« 2015 minimum overview, part 1 | Main | Arctic sea ice -- in pre-election perceptions »


Feed You can follow this conversation by subscribing to the comment feed for this post.


In a couple of days, on October 6th, the Sea Ice Prediction Network is hosting a webinar:

This webinar, entitled "The 2015 Sea Ice Outlook: Post-Season Discussion" will provide discussion of the 2015 summer sea-ice conditions, a review and analysis of outlooks contributed from 2008 to 2015, and discussion of the challenges and successes of predictions at the local scale.

Julienne Stroeve, Larry Hamilton and Cecilia Bitz will be speaking. Be sure to register.


NSIDC has posted the September extent anomaly: 4.63 million sq km


What did you make of that recent report concerning heat being dumped upwards by eddies in the Arctic Ocean?
It seems to be either an unknown contributor to sea ice melt, or potentially an exaggerating factor to the long-term decline.


bobcobb, I haven't yet looked into it as much as I should have, but yeah, this could be a serious positive feedback. There's enough heat in the lower layers of the Arctic Ocean to melt all of the Arctic sea ice many times over.


A few commenters predicted late June, iirc, that losses in August would be significant because the early lack of land snow cover would provide late feedback of heat into the Arctic. I don't understand the mechanism but looks like they were dead on.


Neven, Navegante, Is it possible that perhaps that the heat stored in the Arctic Ocean has been contributing to the melt all along, and they just haven't documented its effect until now?


It's clear - and always has been - that heat in the ocean plays and has played a big part in the long-term thinning of the ice pack (read my blog post on Ocean Heat Flux for instance). They haven't documented the effect per se, but rather one of the possible reasons of the effect, the others being ocean currents bringing heat into the Arctic (again, see that blog post), and heat coming in through the ice-albedo feedback.

Unfortunately it's very difficult to quantify, but scientists are working on that. It's important stuff, not just because of the implications for AGW etc.

Hans Gunnstaddar

"NSIDC has posted the September extent anomaly: 4.63 million sq km"

shwarzbarry, what is the extent anomaly and do you have a link? What I found on their site was the following:

"On September 11, 2015, sea ice extent dropped to 4.41 million square kilometers (1.70 million square miles), the fourth lowest minimum in the satellite record."


Hans - September monthly average. It's listed on the data pages.



I think Barry meant September average or monthly minimum instead of 'anomaly'.


Neven, So they essentially quantified a reason for a feedback that they already knew existed and was affecting the long-term trend?


September average is more precise, but as I mostly deal with the monthly sea ice data values, I'm used to calling them anomalies - which can be a single (ie averaged) value for any period (eg, ENSO indices have 3-month anomalies, or monthly or annual anomalies for global temperature data).


Neven, So they essentially quantified a reason for a feedback that they already knew existed and was affecting the long-term trend?

Bob, my understanding - based on this BBC article - is that they didn't really quantify all that much, but rather found evidence for the heat eddying upwards.

Like the mission's chief scientist, Jennifer MacKinnon, says:

"The strength of [these currents] has been incredible," Dr MacKinnon said. "We now need to disentangle what the contribution of that process is to the multi-year, inexorable decline of the sea ice."

And the NSIDC's Dr Julienne Stroeve comments:

"I think it is quite important to understand this type of mixing of warmer ocean waters at depth with the sea ice," she told the BBC.

It will be crucial, Dr Stroeve added, to quantify exactly how much heat is reaching the ice and how much melting it has caused.

"In 2007 more than 3m of bottom melt was recorded by [an] ice mass balance buoy in the region, which was primarily attributed to earlier development of open water that allowed for warming of the ocean mixed layer. But perhaps some of this is also a result of ocean mixing."


Neven, That sounds about right to me. My prediction for a partly ice-free Arctic summer is late 2020s-2030s. Does that sound reasonable, considering there's a lot of room either way?


I think it's reasonable, although mainstream opinion (I believe, haven't done a survey) is between 2030 and 2040.

I personally believe it could happen at any particular melting season under exceptional circumstances, like weak initial ice state because of a cloudy winter, heavy preconditioning during May and June, followed by a July we saw this year (or warmer), and then a warm, stormy August.

At the same time some negative feedback might kick in, or some unforeseen natural variability keeping things on a plateau*. Of course, that would be a reprieve, nothing more. Whether the Arctic practically melts out before 2030 or after 2050 is not all that relevant (especially if human societies keep opting for business-as-usual).

It's about the process, not the end result, as the consequences are most probably already upon us and becoming more and more visible.

* I don't want to think about it, but it could just as easily be the opposite: positive feedback kicking in, unforeseen natural variability kicking off the turbo.


Great discussion as usual.

I have real from several sources about the heat content of the deep Arctic Ocean and how if this heat ever made it to the surface in enough quantity the era of Arctic Ice would basically be over. Due to saltiness, layering, and lack of the water being disturbed this heat has been stuck at the bottom.
So, what I am hearing from this discussion is that wind and storminess during the low ice conditions we are having now is stirring deep waters and bringing this trapped heat to the surface in greater and greater quantities. That is really a scary thought.

Hans Gunnstaddar

I figure timing of an ice free Arctic from the standpoint of approx. 5 yr. spans between new record minimums, which puts it in the 2027-2037 range.


A new weather game is a foot, where instead of having a massive High at about the North Pole, there is a massive very large pan-Arctic Low, the result: a shrinking, yes reduction in extent again way pass the minimum. From 5069941 to 5053825 Km2 since yesterday. Amazing and expected, the state of the ice, as portrayed above, is very poor, and vulnerable even though it should expand and solidify.
2016 is already shaping up to be a severe melt regardless of lack of compaction.


I would be hesitant about making a 2016 prediction right now. It's far too early.

Pete Williamson

From the conclusion

"..... a more or less average 2015 melting season ....."

The real issue for me is trying to put that particular comment into a long term context. From the (small amount of) science done in this area the tendency seems to be that the circulation patterns in the recent past ( say ~ post 2000) have tended to be at an extreme, showing patterns that favour ice loss.

What represents " more or less average" in recet times does not necessarily mean the same thing when compared with longer term reconstructions(?). At the same time there seems to be competing evidence about whether these changes in circulation patterns are natural or CO2 forced or a bit of both.


Pete Williamson said:

"whether these changes in circulation patterns are natural or CO2 forced or a bit of both."

I like to think of these "changes" in the context of a "cascade of events." The initial causes of human released carbon dioxide and other greenhouse gasses into the atmosphere set off a series of events that result in changes in Arctic Ocean circulation patterns. To oversimplify: Extra carbon dioxide, methane, nitrous oxide, etc., cause extra heat to build up in the atmosphere some of what the ocean absorbs. Water vapor then increases in the atmosphere warming the atmosphere even more. Air circulation changes as does non-Arctic Ocean circulation. More heat invades the Arctic, more ice melts, less ice refreezes and as a result there is more open water in the Arctic. Wind and storms move water through waves and start upwellings and downwellings in the Arctic similar to the other oceans causing even less ice cover in the Arctic. Repeat and repeat as long as additional carbon dioxide has an effect.

The closest parallel example I am thinking about is similar to biological concentration of a toxin in the environment which may be very low. However, as this toxin moves up the food chain and becomes more concentrated at each step, eventually arriving at tertiary carnivores, the effects of this toxin may become lethal. Similarly, a seemingly small increase in carbon dioxide can have large effects up the line as heat energy becomes more concentrated.

I am sure someone here has a better way to explain or deny this thinking pattern but I though I would take a stab at it.

Pete Williamson


I dont doubt the plausibility of the ideas you've laid out but the question is really how much they dominate.

I like this paper which seems to try to put the recent circulation changes into a 100+ year perspective.
I think its important because although others have done a similar analysis for the recent period(with similar results), this seems to be the first work to give the longer time frame.

Most of their attribution discussion focuses around the recent period but I think there are interesting insights from earlier times.

Their fig2 is the main finding and circulation patterns 2 and 4 are the interesting "ice melting" patterns. In the 1930s/40s there was some warming of the NH with an arctic amplification pattern but their results suggest the "ice melting patterns" were less frequent. In the 1950s/60s the frequency of "ice melting patterns increased in a similar way to the present but the temp records suggest temperature in the NH cooled somewhat and the arctic amplification pattern disappeared.

Essentially these two earlier periods appear to be behaving in the opposite way to how you outline warming may have been working in recent years.

This is the type of work that makes me wonder

a) The extent to which the warming pattern is influencing the circulation changes
b) The extent to which the warming is attributed to forcing and dynamical circulation changes.

Clearly as you suggest the two processes are likely to influence each other but it also seems plausible that circulation changes have 'natural' origins. And clearly a warming world can influence ice loss more directly and thru other feedbacks. I struggle to know just which processes are dominant.


Pete, I tend to think the extra warming as a result of emissions tends to aggravate melting conditions and make them worse than they normally would be. The drought in California is a good example. It's more likely to happen and be worse than if climate change wasn't a factor. So I would say manmade warming is the dominant force. How much it is dominant is beyond my expertise.


Quick question here for the veterans of this blog, what do you guys make of the spike in atmospheric ch4 levels from mid-2014 to this point in 2015? Is it mainly wildfires and fracking?


A double Polar atmospheric squeeze, one typhoon and one hurricane heading North Polewards will enhance an anticyclone over the Arctic Ocean gyre next weekend. Likely another drop in extent again in the cards if ECMWF plays it right.


Thanks Pete, I also wonder about the dominant force. I also agree that the "aggravate" philosophy is also part of a plausible explanation as well.

Rob Dekker

Thank you Neven, for yet another comprehensive overview of the factors involved in determining the September minimum.

I am pleased that you prominently mention Northern Hemisphere snow cover in early summer (and June specifically).

And for good reason : Snow cover (or more accurately the "lack of snow") determines how 'dark' the northern hemisphere is, and with that, serves as a amplifier of ice melt.

Quick math : in June, when the sun is high in the sky, snow free soil absorbs something like 200 Watt more per square meter than snow covered land.

The snow cover anomaly of 4 million km^2 in June 2015 (Rutgers), thus resulted in some 2 * 10^21 Joule (2000 exajoule) extra energy inserted into the Northern Hemisphere system, and close to the ice. That is a lot of energy. In fact, enough to melt out 4 M km^2 of 1.5 meter-thick (FYI) ice. Even if only half of that energy blew north and only half of that resulted in ice melt, the snow anomaly of June 2015 would have resulted in 1 million km^2 Arctic sea ice melt.

This is why I believe that no prediction of Arctic sea ice extent is complete if it does not include Northern Hemisphere snow cover as a variable.

As you know, I did use snow cover in June for my submission to SIPN July report, where I predicted 4.61 M km^2 for the Sept minimum.

The final NSIDC number for the Sept minimum is 4.63, which makes my entry the closest of all SIPN entries, and makes it look like I know what I'm talking about :o)

In reality, I had luck on my side this year to get this close, but the main message is clear : snow cover in spring and early summer is important when predicting sea ice extent.

Rob Dekker

I was a bit sloppy with the wording. With "September minimum" I mean the "average September sea ice extent", as reported by NSIDC, and used as the target for the SIPN entries.


As you know, I did use snow cover in June for my submission to SIPN July report, where I predicted 4.61 M km^2 for the Sept minimum.

The final NSIDC number for the Sept minimum is 4.63, which makes my entry the closest of all SIPN entries, and makes it look like I know what I'm talking about :o)

I had noticed that, Rob. Well done!


Rob because clearly you know what you talking about.

Andy Lee Robinson

3 days of rendering later, I've just published my latest "Ice Cube" video:

Arctic Sea Ice Minimum Volumes 1979-2015

Announcement tweet for any RTs: https://twitter.com/ahaveland/status/650730307302375424
2015: 5,670 km³ (1,142 km³ less than 2014).

Arctic Death Spiral also updated:


'Even if only half of that energy blew north and only half of that resulted in ice melt, the snow anomaly of June 2015 would have resulted in 1 million km^2 Arctic sea ice melt.'

... this is phrased to sound like two conservative assumptions, but are they? naturally I have no idea, but does anyone?


Some hot news from the Obuoy Front:

Obuoy 9, Obuoy 10, Obuoy 11 and Obuoy 12 have been deactivted.

OTOH, Obuoy 8, Obuoy 13, Obuoy 14 and Obuoy 15 have been installed and are fully operational now.

Feel free to download the Obuoy collect file

Ghoti Of Lod

The movies for Obuoy 13 14 15 are interesting at this point because you can see the people setting them up. Gives a bit of perspective of the size of the yellow bits we see from the camera the rest of the time.

Obuoy 13 movie includes the ice breaker that carried everything there and the people away.


Ghoti Of Lord wrote:

... Obuoy 13 movie includes the ice breaker...

So it is at Obuoy 8 movie, Obuoy 14 movie and Obuoy 14 movie too.


I don't know if the Arctic's in a death spiral quite yet. It's getting there, but I think that terminology is a bit strong.

Rob Dekker

sofouuk, conservative or not, these assumptions (half of half of the heat from snow cover loss goes to ice melt) at least appear to be realistic :

As I state in the July report, in the 1992-2012 data set, the best correlation (R=0.93) is obtained if for every km^2 of snow cover loss in june, 0.24 km^2 of extra sea ice melts out by september.

So the statement that 4 million km^2 snow cover loss in June leads to 1 million km^2 sea ice loss by september is realistic, both from a physics and an empirical perspective.

Rob Dekker

I may be going out on a limb here, but this increasing summer snow cover loss may have more effects than just amplify sea ice loss during the melting season :

As the snow line moves further north, the heat absorbed by the dark soil will rise, and thus shift the boundary between the Ferrel cells and Polar cells further north as well.

That could mean that snow cover loss could actually move the jet stream further north, most notably over land...

I really do not know enough about such effects, but it is food for thought..


I can see the empirical argument, of course, but I'm not sure what the 'realistic physics' is. most of the directions wind can blow in from previously snow covered areas in the northern hemisphere won't take the extra heat over the ice, so 'even if only half that energy blew north' seems a bit misleading, and warmer inflowing air might tend to rise over the cold dome of air over the ice - yes this will result in some melting but I'm interested to know how much, partly because I've always been frustrated by how vague the information on arctic heat budgets always seems to be. some of this could be correlation rather than causation - a warm June over northern Siberia or North America, or both, might lead to both early snow loss and melt ponding, leading to a bigger ice melt?

Susan Anderson

@Andy Lee Robinson, thanks.

wrt "Death Spiral" I think the phrase was coined before the ominous 2012, but in general the spiral format is a ready reference for a progression that is not perfectly continuous, but whose direction is clear, and as such is quite appropriate.

I'm worry that people who counsel not mentioning the very obvious dangers of our current path and practices are encouraging wishful thinking and donothingism.


Yes, Susan, but there's a fine line between that and being a doomsdayer. Using that sort of rhetoric is never helpful, and a death spiral implies the sea ice is just collapsing in the next few years, which is very much in doubt.

Rob Dekker

sofouuk said

I'm not sure what the 'realistic physics' is. most of the directions wind can blow in from previously snow covered areas in the northern hemisphere won't take the extra heat over the ice, so 'even if only half that energy blew north' seems a bit misleading, and warmer inflowing air might tend to rise over the cold dome of air over the ice - yes this will result in some melting but I'm interested to know how much, partly because I've always been frustrated by how vague the information on arctic heat budgets always seems to be.

I understand that you are skeptical of the physics of ice melt, but your response is a bit of a Gish Gallop, which makes it difficult to address in full.

Let me state the issue the other way around :
We know that a 4 million km^2 loss of snow cover will cause the exposed dark surface to absorb something like 2000 exajoule of heat during June alone.

Can you tell us where you think that heat may finally end up over the months ahead, knowing that any place north of that melted snow area is packed with snow and ice, and cannot become warmer than 0 C. ?


Rob Dekker - I think your answer lies less in where it goes and more in what doesn't leave

I'm not sure about the 2000 exajoule figure, but increased albedo and uptake of sensible heat by various buffers - sea water, atmospheric moisture, sea ice melt - are the likely culprits, and the phase changes which take place close to that -2C to 0C temperature threshold will keep the system close to that temperature.

As far as it not becoming warmer, the arctic doesn't have to in order to become far more energetic than it is; there's still a lot of room in the 'buffers' before that heat shows up as something sensible through temperature increases.

Rob Dekker

jdallen, I agree.
The 2000 exajoule (or whatever the number is) will go into the various buffers (land warming, atmosphere warming, ocean warming, and ice melt), or leave the planet by radiation.

Since radiation increases only by something like 1 W/m^2/Kelvin, I think the majority of that heat will go into these buffers.

And considering that the ice/snow covered surface of the Arctic is pinned down at 0 C during the melting season, the 0.24 factor into ice melt that we obtain from empirical results does not appear to be unrealistic.

Unless of course somebody shows that the alternative buffers can store the 2000 exajoules somewhere else...


I made two points - 1) half of the extra heat won't go north over the ice, partly because the area south of northern snow covered area is much bigger than the area to the north (isn't that obvious?), and 2) of the heat that does go over the ice, it's not clear how much of that will actually result in melting. if you interpret that as a Gish gallop, it's probably not worth continuing the discussion. of course I think the heat will go into the atmosphere (wasnt that clear from what I said?) but the question is then whether it is then eventually lost by radiation - I don't have a clear understanding of what happens next, which was the reason for my original question. but never mind

Rob Dekker

sofouuk, let's talk about your first point first :

half of the extra heat won't go north over the ice, partly because the area south of northern snow covered area is much bigger than the area to the north (isn't that obvious?),

OK. So imagine you are in this pocket of warm air that was just released from the now snow-free soil, and you need to choose to go south or north.

According to you, you choose to go south, since the area south of the northern snow covered area is much bigger than the area to the north. Did I get that right ?


no, and the inane sarcasm is not appreciated - the most likely direction a packet of warm air will travel is vertically upwards (duh). because however lines of longitude converge at the North Pole, any horizontal movement at ground level is much less likely to take it over the ice pack than it is to take it away from the ice pack - one area is vastly bigger than the other.

as an aside, it is worth pointing out that a June negative anomaly in snow cover is likely to lead to extra ice melt, either because 1) warmer than average land temps in high northern Siberia/N Am are also likely to mean that temps over the nearshore ice are above average, or 2) bcz warmer than usual air moves from over the land to over the ice (which is what you believe), and in either case this might lead to more widespread melt ponding than usual; if early melt ponding is greaterer than usual, the albedo effect kicks in and ice melt proceeds from there. the point is that once ice albedo is lowered at an earlier stage than usual, it really doesn't matter where the extra heat from lower land albedo goes for the remainder of the summer - the ice will melt faster all by itself. I suspect this is the cause of a large part of the correlation, but as I keep saying, I really don't know.

just there is a lot of anomalous heat in the northern hemisphere doesn't mean the ice will melt, unless the heat comes in contact with the ice. the Pacific blob, for instance, is irrelevant for predicting arctic sea ice melt, and even anomalously warm waters in the peripheral arctic seas won't lead to greater than average melt unless the warmer water is driven under the ice.

Rob Dekker

sofouuk said

no, and the inane sarcasm is not appreciated

That's OK. I do not appreciate your patronizing "Duh" and "isn't that obvious?" and unfounded accusations of "misleading" either.

Now that we got that out of the way, let me say that I appreciate the second part of you last comment. You clearly have been thinking about your second point, which is encouraging.

This is Neven's Arctic Sea Ice blog, and not some arbitrary MSM article, so we talk science here, and we BOTH can learn something by discussing and understanding the physics involved in this process.

So, I do have some comments on that second point, mostly because the variable ( melt ponding and water-on-ice) is actually measurable, in the form of "extent minus area", which is a variable in my regression equation (please read my ISPN report entry). There is a time lag there, which I would like to talk about later, and that discussion may actually end up answering some of your earlier questions.

But before we go there, we need to come to a common understanding about what is happening with your point (1). What first happens when 4 million km^2 of snow melts out, and gives way to dark soil, absorbing something like 200 W/m^2 of extra heat, some 2000 exajoules over the month of June alone.

You write :

the most likely direction a packet of warm air will travel is vertically upwards "

In the absence of ground wind, yes, that air column will most likely travel upward. Now, imagine the column south of it (which is equally fueled by that 200 W/m^2 heat source) and the column north of it (which is not fueled, because it is still snow covered).

As the pocket of heat moves upward, which direction is it most likely to take ? South, where the air is equally warm or warmer, or north, where the air is cooler, and thus more compact, and thus the atmospheric pressure is lower ?

Isn't it "realistic physics" that that pocket of air would slide down northward following the pressure curve, and become part of the polar cell ?

Why on Earth would that pocket of air go South (as you assert), where the air is warmer and pushing back ?


I have nowhere asserted that it 'would' go south, only that the potential destination area away from the ice is much larger than the area over the ice. and even if it does go over the ice, how much effect will it have on ice melt if it is at altitude? I haven't read your ISPN entry, but I will

Bill Fothergill

@sofouuk & Rob D

Perhaps it might be educational to - quite literally - bring this debate down to earth, so here's a little thought experiment...

Imagine a series of 1km square areas at ground level stretching north-south along some arbitrarily selected Line of Longitude, and let's assume that the snow cover starts at, say, 70N.

As we progress north (ie poleward) from mid-latitudes towards this albedo discontinuity at the 70N boundary, the air will be getting warmed by the ground below. At the albedo discontinuity, this ground-to-air warming will markedly decrease. Polewards, the air at ground level will be colder and denser, whilst equator-wards, the air will be warmer and less dense. (As the warmer air has a greater propensity to rise, it will, by necessity, leave less dense air in its wake.)

Therefore, at ground level at the albedo discontinuity, we not only have a temperature gradient, but we also have a pressure gradient.

Given that the pressure is higher to the north, it should be pretty obvious which way the wind blows.

Of course, since we live on a rotating sphere, there are some complications. Newton's First Law enters the fray in the guise of the pseudo force known as Coriolis acceleration. It should again be pretty obvious in which direction the ensuing deflection manifests itself.

Now expand the thought experiment to encircle the entire globe, rather that just along one Line of Longitude. In this greatly idealised example, the albedo boundary would occur all the way around the 70N Line of Latitude.

However, at each point along the boundary, the wind direction would still generally be from the north (along with the aforementioned Coriolis deflection). Now, as sofouuk points out, the area poleward of this discontinuity is much smaller than the area lying closer to the equator.

It should therefore be pretty clear that, in the absence of some replenishment mechanism, this situation cannot continue.

However, Rob has already mentioned the mechanism, and it is the Polar Cell. In the polar regions, cold, low level winds from the north go hand in hand with energy bearing high altitude winds from around the Ferrel/Polar junction.

This is a fundamental process in the Earth's energy balance, such that the net outgoing radiation in the polar regions is actually in excess of the received Insolation. The polar regions are therefore acting as a heat sink, with the lower latitude regions acting as a heat source, with convection and advection acting as the energy transport mechanism.

So what happens when the high albedo region decreases in size due to snow/ice melt?

In a nutshell, the heat sink reduces in size and the heat source increases in size.

The reader is invited to work out for themselves what is going to result from such a situation.

cheers billf

John Christensen

Slightly OT, but very much Arctic sea ice:

I am noticing from SST and sea ice concentration charts that sea ice is becoming present in areas, where the air temperature is not low enough to support freezing of the sea surface.

In fact it seems like this is happening in more of the adjacent seas right now.

So I am speculating that sea ice right now is primarily generated deeper within the ice pack and then as leads freeze over and the central pack increases in thickness, the ice pack spreads from the center (Either by the increase in ice thickness or by winds moving it into areas of open water) rather than growing at the edges.

Does anyone know if this is normal for this time of year and also if this is actually what is happening?



"So I am speculating that sea ice right now is primarily generated deeper within the ice pack and then as leads freeze over and the central pack increases in thickness, the ice pack spreads from the center (Either by the increase in ice thickness or by winds moving it into areas of open water) rather than growing at the edges."

Not bad, but once the main pack is strongly consolidated frozen as it is ONE MONTH LATE now, drifting snow spreads outwards and basically is a precursor to more freezing at its outwards limits, likely because sea water convection is stopped.

The present anticyclone over the pack is very late:


A sure sign of a coming mild winter for most of the Northern Hemisphere.

John Christensen

Hi wayne,

As is often the case, I have a hard time understanding your argument, which is probably caused by the lack of reference to something substantial - and the use of capital letters doesn't help much.

Regarding temperatures and consolidation of the main pack: DMI 80N temps were on average for the whole month of September, enabling the main pack to consolidate to a normal degree. Cyclone activity starting about 14 days ago then brought in warmer air from the Atlantic, thereby causing the spike in DMI 80N temps, but with only a few days of temps higher than -10C, I do not see how this could have delayed the consolidation of the pack by a month.

Please elaborate, as I may be missing your concept of consolidation.

John Christensen

Secondly, I will add that the AO index has been largely negative since the last week of June, and the forecast for the next two weeks is for the AO to stay negative:


The consequence of this pattern is that the jet stream moves further south and becomes more wobbly, enabling cold Arctic air to move further south. The predominance of high pressure in the high north (I deliberately say 'high north', as we know the AO index covers a large area) enhances the heat sink effect and therefore should cause NH temperatures to stay on the average or slightly colder than average.

The relative lack of sea ice evidently counters the effect of the AO index, so the favorable atmospheric conditions are likely to be balanced out by above-average degree of open waters in the Arctic region.

John Christensen

Related to the Arctic heat sink effect and how this may change in the future, another item has puzzled me for a while - the ACE..

The ACE is an index for Accumulated (tropical) Cyclone Energy. This index is calculated based on the number, strength, and duration of tropical cyclones around the world.

What is puzzling is this: In recent years the index has dropped precipitously:


2012, 2013, 2014, and no doubt also 2015 will have the lowest annual, global ACE index since these started getting calculated globally around 1970.

This could mean:
1) AGW has no impact on the number, duration, or strength of tropical cyclones, or the impact is countered by another factor stronger than AGW. This would be surprising.

2) AGW has a negative correlation with tropical cyclones, due to an increase in wind sheer at altitude that breaks down the tropical cyclones (http://www.wunderground.com/blog/JeffMasters/comment.html?entrynum=1458). This seems most plausible.

If scenario 2 is correct, does this mean we still will have the same level of energy exchange between the tropical and sub-tropical zone, or could this cause the energy exchange between tropical and sub-tropical (and in the end the Polar regions) to be reduced, so that less heat energy will enter the Polar region?

If that happens, will it cause the Polar region to expand (I know this sounds crazy) and the tropics to become even warmer, or will it just accelerate the overall global warming due to reduced heat sink effect?

With four years of extremely low ACE, one would think the scientific organizations would notice and start researching.

John Christensen

And this is my own favorite Doomsday scenario (Making me wonder why we as a species are so engulfed by tales of our own demise..):

What if the extremely low ACE index indicates the onset of a separation between the tropical and Arctic atmospheric zones (Either caused by AGW or whatever natural phenomenon would cause this), and that this separation allows ice to accumulate and spread, leading to the next cold period of our current ice age?
If so, it would be truly disastrous to our modern civilization, so again, I hope someone have noticed and are looking into the ACE..


You crazy alarmist, you. ;-)

John Christensen

If the ACE index does not go back to normal next year I will sell my house in Denmark and buy property at the foothills of Toubkal in the Atlas Mountains - while prices are still reasonable.. ;-)


Hi John,

I am as clear as it gets especially if you go to the link presented.
Sea ice Consolidation is confirmed when Highs subsist at about the Pole especially at this time of the year. Previous Heat sources were multiple as you cited just one, they've delayed the onset of winter. You must visualize the reflection of a solid consolidated pack twinned with very clear and cold air right above it. If you can't see it you wont understand. You must look at the image without complicating it with causations. The capitals used is to emphasize the lateness of consolidation and clear air (that was obvious wasn't it? Or shall I use colour characters :) )....

John Christensen

Hi wayne,

High pressures in the Arctic come and go as is the case in other parts of the World. I have not seen any data or other reference that a HP will actually stabilize over the pack once it has consolidated.

Denmark had a record high air pressure for September one of the last days of the month (1041,9), but with no ice in sight, so please share some data or other reference for the correlation between Arctic ice consolidation and presence of a high pressure.



Visualize this: no stable Highs over the much weakened Arctic Ocean pack until the pack freezes solid.

Sorry I can't describe Denmarks High origin because I did not study it , but I can do the Arctic . It is a great weakness of most who use AO's to describe things, it does not explain anything but vague generalities, without knowing the source reason.

So picture this:

Arctic Ocean loose pack with warm water "invites" Lows to penetrate the pack surface area until its cold enough by darkness, in the case of 2015 the water was much warmer, the sea ice much weakened, the pack consolidated much later.

Once a near North Pole large patch of consolidated pack can acquire much colder air than surrounding areas, its denser surface air flows Southwards, Coriolis effect makes it flowing towards warmer air turn Right, voila : anticyclone-genesis. Don't you like knowing why there is a High pressure rather than describing that it is there? Work on understanding the reasons why AO is + or -, then you will know when winter starts!


Question. Why is most of the ice retreat on the Pacific side? Is it Greenland or what?



I had hoped the comparison would copy. The other half is here:



Dang, my apologies,



Melting of MYI was not complete and its "debris" is catalizing the refreezing? A plausible explanation I read on the Forum.

The comments to this entry are closed.