Arctic Sea Ice Retreat: When Will the Arctic Ocean be Ice-Free During Summer?

Asks climatematters@columbia. But they ask it in a way that suggests they think the trend is going to be steep. So I offered them the standard bet. We'll see.

Meanwhile, anyone interested in whether 2009 is likely to be a record can get some action over at ipredict (thanks Gareth). I've bought some; current price is about 0.23. I'm not really sure what a fair price would be; I have some buy orders in. Its quite educational.

More like this

You that do calculations and such... How would the thickness decrees affect the extent? It seams to me that it would be easier to melt a big part of a not so thick ice... however I have no clue what so ever about what the 3D spread looks like...

What do you mean by 'ice-free'? And does 'Arctic Ocean' refer to the bit in the middle, or the whole Arctic Circle?

If the question was 'when will the Central Arctic (Ocean) be 90% ice-free during Summer?' I might make an offer.

BTW: I'm with Bob Grumbine: I think 2009 is most likely to compare with 2008, may be slightly lower, probably won't exceed 2007; 2010 may well exceed 2007; I'll offer a 60% probablility on the latter, and a 75% on the former. £5 each way?

[Since I'm on the side of more ice than 2007, I'm not quite sure if we disagree. Are you offering to take the less-ice side in 2009 and 2010, at even odds for both? -W]

Kepp up the good work; I'm a bit too busy these days, but still watch with interest.

PS if you want a laugh which has nothing to do with this, check out 'willthetruth93's video on youtube; I chuckled. :)

No, we're on the same side re 2009. we can keep 2010 for later; it was really a generic offer, not especially aimed at you. Would you take evens on 2009's minimum being lower than 2008's, but not lower than 2007's?

[I'd certainly give you evens on A, where A= [ice(2008) >= ice(2009) >= ice(2007) ], and I get not-A. But I'm not sure you mean it; I think you mean just B=[ice(2009) <= ice(2008)]? You can have evens on that, too -W]

When they say "ice-free during summer", do they actually mean 'during' or do they mean 'at the end of'?

1. throughout the duration, continuance, or existence of: He lived in Florida during the winter.
2. at some time or point in the course of: They departed during the night.

By Alan Woods (not verified) on 18 Jan 2009 #permalink

Fergus, Woods;

I'm not sure what climatematters means by 'during' and 'ice free', but as a matter of record William's 'ice-free arctic' bet with both Joe Romm (USD 333, based on USD 1000 split 3 ways) and myself (EUR 200) was formally agreed as:

at no time between now and the end of the year 2020 will the minimum total Arctic Sea ice extent be less than 10% of the 1979-2000 average minimum annual Arctic Sea ice extent, as measured by NSIDC data or any other measurement mutually agreed-upon; provided, however, that if two or more volcanic eruptions with the energy level equal to or greater than the 1991 Mount Pinatubo shall occur between now and the end of 2020, then all bets are voided.

Being a notorious global warming denier, William is on the cold side of that bet (obviously).

[Ooooohhh. But yes, whatever may be waved around in the media glow, the measure everyone uses in practice is minimum extent during the summer. Min averaged over the month of Spetember would be statistically more stable -W]

At December's U.N. Global Warming conference in Poznan, Poland, 650 of the world's top climatologists stood up and said man-made global warming is a media generated myth without basis. Said climatologist Dr. David Gee, Chairman of the International Geological Congress, "For how many years must the planet cool before we begin to understand that the planet is not warming?"

I asked myself, why would such obviously smart guy say such a ridiculous thing? But it turns out he's right.

The earth's temperature peaked in 1998. It's been falling ever since; it dropped dramatically in 2007 and got worse in 2008, when temperatures touched 1980 levels.

Meanwhile, the University of Illinois' Arctic Climate Research Center released conclusive satellite photos showing that Arctic ice is back to 1979 levels. What's more, measurements of Antarctic ice now show that its accumulation is up 5 percent since 1980.

In other words, during what was supposed to be massive global warming, the biggest chunks of ice on earth grew larger.

Were Trent versions 1 thru 141 banned for trolling?

[Hank Roberts mode: on] Googling Trent's second sentence gets 468 hits, mainly from the rabid right in the US (Planet Gore, etc), and is a straight lift from this column in that well-known organ The Flint Journal. It appears to be text for the day for trolls.

Regardless of where Trent got his info, the facts are the facts. So rather than belittle him for adding to the debate maybe the points he raises should be discussed.

In regards to the "ice free during summer" statement i would consider it to mean when you can travel through the Arctic in anything less than an ice breaker. Now i would be prepared to bet all the tea in china that this will not happen this summer.

Of course in 1922 the Artic was "ice free" way before AGW existed can anyone give me a plausible explanation of why that was?


Regardless of where Trent got his info, the facts are the facts. So rather than belittle him for adding to the debate maybe the points he raises should be discussed.

Given that every one of his "facts" is wrong, where would you like us to start?

And you're wrong about 1922 as well.

Probably could have been more precise,

This fact is correct is it not:

"Meanwhile, the University of Illinois' Arctic Climate Research Center released conclusive satellite photos showing that Arctic ice is back to 1979 levels. What's more, measurements of Antarctic ice now show that its accumulation is up 5 percent since 1980."

Therefore if GW is such a big scarey thing why did the ice grow back when the IPCC predicted it would not? In fact the 30 year trend (79-08) shows approx zero deviation.

I have no wish to enter a debate on temp because it usually descends into a statistical argument so we can leave that one alone.

As far 1922 is concerned, did i get the dates wrong? did it not suffer a large summer melt bigger than 2006-07? Your thoughts on this would be nice as i asked in my previous post.

Arctic ice is not back to 1979 levels. You can check this very easily at Cryosphere Today.

Antarctic sea ice shows a very slight upward trend over the period since 1979, but it's not 5%. Again, CT has the detail.

And they don't deal in "satellite photos".

Check CT's graph of seasonal ice extent back to the beginning of the last century. It looks as though the summer minima in the early 1920s were around 10.5m km2 - more than double the last two years.

This stuff you quote is so wrong, and so easily shown to be wrong, that you ought to be asking why you've been told this nonsense.

Would you give 2/1 on 2010 < 2007 ?

[I'm not too keen on going too far out. I'm not trying to pretend I know whats going on here; I feel that 2007/8 may be "1998" but I have no kind of proof -W]

You mean these photos from CT?…

Not a huge difference between the two.

But yes ok i got it slightly wrong, slight misread and then quote, i should have said TOTAL sea ice not just arctic sorry about that. Anyway here is some info about that.

Here is a link of interest (in reference to ice free summer)…

there is a link to original report at the bottom.

[Roger has confused "the summer melt season lengthensâ with the difference between min/max; so has contributed nothing but noise -W]

So in summary my objection to the prospect of an ice free Arctic still stands as the above info will attest to.

I'll give 5-1 odds that the Arctic isn't "ice free" before 2030. If we're really entering into a pro-longed solar COOLING period as it seems we are, then we're going to see LOTS of cold temps and lots of ice build. The question I have is whether we will be more comparable in temps over the next 20 years to the 1960s or the 1800s?

[How about by 2020? -w]

5-1 odds against "ice free" before 2030! Interesting. 2030 is a long way off so if you win I don't have to pay up for a long time. If you lose your opponent might just collect rather earlier. I don't know if that has been factored into your odds calculation.

I think there is more than a 30% chance of arctic being "ice free" by 2030. So there would seem to be some difference of opinion. Probably not enough of a difference to bet on because of the timescale and default risk. I am not sure if William is hinting the timescale for this bet is too long or just pointing out his views or both. His views are against a rapid disappearance so you are more likely to get better odds against someone who believes in rapid disappearance. Professor Wieslaw Maslowski springs to mind but please consider other potential bettors not just William.

If you are willing to provide information about yourself (I am), would you consider reworking it to a shorter timescale?
Eg Jan 2009 to 2030 is 22 summers for 90% reduction. Would you offer odds against a 25% reduction from 2004-2008 five year average of the minimum five day extent occuring in any of the seven years 2009 to 2015?

Can anyone point to scientific case for CO2 causing climate change rather than global warming? Surely an insulating blanket which was claimed to be the mechanism for global warming should reduce rather than increase regional variation.

[An insulating blanket is a poor metaphor. As it happens, GW *is* predicted to reduce equator-to-pole differences, since the poles warm more. You should be aware that the claims of more weather extremes are often over-hyped; our weather will change somewhat, with the climate, and hence become more unexpected; but there is no great evidence for it becoming more variable -W]

I am also confused by the way the models appear to use actual fossil fuel consumption as the baseline for man's emissions which will be the gross figure but use the percentage of CO2 type gasses in the atmosphere which is clearly a net one.

[Which models? Climate models, the AOGCMs, are typically forced by actual (and projected) atmospheric CO2 levels. In turn, those projections are typically generated by other models, which may take economics into account. By "% of GHG in the atmos" I'm not sure exactly what you mean - do you mean the airbourne fraction? -W]

Constraints on future changes in climate and the hydrologic cycle
Myles R. Allen* & William J. Ingramâ

"Changes in precipitation extremes
Long-term mean precipitation is a useful summary indicator of the intensity of the hydrologic cycle, but details of the distribution of precipitation over time, including the peak intensity of precipitation events and duration of prolonged droughts, are likely to be the most important issues in determining impacts of precipitation changes9.
Although global-mean precipitation is primarily constrained by the energy budget, the heaviest rainfall events are likely to occur when effectively all the moisture in a volume of air is precipitated out, suggesting that the intensity of these events will increase with the availability of moisture7 (that is, significantly faster than the global mean). Thus we might expect the uppermost quantiles of the rainfall distribution to increase by about 6.5% per kelvin (ref. 44) if the ambient flows change (most likely at higher latitudes). In the tropics, where the flows leading to precipitation are themselves driven largely
by the latent heat released by precipitation, larger increases still might occur7. In particular, the maximum thermodynamically possible rainfall and winds in hurricanes are predicted to increase rapidly with warming56.
The red and purple curves in Fig. 4 show the magnitude of daily precipitation as a function of percentile of the precipitation distribution in one AOGCM under climate change and control conditions, respectively. The green curve (and right-hand axis) shows the ratio between them. At the highest end of the distribution (all tropical
cases), it appears to be converging to about a 25% increase, which is indeed slightly more than we would expect from the ClausiusâClapeyron relationship.
Because such increases are more than double the increase in
global-mean precipitation (that is, the change summed over all percentiles, which is constrained by the energy budget, not by the ClausiusâClapeyron relation), there must be decreases lower down the distribution. Indeed, the increase at the heaviest rain events is large enough that the energy constraint on the total implies that only on one day in ten does precipitation increase (the control and perturbed distributions cross around the 90th percentile in Fig. 4). It would be interesting to know how other models compare â if this were be another emergent constraint generic over all models in regions of interest, then we could use forecast temperature changes to constrain extreme as well as mean precipitation. Given the acknowledged difficulties in relying on model simulations of extreme events and in observing changes in extremes directly57, this could be a
powerful result. In particular, this convergence to a particular fractional increase, possibly related to the ClausiusâClapeyron limit, could even improve as we move to the highest percentiles, which are generally the least tractable under more direct approaches."

There is probably better papers to quote than that.

>[You should be aware that the claims of more weather extremes are often over-hyped; our weather will change somewhat, with the climate, and hence become more unexpected; but there is no great evidence for it becoming more variable -W]

Regarding temperature extremes and wind extremes, I have seen nothing to disagree with what you are saying. However with rainfall, I thought there was agreement that warmer atmosphere holds more water and due to some constraint (energy budget?) it was fairly clear that the heavy downpours are going to get heavier. Have I got that wrong, is this highly contested?

[I agree; and so does WJI, given the paper you've quoted above. But there is another way of looking at it: most stuff is approx normal, but rainfall isn't. If you look at extremes of its log distribution, which is approx normal, you may find its changes revert to what everything else does, ie the distribution just shifts -W]

See Barry Brook for a discussion on the current Aussie heatwave and weather extremes here.

With respect to water vapour, having more in the atmosphere doesn't simply mean more available to fall as rain, it is also the primary "fuel" for weather systems, so there should be some expectation for an intensification of some kinds of weather event (but that's an unreferenced arm wave, sorry...).

Have I understood what you are saying William because it seems bizarre.

A shift in the log distribution (without increasing its variability) would mean more heavy downpours and greater probability of no rainfall leading to more floods and more droughts. In laymans terms these increasing impacts are an increase in variability and on an ordinary frequency distribution plot would be an increase in variability.

Now merely because you prefer to look at more normal shaped distributions you want to convert to a log distribution which allows you to say anti-alarmist things like 'claims of more weather extremes are often over-hyped'.

I must be misunderstanding because to me that seems more like I want to say anti-alarmist things so I will start making adjustments to distributions to suit me rather than demonstrating impacts so that might it might allow me to get away with saying those anti-alarmist things.

[It depends by what you mean by "more extremes". If you take temperature, then certainly we expect more "extreme" high temperatures, as measured by the distribution we're used to nowadays. But we don't expect more "extremes" as measured by the same distribution, with the mean shifted.

In the case of ppn, thats not true. But arguably thats because the ppn distribution is so obviously non-normal. I'm saying that if you transform it to a normal distribution, then maybe it is so (I haven't done the calculations) -W]

If you haven't done the calculations then I would doubt it. I am judging this by "the increase at the heaviest rain events is large enough that the energy constraint on the total implies that only on one day in ten does precipitation increase" quoted earlier. Other models might show different figures to this 90% of days with less precipitation and 10% of days with (lots) more. Even so, how would anything like this not be an increase in variability even on a log distribution?

IMHO, just because the ppn distribution is so obviously non-normal should not allow you to ignore ppn or handwave it away with some weird log conversion that you want to do to make the distribution more normal, even if that does have the effect you suggest it might but I doubt.