Sunday 19 July 2015

Hello Dipole My Old Friend

I have previously written about the Arctic Dipole, and note that it now has a Wikipedia page. Over the last two years it has been largely absent, leading myself and others to ponder whether it was a transitory phase over the 2007 to 2012 period.

It doesn't seem to have been transitory, it has dominated July so far and looks set to continue throughout the rest of July. My expectation is for it to continue into August, that however is just an expectation based on behaviour in 2007 to 2012.

I have previously defined some sectors of interest in watching sea ice, they are shown in the following graphic, based on Cryosphere Today.

The following plot shows anomaly of extent for various sectors of the Arctic region.

Since the start of July the Siberian sector (red) has finally started to fall at above average rates, leading to an above average loss of extent in Northern Hemisphere extent, as shown in the anomaly plot for the Northern Hemisphere.

So the race to the minimum is finally on. But what has caused this?

In late June sea level pressure (SLP) showed a high over Greenland and the Arctic Ocean dominated by low pressure around the landward periphery.

I commented in my June Status post that this was looking similar to the 2007 to 2012 SLP pattern which caused an Arctic Dipole in those years.

Then in July the pattern became far more typical of the 2007 to 2012 summer average.

The interaction between the low across the Eurasian coast and the high from Greenland and over the Arctic Ocean causes the winds associated with the Arctic Dipole. Low pressures circulate anti-clockwise, high pressure circulates clockwise. In the following graphic I have illustrated the effect on the SLP pattern for 2012. The thick black arrow being the Arctic Dipole driven flow of winds.

Returning to July so far, here is the vector wind for 1 to 16 July 2015, note that this is rotated such that the Eurasian coast is on the left hand side.

The dipole flow in July is clear as a strong (reds/yellows) band of winds blowing from the Bering Strait towards the Barents Sea. 

If this dipole persists through the summer such that the July August average shows such  flow, then 2015 should fairly easily fall towards levels similar to the September minimum of 2007 or 2011.


Neven said...

Nice work, and nice title, Chris.

I keep having that feeling that there hasn't been much of a pressure gradient, and with the high pressure mainly centred over Greenland or to the North of it, there hasn't been much of Beaufort Gyre either. And so a weakish dipole and not much transport, or at least compared to years like 2007 and 2012. But your last plot seems to contradict this feeling.

Dreessen said...

He really does capture the beating the ice has been taking the past week. Hopefully, the dipole doesn't last through August.

Chris Reynolds said...


It was quite a good title. You'll know yourself how the best titles jump at you quickly, most of the time you're left with something not very good.

Remember my post "May 2015 Status: Four Dipoles", I left it overnight, trying to tie it in with Four Weddings and a Funeral... 'But Four Dipoles and a...' What could rhyme with funeral and be relevant? I gave up and the post got a boring title.

You're not the only one who was surprised. I started off to do a mid month status post, when I saw how strong those winds were I scrapped the rest and just posted that.


I want it to continue, I love a good low September extent, it makes for a far more exciting summer than years like 2013 and 2014. In my opinion the summer loss of ice is one of the greatest spectacles of the natural world. The human driven global warming will only make it more spectacular and as the ice thins greater variability will result.

dreesen said...

We all have our own passions in life :) Anyway, I wanted to get your take on this post by SLR:

The linked open access reference examines the coastal dynamics and creation of new submarine permafrost in shallow water of the central Laptev Sea (see the first attached image) and concludes: "For this region, it can be summarized that recent increases in coastal erosion rate and longer-term changes to benthic temperature and salinity regimes are expected to affect the depth to submarine permafrost, leading to coastal regions with shallower IBP." (see the second attached image), where IBP means: ice-bonded permafrost.

This research and conclusions may have profound implications if the abrupt collapse of the West Antarctic Ice Sheet, WAIS, in the next few decades leads to rapid increase in sea level in the Arctic Sea. This is because more prior researchers have assumed that the IBP would be covered by a layer of somewhat protective non-ice-bonded sediment. However, as Overduin (2015) make clear, wave driven coastal erosion (which will increase if/when the Arctic Sea Ice extent seasonally collapses) can/will expose the previously buried but new submarine IBP to relatively rapid warming from the sea water; which would likely result in a multi-decadal period (starting with the ASLR possibly around 2050) of relatively rapid methane emissions as the associated methane hydrates in the new submarine permafrost region decompose

Chris Reynolds said...


Maybe, perhaps, I really don't know, get back to me about this WHEN the WAIS goes.

Dreessen said...

Yeah, he's a bit off his rocker I think

Chris Reynolds said...

I'm not saying that.

Consider a tree of possible impacts, with probabilities. Let's say at each branch you have 2 possibilities, each equiprobable, i.e. 50/50 chance.

b1 b2
d1 d2 e1 e2

c1 c2
f1 f2 g1 g2

So for a1 vs a2 you have 1/2 possibility of an outcome. Each of those splits two ways b1/b2, c1/c2, that's a 1/4 probability. For the next tier we have d1, d2, e1,e2, f1,f2,g1,g2, or 1/8 probability. i.e. the number of options would increase as 2^n for n levels of dependence, with the probability decreasing as 1/(2^n).

Now in reality only one pathway happens, say a2 -> c2 -> g1. In the world where that happens the probability of that path is 1. But who is to say what the actual path will be?

The problem I have with this kind of 'what if' stuff is that:
a) it is hard to ascertain likelihood.
b) the more levels of supposition one involves the greater the uncertainty.

And because of a) I tend to suspect bias when someone always finds either the best outcome (like AGW denialists) or the worst outcome (like AGW alarmists).

Yes if the WAIS goes it will raise SLR and this will increase coastal erosion. But how will that affect weather? e.g. Not all all of the land will be inundated. If the land permafrost thaws dry you get mainly CO2, if it thaws wet you get a lot of CH4. If it thaws wet and there is a lot of CH4 with an increase in GW how does that affect weather - does it make it drier or wetter - which feeds back on the expectation regards a wet or dry thaw.

If the WAIS really goes then this will put a lot of low lying coastal land under water. That might be thought to increase thaw by limiting winter freeze, but it will also reduce summer warming - which is the most critical factor?

If the WAIS goes which will be the biggest practical problem for humanity? GW, or the inundation of coastal cities and farmland, together with the fact that a lot of ports will be put out of action. Together with increased penetration of sea water into coastal aquifers. Together with changes in the tidal limit of rivers, impacts on flooding inland, impacts on nearby farmland, impacts on aquifers under such newly tidal stretches. Together with the loss or near incapacitation of major financial centres like London, New York, Hong Kong...

Really, isn't it enough to say that "The collapse of the WAIS will be problematic at least, it is likely that such an event would have consequences that are not forseeable, however it is not known when the WAIS will collapse or even if this will be a rapid loss of the sheet, or if it will play out over timescales of a century.

Timescale is absolutely critical. If as some doomers maintain the WAIS goes in a few decades the result is chaos. If it starts to go, the scientists agree it is going, but it takes 100 years to play out, adaptation is rather easy - you could simply stop new build in london, let the low lying areas fall derelict and build on higher areas around the city. 100 years is enough time to move an entire city without major financial catastrophe.

dreesen said...

Sorry, my bad. I find it very hard to believe it will go in a few decades. Even the Hansen paper (which has been heavily criticized by some, including Richard Alley) said 50-200 years for WAIS collapse. It will be a problem for sure, but there's little observational evidence to indicate a crash as SLR seems keen to think.

Chris Reynolds said...

No, the apology should be from me.

I wasn't really paying attention. The WAIS is something I haven't paid attention to for years, from memory: Parts of it are ice sheets grounded on the bottom, in the right conditions they might go pretty fast. But you're probably right talk of the whole WAIS going in decades is bonkers.

Blaine said...

I think there's some confusion here about "in a few decades" being used to mean both "starting in earnest in a few decades" and "taking place over a time span of a few decades". If you're using it to mean both, you have two different sets of a few decades before completion, which isn't a just few decades anymore.

Parts of it are ice sheets grounded on the bottom, in the right conditions they might go pretty fast. But you're probably right talk of the whole WAIS going in decades is bonkers.

Most of it is grounded on the bottom, and large parts are grounded more than 1 km below the surface. This is important because, as far as we know, there's no such thing as a calving front over 1km high.

The maximum thickness possible for S3 ranges from about 1100m without crevasses to a little over 900m with crevasses. This not only provides a convincing relationship between ice thickness and the maximum ice thickness permissible, it also provides an (approximate) upper bound on the thickness of grounded and floating termini. To our knowledge, glaciers that terminate in cliffs exceeding approximately 1km in height have not been observed, hinting that this prediction may be robust.

A calving front over 1km thick will probably immediately crumble into icebergs. Remove them and it will keep crumbling until it retreats back to be less than 1km high. A pile of ice rubble in front of the glacier could stabilize it, but Thwaites would be in warm water, and quite possibly open to the ocean along a ~500m sill of around 125km in length. The icebergs might float away very quickly, allowing quick retreat to areas grounded less than 1km deep. This is a lot of ice. After this, most of the rest of the WAIS would probably melt as well, but at a more normal glacial time scales.

Since we haven't observed anything like this happening, we have to infer the time scale from rapid sea level rise near the end of the Eemian. This is the source of the estimates of a few decades. It was much like now, and there's really nowhere else for that kind of sea level rise to have come from.

Blaine said...

Of course, it doesn't really interesting until Thwaites retreats around 65km into deep water, and its grounding line isn't moving much at the moment, even if it's thinning around 8m per year.

get back to me about this WHEN the WAIS goes
If there's a long slow rise in melting, when do we call the WAIS as having gone? When we see massive rafts of icebergs coming out of Thwaites? I'm not sure we'll have real scientific certainty before then. I don't think I'd call it as having gone yet, but I'm not that far from it. I'd call Pine Island Glacier as having gone, but it's relatively unimportant.

Increasing Antarctic melt should drive freshwater input, which should drive increased stratification, which should drive colder SSTs in the Southern Ocean and warmer intermediate temperatures, which should drive increased Antarctic melt. This pattern is apparent is already apparent, but I would be more confident that it would continue if I saw more symmetric melt around Antarctica. This discrepancy is significantly explained by correlation with ENSO, but this means that the WAIS melting is in part due to a string of years without an El Nino. So I'm going to wait a few more years before calling the WAIS destruction as inevitable in a relatively short time period.

100 years is enough time to move an entire city without major financial catastrophe
It was clear more than 100 years ago that we need to abandon Venice and New Orleans, but they're still there. I think we'll abandon our coastal cities when they're too flooded to live there any more. It's the way we humans usually do things.

Chris Reynolds said...


Thanks for taking the time to post that detailed reply. PIG and Thwaites were in my mind, I didn't know PIG had largely gone...

Yes we do tend to muddle along rather than acting before hand. I am just aware that buildings built within my lifetime are now being considered beyond their lifetime. The potential for adaptation is there. A critical caveat here is that the world's poor often don't have that potential.

Chris Reynolds said...


Having just caught up on PIG, It's not largely gone. There's a good page on it here:
Seemingly by someone working on the Antarctic ice sheet.

That states "Pine Island Glacier could collapse – stagnate and retreat far up into the bay, resulting in rapid sea level rise – within the next few centuries, raising global sea levels by 1.5 m"

But the glacier is well into the behaviour of massively accelerating flow, so it seems likely its fate is already sealed (e.g. given thermal lag of GW in the southern ocean).

I'd like to say I will make the time to read up on this, but unless I give up walking and films I don't see how...

dreesen said...

So what exactly are you saying in terms of timing of large sea level rise. Decades? Centuries?

Blaine said...

After asking you to define what you meant, I didn't define what I meant. Yes, the vast majority of the volume of Pine Island Glacier is still there. I think we've already seen most of the flow increase we're going to see from it as a result glacial dynamics. Maybe pulling off its recently pinned grounding line could up to double the flow, and the flow will continue to increase with warmer water, but barring other major change, it should just keep plugging along at roughly the same volume loss for a long time.

Thwaites is still slightly stable with respect to volume loss, but continuation of the current melt conditions will easily cause it to melt back to where it becomes unstable. I see around 60% chance that the current high melt, cold surface and warm water at 50-700m will continue indefinitely.

For the main WAIS disintegration, I'm expecting a better than even chance of a sea level rise of around 1.0-1.5m over 10-40 years (best guess 20) starting in 30-300 (best guess 70) years. It's possible, I suppose, to have a small open channel with an iceberg jam to make it take a century or so to melt. There are big uncertainties in the warming of the incoming seawater and the glacier response of Thwaites, namely whether it retreats or accelerates. On a time scale of hundreds of years, the whole deepwater of the Earth heats up, so it's hard to see how it can not melt out. After initial melt out, probably we'd see another ~1m over ~300 years.

I'm expecting a jump to around 3.5mm/year of melt from Antarctica over the next 10 year or so, followed by only slow increases after that until Thwaites accelerates into WAIS collapse. Warming of the intermediate layers is generally believed to trail cooling of the surface layer by around 10 years or so. I'm guessing a strong response soon followed by saturation, with basically a small step-function jump of melt. The response should be evident in GRACE data or falsified by it in 5-10 years.

The are around 19m of sea level rise in ice grounded below sea level in East Antarctica, including some large very deep areas and some retrograde slopes which could do odd things. To guess, under current forcing (which isn't going to happen), around 0.4mm/year from East Antarctica, jumping to 1.5mm/year after around 200 years as some of the larger glaciers melt back into deeper water, and a total around 6m at around 4000 years. There's certainly the potential for a weird and unexpected response, but the higher sills keep my default expectation low, at least until we throw more carbon into the air.

Dreessen said...

That sounds fairly gradual to me.

Chris Reynolds said...

Thanks Blaine,


It sounds more gradual than what some are saying, and more in line with what my expectations were when I was more up to date on the matter. But, once we start it we probably won't be able to stop it.

I must admit I read the first 15 pages of the new Hansen paper over the weekend and have got no further since...

Dreessen said...

That sounds about right to me. I also think Hansen's paper is going to be ripped apart on peer review

Chris Reynolds said...

David Archer has been quite complementary about the Hansen paper, but makes some suggestions for improvement. There are a few other comments from professionals that are moderate. The most cutting I have read is from Gavin Schmidt who has said it is just one scenario, i.e. there are others (and I suspect most are not as pessimistic).

However there are a few, frankly embarrassing, review comments. IIRC one says that CO2 doesn't cause warming, the hubris required to post something so daft is staggering.

But I haven't finished Hansen's paper, so I cannot offer comment on it. However I am going to finish reading it because Hansen is no Lovelock or Wadhams, he really knows his stuff and from my experience is worth paying attention to.

dreesen said...

I agree with you that it's worth monitoring, but it's also notable that Richard Alley doesn't think terribly much of it. I've heard that the authors of the study are heavy hitters, but so are those criticizing it. My personal opinion is that the true scenario lies between Hansen and the IPCC.

Blaine said...

As I noted, my predictions above assume atmospheric forcing comparable to today. Really I expect that in 200 years CO2 will be at 700-1200 ppm, and my central estimate for sea level rise is 3m/century from East Antarctica alone. The difference is in predictions of politics and not in predictions of geophysics, and politics is a sufficiently soft science that any prediction necessarily amounts to hand-waving.

It's important to realize that Hansen et al. is not making any kind of prediction about glacial dynamics whatsoever. Since glacial dynamics remains very poorly understood, it instead drives a modern climate model (one specifically selected to match modern Antarctic cooling, unlike most of CIMP5) with Eemian paleoclimate freshwater runoff, and shows that this results in a warm mid-level Antarctic ocean which is consistent with the assumed runoff, given our large uncertainty in glacial dynamics.

I find the most relevant criticism of the paper to be that Eemian paleoclimatolgy tells us that at current forcing levels, we can expect a sharp meltwater pulse of around 1.5m sea level in rise in ~20 years and around 3m in ~200, but the timing of this pulse remains very poorly constrained, and hence we cannot be confident that it will occur as quickly as Hansen proposes.

Gavin Schmidt and the rest of the conservative modeling community that he represents have repeatedly had their lunch eaten by the predictions of the observational and paleoclimatology guys. Remember this? Remember the pedictions of AMOC stability? How about this post, showing complete confusion as to what is going on in Antarctica as of Dec 2014?

Higher resolution models show a sharply higher dependency of overturning circulation on freshwater input. One would think that basic physics would indicate that the results of the higher resolution models are much more likely to be correct, but, no. the majority of the modeling community is steadfastly holding on to the lower resolution results.

Look, we're approaching peak Eemian forcing levels, and will shortly be at greater forcings than ever seen during during the Eemian. If you want me to take your model seriously, give me a model which doesn't fail validation against the known sea level response of the Eemian. This condition excludes almost all of CIMP5.

Chris Reynolds said...


I think the reasonable conclusion regards the MOC is that it is in decline well ahead of the models (well the coarse ones). There have been references over at Neven's in the comments to a 'cool pool' in the North Atlantic, exactly where the MOC slowdown should be apparent.

I don't disagree with anything you say. On this blog and elsewhere, you will find me saying repeatedly that the consequences of AGW are most serious because of the long term consequences. I suspect that Dreessen is in a similar position to mine - scepticism regards claims of an imminent catastrophe within decades due to AGW, not a dismissal of the potentially serious long term impacts. Dreessen can correct me if I have misunderstood here.

I have now managed to finish Hansen, I find this paper as persuasive and problematic as I have his past output. But I don't consider the warning to be alarmist. Note that the definition of alarmism is the unnecessary and overstated raising of alarm, not the justified raising of alarm.

The problems with Hansen? I can quite appreciate why Gavin Schmidt said it was just one scenario. I had a few moments where I wasn't convinced, such as regards tsunamis not being the agent behind movement of boulders in the Bahamas. But I have never been one to throw out a wider argument over small technicalities. In engineering the first 'proof of purpose' is often 'full of holes', what one needs to do is to judge whether the gist is there. The question then becomes is it unlikely, or worse; inherently implausible. I consider still SLR of more than 1 to 2m this century to be rather unlikely. But have never had a problem with the implication of multi meter SLR over the coming centuries, this is hardly novel in the literature.

dreesen said...

Chris and Blaine,
Gavin and the modelling community may be correct about methane clathrates, as recent studies by Krill and Overduin demonstrate (I certainly hope so). That being said, I find Hansen's argument to be worthy of note and careful monitoring must be continued

Dreessen said...

I wanted to get your thoughts on the Discussion section of this methane paper. I don't think they're saying anything concerning, but I'll let you be the judge:

Chris Reynolds said...

VBG, I read the abstract and thought 'Why the hell he is dumping a genetics paper on me? Anyway that was interesting because I was listening to a podcast at work: Melvin Bragg's In Our Time, from BBC Radio 4. It was about extremophiles and covered the evolution of archea, bacteria, and eukaryotes, that paper tied in with the subject.

BTW - that paper says atmospheric CO2 has tripled since pre-industrial, taking 1880 as pre-industrial we haven't yet doubled. But their references might use a different baseline. Even if you take the roughly 150ppm from the bottom of the glacials it's still not 3X.

So I guess you're wondering about the possibility of a re-run of an End Permian event? I don't know. I don't know if we have enough carbon to burn to bump atmospheric carbon dioxide up by the same sort of relative amount, and I don't know if there is enough methane inventory in the oceans and permafrost to do the job of the second phase of the End Permian.

dreesen said...

Thanks for that. I sort of pay attention to the methane studies because the feedbacks surrounding that gas are the only ones that really matter. Do you remember the Kara Sea study by Portnov that came out in September? He postulated that explosive releases could occur if the bottom waters warm by 2 degrees. I would assume he means of local importance, but I'm not sure.