2013 Arctic Sea Ice Extent Minimum Poll

[BornFromTheVoid]

Well folks.

The annual maximum extent has passed and so it's time to start looking ahead to the melt season and to set up the first poll for the annual minimum.

Last year had the lowest minimum extent on record by almost 800,000km². Since then, the ice has extent has climbed to 6th lowest on record. As of March 31^st, it still sits at 6^th lowest on record, starting off April 341,000km² lower than last year.

In the last 5 years, the highest minimum has been 5.05 million km², the lowest being 3.37 million km² and the average being 4.38 million km²

2008 ... 4.55 million km²

2009 ... 5.05 million km²

2010 ... 4.60 million km²

2011 ... 4.30 million km²

2012 ... 3.37 million km²

The below graphs show the trends in the annual minima based on the daily extent values, with linear trend in green, the 2^nd order polynomial in red and 3^rd order polynomial in purple.

The graphs seem to suggest a minimum this year of between 3.3 and 4.6 million km². There has never been two consecutive record extent minima before, so could this be the first year? Or could it be the beginning of a recovery?

This poll won't be submitted to the ARCUS sea ice outlook. They don't have their first report until June, so I'll set up a separate poll for that later in May.

[knocker]

Okay I'm going for 3.0-3.1 million km2.

[BornFromTheVoid]

I'm slightly caught. The Arctic has had what seems like its coldest winter since about 1997/98, but the volume is still lowest on record for the time of year.

So for now, I'll go 3.3 to 3.4 million km2, similar to last year. If the dipole decides to return this summer though, sub 3 million would be much more likely I think. With more benign conditions, maybe something closer to 4 million.

[Methuselah]

I'm making a stab at 3.5-3.6 million.

[Polar Maritime]

3.2 to 3.3 million km2

[songster]

3.6-3.7 million, i.e. reversion to the mean (but accelerating) trend and therefore a similar size "recovery" to what we saw from 2007-2008. Main difference is that 2012 is less of an excursion below the long term trend, and so the apparent recovery won't be sustained for a second year, unlike 2009.

[gagerg]

I would go for an Ice Free Arctic, but at this point it does not seem to be doing too bad, but the situation still seems dire so I think it would be around the 2.5 million area or less.

[be cause]

In the hope there will be some ..2.42 million sq Km ..

[jonboy]

The pattern has changed 4.2m

[pottyprof]

I'm going to have a go at 3.7 million km²

[Gray-Wolf]

I'll have a hopeful 2.6. Better hope that we do end up with a similar summer pattern to the past few years though.

If last years losses have 'pre-programmed' part of this years season, and we get high early losses or float out then things could get very bad, very quick through Aug/Sept.

[DAVID SNOW]

3.4 to 3.5.

[songster]

I'll have a hopeful 2.6. Better hope that we do end up with a similar summer pattern to the past few years though.

If last years losses have 'pre-programmed' part of this years season, and we get high early losses or float out then things could get very bad, very quick through Aug/Sept.

Well yes, but they said that after 2007, and over the following two years the losses bounced back to the long term trendline. Tietsche et al showed the same for a very wide range of different ice levels. Arctic sea ice summer minimum seems to be governed by overall climatology, and not by hysteresis effects (pre-conditioning, "tipping points", or what have you). I don't see any reason for this season to differ. However, as and when it does bounce back to trend, it's important not to be misled into a belief that the trend itself is recovering.

[UV-RAY]

Above 4.8, not that I'm that bothered either way as there are far, far more important things in life IMO.

[Don]

What the hell, I'll go for 3.8 million!

[Gray-Wolf]

Well yes, but they said that after 2007, and over the following two years the losses bounced back to the long term trendline. Tietsche et al showed the same for a very wide range of different ice levels. Arctic sea ice summer minimum seems to be governed by overall climatology, and not by hysteresis effects (pre-conditioning, "tipping points", or what have you). I don't see any reason for this season to differ. However, as and when it does bounce back to trend, it's important not to be misled into a belief that the trend itself is recovering.

How much of that 'bounce back' was the 'collapse and Spread' of the last of the paleocryistic (as Prof Barber and his 'rotten ice' showed us)? How much did that event 'prop up' ice by keeping the waters 'ice friendly' as the sloughed into the basin? The re-freeze certainly showed how fast the new ice bound together the collapsed stuff didn't it?

We have no 'collapse and spread to come to our aid this time and the last few years have showed us , even with higher percentages of thicker ,older ice, that average summers will take massive amounts of ice? Factor in the emerging trend for past 'ice friendly synoptics' to turn traitor and now kill offf ice and you can see why I had to shy away from notions of 'rebound' this time around.

All of this was mulling was before I was further shocked by the wholesale collapse of ice integrity, this Feb/March, by the Crackopalypse event.

Now I'm actually entertaining the possibility, should synoptics turn into anything like the 07' perfect storm mix of melt and export, the drop below 1 million as a potential for the year!!! High losses prior to the 'bottom melt' end of the season will allow water temps so high that will be like an aquatic Chinook across the remaining ice!

TBH I am more concerned about the impacts of high ice losses on the Northern Hemisphere Summer than the ice loss itself. Last year was bad enough but the agricultural losses cannot be sustainable should we see Russia,China and the U.S. again struck by heatwave/drought/fire?

[songster]

How much of that 'bounce back' was the 'collapse and Spread' of the last of the paleocryistic (as Prof Barber and his 'rotten ice' showed us)?

Effectively none. I've pointed this out to you a few times and you've never once given an indication of trying to think through the basic geometry of the situation.

Your model is essentially one whereby a 3+ metre thick chunk of multi-year ice somehow spreads out and becomes a much wider area of 1-2 metre ice. The problem is that it's blatantly unphysical. Ice floes have a vast area in comparison to their thickness, they're tens, hundreds or even thousands of metres across. If you take a small multi-year ice floe, say 100x100m at 3m thick and then fracture it even further into a hundred individual pieces, each remaining fragment is still 3m thick and 10x10m in area. Collapse, yes, but there is no spread. "Spread" can't occur unless your floe breaks into shards that are less than 3m across.

You're modelling it like a glacier calving, but an ice floe is not a glacier. The multi-year ice (call it "palaeocrystic" if you like, but please check the spelling!) is thinning in situ and becoming less concentrated. It is not increasing in area. Barber's "rotten ice" was not multi-year ice that had magically spread out and taken up more area. Rather, it was low concentration chunks of multi-year ice embedded in first-year ice. The point of his observations on the ground was to show that the remote observations were overestimating the remaining multi-year ice area, because (for various physical reasons) a 10:90 mix of multi-year:first-year ice was giving the same radar signature as 100% multi-year ice. That is very different from the process you outline as "collapse and spread".

[songster]

To be even-handed, I'll say that anyone going for over about 4.5 is similarly out of touch with reality. In the entire record, the largest year-on-year gain is slightly over 1 million, so 4.5 represents an absolute ceiling for sane prediction. I have no idea who the three going for over 4.8 are, but if they want to contact me by email I'll happily bet real money against any of them.

[Radiating Dendrite]

To be even-handed, I'll say that anyone going for over about 4.5 is similarly out of touch with reality. In the entire record, the largest year-on-year gain is slightly over 1 million, so 4.5 represents an absolute ceiling for sane prediction. I have no idea who the three going for over 4.8 are, but if they want to contact me by email I'll happily bet real money against any of them.

To be fair though, things are currently very unpredictable in the Arctic, a record gain could be recorded (even though it is very unlikely).

I would have said the same 5 weeks ago in regards to anyone putting down a CET figure for March of 2.7c and we know what happened there!

[Gray-Wolf]

If I remember correctly Barber saw a multi km floe calve into shards the length of the wavelength that hit the floe. much the same as the slices that fall off the front of a glacier? If it's 5m thick and the wave was 2m long then it wouln't stand would it ? It would collapse over so the '5m would be the surface and the 2m the depth? When a floe does this does it not over double the area it was covering? And then the 'slabs' fracture and float apart a few cms and the floe grows even more?

I've always envisioned the above when I've spoken of 'collapse and spread'? You can effectively treble the surface area of the floe under collapse with no 'freezing' of sea surface involved?

As we know the sats doing the 'extent/area' calcs will not see a distiction between this type of rubble and solid ice covering the whole grid.

The old Paleo did not just sit in the water and fade away!

[songster]

If I remember correctly Barber saw a multi km floe calve into shards the length of the wavelength that hit the floe. much the same as the slices that fall off the front of a glacier?

The only quote from him I can find says nothing about the shards being related to the wavelength. If true, it's worth bearing in mind that he'll have been talking about swell waves, not surface ripples (the latter won't have enough energy to fracture ice). Swell waves have a wavelength from 150-500+m depending on the size of the ocean basin.

If it's 5m thick and the wave was 2m long then it wouln't stand would it ? It would collapse over so the '5m would be the surface and the 2m the depth? When a floe does this does it not over double the area it was covering? And then the 'slabs' fracture and float apart a few cms and the floe grows even more?

Think, GW, think. Barber says a 100-mile floe collapsed in a matter of minutes. If it spread out the way you describe, then the outer edges of the fractured ice would have to have moved apart by 100 miles or so in that time frame. Short of a freaking nuclear bomb, you can't fling tonnes of ice that far that fast.

I've always envisioned the above when I've spoken of 'collapse and spread'? You can effectively treble the surface area of the floe under collapse with no 'freezing' of sea surface involved?

I know that's what you envision. It's wrong.

As we know the sats doing the 'extent/area' calcs will not see a distiction between this type of rubble and solid ice covering the whole grid.

But that's not what Barber saw! What he saw was "a surface of heavily decayed ice composed of some small MY floes (1 tenth) interspersed in a cover dominated by heavily decayed FY floes (1 tenths) and overlain by new sea ice in areas of negative freeboard and in open water between floes."

i.e. small chunks embedded in first-year ice, fragmented but not "spread" in the manner you envision.

Here's the PDF: http://www.arcus.org/files/search/sea-ice-outlook/2009/10/pdf/regional/barber-etal-2009-regional-summary.pdf

The old Paleo did not just sit in the water and fade away!

Yes. Yes it did. It thinned in situ and then got broken into smaller chunks that were glued back together with first-year-ice in between the remnants. Barber wasn't saying that the multi-year ice was collapsing and spreading out, he was reporting that it had already gone!

[Gray-Wolf]

If you look at 'air melt v's bottom melt' you'll find the best way to melt out lots of ice very quickly is to have it in the Arctic ocean over the latter part of the melt season.

What you are suggesting would mean that the years that took the Paleo ice had an awful lot of energy around to 'wither away' the huge floes? Have them collapse into the water and it's job over in less than a month (as we see each year).

i'd rather take the 'moderate line' than this flamethrower approach to the demise of the last of Beaufort/arctic ocean's paleo ice?

also note that 'all' of the thicker ice went in the manner I suggest not just the ice post 07' so we have ice maps of the noughties not really reflecting the conditions on the ground? Had we the real data 07' might not have been the surprise it proved to be?

today we have a similar spectre of 'old ice' with a thick keel of FY ice? The map shows more resilient ice yet we see it bow out in a matter of days over high season? We need ice thickness measures at ice min to fully work out how the ice will survive the next melt cycle? Much of the periphery of the central older ice must have been very close to melt out last summer , yet clung on, to have a new growth of FY ice (with it's poor survivability qualities) grow on it's base over winter? Folk should take at least 1/3 of this ice off the maps if they are looking at how low the ice will go this summer.

We saw 3m+ FY ice melt out from the C.A. last July so we know that even that thickness of ice will not survive over most of the basin (i.e. most all of the ice!) so only geographic position looks to be the saving grace for today's pack? Synoptics drifting central ice ,conveyor belt like, into warmer southern regions will put us in real danger of mega melt during July/Aug and how do we forecast such???

If last years melt allows for a constant H.P. to the south of Greenland then we may well see GAC12 type incursions , on the Siberian side, bringing ice cover to an end this summer.

[BornFromTheVoid]

Taking the middle values for most votes, 5 million for above 4.8, and 2.3 million for below 2.5, the average guess so far is for 3.7 million km2, which would be the 2nd lowest on record.

[stewfox]

Taking the middle values for most votes, 5 million for above 4.8, and 2.3 million for below 2.5, the average guess so far is for 3.7 million km2, which would be the 2nd lowest on record.

Given the 2000s average was cira 5.8m would have like a few options above 4.8m so the 'average' doesn't have a inherent bias.. Anyway 5.8m for me.

[BornFromTheVoid]

Given the 2000s average was cira 5.8m would have like a few options above 4.8m so the 'average' doesn't have a inherent bias.. Anyway 5.8m for me.

Sorry Stew, tried to keep it within a reasonable range, given the number of options I was allowed to add. I would have added more if I could.

If people were up for creating a CET styled voting thread, I'd have no problem creating the spreadsheet for it. But from past experience, it seems many people in the climate and environment area would rather remain anonymous when voting...

Anyway, I don't appreciate the "inherent bias" comment, especially seeing as I have twice as many options above last years minimum than below it, despite the strong downward trend.