation, a bit of a big issue when attempting to model something new, like a gigantic floating ice shelf.

GregK

There is empirical evidence however……
“The results are consistent with a recent discovery of ice scours – a narrow ditch on a seabed caused by the movement of pack ice – on the Lomonosov Ridge, which runs across the middle of the Arctic Basin. This indicated that a floating ice shelf might have formed in the Arctic Ocean during some of the Earth’s Ice Ages.”
Something made the scours.
When is a bit difficult to work out.
Still, 20,000 years ago New York was covered by 2 to 3km of ice

WXcycles

It’s not about computer models.
It’s about field data, (evil full-scale drill-coring of carefuly designed survey transects) to describe the palaeo-environment of sediment depositions and discontinuities, biota, erosion indicators, dating of all sed units, crustal loading indicators, seismic reflection profiles, which are all used to MAP the actual extent (or not) of what was (or was not) there periodically.
i.e Repeatable and improvable stuff.
Once the field data is gathered and mapped out geographically, and temporally, the existance (or not) of ice sheets, and where and when they occurred, can and would be physically proven, or else, more or less falsified.
Research of deep ocean sediments is slow and horribly expensive ( often multinational). Double the cost and the difficulty for doing that in the polar Arctic and margins, at the scale required, but this is just bread and butter geological basic research.
But you sure don’t want some warped ideological bonehead with no understanding of marine geology, involved in that, to ‘fit’ a pet computer model.
Simple spreadsheet calcs will give you the implied numbers—computer ‘models’ are not required for this.
4D geo-mapping of the data layers is all that you need. Classic geological survey stuff.
Models are spin—Mapping is Science.

Lars P.

Interesting. This would mean
1) sea level was even lower 140 k years ago, maybe another 20 m to the 100 m lower during last glaciation?
2) ocean currents were different in that period – wonder how different?

kokoda - AZEK (Deck Boards) doesn't stand behind its product

Lars….what it means is that it could mean anything.
I’m beginning to think that to be a scientist only requires a PhD in the use of ‘may’ and ‘could’.

Taylor Ponlman

I would think that unless it was significantly grounded over a large area, that the sea level impact would not be that great.
Analog would be Antarctic ice sheets, sea level effects only driven by the grounded portions.

Latitude

I thought sea level was supposed to be a lot lower 140,000 years ago?

tty

Yes, it was a bit lower than during the latest glaciation, but since an ice-shelf floats it does not directly affect sea-level. However the sea-level can’t have been more than c. 140 m lower than now since the Bab el Mandeb was never completely dry during the Pleistocene.

RobR

If the ice shelf was mostly floating the impact on global sea level would have been trivial to minor as the most.

AndyE

Where would all the necessary precipitation come from??

2hotel9

Please smack me around if wrong, would not a mass of ice that large press unfrozen water out from underneath itself by virtue of it’s weight? Parts remain floating and other parts becoming grounded on sea floor? Weight displaces water, right? Some old guy back in the long ago figured that out, Archi-somebody? Help a brother out here!

charles nelson

You are correct, a shallow sea would be displaced by the heavier ice.
But you’ll find that most academics do not have a good grasp of practicalities like that!

WXcycles

Yes, and the 4 to 5 km thick ice sheets, on the arctic fringing land masses would be continually donating giant bergs via gravitational (down hill) flow/creep alone (carving bergs that don’t melt untill an interglacial) into the now frozen closed-off arctic basin, thus forming a trapped core of adjacent welded bergs, which form the basis of a soon grounded ice sheet to develop.
i.e. glaciers did not just advance towards the equator, they would also have advanced north, off of the arctic land masses as bergs, into the central arctic basin. —> proto ice sheet core growth.
My guess is that process would begin to occur early, even as sea level is falling.

2hotel9

A lot of commenters addressing this point, the beu-tay of WUWT! I have often considered the effect on sea level from glaciation on land and have asked a few geologists about it over the years, received varying answers. An oceanographer has explained it is a yes and no situation because so much of ice accumulation during periods of heavy glaciation would occur over ocean/sea areas. Sea levels would go down with ice growth, not as much as the estimated amount of ice, overall, would indicate on paper. Paper, since when he was a working oceanographer computers were still huge contraptions filled with tubes and servomechanisms and whatnot. He would be highly amused by the globall warmining crowd, since at the time of his passing everyone still “thought” the Earth of cooling from human activity, a supposition he did find highly amusing.

rocketscientist

If the weight of the ice becomes greater than the available depth of the sea to support the ice then it becomes grounded. This happens all the time.

2hotel9

Have read about that in articles about Antarctic ice sheets, was being only a bit facetious. That Archi guy was pretty sharp! I use some of his work in building things to this very day, going to be building a spiral staircase next month thanks to him.

J Mac

RE: The results are consistent with a recent discovery of ice scours – a narrow ditch on a seabed caused by the movement of pack ice – on the Lomonosov Ridge, which runs across the middle of the Arctic Basin. This indicated that a floating ice shelf might have formed in the Arctic Ocean during some of the Earth’s Ice Ages.
The ice scours on the Lomonosov Ridge, at depths up to 1280m, are real physical evidence of glacial abrasion/erosion. The numerical simulations sought understanding as to how ‘big’ the glacier would have to be to create such features.

+100

John B

That would also be assuming the sea floor hadn’t fallen since the scours were made.

rocketscientist

It very well may be rebounding.

richard

Might! zzzzzzzzzzzzzz

Bill Illis

Are there ice gouges on the Lomonosov Ridge?
It is an ancient stranded piece of continental origin and has ancient rift action features on it but it seems to be too deep now to have ice gouges in it. (Note that small stranded continents like this eventually sink below sea level since the oceanic crust they are riding on sinks deeper and deeper in the mantle as the ocean gets older over 100s of millions of years. There are many of these small continental pieces that are in the same situation being stranded sunken continents.)

tty

“Are there ice gouges on the Lomonosov Ridge?”
Yes, extensive ice gouges down to about 1000 meter depth. For details refer to:.
M.H. Edwards, B.J. Coakley: SCICEX investigations of the Arctic Ocean System. Chem. Erde, 63 (2003), pp. 281-328

Bill Illis

I don’t think it is physically possible.
The glaciers coming off Ellesmere Island and northern Greenland show clear outflow channels. But these are not 1 km deep.
Furthermore, to build out all the way to Lamonosov Ridge, those glaciers would have to move across a channel that is 2 kms deep.
Now in the ice ages, the sea ice is obviously not melting. Winter snows are going to build up and build up as long as the sea ice below is stable in place. You can get large in-place ocean glaciers building up as a result of that. Indeed, the North Sea, the Barents Sea, the Kara Sea and at different times, the entire East Siberian Sea were glaciated over from this process and the land glaciers. But this is all shallow continental shelf (which before the ice ages 3 million years ago, were probably above sea level but have been pushed down by ice age glacial loads). Lamonosov is a different situation.

Chimp

Bill,
The idea is not land-based ice sheets, as on the then exposed continental shelves, but floating sea ice thick enough to reach down to the ridge.

Eric Stevens

But in which ice age were the ice gouges formed?

Chimp

From Tty’s link:
https://nsidc.org/sites/nsidc.org/files/files/scicex/Edwards_Coakley_2003.pdf
Another startling discovery from the SCICEX-99 SCAMP data was evidence of
grounded ice to depths of almost 1000 m on the central portion of Lomonosov Ridge.
Although erosional truncation has been previously reported for Lomonosov Ridge
(Jakobsson 1999), the SCAMP chirp sonar records, showing erosionally planed ridge
tops with rough micro-relief that includes parallel lineations and sub-parallel gouges
(Figure 4) represent the first compelling evidence for widespread scouring and molding
of seafloor by grounded ice at these depths in the central Arctic Ocean (Polyak et al.
2001). Based on the SCICEX findings, Polyak et al. (2001) suggest that a vast ice shelf
advanced from the Barents Sea shelf and eroded parts of the top of Lomonosov Ridge to
depths of almost 1 km.

WXcycles

Chimp on April 17, 2018 at 1:09 pm
Bill,
The idea is not land-based ice sheets, as on the then exposed continental shelves, but floating sea ice thick enough to reach down to the ridge.
—-
He knows that, he’s making further points about the process of getting it pushed there, under gravitational flow presure, from high glaciers on surrounding land, pushing it out compressionally, into the arctic basin as non melting accreting welded bergs, and the geodynamic basin margin response to that. He’s right.

WXcycles

‘Continents’ form from less dense rock cores of defined primary global ‘plates’. The submerged continental terrain blocks are just sunken continental crust ‘crumble’, off of the edges of distended continental crustal margins. i.e. they do not define the core of a primary Earth crustal plate. Sunker blocks of dessicated continental crust, yes, sunken continents, no. Even Greenland isn’t a continent, just a dessicated island ‘crumble’ within the primary North American plate. There are second order plates, and micro plates, but none define a continent at their core. But many contain partially altered (metamorphosed and made denser, thus submerged) continental bits and peices.

Gary

It would be nice to see an analysis of Arctic ocean sediment cores that are sufficiently long and widely distributed throughout the basin.

tty

Unfortunately deposition is very slow and current- and ice-scour has often removed what deposits there are, so it is very difficult to find “sufficiently long and widely distributed” core-sites.

WXcycles

Which is why you design a transect survey.
And keep in mind, sediment discontinuities are VITAL information in palaeo studies. In fact, if you’re looking for ice sheet grounding, that’s exactly the data you’re looking for, large area discontinuities, and the seds, and their dates and derivations (marine or terrigenous) either side of the discontinuities.
That IS the data that you want—discontinuous.

beng135

The proposed Arctic ocean ice-shelf would be dang massive, so I’d think they could get a good idea by knowing the sea-level and the best-estimates for the land-glacier masses at the proposed time.

tty

An ice-shelf floats and so does not affect sea level.

charles nelson

An ice shelf would only ‘float’ on a body of water of equal or greater density.
If the ice mass was was large it would be heavy enough to displace the ocean.

tty

“If the ice mass was was large it would be heavy enough to displace the ocean.”
That is the definition of “floating”.

charles nelson

…displace the ocean completely.

beng135

I assumed the ice-shelf they are talking about is like Antarctica — an extension of land-glaciers & parts of it pinned on top of islands or on the bottom of very shallow water & thus not entirely floating. But whatever…..

OK, what they are saying is that they have modeled an ice sheet > 1 km thick at 140 kya.
That doesn’t mean there weren’t *thinner* ice sheets during other glacial periods.
From the paper: “It is possible that large Arctic ice shelves also formed during other glacial stages, although interestingly no erosional features deeper than ~600m water depth have yet been dated to the Last Glacial Maximum (LGM)”
600 m of ice still excludes life for any marine mammal in the Arctic Basin during the LGM, including polar bears. http://polarbearscience.com/2015/04/21/polar-bears-barely-survived-the-sea-ice-habitat-changes-of-the-last-ice-age-evidence-suggests/
All of the literature I’ve read indicates the MIS 2 glaciation (160-140 kya) was colder than the LGM, which would explain thicker ice over the Arctic Basin. Part of the problem with knowing for sure what went on during MIS glaciation is that in many regions, ice from the LGM removed evidence left by previous ice ages.
There would have been a huge “Beringian” plain of dry land during MIS 2 because of the lowered sea level, as there was during the LGM. The question seems to be whether there was an ice sheet on top of that land or not. During the LGM, there was not.

tty

You mean the MIS 6 glaciation. MIS 2 is = LGM.

Yes, sorry. Right you are

WXcycles

” … is that in many regions, ice from the LGM removed evidence left by previous ice ages. ”
—–
ayes, and those (very large areas) regions are exactly the places geologists will need to core with a drill ship, and map thosecdiscontinuities in detail, to constrain timing and extent of each and every event, as they are found identified, characterised and dated.

kalsel3294

What effect would the weight of such an accumulation of ice have on the rotational balance of our “spinning top”?

tty

None, since shelf-ice is floating and displaces an equal mass of water it does not affect the Earth’s rotation. Well almost none. The centre of mass would be very slightly further from the Earth’s centre compared to an ice-free ocean.

J Mac

If the ice was mostly displacing equivalent liquid water, the mass effects on planetary rotation would be negligible. The article discusses a purported approximately ‘kilometer thick floating ice shelf’ that left parallel scour marks on the Lomonosov Ridge at a depth of 1280m below the current sea level. ‘Floating’ is the key word…
In addition, Arctic sea ice is roughly distributed around the north pole of Earth’s spin axis. A distributed mass around the spin axis has little perturbing effect on ‘rotational balance’.

tty

This is by no means a new idea. Here is a rather more evidence (not model) based review of the concept:
https://www.sciencedirect.com/science/article/pii/S0277379113002989

Chimp

Good survey and discussion. Thanks.
The modeling effort reported in this post responds to some of the issues raised by your linked paper. If warm Atlantic water were indeed kept out of the Arctic, and the wind patterns changed, then the whole ocean most likely would ice over. The formation of a kilometer-thick floating shelf during MIS-6, capable of scouring Lomonosov Ridge, seems plausible.
More data needed.

hunter

An ice shelf is fixed on lamd floating into the ocean.
Ice floats, barely.
So 1 kilomter thick presents some very strange requirements.
Are there Antarctice ice shelves sticking hundreds of kilomoters into southern ocean?
No. and the shelves don’t rise amywhere close to a kilometer high.
This a strange report.

tty

“Are there Antarctice ice shelves sticking hundreds of kilomoters into southern ocean?”
Yes, several.
“the shelves don’t rise amywhere close to a kilometer high.”
True, typically they are 200-300 meters thick, of which 10% is above water. Normally an ice-wall is not dynamically stable if it is >50 meters high which limits the thicknes near the ice-front to c. 500 meters. Ice scour at greater depth is therefore normally ascribed to more irregularly shaped ice-bergs which can have deeper “keels”. But there is nothing to prevent an ice-shelf away from the calving front from being thicker than 500 meters. The innermost parts of the Filchner and Amery ice-shelves are about 1,000 meter thick.

WXcycles

It does not even need to ‘carve’ off though, as the entire basin will fill quickly with ice from the surrounding glaciers 3 to 5 km higher. In other words, even when the arctic is full of bergs, packed in tight, the compressional energy (potential energy) of the elevated downhill flowing land glaciers will keep on compressionally stacking the ice flow deeper in the badib, until it grounds.
Where’s the space to carve? Or even the need to carve? No, it would just be a giant compressional ice bowl that keeps getting thicker. No carving.
So once it grounds the strong gravity-driven compressional-flow can and wll make it’s elevation higher than you’d think within your Antarctic grounded or open water example.
In other words, the hight of the arctic basin’s ice sheet will try to equilibrate with the elevation of the higher feeder glaciers. Only then will the inflow of ice pressure slow then halt.

tty

“as the entire basin will fill quickly with ice from the surrounding glaciers 3 to 5 km higher”
No it won’t since the ice is lighter than the water. The only way a ice-shelf from a glacier can fill a deep ocean basin is if there is a threshold on which the ice-shelf can get stuck and cause “back-filling” of the basin behind the treshold.

WXcycles

You’re not getting the implication TTY, the source ice sheets are well above the elevation of the basin, thus it keeps ‘falling’ into it, under gravity-flow, it does not stop flowing until the continental located icesheets melt, in 100 ky time.
So you get ~100 K years of ice sliding off of the northern land masses, down into the much lower elevation arctic basin, where it is trapped and continually compressed and deformed.
Sorry, you are quite wrong if you think that won’t fill the entire basin with ice with 100 k years of poleward flows of kilometers thick glaciers, off the vast surrounding land masses.
The basin will fill and choak with welded floating large bergs, early on, with glacial conditions’ onset, then this will compact and get progressively deeper (from the margins) as the glacial proceeds, due to the increasing weight and thickness of ice inflow, from much higher elevations on land, pushint thicker and thicker ice flows, on a broard front, towards the polar area, proper.
That will happen. It’s illogical to suppose it won’t.

Chimp

WX,
IMO you don’t need calving land ice to get the Arctic Ocean covered with thick sea ice, although I guess it couldn’t hurt. What does work however is cutting off or drastically reducing the flow of warmer water from the Atlantic into the basin, then surrounding it with high albedo land ice sheets and altering wind patterns such that it gets really, really cold in the Arctic.
During the LGM and even more so in the previous glacial maximum, the Arctic Ocean was cut off from the Pacific and to a large extent the Atlantic’ warm Northern Drift. In this climate regime (the coldest of the three regimes of the Pleistocene), the Arctic Ocean freezes over permanently and the North Atlantic acts as the Arctic Ocean does today, with a fringe of permanent sea ice around a seasonal melting center. If this situation goes on long enough, then, yes, the central Arctic Ocean sea ice can become 1000 meters thick or more and its keels can scrape the Lomonosov Ridge.

Bruce Cobb

In keeping with Warmist ideology then, all that extra ice was a good thing, helping to keep the planet nice and cool. No one likes a hot planet. We could be like Venus by now.

DonK31

I would be amazed if this didn’t happen. The Laurentide Ice Sheet during the most recent glaciation was up to 1.5 kilometers high as far South as New York. I can’t see that there would be summer ice melt in the Arctic. There would have been multi century ice that would have continued accumulating to the point that if the ice accumulation was 100 meters above the water line, there would have been a kilometer of ice hanging below the water line. While 300 ft at the face of the glacier would promote calving, anywhere away from the face would not allow calving. Eventually the ice sheet would ground just as it does in the Antarctic now.

There was a mile of ice on Detroit, and this is surprising?

NZ Willy

Ice scours on the Arctic Sea seabed indicate not a large floating ice mass but that the entire Arctic Sea was frozen solid with concommitant glacial flow. I’ve previously written that the Younger Dryas cooling period (at the start of the modern warm era) could have been caused by the the re-opening of a water channel beneath the frozen Arctic and the 9000 years needed to melt the whole thing out, spilling Arctic waters into the global oceans in the meantime.