Polls 2012: daily and monthly minimum

Now that the melting season is underway and we have had the opportunity to study the prelude, this is a good moment to start guessing the minimum. But there are different kinds of minimum: area, extent, daily minimum, average September minimum. Of course, the difference between extent and area is well-known and area tends to be lower than extent. There's also a difference between the day in September with the lowest extent/area, and an average of all September days.

Personally I'm a big fan of minimum daily extent/area polls, because that makes it more exciting around the time of the minimum (not having to wait until October for the final result to come in). On the other hand, the monthly SEARCH Sea Ice Outlook reports, where science pros and knowledgeable amateurs submit their guesses, garner more interest every year. And these are for the NSIDC September (monthly average) minimum extent. Okay, so let's do two polls then. But which datasets?

Last year the polls were focused on IJIS minimum daily sea ice extent, but because of the switch to WindSat it's not sure yet how 'trustworthy' or consistent it is. That means there's only one option left and that's the Cryosphere Today sea ice area numbers (to be found here).

These are the daily minimums in the 2005-2011 period:

• 2005: 4.09 million square km
• 2006: 4.03 million square km
• 2007: 2.92 million square km
• 2008: 3.00 million square km
• 2009: 3.42 million square km
• 2010: 3.07 million square km
• 2011: 2.90 million square km

The other poll will then obviously be focused on the NSIDC September minimum sea ice extent numbers (to be found here), so that we can compare the blog community's estimate with the estimates submitted to SEARCH.

These are the September monthly average minimums in the 2005-2011 period:

• 2005: 5.57 million square km
• 2006: 5.92 million square km
• 2007: 4.30 million square km
• 2008: 4.73 million square km
• 2009: 5.39 million square km
• 2010: 4.93 million square km
• 2011: 4.61 million square km

Both polls are in the right hand bar and will be replaced every month by new ones. These first polls run until June 30th.

Edit: I've made a mistake in the question of the NSIDC poll. It says "The NSIDC minimum monthly/September sea ice area for 2012 will be", but of course 'area' should be 'extent'. Sorry for the confusion. I've changed the poll, but it could take some time to show up.

Comments

[Neven]

I had to put this post up quick, because I have to run. Any mistakes will be corrected tonight. I'll make the NSIDC poll look a bit different.

[Seke Rob]

Tough one when looking at chart http://i137.photobucket.com/albums/q210/Sekerob/Climate/ArcticSIASIESeptember.png

3.2 MKm2 it was last year including the 0.31MKm2 assumed to be ice for that black-out pole area. Voting for 3.0-3.2, but 2.8-3.0 is not in the realm of impossible... Winds of Change.

-- Rob

P.S. Time to dig out where the PolarStern is and up to.

[Seke Rob]

Updated the NSIDC Extent Quality Indicator chart adding August-Dec 2011 and only now see how deplorable the August number was... who's ever going to say out loud there's recovery and not choke on those words

http://i137.photobucket.com/albums/q210/Sekerob/Climate/ArcticSIA-SIERatio2.png

-- Rob

[Aaron Lewis]

I fear that the Arctic has tipped into a new era. Note the consistent winds south across the Fram Straits during the last week (eg http://virga.sfsu.edu/pub/jetstream/jetstream_norhem/1206/12060112_jetstream_norhem.gif )
And the melt across nothern GIS (eg, blue ice). This suggests ice export has increased, which begs the question, "Why aren't there big gaps in the sea ice cover?" That would suggest that this early in the season, the ice is already so soft that it spreads out to "fill the gaps."

I expect both the Daily and the Monthly minimum to obliterate their old minimum records this year.

[Neven]

FWIW, I voted between 3.0 and 3.2 million km2 for CT SIA daily minimum, and between 4.25 and 4.5 million km2 for NSIDC monthly minimum.

I am basing myself mainly on what I've described in ASI update 4: CT SIA is already pretty low, despite having still quite a bit of ice cover in the Bering Sea, and Hudson Bay being slow to start melting out. Things aren't looking great on the Siberian/Atlantic side of the Arctic, but not so great either in the Beaufort Sea where I expect ice to be thicker.

I might go for a lower estimate in the next two polls, unless the melt stalls completely in July or August for two weeks or more because of a switch in weather patterns, like we saw in 2010 (start of July - mid-August) and 2011 (mid-July - mid-August).

But to be clear: If this melting season resembles that of 2007, all records will be broken.

[L. Hamilton]

CT SIA for 5/31 is by a small amount the lowest for that date.

DMI SIE for 6/2 is second-lowest, above 2011 (although I think it's still provisional).

This morning's MODIS view of the Kara Sea looks washed-out,
http://rapidfire.sci.gsfc.nasa.gov/imagery/subsets/?subset=Arctic_r03c05.2012155.terra

in contrast to its mostly-white appearance for NSIDC SIE.
http://nsidc.org/data/seaice_index/images/daily_images/N_daily_extent_hires.png

[Bob Wallace]

The top poll (CT area) either did not record my vote the first time or let me vote twice.

I 'cast', it changed to results with my vote marked on the left hand margin. When I refreshed the page it presented the choices, not the results graph as did NSIDC area.

--

You (Nevin) might not have the ability to modify those aps, but it would be nice to have ones vote permanently recorded for personal use.

Yeah, I could write it down....

[Seke Rob]

Was confused, but the second poll on NSIDC data based says "sea ice area", when it's "extent" data. Voted below 4 million which is not probable as a month average for extent... recasting for 4.25-4.50

-- Rob

[Neven]

Bob, I believe you can change your vote by voting again. If you're afraid of forgetting what you voted for, you can share with all to see, right here in this comment thread. :-)

Was confused, but the second poll on NSIDC data based says "sea ice area", when it's "extent" data.

Damn it, that's what you get for doing things too quick! Yes, it should say 'extent' and not 'area' (I had copied it from the other poll). Sorry for the confusion. I've changed it, but it's not showing yet. I'll update the post.

[Llosmith57]

September Monthly minimum between 4.5 - 4.75 and CT SIA at 3.4 -3.65. Volume of 2.5-2.75? When will the May volume numbers be updated? My volume guess may change depending on whether we get any rainfall across the CAB this summer. Rainfall would be a game changer.

[Neven]

You can re-vote by clicking 'comment' in the lower part of the poll widget. This will take you to Vizu where you can click on the 're-vote' tab next to 'poll results'.

[Matthew Opitz]

The Kara sea ice is trash according to Lance-Modis, even while Cryosphere Today is counting that whole area as full of ice. A similar thing is going on in parts of Hudson Bay, it seems. In reality, the ice is even worse than the CT numbers let on just yet. This will become more apparent as those last little ice floes melt out of the Kara Sea, and all of the sudden that area gets subtracted from the total, sending 2012 racing ahead of the others. I voted <2.8 million km.

[Diablobanquisa.wordpress.com]

CT minimum daily SIA: between 3.2 and 3.4
NSIDC september SIE: between 5 and 5.25

[Jim_pettit]

I went with a minimum area of 2.8 to 3.0, and minimum extent of 4.25 to 4.5. FWIW, though, I may have just voted conservatively because I'm still stinging from my horrible guess on maximum; I've a hunch the final numbers may actually come in a bit lower.

[fred]

I went lowest on both polls. Comparing the ice so far with the ice of 2007, the only reason this year is not already ahead of 2007 is because of ice that is doomed to melt for sure, i.e. the southern ice. Take away that ice and the slush, and there are open fingers of water for solor heating extending much farther north than normal. Add in warm run-off water and it looks pretty dismal.

[JackTaylor]

I also voted the lowest on both polls.
Mainly because of the "Jet-Stream" effects-excursions of the "Hot is hotter" and the "Cold is colder" along with reasons expressed above by fred.

Good Luck to "All Guess-timators"
Enjoy your count on the polls with your comments.

[Account Deleted]

I put on the minimum values.

Last winter and spring stands out very cold anomaly in the Eastern Arctic. Therefore, in the summer and fall I expect in the Eastern Arctic, a similar positive anomaly, such as 2007. It will be catastrophic for the ice.

[dorlomin]

Every season I have watched the ice it always seems around May/ early June that this year is going to be huge.

From my musings it is far more about the weather in July and early August than anything else. Guessing the ice for one year is really a bit of a guess about weather.

Others though are far better qualified than me to offer an opinion.

[Werther]

OK I supposed 3,5 MK as extent minimum. Though my cardboard calc was flawed, I’m not backing off. I know there has never been more melt than 5 MK up to half august. First I dismiss what’s left in Bering and SOO 0106. Second, I hold on to 6 MK losses up to mid august. That makes 5MK with five weeks left to go. Based on 2007-2011 that should result in an extra 1,1MK. What’s my discriminant? Almost a million km2 extra losses on the Atlantic/Siberian side, prepared by what I see as 'Atlantic influx’ last winter. My estimate: SIE minimum 3,7 MK, SIA minimum 2,4 MK.

[Stevemosher.wordpress.com]

2.9
4.29

records falling this year.

[L. Hamilton]

One-day minimum seems inherently more random, but getting into the spirit of things perhaps 2.7 is as good a guess as any:
http://img.photobucket.com/albums/v224/Chiloe/Climate/2012_Gompertz_CTarea1.png

Earlier I'd looked just at the September *mean* area, 3.0:
http://img.photobucket.com/albums/v224/Chiloe/Climate/2012_Gompertz_area1.png

With the usual caveats:
http://img.photobucket.com/albums/v224/Chiloe/Climate/2012_Gompertz_area4.png

[Rlkittiwake]

I voted lowest for both area and extent.

Wipneus' most awesome sea ice volume graph (which is pretty much my favorite thing ever) indicates that since 2001, we've had TWO years that weren't the worst years ever recorded.

At some point the low volume has to be reflected in area and extent in a huge way, and I think the odds are good that this will be the year.

[Tor Bejnar]

Remaining pessimistic (blue?) about Arctic sea ice, I'm also checking both minimum area and average extent options. At this time I'm guessing 2.79 and 3.99, respectively, mostly influenced by my attitude and comments about how bad the ice is looking, and a smidgen influenced by a glance at historical numbers. If only one of my votes is wrong, I'd guess I'm too low on extent.

[Remko Kampen]

Chose the lowest options for both polls.

"If this melting season resembles that of 2007, all records will be broken."

For the last time ever maybe, because that season might well wipe out all.

[RunInCircles]

I have been watching the snow melt on the Cryosphere Today graphic. The siberian snow melt was early which leads me to think that the albedo dropped early for at least that section of the arctic. That combined with the early open spots leads me to the conclusion that the melt will be greater this year over the last couple of years.
I vote for a area min around the 9th of Sept at 2.75 and an extent min for the month at 4.45

[Wipneus]

My predictions are:

Min CT area: 2.6 [10^6 km3], range: 2.2-3.0
Sep NSIDC ext: 4.1 [10^6 km3], range: 3.5-4.6

Both values can be seriously lower than the above low-ranges when the ice mass gets too low to thin even further. I estimate "hard" low ranges 1.6 (CT-area) and 2.6 (NSIDC extent).

Short method description:

1) use PIOMAS to predict ice volume. Volume appears to have a lower noise/signal ratio making it easier to get a good estimate. I estimate minimum ice volume in 2012: 3.0 [10^3 km3] using an exponential regression: https://sites.google.com/site/arctischepinguin/home/piomas/piomas-trnd6.png
2) regress area/extent as function of ice volume. The functions appear to be linear, no need/justification to do non-linear extrapolations.
3) use 1+2 to get estimates and from the standard errors a range (2*sigma).

Note that the extrapolation of the extent/area function in 2) returns non-zero value for zero ice volume. Either the linear functions will break down at very low volume, or the whole basis of this (PIOMAS) is not as reliable as I thought.
Assuming a break, making extent/area to cross zero is the basis for the estimates for the "hard" lower bounds.

[Artful Dodger]

Shrewd estimate, Wipneus! Tough to argue with your approach! Should be an excellent test of PIOMAS trends this year.

Has there been any work done to reduce those error bars? Specifically I mean has anyone identified any independent variables that correlate well with the 'out-of-bounds' years, 1981 and 1995?

Candidates like El Nino come to mind...

[Wipneus]

With a bit of more calculating (instead of eye balling from graphics) I have revised my estimates to:

CT minimum: 2.50 +/- 0.41 [10^6 km2]
NSIDC Sept: 3.98 +/- 0.67 [10^6 km2]

A graphic of the method is here:
https://sites.google.com/site/arctischepinguin/home/piomas/aib/AIPoll2012.png

[FrankD]

Lodger, the search for a global climatic variable that aligns with SIA/E/V continues, but is elusive.

I don't think the El Nino Southern Oscillation is it. I cannot see any consistent correlation with this oscillation, either in real time or with a delayed effect of up to 5 years. It weakly correlates with Antarctic sea ice area with ~a 2 year offset - determining the strength of the correlation and the offset parameter is beyond my skills. Might be a nice little time-filler-in for someone like Tamino.

But for Arctic Sea Ice, I think you have to look elsewhere (or possibly on a longer time scale...)