Naive Predictions of 2013 Sea Ice

These predictions are naive in the sense that they are not based on a physical model, nor other measurements apart from the 30-odd year history of the index in question. Moreover, they are made a year in advance as winter freeze-up is just starting. The predictions are simply If ... Then statements: If trends from the recent past continue ... Then we could expect this much ice next September. Those past trends appear reasonably well characterized by Gompertz curves, fit by nonlinear least squares to the data. Error bands in Figure 1 shade a range plus or minus twice the standard deviation of the residuals (observed variation around the curve).

Figure 1

The resulting naive prediction for mean Arctic sea ice volume in September 2013 is 3,100 km³, plus or minus 1,900. Extrapolated beyond 2013, again as an If ... Then exercise, the curve drops below a "virtually ice free" threshold  of 2,000 km³ by 2015, and approaches zero volume after 2020 -- with uncertainties that could go either way, of course.

All indicators of Arctic sea ice reached satellite-era low points in September 2012, below the predictions of most experts. Sea ice volume, estimated by the PIOMAS team, fell more steeply than ice area or extent. Although the two-dimensional indexes, area and extent, declined less steeply, the rapid volume decline implies thinning that could allow area and extent to drop precipitously in the future. If so, the trend-extrapolation approach would fail abruptly for these measures. With that important caveat, Figure 2 shows a model for NSIDC September mean sea ice extent, which is the target of SEARCH Sea Ice Outlook estimates.

Figure 2

Daily updates in Cryosphere Today sea ice area receive close attention during the melt season. In an earlier analysis I was surprised to find that the 1-day minimum area seemed no more unpredictable than the September mean. In that spirit, Figure 3 offers a prediction of CT 1-day minimum area.

Figure 3

Historical Notes

The predecessor to this post, Naive Predictions of 2012 Sea Ice, offered similarly based predictions for September 2012. Observed 2012 values in every case fell below the central prediction, though still within error bounds:

                Sep 2012         Sep 2012
                Predicted        Observed
NSIDC extent    4.3 (3.4-5.1)    3.6
NSIDC area      3.0 (2.2-5.1)    2.4
PIOMAS volume   4.0 (2.0-5.9)    3.4
CT area 1-day   2.7 (2.1-3.3)    2.2

An even earlier post from April 6, 2011, applied Gompertz curves to make naive (although surprisingly close) predictions about September 2011 ice conditions:

                 Sep 2011        Sep 2011
                 Predicted       Observed
NSIDC extent     4.4             4.6
U Bremen extent  4.6             4.6
NSIDC area       3.1             3.2
PIOMAS volume    4.9             4.2

When I wrote that April 2011 post (which became my first SEARCH entry), the idea of applying Gompertz curves to sea ice had been loose in my head for a while, but I was not aware of other publications taking this approach. It seems an obvious idea, however, and I would welcome links to earlier reports.

Comments

[dabize]

Well, at least you have an actual track record of predictions that match reality, which puts you ahead of all of the modelers.

As you say this is naive - which is (apparently) a plus. It implies that the models which keep the summer SI until 2040 are missing something critical.

I must say that I still don't know where the negative feedback item implied by a Gompertz curve would come from, since the reduction in all of these parameters should be amplified (rather than damped) toward the end by the increasing exposure of seawater in the highest insolation period.

So the fact that a Gompertz curve fits the facts best so far is bemusing

[Llosmith57]

Larry, The uncertainties take into account many possibilities and include my own numbers for 2013 without any extra energy inputs from the Pacific if an El Nino materializes. Also, what would the 2013 maximums look like using the Gompertz function?

Do you think that the Exponential function is worth computing? I am uncertain what to think about the current state of affairs in the Arctic. There seems to be at least some element of chaos from unexpected mixing at depth of mid level warmth and low levels of snow pack just about everywhere so an exponential function seems to be a possible alternative.

Excluding the possible but unlikely venting of CO2 or Methane is there a way to assign some statistical chance of a sudden collapse within the Gompertz or Exponential functions?

[Ggelsrinc]

I can think of a couple of things right off that may be considered negative feedback to total melting. We all know how the sea ice has a long history to drift towards northern Greenland and the CAA. It also tends to be in a pack towards the minimum, because alone it doesn't have much of a chance to survive. Sea ice tends to be destroyed along the edge, so as it gets smaller so does it's "circumference". Perhaps the CAA could continue to protect it, but I think those straits will open up, allowing sea ice to escape.

Both of these factors could be considered a negative feedback against total sea ice destruction. The problem I have with that analysis is, when sea ice extent is that low, will the status quo still prevail in it's drift, which is basically wind direction, or will that open ocean cause a storm to push the sea ice away from it's historic nesting place?

[Twemoran]

Neven

Are lower categories possible for the poll? I thought that volume was actually falling a little more slowly than area or extent this last year.

Terry

[Robertscribbler.wordpress.com]

I think it is fair to say that all predictions are naive at this point. We simply have more and less naive. And we could certainly do worse than what Mr. Hamilton has done. I'd say those predicting ice-free or near ice free conditions after 2040 are the most naive of the bunch.

For my part, I'll be watching this year's refreeze like a hawk. There seems to be a lot of northward heat transport at the moment. Will have to keep an eye out.

[Bob Wallace]

" will that open ocean cause a storm to push the sea ice away from it's historic nesting place?"

The real estate below the 'historic nesting place' is not ice-clogged in the late melt season any longer. There are numerous channels among the Queen Elizabeth Island and on each side of Ellsmere Island which are now melting out and allowing Central Basin sea ice to escape into warmer weather.

Then, when wind directions assist, some of that thick ice gets transported out the Fram Straight to the east of Greenland. As the overall extent decreases there will be more and more available space for that thicker ice to travel through the Fram.

With the overall ice being thinner along with storms and wave action stronger the pack ice is likely to get broken into smaller pieces which will be easier to push through the various openings.

Once the ice gets in among the islands it's likely meet some warmer than historical temperatures as the land is loosing its snow cover and absorbing more sunlight.

[Yuha]

There are two obvious negative feedbacks:

1. As the ice extent shrinks, more of the extra heat in the arctic goes into warming the waters rather than melting the ice. Note that this is a negative feedback with respect to ice loss but not with respect to arctic warming.
   
2. As the arctic warms, the temperature difference between the arctic and lower latitudes shrinks. This tends to reduce the net heat flux into the arctic. Note again that this is a negative feedback with respect to arctic warming but not with respect to global warming.
   

These are real negative feedbacks but I'm not sure about their timing and strength. They might be overwhelmed by other effects including new kinds of positive feedbacks like storm damage to thinning ice, but I would not be surprised to see some kind of a Gompertz like tail.

[Ggelsrinc]

Bob Wallace

I spend a lot of time arguing with the Denialistas and reading the science. I tend to agree that the historic rest area will be the last stand for the sea ice and I don't expect much negative feedback from the CAA. Predictions that would be right on the money are still naive if someone thinks they are scientifically acceptable. I can make a safe prediction that the arctic sea ice will be gone before we can make a good model.

The brain is a lot like a computer and perhaps it's just as good at making a reasonable judgment than any scientific model we have. I've inputted most of the factors commonly talked about on this site, but there are a couple that aren't mentioned much. One of the things I really noticed this year was changes in salinity. The factor that I find most overlooked is the changes in snow cover in the NH. It would be nice if we had better data to determine how much tundra, wetlands and barren land are involved, because those areas are closer.

We lost nearly 6 million sq km of snow cover this year and broke the record in June. That's almost 3 Greenlands or 3 arctic sea ice minimums and it's at lower latitudes. The trend for many years has been for less snow cover and I believe it's big enough to be a major factor. It's hard to know the actual volume of the sea ice based on PIOMAS, but considering the 2007 and 2010 melts, it's possible under the perfect storm scenario for the arctic to be ice free next year, but it's unlikely. If the current trends hold for snow cover, my naive prediction would be 2015.

[Chuck Yokota]

Speaking toward Yuha's two proposed negative feedbacks:

1. How much warming can the Arctic Ocean waters take before the halocline breaks down, and the warmer deep waters mix with the surface? When that happens, wouldn't that be a huge positive feedback toward melting sea ice in the central basin? Could the shallow seas on the periphery still freeze seasonally?

2. Do the climate models offer any guidance about how much the heat flow would reduce with a warmer Arctic? How would the slowdown of heat flow be divided between the atmosphere and the water? What would that look like in terms of changes in weather patterns? Would the change be gradual or would there be wild oscillations?

I pose these thoughts as questions, as I am a novice at this. I confess I had not paid enough attention to climate change until this summer's amazing melt. That came like a kick in the teeth, and I realized we were falling off a climate change cliff right now. I thank Neven and all of you for the wealth of information, discussion, and speculation provided by this site.

[NLPatents]

Interesting.

How about with exponential melting of ice volume? That seems to be the better predictor. Problem with gompertz, is that 0 ice is not a true 0, water keeps getting warmer after it melts.

[Seke Rob]

Posted this in the Snow thread over at Tamino's

http://tamino.wordpress.com/2012/10/05/snow-2/#comment-71099

The animation is now in 3rd generation, having removed Greenland which so oddly distorts the rest of the hemisphere data and anomalies [10+ meters thickness is considered snow covered] and slowed down the frame speed at end to try allow a little focus on where the Apr-Sep period is heading.

Carefully consider that if there's 5% missing in February on 45-50 million, there's a good chance the spring/summer also shows a shortage. Where it's in the last 5 years getting ugly is, that more winter-snow [5% more vapor going around than 3-4 decades ago] it is not helping to stave off summer melt out with the new July record set at 86% anomaly [though it's *only* 2 million off from what there's supposed to be]. Think farmers have been seeing deeper and deeper reduction of soil moisture. One year they could survive, and plants still able to get some, but multiple years and you'll have disaster as was seen in the corn belt (Last I heard from someone in Illinois a week or so ago, it was still dry]

Chart added to the Snow page on the ATCC site [see daily graphs like]

[LRC]

Your heat flux coming into the Arctic is penetrating deeper as there is far less ice to moderate it. You have weather systems that swing from extreme to extreme more often but at the same time stay in the extreme mode for longer periods of time. On top of that even small weather changes are starting to make exponentially far greater impact on the ice then ever before. All this makes predicting how much ice will be left very difficult.
I fall in the exponential rather then Gompertz curve camp. Having lived in the Maritimes you could find ice holding out for a very long time only if it was along a quiet stream deep in the forest very well protected from everything. In other words perfect circumstances. In the Arctic I do not see anywhere left that can give you that protection. The CAA has been smashed to pieces. The MYI that is left is so rotten that almost any boat can brake it up. Swells brake up large flows 100's of mills from the edge. Where is your protection from the last bit of ice? Greenland? Once the open ocean is free that will go fast because there will be non left to replenish what is being sent into the Atlantic.

[Protege Cuajimalpa]

I was looking to see your new set of graphs, Larry. Thanks for sharing them with us.

[P-maker]

LRC,Rob,NLP,Chuck & Neven,

I suggest that this thread changes into:

Evidence-based projections of the Arctic in 2013

As you have all pointed out, snow melt, marine and atmospheric advection, albedo changes and many other factors are at play. Analysing any of these - or any combination - is not for the naive.

I could also offer a few new thread titles:

1) Will the IPCC report, or will the Arctic sea ice disappear first?

2) Hidden declines at a glance

3) World Championships in misleading quotes

The first race is self evident, since the IPCC has passed it's cut-off date for scientific publications already, and all new evidence about vanishing sea ice will in fact be new.

The second headline covers a number of known unknowns in the Arctic system: Hot water under the halocline, Methane under the seabed, frozen Carbon in the thawing permafrost or burning dead forests in the NH mid-latitudes.

The third WC-headline addresses the FRAUD-crowd (Fact-Resistant Anti-University Denialists). I suggest a simple compilation of outrageous statements backed up by plain statistics on misleading use of quotes from scientists. Some media outlets may have a head start, but eventually this indicator will decline faster than the Arctic Ice Volume, I hope.

[Bruce Worden]

I, too, fall in the exponential rather than Gompertz camp. My reasoning is simple: The ice has been thinning. As the (surface area) / (volume) ratio increases, the ice will melt faster. This is borne out by the acceleration in volume loss over the past couple of decades as seen in your first figure (vis a vis the area/extent losses).

I see no reason for the volume loss to decelerate without a significant change in climate/weather.

P.S. It would be helpful if you were to make the x-axes of your plots cover the same time period and extend the area and extent projections to cover the time out to 2025. If volume drops to (near) zero while there is still significant area/extent, that will suggest that the projections are not consistent and need some revision.

[Aaron Lewis]

Melt water from the ice tends to be colder and more dense than the sun warmed water outside the ice pack.

The cold melt water of the ice pack sinks as a plume, drawing the sun warmed water into the edge of the ice pack. (And a volume of much warmer surface water being pulled into the ice will result in localized erosion of the ice pack.)

This circulation facilitated by more clearance between the bottom of the ice and the halocline.

This means there is more local interaction between the water, sea ice, and the atmosphere than has been captured in the general circulation climate models.

[L. Hamilton]

Our "what shape?" discussion focuses on the end game, which of course has importance. As you can guess from Figure 1, however, the Gompertz does not become distinguishable from other steep curves until we're below 2,000 cubic km or so -- in a transition that could take just a few years. There's no historical experience telling us exactly how that will happen, and weather could be a complication as well.

One thing I liked about Gompertz curves in the first place is what they do at the other end. In contrast to linear, exponential, quadratic or polynomial alternatives, a Gompertz does not have to start at 1979 to look reasonable. For example, here's mean September extent back to 1972, with both 1972-2012 and 1979-2012 Gompertz curves drawn in. They appear to be identical, although numerically they do differ slightly.

http://img.photobucket.com/albums/v224/Chiloe/Climate/NU_gompertz.png

[Syddbridges]

Whichever way you look at it, Gompertz or exponential, it's only a few years differet. For someone of my age, early 60s, it might mean I can avoid the consequences if a slower decline occurs. Hiwever, I suspect I will see the Age of Consequences. Even if I don't, my 4 month old grand niece will, no matter what the final death curve is. A few years in an interglacial that has lasted 10,000 years, is of lttle longterm sgnficance.

An expnential decline would match ever-increasing open water in the summer. With more warmth likely to be advected into the Arctic, I thnk a slowdown less likely, though I hope I'm wrong.

So I expect exponential decline but hope for Gompertz or slower. Given a choice between what I expect and what Watts claims, I would prefer the latter to be correct. Sadly, I learned as a child that reality always trumps fantasy

[Jim Williams]

I will say one point in the Gompertz' favor -- it's got a big increase in ice flow from the GIS covered.

[Protege Cuajimalpa]

Regarding negative feedbacks, I agree with Jim. That is:
1. We have warmer summers at the Arctic and there is a huge inland ice melt, specially at Greenland
2. Some of the water enters the Arctic Ocean at summer, but the other portion reach the Arctic in Autumn, so we have new thick floes, made of fresh water coming from inland melted ice.
3. Some glaciers collapsed and bring icebergs to the Arctic Ocean

So at the end, the Gompers curve reflects the influence of the inland ice melting, especially the new ice that forms north of Greenland.

Would you agree?

[johnm33]

I lean towards exponential, but if we have a fairly normal winter Gompertz seems plausible. If however we get massive [30ft] snowfalls around the great lakes lattitudes, or the alps in europe, or even floods in the saharan highlands we'll know the age of consequences is upon us.

[Bob Wallace]

One negative feedback I don't see addressed is smoke from massive fires.

I get my power from solar panels. A few days ago we had a layer of smoke-filled air blowing in from about a thousand miles away - a large wildfire in Idaho. That smoke drastically cut the output from my panels, it was clearly blocking incoming sunlight.

We're thawing out and drying out large areas of permafrost. Alaska had a large permafrost fire last year. Siberia experienced massive fires this year. We could see imported smoke slowing melt rates.

[R. Gates]

Either Gompertz or exponential...it all heads to the same 2015-2017 window to be virtually ice free, and after that, isn't it truly academic whether little pieces of the former Arctic ice cap manage to hide in the CA or Greenland area for a few years, to await their eventual melt? The change in weather patterns and the increase in other positive feedback effects are already in full swing, with the rather puny negative feed backs only of academic interest.

What the ultimate natural negative feedback for these levels of greenhouse gases (the source of all this melt, afterall) is to reduce the concentration of these gases, and the ultimate natural way to do this is through the rock-carbon cycle, specifically through the acceleration of the hydrological cycle and increased weathering of rock, which aturally sequesters CO2 from the atmosphere, washing it downstream to the ocean and turning it into safe limestone.

Unfortunately, nature operates on much longer time frames than the sudden rapid pulse of anthropogenic CO2, methane, and N2O to the atmosphere can accommodate. The feedback processes are overwhelmed, and the greenhouse gases continue to grow. This overwhelming of the natural negative feeback processes is the source of this Arctic tipping point we've seen.

So while we currently have the luxury of discussing Gompertz vs. exponential ends to our wonderful Arctic (and make no mistake, I find it interesting as well), what should be clear is that natural negative feedbacks no longer make much of a difference, as the system has become overwhelmed and unstable as positive feedbacks have led to a new Arctic regime already.

We are, as it were, simply sitting by the patient's bedside, waiting and wondering when the final breath will be exhaled, and wondering if it will be midnight or 12:02 a.m.

And God help us, really, if some of the extreme Arctic prognosticators are correct and the methane genie is starting to arise from her formerly frozen lamp. http://www.atmos-chem-phys-discuss.net/12/26477/2012/acpd-12-26477-2012.pdf

[Ron Broberg]

Another naive prediction, this one from 2010
http://hot-topic.co.nz/five-years-threnody-for-arctic-sea-ice/

[Ggelsrinc]

R. Gates

The ocean is the frontline in carbon capture, but it's been overwhelmed. The ultimate natural negative feedback are clouds, as evidenced by a 22 degree C Earth, with only one other possible negative feedback to keep us from getting there and that is mankind.

At some point with our continents in their present position and with a certain amount of greenhouse gases, the Earth will keep warming to Hothouse Earth.

[Arcticio]

Seasonal volume loss was mostly in the 16 tkm³ ballpark and increased slightly to 18 tkm³ in recent years. Relative loss this year was 85% - another record.

http://www.arctic.io/2012/7/yearly-sea-ice-volume-loss

I would translate that into a quite constant amount of energy available every year to melt ice. Looking at the Gompertz fitting I wonder what this energy does once ice gets rare in the Arctic. Apparently it's neither heating the ocean nor melting the ice.

I think, which ever of the upcoming Winters fails to build ~17 tkm³, next Summer the Arctic is ice free. And the only good looking fit is just a vertical line.

[Wayne Kernochan]

I'll just delurk to make a few points, some repetitions of past years. First, Gompertz seems to me a superior way of projecting volume, as it handles the idea that ice thickness varies in somewhat of a normal fashion around the average. Thus, an exponential decrease in thickness year to year would evidence as a slower decrease in volume until the average is approached, after which there is a sharp drop followed by a multi-year (but not decadal) tail. In plain English, we're on the steep part of the volume slide, with thicknesses even in the CAB of a meter or less at minimum, but there's enough variance that perhaps 10% of the thickest ice can last for another few years.

However, it seems plain to me that a similar Gompertz curve just doesn't capture the fact that in this model, area and extent are functions of volume. They will tend to stay flatter longer, and then dive more sharply. I find the idea that both will increase next year possible but unlikely. How do you get a fair amount less volume and yet more area and extent once you reach this point?

It seems to me that a key test of the idea that area and extent will increase is what is happening right now during refreeze. A "bankable store" of energy from this year in the water should keep the area anomaly from shrinking rapidly -- else you would expect air temperatures below freezing and roughly comparable to last year to freeze up the ice at the same times and the same places as last year -- and we could expect greater area and extent as part of normal variation. So far, that doesn't seem to be happening at all. Am I wrong, or are both the NW and NE Passages open later in the year than ever before? What's more, the NE Passage at least seems nowhere near freezing.

I would note, in passing, that the idea that the Antarctic ice area can vary widely at maximum seems to be proving out -- a temporary bulge seems to be going away rapidly, and the anomaly has gone from 1.16 to 0.45 in nine days. If both trends continue, we should be approaching record combined territory iirc in about 6 days.

Anyway, I'll stick my neck out again and say that I expect volume to go below 3 mkm3, area to go below 2 mkm2, and extent below 3.2 mkm2 at minimum in 2013. Naive as ever ...

[Seke Rob]

"Record combined territory in about 6 days"

Do these chart bare witness of that in extent and area expression?

http://bit.ly/NSGL01
http://bit.ly/CTGB01

Were the words "high anomaly" maybe missing in that sentence?

Seeing some contradictory information on why Antarctic extent was so large this year, to summary for myself:

1. The Antartic Vortex strenthened [ozone hole origin], which we've known for longer.
2. More mass balance loss [to include increased calving] which adds to extend, noting that the Antarctic is deemed to be a vapor > snow buffer, containing SLR by as much as 80% of what it could have been.
3. More melt water from 2., lowering surface salinity, raising melt point by about 2C.

The circumpolar current ripping [extent apart due it's force] or keeping extent together, my reading was that it goes by 2 clicks / hour to 2000 meters depth. I'm not sure of it's "overwhelming" influence on raising or containing the Antarctic SIE as is postulated here and there. Am simply not well read up on the matter.

[Seke Rob]

For the spellcheckers: "strengthened" and extent and "to summarize"for myself". Point 1. was recently repeated by Dr. Mark Serreze, as having the function of stronger winds pushing out the extent, spreading it thinner [though the area-extent ratio would not support that]. NSIDC reports proportionally more area for September. That more vapor translates to more snow over the Antarctic, and being less prone to "summer" melt off i.e. somewhere ice the [land] ice sheet is thickening I've got as yet no doubt.

[Misfratz.wordpress.com]

Observed 2012 values in every case fell below the central prediction

Larry, thanks for presenting these statistical predictions. I find it interesting that the 2012 values were below those predicted. This can mean one of two things:

1. Overall, the weather of the melting season was more favourable for sea-ice melt than on average, so that 2012 is below trend.

2. The trend is increasing at a faster rate than predicted by the Gompertz curve, so that a weather-neutral year like 2012 ends up below the estimated trend.

I know there are papers in the literature that attempt to decompose the contribution to sea-ice melt from the weather, and I wonder whether the metrics that they use - based on wind patterns, etc - would help to determine what effect the weather had on the melt this year.

The Great Arctic Cyclone is interesting to consider in this context, because it likely would not have had anywhere near as large an impact if it had occurred during the 1980s, because the thicker sea-ice at that time would have been less susceptible to being churned-up. Also, one might consider the very existence of such an anomalously strong Arctic Cyclone as itself a feedback resulting from the very low sea-ice extent.

This tentatively leads me to suggest that #2 may be closer to the mark, but really I'm just guessing.

I have one direct question for you. Is it possible to calculate the probability that PIOMAS volume in 2013 will be below 2,000 cubic km?

Obviously, one can tell that it is somewhere in the range 5% - 45%, but it would be interesting to know where.

[Remko Kampen]

The thing I don't get in those graphs like Gompertz' is the tapering off of the melt in the last phase/2 M km^2.

The reality hitherto is simply sudden drop en well 'gravity' will have the curve going down virtually perpendicular. End of ice 2017 at the very latest.

Of course 'gravity' is a metaphor, bad one even because the curve would be parabolic. Not (hyper-)exponential as is the real case. The theory involved is catastrophe theory.
http://en.wikipedia.org/wiki/Catastrophe_theory

[Wipneus]

How do you get a fair amount less volume and yet more area and extent once you reach this point?

This is the point of course. This prediction holds the implied assumptions that 2012's low extent and area are statistical lows by chance.
In other words caused by weather more than trend.

If you don't believe that, perhaps by knowledge that weather was not that favorable to melting, you should reject the model.

Not just this model of course, I am very curious to know the answer:
Was weather favorable or unfavorable to 2012 extreme declines in area/extent?

[Glacierchange.wordpress.com]

Major calving event on Thwaites Glacier, Antarctica.

[johnm33]

This map shows just how much grounded ice there is to melt out down south.
http://en.wikipedia.org/wiki/File:AntarcticaRockSurface.jpg
when compared to the march extent on this page http://nsidc.org/cryosphere/seaice/characteristics/difference.html
I assume that the extent is enhanced[?] by the undermining of ice sheets rendering them shelves so consider the growth in extent just as concerning as arctic ice loss.

[johnm33]

I was going to add that we'd be lucky to see these shelves last 'til 2023/5 the next extreme of the moons cycle, and a miracle for them to survive after that, but decided not to until i saw the above.
This taking place now presages an interesting summer for the west antarctic peninsular summer melt.

[Robertscribbler.wordpress.com]

My own naive prediction is based on percent totalling of exponential trends (I believe the exponential melt will bear out and that we don't have much in the way of negative feedbacks).

The analysis is based on the fact that we have both a growing forcing (increasing CO2 levels) and amplifying feedbacks (albedo loss, methane increase etc). So the scale is heavily weighted toward the exponential.

The idea is that we have a 10% chance of no sea ice at the end of 2013 (Summer). This chance may increase if refreeze this year is weak (About a 50% possibility).

By 2015, the chance of no sea ice end year is up to around 35% (again, this will depend on how the ice responds over the coming years). In general, I think this estimate is conservative and may need revision.

By 2018, I'm putting the estimate for no sea ice at end summer by 50%. This hits the mid-line for the Gompertz estimate and, is probably the best current estimate.

In the out years, I'm estimating that it is almost certain, 90% probability, that all the sea ice will be gone before 2030.

What's concerning is the high potential for early melt out. And though 10% by next year doesn't look that high, it really is quite concerning that there's even a possibility of total melt out by next year. The 35% confidence for 2015 is very high as well and is also a huge concern. In all, my view is the most likely year is around 2018. But I'm very concerned about an earlier melt given some rather scarey ongoing melt factors.

[Chris Reynolds]

Wipneus,

Weather?

The Arctic Dipole index was low for June, although not as low as in previous years graph. Nonetheless CT area anomalies still showed record low values from June onwards, graph. As shown in that plot, it was only later that the August storm lead to the rapid drop in anomalies that characterised the record low. Even without that storm the anomalies indicated a new record would have been set, although not as extreme as with the August storm.

That said there was a clear role for thinner ice. Kaleschke's SMOS suggests thinning from 2011 to 2012, as do the extremely low concentrations over the Arctic this summer.

I suspect that the ice is now critically thin, and that regardless of forseeable weather, given the great likelihood of a continuation of the AD and Greenland High anomaly, we will see thinness being the driving factor from now on. This is why I think we're now in the early stages of a Rapid Ice Loss Event (RILE), one that I suspect could end in the loss of summer sea ice. I don't expect a re-run of 2007 - a drop to a new low baseline and stabilisation of area/extent for some years after - precisely because I think the ice is critically thin.

As for the tail of the Gompertz projection: Gompertz is a logistic function that has a sigmoid form. So a tail is to be expected. Exponential crashes to zero. Projections using either function tell us nothing about the real physical processes.

P-Maker suggested reasoned discussion of the likely behaviour for the future. The reason I've not chipped in on this discussion is that I cannot now see further than the next twelve months. All I can do is restate the milestones I laid out in this blog post, bullet points at close of post.

After that the crystal ball is very murky.

[Paddy]

"As for the tail of the Gompertz projection: Gompertz is a logistic function that has a sigmoid form. So a tail is to be expected. Exponential crashes to zero. Projections using either function tell us nothing about the real physical processes."

This. Also, the ice following such a projection at one stage in its decline does not mean it will continue such a projection in future stages in its decline when conditions are altered by the absence of ice.

[Anthonywobrien]

Why exponential rather than linear?
Because that is what is happening. The trend is so clearly non linear.

Why Gompertz? Because the refreeze starts well before the melting has finished, the refreeze is fast in extent and as the minimums get ever lower the refreeze gets faster.

If this is correct then we may see the minimums get even lower, but also get earlier. The eventual zero sea ice could be Early August.

The thinning of the ice will be even faster than the loss of extent. Huge areas of very thin ice are possible in the dark. As someone else here has already noted that is what the trends tell us.

In truth I am only guessing and know far less than many here. But do think of the geometry. (I can see no mechanism to discount an exponential fit of the July figures.)

[NLPatents]

I stand my my prediction of first ice free day of August 24, 2016. By ice free, I'm saying less than 250km^3 ice per PIOMAS.

I'm backing the exponential curve for August, but just taking the guess that it doesn't quite make it until late in the month.

I'm convinced by Anthony, that the ice free in September will be very tough given that some ice will likely start to reform somewhere in that month.

[Aaron Lewis]

If we had a statistical population of ice packs, then then there would be a PDF for the melt date of one of those ice packs, and that PDF would have a Gompertz shape.

However, we have one Arctic ice pack and time to melt that ice pack depends only on the amount of heat in the Arctic and how fast that heat can be transferred to the ice. Rate of heat transfer to the ice depends on the ratio of surface area of the ice to the volume of the ice and the effective temperature differential.

Since the ice is mechanically weak and is likely to be subjected to mechanical stress from storms, we can expect the ice to break up into smaller pieces and the ratio of surface area to volume will sky rocket.

Thus, the volume of Arctic sea ice will approach zero as an exponential function.

Nevertheless, the plots of Gompertz are extremely valuable in considering the variability of weather and thus the variablilty in heat transfer. However, it tells us how the ice would behave under past weather conditions, rather than the weather we expect next year. (Warmer!)

[Kevin McKinney]

"If this is correct then we may see the minimums get even lower, but also get earlier. The eventual zero sea ice could be Early August."

No. Refreeze won't happen any earlier, even if, as seems possible, it may become more rapid in the sense that larger areas are available to freeze (and do so.)

While I have no doubt that early August metrics will continue to decline, I'm very confident indeed that melting will not somehow cease earlier--why would it? Temps will be warmer, insolation will be greater, and the ice will be just as thin, if not thinner.

What I expect is that the slope of the refreeze line will become steeper, and the bottom of the 'melt trough' in the various graphs will widen as well as deepen. The annual cycle will look much less sine-like, I think, even before the ice melts out completely for the first time.

[Kevin McKinney]

Hm. My comment looks as though I have a bad HTML tag in there, except I don't remember using one...

[Kevin McKinney]

...maybe this will help.

[Kevin McKinney]

HTML test...

[Kevin McKinney]

...good.

[LRC]

Farther to Kevin's comment. Early refreeze needs ice nodes to build up from. The more scattered the ice becomes the more ices nodes you have the greater area the early freeze can work from. Now when you get near 0 area the early freeze has very little to work with so that freeze up must then wait until things get even colder then before. Add to that the likelihood that the surface water salinity levels will become higher things have to become even colder. On top of that will be the higher probability of more frequent violent storms. So IMO the opposite will happen. Freeze up will begin latter, but once it does take hold will progress much faster. Although the ice at the maxima will be very thin and broken up.

[Jim Williams]

LRC: "Freeze up will begin latter, but once it does take hold will progress much faster. "

Why much faster? I can see slightly faster, though I think the cloud cover will soon prevent that, but the only negative feedback I see is that open water isn't protected from the cold night air, and positive feedbacks of several kinds will lead to warmer waters needing to be cooled.

If the thaw is late in the Summer then a normal looking refreeze is possible. If the thaw is early in the Summer I don't see how the ice can even largely recover in extent. The start up of refreeze will cause massive downwelling of salty water in an already weakened fresh water lens, with the churning bringing up warmer AW water.

[Anthony]

Perhaps the melt/refreeze history of Hudson Bay, Great Slave Lake, Lake Baikal etc.. . might shed some light on the subject. Hudson Bay shouldn't be so hard to create.

[dabize]

Anthony, here's a link to the AmWx thread on Hudson Bay freeze-up last year

http://www.americanwx.com/bb/index.php/topic/26831-hudson-bay-winter-of-2011-2012-tracking-thread/page__st__245

There are also links to Environment Canada's ice cover sites for HB that you might find useful

[Anthonywobrien]

Perhaps I should have made myself clearer. When I say the refreeze will start earlier, this will happen when what is now late melting ice has gone earlier.

Presently the freeze up begins long before the melt has finished. As the ice melts out earlier further and further north, eventually the freeze will overtake the melt earlier.

In any one position the freeze up will not be earlier.

Remember the Arctic is huge and goes right to 90 degrees north, Although the North Pole sunset is not until September 24 it is so low on the horizon long before that that even tiny bumps of ice cast a huge shadow. So while the bigger lumps of ice may be still melting, large areas of surrounding sea are freezing.

Towards the poles it is surfaces that tend towards the vertical that receive the most light.

[Ggelsrinc]

I'm not trying to go off topic, but has anyone discussed the possibility of man making sea ice?

Let me stay on topic to state, I believe the arctic sea ice will collapse exponentially and will do so by 2015. I agree with a lot of the analysis of what can follow, although cloud cover isn't discussed much when it comes to warming ice free areas and I've never found good data on the subject.

I've been fascinated with sea ice for many years and I like all the subjects pertaining to the dynamic processes of it forming and melting. What I was wondering is, since man caused this problem, is it possible he can solve it? Has anyone ever discussed a way to stop sea ice from continually reaching a record minimum, even with our present greenhouse gases, which I surely think need to be reduced?

Reading what many have posted lately gave me an idea I had a long time ago to solve a different problem with stagnant water, using windpower to power air pumps and aerate it. What I was wondering is, is it possible to drill holes through the sea ice and thicken it in winter, using windpower to pump the water? I understand how MYI is not only thicker, but has lost most of the salt, but can MYI be mimicked well enough to last a season and many more seasons, as it loses salt?

I have no delusions that rescuing the arctic sea ice would be an easy task, but what I was wondering is, do you think it's possible, or would the salt water just rotten the ice?

[Jim Williams]

Ogelsrinc, I think the only even slightly plausible approach to saving the sea ice would be to install solar panels in the Arctic Ocean covering it. (No, I don't think it would work.)

[Anthonywobrien]

The sea ice is going and soon we will see a zero minimum. Even after this has happened we may still be arguing over what is the best fit and it may well be, that the answer is not at all clear.

What will be clear is that there will be consequences.

I am not at all sure that a Gompertz curve is the best fit. But I suspect that it may be. Regardless after we see a zero minimum a zero month and then a zero season will not be far behind.

Perhaps even more frightening is that a zero maximum may only be a few decades away. I do not want to believe that, but look at the fit.

[Lynn Shwadchuck]

I don't want to believe a zero maximum either, but think of the methane plumes coming up from the Laptev Sea and how much worse that will get as the Arctic Ocean is black longer and longer. It's going to be a pressure-cooker up there, relatively speaking.

[Chris Reynolds]

Gelsrinc,

To get your head around the figures go to an open area of fields with a map, start off at some landmark like a church steeple and walk 1km ou from it. Then imagine a square with sides of that length to get an idea of a square kilometre. Once you've got your head around that imagine going up a kilometre to make that a cube.

The Arctic Ocean is 13,990,000 squared kilometres.

Current peak volume for PIOMAS is around 25,000 cubed kilometres - the bulk of which is in the Arctic Ocean.

Sorry but there is no attainable option for stopping the process in operation.

[Chuck Yokota]

Ggelsring,

My amateur understanding is that after freezing out some ice from the seawater, you would have to get rid of the remaining salt water by some means, as brine inclusions would make the ice weak and "rotten". Rinsing it away would mean pumping at least 4 times as much seawater as the final amount of ice formed.

I think the scale of the undertaking would make it impossible. A cubic kilometer of ice masses a billion tons, and with annual melts being 16,000-18,000 cubic kilometers, you would need to freeze 1000 km^3 to even be visible above measurement uncertainties. If you pump the seawater up 10 meters in making manmade MYI, the energy required would be 10^15 Joules per km^3, or 270 million kw-hr. If you pump 4 times as much (per above), you would need a 500 MW installation to pump that much in 3 months of winter. Multiply that by 1000 km^3 and it would be 10 times the total US windpower installed capacity, or more than a third of total power capacity. Considering the difficulty of operating in open ocean in the Arctic winter, switching us to all renewable energy would be cheaper and easier.

[Seke Rob]

Snow, the daily [anomaly] chart to follow in a simple "It's on / It's Off" depiction: http://climate.rutgers.edu/snowcover/chart_daily.php?ui_year=2012&ui_day=282&ui_set=2. I've not counted the grid boxes, but on eyeball, there's more off than on for the time of the year.

[Ggelsrinc]

Chris Reynolds

I have my head around the figures for area, extent and volume and the population of the Earth.

I asked a simple question whether sea ice made with high saline has enough strength to last the melt season. It isn't hard to drill through ice. Water pumped from below the sea ice would freeze in the winter and thicken the sea ice.

I also asked if anyone has ever suggested a way to stop the sea ice from continually reaching minimums.

That sea ice is just one domino in the problems of global warming, but it isn't the only thing happening. The sea ice minimum is about the size of Greenland when measured in area. We lost snow cover in June equal to three Greenlands. The trend towards Hothouse Earth has already begun and the only negative feedback, I can think of, is mankind, who caused the problem.

It would be nice to prevent an ice free arctic and I'm just tossing out ideas, like could we build ice bridges at Nares and the CAA with some liquid nitrogen bombs in the winter?

If we don't want a 22 degree C average temperature Earth, we better start thinking of a good way to geoengineer this one and stop carbon emissions. I don't think nature has the negative feedbacks in her to prevent what we started until the Hothouse effect takes place.

[P-maker]

Chris Reynolds:

I somehow get the impression, that you do not consider North Atlantic water influx as a critical factor in the end game of the Arctic sea ice.

Please have a look at these maps: A) Beaufort Sea and Baffin Bay - http://polar.ncep.noaa.gov/sst/ophi/color_anomaly_NW_ophi0.png and B) Kara Sea and Laptev Sea - http://polar.ncep.noaa.gov/sst/ophi/color_anomaly_NE_ophi0.png.

Or alternatively, for a general overview of the Arctic, please have a look at the DMI SST anomaly map here: http://ocean.dmi.dk/arctic/satellite/index.uk.php

I hope you also see four orange pools of warm water in the Arctic:

1) Beaufort Sea (see upwelling mechanism explained here: http://www.whoi.edu/science/po/people/mspall/pdfs/ProgOcean_2011.pdf) 2) Baffin Bay (see my earlier post here: http://neven1.typepad.com/blog/2012/09/models-are-improving-but-can-they-catch-up.html?cid=6a0133f03a1e37970b017c3234fa38970b#comment-6a0133f03a1e37970b017c3234fa38970b
3) Kara Sea and 4) Laptev Sea (see brief explanation of upwelling mechanism explained here: http://www.whoi.edu/fileserver.do?id=92424&pt=2&p=44107)

These massive upwelling phenomena by mid-October this year in four corners of the Arctic brings up loads of “old and warm” North Atlantic water to the surface near the continents. I see a number of severe implications:
X) winter freezing will be postponed,
Y) Sea ice will not grow from the continents outwards and
Z) there will be more room for new warm North Atlantic deep water in the Arctic abyss.

Is the driver of all this the “Katabatic Heat Pump” mechanism, which I tried to describe in my previous post, or is it the “North Atlantic Waterfall” mechanism over the Arctic abyss - as described by you in an earlier post - that leads to this new pattern?

I am sorry to say Chris, but not even your first blog post statement seems to be valid, if this upwelling pattern continues into the autumn.

[Tor Bejnar]

I’ve been playing with the CT area data at http://arctic.atmos.uiuc.edu/cryosphere/timeseries.anom.1979-2008, organizing it in different ways, having been inspired by a graph by Seke Rob.

I plotted Melting Season Day-of-Year against First-Day-Below-X-million-Sq.-Km., with a data set (zig-zag quazi-linear line) for each year. The graph shows that melting (area reduction) is slowest at the beginning and at the end of the melt season, and fastest in between: third order polynomial best-fit curves approximate the data with R^2 values on the order of 0.99 (according to Excel). Visually, however, the minimum extent/last days of melt data points usually indicate slower melting than the curves would ‘predict’. Using linear projections of these data sets clearly do not provide a good basis for extrapolating phantom 1M sq. km. points, as I had done earlier this year from the data organized in a different way.

I've never learned how to put a graph onto a web page, so cannot link to one here.

[Tor Bejnar]

A plot of Area against Year has melting season data sets for each selected day-of-year (I used 5-day intervals) suggests a slope change in the late 1990s or earliest 2000s for the various data sets. The slope change is quite small early in the melt season (e.g. Day #150 has line and 2nd order polynomial best-fit curve that aren't far different from each other) to noticeably larger by Day #215. Day #250 (last day with near-continuous melt-season data over the satellite record) was graphed with all data, the last ten years of data, and the last five years of data. The best-fit lines for 10 years of data (actually 9 years of day #250 with one year’s day #247 – 2005’s min.) and 5 years of data are virtually collinear. Best-fit line and 2nd order polynomial curve for the entire Day #250 data set show that the data earlier than about 2003 isn't anything close to collinear with the last 10 years.

Projecting Day #250’s last 10 years (or 5 years) line forward, 1 million sq. km. (“ice free”) will be reached in 2019 or 2020. These best-fit lines suggest 2013 will reach 2.35 M km^2 about on Day #250.

Because of Arctic ice volume declines (or just because I’m pessimistic), I rather suspect an "ice-free" Arctic will be seen prior to 2019.

[LRC]

@ Ggelsrinc: You could probably easily setup a nuke pump and spay such as they do for ski slopes. Weight alone should compress things so the the salt would get pressed out. The only issues would be that the ice would be very fragile at 1st and therefore would there be enough time in the freeze season for it to compress hard enough to hold together and the biggy. You would need a whole lot of pumps to cover an area that would help things last the next melt season. then of course you are back to the problem of all that heat.
As the heat is also coming from the lower latitudes that solution I do not think would even be a starter.
Other solutions Sulphur Screen http://en.wikipedia.org/wiki/Stratospheric_sulfate_aerosols_%28geoengineering%29
Carbon dioxide removal: http://en.wikipedia.org/wiki/Carbon_dioxide_removal
2 would be the best method but would be very energy expensive (costs more energy to go from CO2 to C the from C to CO2.
Man will probably end up going for 1 and that would end up with who knows what.

[Lewis Cleverdon]

Ggelsrinc - While I too would doubt the practicality of directly manufacturing ice in significant volumes, I've no doubt of the need to reverse its decline, or of the confidence of scientists of renown that there are indeed attainable options for doing so. The writings of professors Wadhams, Salter and Caldiera are worth studying in this regard.

As you may know, Professor Salter was the originator of the 'Cloud Brightening' concept (by means of lofting seawater mist several thousand feet from wind-powered vessels) which may be particularly appropriate for targetted use over the arctic ocean in summer. He and a Dr Latham have worked on the dynamics and engineering for over a decade, but it's still some years off.

Another possibility (merely a surmise on my part) on which I'd be glad of feedback, is that of robbing heat from the Gulf Stream before it reaches the arctic, by means of barrages of Marine Heat Pump Energy Technology and possibly Ocean Thermal Energy Technology (if the temp differential were enough), with the potentially massive power output used for liquid fuel production, apart from a fraction transmitted to urban blots in Europe.

This benign larceny would offer four main benefits:
- it's active year round against the most advanced mega-feedback;
- it's potentially self-funding in deployment after R&D;
- it's providing non-fossil energy;
- and it's mitigating the rising impact of Peak Traded Oil on geo-economic and thus geo-political coherence that threatens the entire mitigation effort.

If it could be done, 'twere well it were done well, and soon.

Regards

Lewis

[Lewis Cleverdon]

Chuck - with regard to viewing rapid renewable energy investment as an alternative to the requisite Albedo Restoration mode of geo-engineering, I'd respectfully differ with you for three reasons.

First, renewable energy offers a partial solution to the energy fraction of our future CO2 emissions, which are only a fraction of our future GHG emissions. No amount of investment in them can address more than a minor fraction of the problem we face.

Second, any fossil fuels locally displaced by renewable energies are being bought and burned elsewhere, and despite major REs' deployment-spending, the FFs' volume burnt is still rising inexorably (despite a slump arguably worse than the '30s since we lack the cheap energy and collective politics to climb out of it). In my view FF usage is likely to rise until after the agreement of a stringent global climate treaty that caps and allocates tradable declining national emissions budgets. Only that treaty can end the bought-&-burnt-elsewhere syndrome.

Free Market shills and dupes, and the latters' sites such as Grist, HuffPo, Climate Progress and others, may ignore both that syndrome and the question of how cheap fossil fuels would become if renewables eventually took say a third of the world energy market - but that is their problem, not mine: I merely point it out now and again.

Third, our best efforts at emissions control, say an early stringent treaty that cuts global GHG outputs to near zero by 2050, entailing full scale renewables' adoption, is not remotely commensurate with the problem we face:-
- Summing the realized warming of 0.8C,
plus the timelagged pipeline output of 0.7C,
plus the phase-out emissions' output of 0.6C,
equals 2.1C.
- Ending emissions also ends our upkeep of the cooling 'sulphate parasol', which Hansen et al report will unveil a mean rise of 110% of warming (+/-30%), which would yield 4.41C (+/-0.6C), timelagged after 2050 to be fully realized around 2080.
- In the continually intensifying warming over the intervening 68 years, we can be very confident that the six out of seven interactive mega-feedbacks that are already accelerating would contribute major additional warming, as would the methyl clathrates under those conditions.

As a measure of just how remote those 'best efforts' at emissions control would be from a commensurate response to our predicament, it's worth considering just how much global food production is being affected by the climate destabilization due to a mere 0.8C of global warming.

In short, a failure to include the mandating of accountable scientific supervision of both Albedo Restoration and Carbon Recovery in an equitable and efficient global climate treaty would be the pivotal failure to address the problem. We've reached the point where it needs saying that those who claim otherwise either haven't been paying close attention (no blame is implied: it is a foul issue to study) or in some cases have some political motive driving them to mislead people.

Fortunately it seems that, due in large part to the feedback's role in sea-ice loss, there is a wave of recognition under way that poses a step-change in the motivation for collective action, as well as for a fresh assessment of just what demands need to be focussed where - so maybe as individuals we can engage with others in advancing the common purpose.

If you or others have thoughts on aspects of the above - I'd be glad to hear them.

Regards,

Lewis

[R. Gates]

The notions related to geoengineering the Arctic are very interesting but I fear might be misplaced use of resources and vastly impractical. The rapidity of the decline in sea ice coupled with the the very high total energy content of we are now seeing in the oceans (something on the order of 25 x 10^22 Joules added just in the top 2000 meters alone in the past 50 years) should be all the signs we need that a huge climate shift is just now getting underway. The tipping point has come and gone.

What we need to be putting resources into is adaptation to a new climate. something akin to the Mid-Pliocene is certainly coming within a geologically very short period of time, and further down the road, perhaps in less than five or six centuries, we could be facing a Miocene climate. We need to get about the business of putting full resources into hardening our food and water supplies on a global basis. Without these basic efforts, the struggle for resources and competition between countries will increase yearly until war grips many regions and touches everyone.

It is a noble cause to want to save the beautiful Arctic, but I feel the time to begin those efforts is now many decades too late. The climate changes we are seeing in the Arctic will spread (and are already) to every corner of the planet. The momentum of near 400 ppm of CO2 and rising is too great. But humans are quite an adaptable species. We can survive climate change and have several times in our history, but the humane way is too assure that the certain disruptions to the food and fresh water supplies do not bring conflict, war, and the specter of famine and mass suffering that usually
follow.

[Werther]

During the last ten days I’ve not been very active in following the discussion. Partly because I am busy to participate in city-farming where I live. Partly because I find not much joy in elaborating on the near future of sea ice; I think the Rubicon has been passed…

What I keep looking for is anomalous change in synoptic scale weather patterns and hemispheric circulation. It’s rather daunting to filter anything out of GFS or ECMWF.
But what is interesting is the recent faltering of what seemed to become the next El Nino phase. What is happening in the specific region is probably very different from the usual pattern as we have witnessed through the last few decennia.
In fact, atmospheric patterns never really reflected El Nino conditions, although relatively warm waters were upwelling and shopwing up op the maps as the typical El Nino SST configuration.

I wonder if this is related to other anomalous features elsewhere on the Northern Hemisphere. There are large, relatively warm pools on the western side of both ocean basins. Together with the anomaly in the Arctic Ocean, they add up to about 20 million km2 of relatively warm ocean surface.

Coupled ocean-atmosphere mechanisms could well be active here in ways I do obviously not understand.
The effects of these configurations will probably evolve further as soon as excess heat is released from the still ice free Artic waters and a snow cover has formed over Siberia and North America. What I imagine is the Polar Jet will start undulating in November/December. Perturbations in the lower troposphere could get nasty through lows on the fringe of the temperature anomalies in the same period.
That’s what I’m looking for… and it won’t bode well for sea ice through 2013 at the same time, when this comes to pass.

[Misfratz.wordpress.com]

@Ggelsrinc - Have you read the trilogy of books Forty Signs of Rain, Fifty Degrees Below, and Sixty Days and Counting by Kim Stanley Robinson?

I think you'd find them interesting. I think most people would. As part of the narrative there are various large-scale geo-engineering efforts to repair the damage created by global warming, which are not too dissimilar to your own idea in terms of their scale and feasibility (ie the scale is gargantuan and the feasibility low).

Theoretically, I think what you propose is possible, though there would be several challenging technical problems. What I would envisage would be large buoys which could be deployed to the Arctic during summer.

If we assume that each buoy has a very powerful pump then one might be able to thicken a square kilometre of sea-ice per buoy.

The technical challenges that would need to be overcome would be:

(1) The pump. To cover a square kilometre it would have to be very powerful - farm irrigation systems can cover about one-third of a square kilometre with a "big gun" style sprinkler - so this is feasible.

(2) The temperatures. You would be operating in temperatures of below 30C at times, maybe for four months of the year. This would make pumping large quantities of water very hard.

(3) The energy. The buoys would need a dense, high output power source for the pump. Something nuclear, inevitably. I doubt that the systems designed for satellites would deliver enough power to operate (and heat) the pump, so you might need to use something derived from nuclear-powered submarines. This would be quite large (and very expensive).

(4) The ice. One of the big risks to these buoys would be that they would be crushed by movements in the ice-pack, so they would have to be quite strong, particularly if they are nuclear powered.

(5) The numbers. If we assume that we have a modest aim, to thicken just 1 million square kilometres of sea-ice, and that we have a very powerful pump as above, then we would require 1 million buoys. By comparison, the international ARGO project has 3594 floats/buoys deployed as of yesterday, so you would need nearly 280 of your large, nuclear-powered, cold-resistant, very high pressure pump buoys for each ARGO float.

I think all these technical problems would be solvable, just about, but it would be really hard, and ruinously expensive. By way of comparison, each ARGO float costs about US$ 15,000.

The annual cost of maintaining a fleet of one million ARGO floats would be about US$ 6.7bn. How many times more expensive would our nuclear-powered Arctic pump buoys be than ARGO floats? I think US$ 15 million would be very cheap for this. So the cost would be, very roughly, at least seven trillion United States Dollars. Every year.

By way of comparison, one estimate I have seen is that World War Two cost the United States five trillion dollars (in present prices), in total.

So, to conclude my lengthy and assumption-filled comment: I can imagine what you propose being possible, but it would be so difficult that there are very many things we should do first, and it's possible that removing CO2 from the atmosphere would be a cheaper way (of repairing the Arctic by reducing temperatures) with the benefit that once it was done, you wouldn't have to do it again.

[idunno]

AMEG (Arctic Methane Emergency Group) have some proposed geo-engineering solutions.

Their top post at present (Saving The Arctic Ice (#1)), which is a HuffPo reprint also on the right sidebar, seems relevent to this discussion...

http://arctic-news.blogspot.co.uk/

[Jim Williams]

I've three words for all the geoengineers out there: "It's too late."

It'd take at least 10 years to get an effort organized, and that's assuming that people were willing and able to replace governments which were less than fully committed to the project. After getting the project started you'd then have to agree upon a technology -- another 10 years. So, at least 20 years before you even start a build-out for remediation effort.

Then in another 10 years or so you are going to have to start dealing with the unintended consequences of whichever solution was chosen.

What's really going to happen is a wild mix of wars and unrelated technological advances. Many people are going to starve, and some are going to turn themselves into cyborgs -- and eventually enter the Matrix.

In the end Mankind won't even notice when Homo Sapiens becomes extinct in a much warmer world.

[Chuck Yokota]

Lewis,
I agree that renewable energy is only a partial mitigation of the global warming problem.

However, my concern about geoengineering is that we don't yet know enough to be confident we wouldn't create more problems than we solve. The recent failure of the models to predict the magnitude of the sea ice melt is just the latest demonstration that we don't have a good grasp of the feedbacks and the magnitude of their effects. I don't like the idea of intentionally mucking about with our climate, but if geoengineering is found to be necessary, then so be it.

[crandles]

I doubt you need a hugely powerful pump to raise water 1m above sea level. Remember 1m above sea level gives 9m ice thickness. If the water was give some horizontal movement in direction of wind, how far would water flow downhill before it freezes?

The pump wants to be below the ice to stop it freezing. If a tube takes water above surface where it loses heat and returns to by the pump then down into deeper water to warm it up before returning to the pump again, how much thereoeleectric power can you generate with the temperature difference? Can that power pumping some water 1m above sea level as well as circulating water through the tube to generate the temperature difference?

[Peter Ellis]

Yes, but remember you have to get your water from underneath the ice, so the minimum distance you have to raise it is the thickness of the ice.

Of course, that factor's irrelevant, because what really matters is throwing it high enough and fast enough that it'll cover the surrounding square kilometer. Have you ever seen industrial irrigation pumps on a farm? That's a lot of energy required.

Then, you need to do all that without the water freezing in the pipes and clogging the pump, so you'll need to heat the water... so it'll be falling as snow rather than ice, thus insulating things and lowering natural heat transmission...

And remember, every joule of energy you use to do this will ultimately end up as more heat in the system, in the Arctic, warming the area around the pumps and the air above. You do not fuck with the first and second laws of thermodynamics.

This doesn't make sense on any level.

[Sam]

I very much agree with R. Gates and Chuck Yokota. The time when geoengineering could have worked is long since past. And even were we to try, our understanding is so poor it would likely fail with large unanticipated side effects.

Back in the 1990s Edward Teller wrote a paper that both asserted that climate change isn't real and then went on to say that if it were that the impacts to business would be so unacceptable that we would have to use large efforts to offset it. He then proposed a $1 billion per year plan to use high altitude sprayers to spread barium, aluminum and other sols to form clouds to increase the earths albedo offsetting the heating.

There were five such papers in all from national labs making this same suggestion and reaching the same conclusion. Then came the 2000s and chem trails. Now the sky's are clear again. Whether it was covertly attempted or not is an open debate. If it was, it did very little and was a failure.

There have been five attempts at seeding the oceans with iron to create blooms and increase oceanic CO2 uptake. Four failed completely. One showed some effect. In the end phosphate was limiting. The effect was insignificant. And we're it tried full scale we would immediately run out of phosphate, which is already a societal limiting resource needed for agriculture and wasted in how we handle sewage.

We as modern peoples don't think we live in closed systems. We think we live in a use it once, throw it away and when we run out, replace that resource with some other system. This is of course foolish thinking and disastrous.

The big key though is the failing permafrost. As the roughly 1,466 gigatons of carbon begins to release into the atmosphere as methane and CO2, the permafrost changes from sink to source. A recent model and paper suggest that crossover will happen about the year 2023. The paper also notes very large positive feedback loops and effects not included in the model, making crossover from sink to source much sooner. In truth, we are probably already there or past there.

Even reducing the size of the sink is a serious problem. We have clearly passed that point. So, it is likely that we have passed the point of state change. The earth alone is now in control. No action by humans shy of blowing up Yellowstone is likely to have any large effect. Even that would only be temporary as the sulfates and aerosols would wash out of the system long before the CO2 already in the atmosphere declined.

We can perhaps alter the trajectory of coming events slightly; nothing more.

What we can do now is understand what is happening, how it is happening and develop better guesses at what is to come with shifting weather patterns, moving climate bands and all the rest. We can work to figure out how to preserve agriculture in a world where the seasonal shifts and biome transitions will likely make that very difficult.

In the near term that means recognizing that our models are woefully inadequate, learning how to accept what is happening is real, learning to see the changing dynamics and the new states so we can use that knowledge to preserve whatever we can of society.

One recent key that seems to have first been realized on this blog (thanks to Neven for creating this place for that too happen) is the shifting northern cold pole. Scientists recognized and reported on the shifting climate in places like England. But they seem to have missed the cause; the movement of the high pressure zone from the central arctic toward Greenland.

As the ice in the arctic fails exponentially (or worse) and not as a Gompertz tail, we will likely see a sudden state transition to a new cold pole over north central Greenland. The northern cycle of low pressure systems and the polar jet stream will then move, dragging weather patterns in the northern hemisphere with them. This will change weather all over the hemisphere.

We can then speculate what that will lead too, e.g. Desertification of the Midwestern U.S. including all of the American heartland, and the absence of summers in England and parts of Europe.

But with the newly lopsided flow creating by misaligning the cold pole with the rotational pole, much larger and broader effects are likely, including large variability making growing of crops seriously difficult or impossible.

Our society is run on an economic model were efficiency is king. Excess is not allowed and is punished. There is not only no banking for a rainy day, instead we borrow from the future and run massive debts. We must shift back to systems that build and bank large resources for those rainy (or drought plagued) days, years, decades and centuries to come.

Sam

[Aaron Lewis]

Sam,
I concur - strongly!

AL

[Lord Soth]

I just did a quick look at the North Pole Web Cam, which is now at 79 North.

I didn't expect to see anything, but we have a clear picture, and and explanation of what happened to Web Cam #1.

http://psc.apl.washington.edu/northpole/NPEO2012/17.jpg

It's my understanding that Camera #1 and Camera #2 were pointed at each other and roughly 100 feet apart.

Camera #1 stopped transmitting on Oct 4. It looks like it sunk as the ice pan broke up.

Camera #2 is probably not far behind. However arctic night is fast approaching and will soon descend over camera #2

As with the Gompertz curve, the way things are going, it probably too conservative, and we will probably see the last of septeber ice before 2020.

[Espen]

Lord Soth,

As you can see from this web cam image, the camera is ow placed on a floe, it is easy to see open water to the left, at least from the image now, it will be up dated soon though!

[Espen]

Healy,

No ice to be seen near Healy:
http://icefloe.net/Aloftcon_Photos/index.php?album=2012&image=20121010-1701.jpeg

[Chris Reynolds]

P Maker,

That paper is interesting and the Beaufort warm patch may be due to upwelling via the same mechanism. The Siberian sector ativity is what Polyakov outlined, I addressed that in my earlier post. Baffin Bay is outside the Arctic Ocean so isn't really relevant in terms of what's going on there.

My point remains that the top of the AW layer is some 100m below the surface. So the bulk of the warm pool is insulated from interaction with the pack, it being saline and dense. People claiming the AW is responsible for the observed sea ice loss are really stepping beyond the evidence - there will be a role, but the evidence shows other processes are dominant.

As for this season. CT area anomalies continue to fall, that is to be expected as we're so far below climatology. The 2007 anomaly record may be broken within the next two weeks. However CT area absolutes are trending sharply upwards. 2012 has gone from being 0.37M km^2 behind 2007 a week ago to 0.25M behind 2007 as of yesterday. This closure of the gap seems to me to support my contention that the oceans weren't as warm as in 2007, there's much less heat to get rid of.

[Ggelsrinc]

Peter Ellis

And remember, every joule of energy you use to do this will ultimately end up as more heat in the system, in the Arctic, warming the area around the pumps and the air above. You do not fuck with the first and second laws of thermodynamics.

This doesn't make sense on any level.

I suggested doing it with windpower, like the kind that used to pump water on farms. The season to make ice is longer than 3 months and doesn't require pumping water 10 meters. I suggested boring holes in existing sea ice, but I never explained that the pumps could be moved around and don't have to spray a large area and yes it would take many of them. They wouldn't be very expensive though. I also suggested plugging up the exits for sea ice in the CAA and Nares. 30 years ago that whole CAA area was covered with sea ice at the minimum and it could be plugged up again, with artificial MYI. If we built an ice bridge at Nares, Fram and the Beaufort Gyre would be the only sea ice exit routes. If we concentrated on those areas and 80N, thickening the sea ice where needed, we could expand next year based on what survives.

The heat insulated and trapped below the sea ice in liquid water is obviously released to form ice, but it's done during a time cold enough so it can't melt ice and is in outer space by the time the next melting season starts, instead of still being trapped under the sea ice.

There is no doubt we are way beyond a tipping point, but it's only irreversible, if allowed to continue. The realities are mankind isn't going to immediately stop using fossil fuels and governments probably view the loss of sea ice in the arctic like a child finding a new toy, that doesn't belong to them. I expect some crying trying to take that toy away. We can only hope they get bored with it in time or events cause them to quickly mature.

I want to thank the many people who made many logical suggestions and I would like to throw a monkey wrench in that Gompertz. I agree they will probably use stratospheric sulfur aerosols and cloud reflectivity enhancement. Iron fertilisation and other ocean nourishment schemes don't show much promise. Biochar and bio-energy with carbon capture and storage does. Stratospheric sulfur aerosols are going to add to acidification and would have to be replaced until CO2 levels are reduced.

[Enno]

" ... thickening the sea ice where needed, we could expand next year based on what survives..."

I suspect there´s a misconception there. If we get to a basinwide seasonal ice - which we will soon; the only thing in doubt is how many or few years it will still take - then there won´t be any winter ice surviving. That´s the definition of the new state: unaided, the winter won´t be able to grow ice thickness anymore that´s able to not melt complelety. So any "exit route blocking" artificial ice would be uselessly blocking the exits of an otherwise ice free arctic summer basin. So, conceptually, if we wnated to make an artificial ice cover we would have to make it over all the areas where we want it.

And; it can be even worse: if freshwater stratification indeed breaks down (which I dont know if it will, but it might) then the entire idea becomes nonsensical anyway: we could just as well try artificial ice growth in the North Atlantic then.

the only (un?)realistic strategy I can think off to recreate the ice would be to stop emissions and actively draw down the already too high CO2 in the atmosphere, so that the winter can again grow ice thick enough to become perennial.

[bluesky]

This is off topic, however since there have been a few comments here and there on how to alert more efficiently the people about Arctic Amplification and its impact ont Northern Hemisphere climate, I was thinking about contacting the climate modeller and risk managers in insurance companies, it seems that not all are aware of the wider risk. As an example, the Lloyds report on Arctic which describes the situation quite clearly, only mentions the impact on winter snowfall (page 17) as a consequence on the Northern Hemisphere climate, although when it was released in April, Francis and Vavrus research were already known (link between the Arctic change and frequency of extreme events). Insurance and reinsurance companies may have a vested interest in helping to change the policy as their risk exposure may increase significantly.
The MEP for the moment don't seem to be very motivated, I contacted the 8 MEP for London 3 weeks ago and still haven't received any reply....

http://www.lloyds.com/~/media/Files/News%20and%20Insight/360%20Risk%20Insight/Arctic_Risk_Report_20120412.pdf

[Chuck Yokota]

Sam,
I am interested in learning more about the shifting north cold pole. Can you provide some links to discussions and research?

[Espen]

Bluesky,

Insurance companies in general, wants more risks, they will then be able to increase premiums, which will then increase their market share in your wallet. Insurance companies dont want less risk!

[Chris Reynolds]

One recent key that seems to have first been realized on this blog (thanks to Neven for creating this place for that too happen) is the shifting northern cold pole. Scientists recognized and reported on the shifting climate in places like England. But they seem to have missed the cause; the movement of the high pressure zone from the central arctic toward Greenland.

Sam, I really don't think what is happening is as simple as this. See this blog post.

[Ggelsrinc]

Enno

I suspect there´s a misconception there. If we get to a basinwide seasonal ice - which we will soon; the only thing in doubt is how many or few years it will still take - then there won´t be any winter ice surviving.

I've looked at the 2012 drift and we are getting MYI entering the CAA. There was a time it couldn't enter Nares, but the ice bridge collapsed. There may be some sea ice left in the CAA that didn't come from the arctic basin, but the sea ice in Fram and Nares are arctic basin sea ice at the minimum. The sea ice at the minimum is arctic basin sea ice. The Beaufort Gyre can expose a lot of MYI to open water and cause it to melt.

The arctic sea ice isn't going to melt up to the North Pole, the last of it is going to get caught north of Greenland and along the CAA. That is the way it circulates in the arctic. If you want to save sea ice, those new exits need to be closed.

[crandles]

>"Insurance companies in general, wants more risks"

Yes in general/overall, more risk = bigger cake to share between them. However, what each wants are risks that they can quantify more precisely than competitors and for the risks to be random rather than potentially containing systematic risks to keep the reinsurance headaches and costs down.

Significant climate change raises systematic risks perhaps in ways that might become difficult to quantify.

Reinsurance companies in particular need the systematic risks to be precisely quantified.

So is climate change good or bad for them? It might still be good for some if it frightens off weak competitors, provided they can still assess the risks reasonably well. Even if it is good in this way, they need to be aware of the systematic risks and that might start changing at increasing rates. Latest reseach might need to be followed and understood more closely and carefully. Perhaps a nice little income top up for more climate scientists?

[Werther]

As a follow up on my post earlier today, here’s the NCEP/NCAR surface skin temp anomaly for the first week of October.

The data are probably not comparable to SST anomaly data used to produce the DMI, NOAA and OSPD maps. However, the graph is easy comparable with the same period in 2007. Both years are, maybe not surprisingly, very similar in the total spread of +2 dC anomaly over the Northern Hemisphere. A little above 15 million km2.
For the +10 dC anomaly; 2007 shows 1,4 km2, 2012 1,6 km2.
Chris, to suggest much less heat to get rid of compared to 2007 depends on what scale you imply with the term ‘the oceans’.
The closing of the CT area gap between the two years is just one way of looking into this. It is a secondary ‘piece of evidence’.
Of course, I’ve given attention to the years in between. None of them has as much anomaly as ’07 and ’12.

[Enno]

Werther, that is one impressive graph. Thank you. (Are there meteorologically versed people here who can spell out what such a heat anomaly will do to the atmosphere in the short run?)

/Ggelsrinc, we may have talked at cross purposes. If you want to artificially save the last 1 MKM2 ice, that´ll be where you indicate; but then the Arctic Ocean would be functionally seasonal already: for the climate effects the damage would still be done./

[idunno]

This Guardian article...

http://www.guardian.co.uk/environment/damian-carrington-blog/2012/oct/10/food-price-rise-uk-crop-harvest?intcmp=122

...has a link to the latest research on the Arctic Dipole (paywalled), here...

http://www.agu.org/pubs/crossref/2012/2012GL053268.shtml

[Chuck Yokota]

The NOAA article about the research:

http://www.noaanews.noaa.gov/stories2012/20121010_arcticwinds.html

[me.yahoo.com/a/nSjChi4X3vr8X3DRw93GkY1.cerja.8nvWk-]

Lord Soth camera buoy #1 was picked up by the research vessel from the edge of the ice. Camera buoy #2 caught the approaching search light on an image!

Phil.

[Lewis Cleverdon]

I'm sorry to see such irrational defeatism here on ASI regarding the efficacy of albedo restoration and carbon recovery, though it's heartening that at least the assessment of the futility of an 'emissions-control-only' strategy stands unchallenged.

With respect, the claim that 'It is too late' to apply geo-engineering effectively makes no sense, given that our incidental release of fossil sulphates is currently masking over 50% of warming - according to the mean finding of no less a pair of scientists than Hansen & Sato. While additional sulphates are patently not the best choice for albedo restoration, it is very plain that global SA temperature is being, and can be, controlled.

Finding and deploying the best means of doing so is already occupying numerous scientists and while it will not be achieved overnight, it is both eminently feasible and patently urgently necessary. Such is the catastrophic warming to which we are now committed that, even with best efforts at emissions control, this is not a free choice; given the alternative of serial global crop failure and geopolitical destabilization and conflict, it is an unavoidable expedient.

That the practice of geo-engineering would need to be maintained (to some degree) while anthro GHG output is ended (2050?), and airborne anthro CO2 is fully recovered (2100?), and excess oceanic heat dissipated (2250?), does not diminish its practicality in the slightest; it merely indicates the duration of the task.

What the scope of the task indicates strongly is that those techniques that offer multiple benefits, particularly including some self-funding potential, should be pursued as priorities. Both the extraction of Gulfstream heat for liquid fuel synthesis and a global afforestation program for biochar and coproduct methanol exemplify this criterion.

Quite how many years it will take for techniques to be applied to restore the PI SAT, and thus how much further the feedbacks will accelerate before most are stopped by lack of heat, is an open question. I'd well agree that there may be a cut-off point - as in a Shakhova-style eruption of clathrates' methane - beyond which albedo restoration would likely be ineffective. To respect that window of viability what counts is just how soon the necessary international adamant demand for very specific effective action can be assembled, from which the decisions over well-supervised trials and deployments flow.

Those who've overlooked these realities to date would do well to reconsider their defeatism - and most definitely to avoid voicing it in public fora - for the propagation of defeatism is at least as damaging to the urgent necessity of assembling effective demand for political action as the ramblings of the paid pseudo-skeptic deniers. Perhaps more so given ordinary peoples' propensity to jump from "It's not happening" to "There's nothing we can do about it." As has been pointed out before, what is said here matters.

Regards,

Lewis

[P-maker]

Lewis:

"...those techniques that offer multiple benefits, particularly including some self-funding potential, should be pursued as priorities. Both the extraction of Gulfstream heat for liquid fuel synthesis and a global afforestation program for biochar and coproduct methanol exemplify this criterion."

Well spoken!

Building resilience into our systems of energy, food, water and transport services is essential. It is necessary to include adaptation and mitigation perspectives in every decision we make.

[P-maker]

On the ”absence of weather”

Enno, you ask what the weather implications are from the current SST anomaly pattern in the Arctic.

Please have a look at: http://www.uni-koeln.de/math-nat-fak/geomet/meteo/winfos/arcisoTTPPWW.gif

You will see three high pressure areas located over the Alaska North Slope, the Siberian North Slope and the Northern part of Greenland.

If you compare these three areas - where pressure gradients are largely absent - with the wind observations on this map: http://www.uni-koeln.de/math-nat-fak/geomet/meteo/winfos/synNNWWarctis.gif,
you will see a striking pattern of strong katabatic winds coming off the three slopes right now.

If you then cross check with the SST anomalies on this map: http://ocean.dmi.dk/arctic/satellite/index.uk.php (please change to anomaly mode), you will see a beautiful example of polynias maintained by katabatic winds.

So the answer to your question may be: The absence of weather leads to open water in some parts of the Arctic right now.

[Jim Williams]

Lewis: "Such is the catastrophic warming to which we are now committed that, even with best efforts at emissions control, this is not a free choice; given the alternative of serial global crop failure and geopolitical destabilization and conflict, it is an unavoidable expedient."

Given the impending global crop failures, geopolitical destabilization, and conflicts your remediation efforts are not going to happen.

This isn't defeatism, nor even realism. This is just accepting the obvious.

[wayne]

Ggelsrinc " If you want to save sea ice, those new exits need to be closed."

True, but it was always closed by fast ice for most of the year, Once in a while in the past huge mutti-year ice escaped through CAI channels, some even coming from Russian river mouths. Now these channels are mostly open during a longer time period. The way to save MYI is to have a very cold winter. As far as Fram Strait is concerned, it is obvious that this ice flowing out serves a purpose, it is needed for the THC, or the THC itself is pulling it out by the current it gives, It would be unwise to tamper with that.

Espen, what is the surface temperature around the Healy? If we can figure out the surface temperature of freezing sea ice around the Healy that would be fantastic.

[wayne]

some surface temperature maps around the pole are not so good. But this one is probably the best:

http://www.esrl.noaa.gov/psd/map/images/fnl/sfctmp_01.fnl.gif

Look for the surface temperature around the growing Arctic sea Ice edge it is between -10 and -15. The true freezing temperature of sea ice is when surface temperatures should be about -11 C or colder. The surviving Pack is like Greenland and shifted the Cold Temperature North Pole
to Northern Ellesmere. Cold air already at sea level is far more potent than cold air above the GIS at high altitude. But the sea ice itself is in a battle to regain its former extent. Since ice begets ice, the wide open water all around it seems to keep surface temperatures much warmer and not favourable for a quick refreeze.

[FrankD]

Lewis, if you put an audacious idea out there and don't want people to disagree, its probably best not to end your post with:
If you or others have thoughts on aspects of the above - I'd be glad to hear them.
It conveys the idea that you are interested in boos as well as the cheers.

My own take? At this point, geoengineering is like grabbing another bucket to bail the sinking boat. If little effort is going into stopping the damn leak, perhaps that's doing little more than delaying the inevitable. If it can keep us afloat long enough to plug the leak then yay!

But my cynicism says that bailing harder while doing little to stop the leak is just an invitation to the fat bastard in the stern who is doing little but weigh the boat down more to say "Meh, the water's not coming in so fast..."

As for whether it matters what we say...so, what, the guy from EPA goes to President Schwarzenegger and lays out five geoengineering plans and the President says: "No! I read on Neffen's blog dat ve are all dooomt anyvay..."

?

:-)

[FrankD]

Somewhat more on topic...

If Wipneus is following this thread, I'd be very interested to see if he's done an early-bird "just-a-bit-of-fun, don't-take-it-too-seriously" forecast for September 2013?

IMO, this years volume tracked his projections, however "naive", like they were on a rail, and following that success, I'm looking forward to his projection for next year.