Possible 1,000-kilometer-long river running deep below Greenland’s ice sheet


Research News


Computational models suggest that melting water originating in the deep interior of Greenland could flow the entire length of a subglacial valley and exit at Petermann Fjord, along the northern coast of the island. Updating ice sheet models with this open valley could provide additional insight for future climate change predictions.

Radar surveys have previously mapped Greenland’s bedrock buried beneath two to three thousand meters of ice. Mathematical models were used to fill in the gaps in survey data and infer bedrock depths. The surveys revealed the long valley, but suggested it was segmented, preventing water from flowing freely through it. However, the peaks breaking the valley into segments only show up in areas where the mathematical modelling was used to fill in missing data, so could not be real.

Christopher Chambers and Ralf Greve, scientists at Hokkaido University’s Institute of Low Temperature Science, wanted to explore what might happen if the valley is open and melting increases at an area deep in Greenland’s interior known for melting. Collaborating with researchers at the University of Oslo, they ran numerous simulations to compare water dynamics in northern Greenland with and without valley segmentation.

The results, recently published in The Cryosphere, show a dramatic change in how water melting at the base of the ice sheet would flow, if the valley is indeed open. A distinct subglacial watercourse runs all the way from the melting site to Petermann Fjord, which is located more than 1,000 kilometers away on the northern coast of Greenland. The watercourse only appears when valley segmentation is removed; there are no other major changes to the landscape or water dynamics.

“The results are consistent with a long subglacial river,” Chambers says, “but considerable uncertainty remains. For example, we don’t know how much water, if any, is available to flow along the valley, and if it does indeed exit at Petermann Fjord or is refrozen, or escapes the valley, along the way.”

If water is flowing, the model suggests it could traverse the whole length of the valley because the valley is relatively flat, similar to a riverbed. This suggests no parts of the ice sheet form a physical blockade. The simulations also suggested that there was more water flow towards the fjord with a level valley base set at 500 meters below sea level than when set at 100 meters below. In addition, when melting is increased only in the deep interior at a known region of basal melting, the simulated discharge is increased down the entire length of the valley only when the valley is unblocked. This suggests that a quite finely tuned relationship between the valley form and overlying ice can allow a very long down-valley water pathway to develop.

“Additional radar surveys are needed to confirm the simulations are accurate,” says Greve, who has been developing the model used in the study, called Simulation Code for Polythermal Ice Sheets (SICOPOLIS). “This could introduce a fundamentally different hydrological system for the Greenland ice sheet. The correct simulation of such a long subglacial hydrological system could be important for accurate future ice sheet simulations under a changing climate.”


From EurekAlert!

0 0 vote
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
Steve Case
November 15, 2020 2:19 pm

The correct simulation of such a long subglacial hydrological system could be important for accurate future ice sheet simulations under a changing climate.”

It will be a disaster, CO2 will be implicated, we need more money

stephen duval
Reply to  Steve Case
November 16, 2020 11:21 am

Did anyone think of measuring the amount of water coming out of the Petrmann Fjord?

Reply to  stephen duval
November 16, 2020 1:50 pm

Measure ? The have modelled and simulated everything, who needs data from the location

November 15, 2020 2:22 pm

Ice is nice, but rivers deliver.

November 15, 2020 2:29 pm

“Computational models suggest…” then goes on to say the parameters used in the model are unknown but “what if”. Just an exercise in fantasy.

Rich Davis
November 15, 2020 2:38 pm

What is the relevance of this? If any ice is melting 2km below the glacier surface, that would have to be due to geothermal heat, wouldn’t it? Nothing to do with human activity and nothing new.

Splitdog Homee
Reply to  Rich Davis
November 15, 2020 3:28 pm

Release the Kraken

Reply to  Splitdog Homee
November 15, 2020 4:36 pm

I’ll get Kraken on that task.

Reply to  Splitdog Homee
November 15, 2020 5:13 pm

Yes, please do.

Ron Long
Reply to  Rich Davis
November 15, 2020 5:02 pm

Rich Davis, you are obviously in need of schooling. You don’t have long to wait as Joe Biden has announced he will explain global warming. In a related story I am on the verge of howling mad.

Mickey Reno
Reply to  Ron Long
November 16, 2020 11:53 am

Can’t you just bust in anticipation of the new science wars that are soon coming over the best treatments for dementia and memory loss? I await them with baited breath. I have a feeling Kamala will be in the winning camp.

Rich Davis
Reply to  Mickey Reno
November 17, 2020 2:05 pm

Isn’t there a supplement that supposedly comes from jellyfish? Somebody was commenting about that since obviously what you associate with jellyfish is extreme brain power. I believe that ol’ Dementia Joe must have been sucking down a lot of that product to get his brain in its current state of gel.

P.S. Mickey, I certainly hope you meant bated breath, unless you have been eating a lot of raw fish or something. In which case a good mouthwash would be recommended, or at least a mint.

james fosser
Reply to  Mickey Reno
November 19, 2020 10:10 pm

A fear is that if that happens then Pelosi will be the VP and if something then happens to Harris!

Reply to  Rich Davis
November 16, 2020 6:50 am

Yep – geothermal is the ignored heat source. It can’t be traced to CO2 (though I presume they are working on that) so it generates no grants.

November 15, 2020 2:43 pm

“The simulations also suggested that there was more water flow towards the fjord with a level valley base set at 500 meters below sea level than when set at 100 meters below.”
Seems reasonable. Water flowing uphill only 100 meters vertical to the sea will go much faster than when the water must climb a huge 500 meters.

“the simulated discharge is increased down the entire length of the valley only when the valley is unblocked.”
No big deal, just the difference between a river and a lake with no outflow.
Details, details. Picky, picky, picky.

November 15, 2020 2:52 pm


“Greenland is thought to have slowly moved over a mantle plume, a source of great heat, which left a diagonal scar of warm, dense rock below the surface as the tectonic plate shifted. Greenland moved from a more southern latitude toward the Arctic over 100 million years, a period when the supercontinent Pangaea was breaking up into the drifting continents of today. Eventually, the plume is thought to have formed Iceland above the surface of the ocean through countless volcanic eruptions ­­— a visible trace of the plume’s existence, in contrast to Greenland’s hidden scar.”

November 15, 2020 3:32 pm

This is pure BS. They can confirm it in 30 days next summer. Take a flow meter and salinity meter to the Petermann fjord and measure if any such outflow exists. More modeling will not tell them anything. They need to get their boots cold and go measure it. No Outflow, No River.

Reply to  OweninGA
November 15, 2020 4:02 pm

Much too logical, Owen.

When airplane and satellite surveys are coupled with extensive computer magic to analyze the data and keep them globe trotting and earning salaries for years, why would they jeopardize that with a logical physical test?

Reply to  OweninGA
November 15, 2020 4:15 pm

And use ground penetrating radar in the few critical places to find out what’s down there.

Dr Dave
Reply to  OweninGA
November 15, 2020 4:17 pm

With all due respect, why does Mr Rotter post this kind of “information” on WUWT?

Rich Davis
Reply to  Dr Dave
November 15, 2020 6:32 pm

To ridicule EurekAlert! I imagine.

Pat from kerbob
Reply to  Dr Dave
November 15, 2020 7:24 pm

To expose the ridiculousness of climatology.

If we don’t read it, is it still stupid?

Reply to  Dr Dave
November 16, 2020 4:19 am

why does Mr Rotter post this kind of “information” on WUWT?

What if one creatively used this information in their own proselytizing against the methodology that produces “this kind of ‘information'” Dr. Dave?

Wouldn’t it then be useful?

Reply to  OweninGA
November 15, 2020 4:36 pm

“They need to get their boots cold and go measure it.” I agree, we need empirical data.

“No Outflow, No River.” … Not true…
there are many rivers around the world (some 100s of miles long) that have no outflow, look up ‘Endorheic basins’ – watersheds which do not drain to the sea.

The Hari River, The Zayandeh river, The Carson River, The Jordan River to name but a few.

Mark A Luhman
Reply to  saveenergy
November 15, 2020 5:30 pm

One problem with that is they said said model showed the river flow into the Petermann fjord. Either it does or the model wrong. The model is not evidence, it is only a guess.

Pat from kerbob
Reply to  saveenergy
November 15, 2020 7:25 pm

Jordan flows to the Dead Sea

Reply to  Pat from kerbob
November 16, 2020 2:16 am

“Jordan flows to the Dead Sea”
The Dead Sea is an endorheic lake, there are no outlet streams.
The Sea of Galilee, Jordan & Dead Sea are a closed system &
are what remains of the ‘Sedom Lagoon’ that formed in the Jordan Rift Valley

Reply to  Pat from kerbob
November 17, 2020 1:16 am

Yes, look up Endorheic Basin. That’s what the Dead Sea is.

Coeur de Lion
Reply to  saveenergy
November 15, 2020 11:46 pm

The mighty Oxus

Reply to  saveenergy
November 16, 2020 1:25 pm

An endorheic subglacial river? In an ordinary endorheic basin the water finally vaporizes leaving a salt lake/flat at the lowest point (cf the Great Salt Lake).

This is somewhat difficult to arrange under an ice cap.

Reply to  OweninGA
November 15, 2020 5:15 pm

Get Elon Musk to design an under glacier submarine.

Highway Engineer
November 15, 2020 3:44 pm

As a Civil Engineer involved in using ground penetrating radar to locate pipes and ductbanks only a few meters below top of ground, knowing the inaccuracy, I can only say that using ground radar to interpret rock lines 2000 to 3000 m deep below ice, is a fantasy. This studies have little basis to offer any reasonable estimate and no understanding of the un reliability of such investigation . Similarly, having interpreted rock lines for highway work, I know that it is impossible to reasonably infer rock surfaces. Such studies are nonsense.

Rich Davis
Reply to  Highway Engineer
November 15, 2020 6:24 pm

Perfect! We love measurements that require massive adjustments and can be molded to fit whatever outcome is desired.

john harmsworth
Reply to  Rich Davis
November 16, 2020 10:41 am

Rich Irony!

Reply to  Highway Engineer
November 15, 2020 10:07 pm

Exactly, and why we don’t do ground-penetrating radar surveys for site investigations, just the trusted borehole drilling.

Reply to  Streetcred
November 16, 2020 1:28 pm

Which has actually been done here. Google “NEEM”. It found cold-based ice frozen to the bedrock.

November 15, 2020 3:51 pm

“Collaborating with researchers at the University of Oslo, they ran numerous simulations”

And selected the self satisfying simulations they liked the most…

“Chambers says, “but considerable uncertainty remains. For example, we don’t know how much water, if any”

“considerable uncertainty remains“, that is “we do not know but we are thrilled to publish our desires and dreams as if they are solid findings…
Please send more grants…

November 15, 2020 3:54 pm

“500 meters below sea level”… does water flow uphill in Greenland? Asking for a friend.

Loren C Wilson
Reply to  Dave
November 15, 2020 5:55 pm

Since the ice in the interior of Greenland is about 2000 meters above sea level, water melting at the interface is under a lot of pressure. If the water makes it to the end of the proposed river, it can be forced out against the sea water’s lower pressure of 500 meters of head.

Reply to  Dave
November 16, 2020 8:48 am

… not just Greenland.

Surcharge works everywhere. (not to say that it is happening on such a scale in Greenland).

November 15, 2020 4:27 pm

If it has to be confirmed with radar, why not just do the radar and skip the modeling?

Carl Friis-Hansen
Reply to  Jeff
November 16, 2020 1:12 am

Answered in Highway Engineer’s comment above.

Smart Rock
November 15, 2020 4:28 pm

These people apparently don’t know much about glaciology. Their puzzling over whether there are topographic “barriers” to water flow shows this.

Subglacial rivers don’t just flow downhill. They are under pressure from the overlying ice and when they come to a bedrock ridge, they just go over it. I’ve seen an esker (a ridge of sand, gravel, boulders left behind by a subglacial river) take a 90° bend for 3 kilometres before finding the low spot in a prominent bedrock ridge, taking another 90° bend, and resuming its generally southward course.

Subglacial rivers don’t just stay on the bedrock surface either. They can move up into the interior of the ice sheet, then go down to the base again. You cannot treat a subglacial river like a surface river; its behaviour is more like that of groundwater flow.

Reply to  Smart Rock
November 16, 2020 8:56 am

they’ll take that and run with it.

IF/WHEN sea surface levels rise 2 mm the compensatory force necessary to push the river out into the sea will result in heating a larger subsurface area, causing more melting, causing faster SLR, causing more melting from below, and so on; like the runaway GW we are seeing. AND southern Greenland will be ice free by September of next year (plant your grapes now).

Reply to  Smart Rock
November 16, 2020 1:51 pm

However an esker does not show how the subglacial river ran. An esker forms just outside the mouth of an subglacial river as it debouchs from the ice-front. There the pressue and current drops abruptly and the sediment in the river settles.
So an esker doesn’t show how the river flowed, only how the rivermouth changed position as the ice retreated:

comment image?cb=1414985381

Incidentally if this river exists one would expect that an esker is forming on the bottom of Petermann fjord. Or perhaps rather a subaquatic delta plateau if the river mouth has stayed stationary for a considerable time. Such marginal deltas are well-known from e. g. Scandinavia and are morphologically very distinctive. It should be easy to check if there is one in Petermann fjord.

Tom Abbott
Reply to  Smart Rock
November 17, 2020 6:11 am

Someone raised a question about why this article was posted at WUWT.

I would submit that Smart Rock’s reply just above is why articles like this should be posted at WUWT.

Thanks, Smart Rock, I learned something today. I didn’t know that under-ice rivers could defy gravity sometimes. I would probably never have known that if this article had not been posted.

You find little jewels of knowledge in every WUWT post, and the most knowledge you will find is in the comments.

Walter Sobchak
November 15, 2020 4:34 pm

I stopped with these three words “Computational models suggest”. Models suggest nothing. Models tell you what the assumptions you built into them are.

Mathematical onanism. They have got to stop it, or they will go blind.

Reply to  Walter Sobchak
November 16, 2020 1:16 am

Onanism, onanist, onanistic…I bin hearing that word much lately. Just looked it up. First result does indeed translate it as masturbation, all the other millions of definitions are about coitus interruptus. With biblical history, no less!
Thought I’ll share that.

Reply to  Walter Sobchak
November 16, 2020 6:31 am

I understand AP style book has changed onanism to Toobinism.

November 15, 2020 5:48 pm

“For example, we don’t know how much water, if any, is available to flow along the valley”

By that logic, I have a torrential river flowing through the gully in my paddock, I’m just not really sure how much water, if any, is available to flow through it.

Carl Friis-Hansen
November 16, 2020 1:20 am

We will know if their wild guess was correct in about 25,000 years where all ice is melted away from the interior of Greenland, if the melting continues at the same rate of 0.004% per year.
Ref.: Fettweise et al. 2008

November 16, 2020 6:28 am

Mildly interesting & not at all surprising. Question is, how will the green-loons try to exploit this to further their attempt at global takeover?

Tom Abbott
Reply to  beng135
November 17, 2020 6:18 am

That *is* the question! 🙂

As I read this article, I kept waiting for the Human-caused Climate Change angle to be brought in, but in this case, it never was. But don’t put it past the alarmists to find CO2 involved in this somehow. They see CO2 in everything.

November 16, 2020 8:20 am

The US should purchase Greenland! As in the first Frankenstein movie, it’s alive it’s alive. Good to know what is under all that ice.

November 16, 2020 1:19 pm

This seems a fairly nonsensical theory. Not only is most of this hypothetical river flowing through areas where the ice is cold-based (=frozen to the ground) to judge from the surface topography, but it passes almost exactly under the NEEM drilling site where the ice is most definitely known to be frozen to its bed:


Subglacial rivers are for obvious reasons only possible under warm-based ice, where the bottom temperature is above the pressure melting point. This does not mean that the purported river valley doesn’t exist. This area as far as we know has been continually glaciated for 2.5 MYA, and cold-based ice has virtually no erosive effect. The pre-ice age landscape is probably fairly well preserved, river valleys and all.

November 17, 2020 9:16 am

So the basement rock bathtub under Greenland has been airbrushed away?

%d bloggers like this: