Sea Ice News Volume 3 Number 6 – Sea Ice Outlook forecasting contest for 2012 is online

First some news on the status of the WUWT Sea Ice Page. The JAXA imagery, which has been DOA ever since the failure of the AMSRE instrument on the AQUA satellite is now operational again. The plot is below:
JAXA AMSR-E Sea Ice Extent -15% or greater – click to enlarge

In addition, they have a new product, which combines the averages with lowest extent years for comparison:

Arctic Sea Ice Extent

And our final bit of news on the replacement satellite is from Dr. Roy Spencer, who reports that:

The AMSR2 Antenna has been Successfully Deployed

May 18th, 2012 by Roy W. Spencer, Ph. D.

UPDATE: AMSR2 has been successfully spun up to its initial spin rate of 4 rpm (it will later be spun up to 40 rpm for normal operations). It will take 45 days for the Shinzuku satellite to be maneuvered into the A-Train satellite constellation; about 3 months for complete instrument checkout.

Now this is a real treat…the first on-orbit image I am aware of taken from the spacecraft of an Earth-observation instrument ON the spacecraft itself (all imagery courtesy of JAXA).

===============================================================

That said, not much else of note has been going on in the sea-ice arena, as happens every year, seen in the JAXA plots above, we are in that time of year when all of the years converge into a tighter grouping. This makes judging where the current year is going a tougher challenge.

But as usual, the race to forecast the minimum is on again. Will it be another Serreze death spiral media opportunity? Or will it be ho-hum- nothing to see here, move along?

Anyone can submit a forecast to ARCUS, all you need is a rationale and you have to put your name on it. Even “SWAG” qualifies as a rationale, though there are many who will use models and statistical techniques to try predicting the sea-ice minimum.

I’m going to give WUWT readers an opportunity to make a forecast for submission, based on voting. See the poll at the end. I’ll run this poll each month in the week before the deadline, and we’ll see how we do as the minimum approaches. The value used by ARCUS is the NSIDC value as they say here:

The sea ice monthly extent for September 2011 was 4.6 million square kilometers, based on National Snow and Ice Data Center (NSIDC) estimates. This was the second lowest extent behind 2007 and 2.4 million square kilometers below the 1979 to 2000 average.

I suggest that you should not be using the JAXA graph to forecast minimums, though it it useful for determining short term trends as it is more responsive than the NSDIC graph below, which is averaged.

http://nsidc.org/data/seaice_index/images/daily_images/N_stddev_timeseries.png

Here is what past forecast submission reports looked like:

Figure 3a. Distributions of Outlook estimates for September 2011 arctic sea ice

Figure 3a. Distributions of Outlook estimates for September 2011 arctic sea ice extent based on May data.
Figure 3b. Distributions of Outlook estimates for September 2010 arctic sea ice

Figure 3b. Distributions of Outlook estimates for September 2010 arctic sea ice extent based on June data.

WUWT readers came in a bit overly optimistic last year, as they did the prior year. Take that into consideration in your forecast for this year.

For the record, I’ve picked 4.9 million square kilometers as the first vote in the poll.

2012 Guidelines for Contributors

The SEARCH Sea Ice Outlook organizers are now soliciting pan-arctic and regional outlooks for the 2012 season. We encourage past and new contributors to participate.

The 2012 Outlook season will be a transition year to an expanded Outlook in 2013; this year, we would like to focus on expanding discussion of ice thickness, expanding discussion of the relative performance of different Outlook techniques, and improving access to relevant outlook data (see “Data Resources” webpage at: http://www.arcus.org/search/seaiceoutlook/data.php). So in addition to the pan-arctic and regional outlook contributions, we invite any information or input to those topics as well.

ALL Outlook submissions should be sent directly to Helen Wiggins, ARCUS, at: helen@arcus.org, with the following subject lines, as relevant:

PAN-ARCTIC OUTLOOK – [YOUR LAST NAME]

REGIONAL OUTLOOK – [YOUR LAST NAME]

OUTLOOK FOR BOTH REGIONAL AND PAN-ARCTIC – [YOUR LAST NAME]

An MS Word document is preferred for ease of formatting to PDF files and extracting images for the website – we will not edit your individual submission and will not post your Word documents.

SUBMITTING A PAN-ARCTIC OUTLOOK

Pan-arctic Outlook contributions should include:

  1. Extent Projection
    Provide a sea ice projection for the September monthly mean arctic sea ice extent (in million square kilometers).
  2. Methods/Techniques
    Provide the type of estimate (heuristic, statistical, ice-ocean model ensemble runs, etc.).
  3. Rationale
    Include a short paragraph on the physical rationale for the estimate.
  4. Executive Summary
    Provide a short paragraph that summarizes your outlook contribution in two or three sentences.
  5. Estimate of Forecast Skill (if available)
    If possible, please include any estimates of forecast skill, uncertainty, or error associated with your prediction. This year, we will add error estimates to the summary bar chart of outlook estimates, as appropriate.

We would also like to expand discussion of ice thickness in the monthly reports, so please include any relevant information on ice thickness (or age), if available.

Submission deadline: Monday, 4 June 2012.

All Outlooks should be sent to:
Helen Wiggins, ARCUS

==============================================================

The forecast poll, pick a value:

Note: the poll is setup to prevent vote stuffing, you can only vote once.

About these ads
This entry was posted in Arctic, Forecasting, Sea Ice News and tagged , , , , , , , , . Bookmark the permalink.

47 Responses to Sea Ice News Volume 3 Number 6 – Sea Ice Outlook forecasting contest for 2012 is online

  1. Harold Ambler says:

    1. I’ve written to Walt Meier to ask whether, as my memory suggests, the Arctic ice extent on NSIDC is the longest streak in the 2-standard-deviation gray area in many years.
    2. My bookgoes into quite a lot of detail on misunderstandings about what ice is melting at the ends of the Earth and what it means and doesn’t mean.

  2. Nylo says:

    Option “4.6 to 4.8″ should read “4.6 to 4.7″, as there exists a “4.7 to 4.8″ option…

    REPLY: Typo fixed, thanks. May take an hour to clear cache to show the update -A

  3. tommoriarty says:

    This data looks alot different than the DMI extent and Nansen extent. See them all three in the right sidebar here…

    http://climatesanity.wordpress.com/

  4. HenryP says:

    Well,
    apparently, according to the records of voting,
    I made the highest forecast of ice
    since I found earth is cooling

    http://www.letterdash.com/henryp/global-cooling-is-here

  5. Brad says:

    Do they offer a decade by decade analysis for the Antarctic or just the Arctic?

  6. Jared says:

    I went with 5.0-5.1

  7. Steven Mosher says:

    Huh. last year is 4.6 and look at your scale. lopsided

  8. I looked at the new JAXA plot and went from highly approving to “fooled me again” in about 90 seconds.

    I like the idea about plotting the averages of ten year windows. 1980’s,1990’s, 2000s. Very reasonable and a good visual with the oldest data in the lightest shade of grey.

    Then, like a illusionist cutting the woman in half, they insert the three lowest years. By definition, this is cherry picking.

    What if all we are seeing as the three lowest years is just natural variability, the std deviation about the mean? Showing only one side of the distribution is incredibly deceptive.

    What if GISS showed just the mean temp anomalies for 1980’s, 1990’s, 2000’s, and then the three hottest years since 2000? “Katie, bar the door and head for the hills!”

    At the very least, JAXA should show the single year(s) since 2000 (the last average shown) with the largest 1, 2, or 3 minimums. I point out that if the 2000’s average minimum is 5.5 MMkm2, and 2007 was 4.2 and 2008 was 4.8, then it is dead certain that some years since 2000 were well above 5.5, possibly even higher than 6.5, close to the 1990’s average.

    You cannot be objective if you look at only one side of the distribution.

    The JAXA plot that shows only the three lowest years next to the averages….
    is NOT sound statistics,
    is NOT science,
    is NOT the right thing to do.
    It is the work of an illusionist. It is misdirection at its best.

    But let’s be charitable and use as a working assumption that the authors of the plot, in an effort to make it readable have fooled themselves without meaning to fool us. Let’s fix it.
    Show the progression of the averages by decade — well done.
    Show the three lowest. Ok, good info.
    But show us, too, the three highest years in the same span since the beginning of the last average.
    Do a /headslap “what were we thinking?!” and correct the cherry picking error.

  9. Dermot O'Logical says:

    Agree with Mr Mosher here. Please allow us all to put in our opinion and get equal weighting. I want to go in at 4.2, but I can’t.

    Can I suggest 0.1 intervals down to 3.5 please?

  10. I think the poll should continue the 0.1 increment down to about 3.8.
    Switiching to a 0.5 increment for the 4.5 to 4.0 range, where the recent years has reached is quite inexplicable.

    Better to use a 0.2 bin size for the entire range 6.2 to 3.8.

    REPLY: This is the sort of suggestion I had hoped for when I announced it was coming last week, too late to change now, many votes are in – Anthony

  11. kadaka (KD Knoebel) says:

    Now awaiting first disparaging remark from Günther Cherrytree against Anthony and/or the site…

    Slightly more entertaining than watching ice melt at this point, almost as much fun as being stuck on an ice floe as a non-drowning polar bear approaches, and about as certain to happen as a sea ice-free Arctic summer isn’t.

    (Here’s hoping he won’t post at all just to spite me…)

  12. Anthony: This is the sort of suggestion I had hoped for …..too late to change now,

    Well there will be another poll in July. The a wider range can be put in then.
    If so, since there will be a change in the poll from June 1 to July 1, you could do another poll with a the July 1 design on June 15 to “split the difference” and see if the set up of the poll seems to have any effect on the results.

    REPLY
    Sure, why not – Anthony

  13. dh7fb says:

    Is this plot better? http://www.dh7fb.de/noaice/max11akt.gif . The data-Basis is also JAXA and it shows the anomaly to the average of the 2000s. You can see clearly the years 2006…2012 and it will be updated nearly daily.

  14. don penman says:

    Jared says:
    June 1, 2012 at 8:50 am
    I went with 5.0-5.1
    Yes so did I, I think we will see some ice extent recovery this summer.It is looking quite good so far

  15. RobR says:

    I picked 5.0 – 5.1 based on the following

    MO___2012____2011____Diff
    01___13.73___13.55___0.18
    02___14.56___14.36___0.20
    03___15.51___14.56___0.65
    04___17.73___14.15___0.58

    Average difference = 0.40
    2011 Minimum = 4.61

    2012 = 2011 + Average Difference = 5.01

  16. John from CA says:

    Where are the rules for the “contest”? They should be upfront so we know which side of the stupidity we seek to embrace.

    I’ll bet in any penny arcade game like the next man but not unless I can see the game.

    Please post the proper Rules.

  17. JAXA looks a lot different to NORSEX which has current extent as highest since 2006.

  18. tetris says:

    Anthony: why is anyone still working with the 1979-2000 average? Why not 1979- 2011? We have that data and any comparison with the 1979-2000 series is by definition skewed.

  19. NZ Willy says:

    Get a load of the DMI graph, it was broken for the past week and now it shows results to 18 June! Gawd, they’re good! http://ocean.dmi.dk/arctic/icecover.uk.php

  20. Shevva says:

    Sea ice extent acting as usual, have a choclate.

    The Gothowitz Deviation model.

  21. kadaka (KD Knoebel) says:

    I’m feeling somewhat pessimistic this year, about 4.7 million sq km, possibly lower.

    Reasoning: Here in central Pennsylvania, our weather seems moderated, we don’t see much extremes in weather or temperature. Between the Great Lakes to the Northwest and the Atlantic Ocean to the East, and Appalachian Mountains, the extremes get moderated out compared to elsewhere in the US. The “noise” is reduced.

    Last summer was hardly a summer, more like an endless spring. Few days broke 90°F, it was cool and wet. The heat was elsewhere. The Arctic sea ice extent crashed.

    Past winter for the contiguous US was mild. Elsewhere in the Northern Hemisphere it froze while here it was an endless late fall. Extent had a late bump-up, however Julienne Stroeve of NSIDC commented here that the late ice was thin thus extent was deceptive.

    Now, on the first of June, there’s still not much garden planting. On the TV weather, between things like the pics of the jet stream and the water vapor loop, I can see the patterns are still off-kilter. The summer is shaping up to be like last year, endless spring. The heat will be elsewhere again, and there’s the thin ice. I doubt the extent will crash as bad as last year. ENSO is neutral, it’s about time for patterns to move closer to normal. But that won’t be until late in the melt season if not later.

    Those more knowledgeable about meteorology should know more of what I’m describing in my country bumpkin finger-in-the-air fashion. In any case, my brain has been paying attention enough that my gut is sure of what it feels. I’m not expecting much out of this round.

  22. Barbee says:

    5.3 – 5.4 listed twice in poll

  23. RE: John from CA post the rules

    What is the value of such a predictive poll when there is no individual accountability? Some of WUWT biggest detractors can enter into the poll absurdly high or low values with no cost to their own reputations, but to soil the reputation of WUWT as a blog site.

    Unless you can stratify the votes by IP address and some level of participation in the site, I see potential for mischief. If you had one distribution from contributors to WUWT, another distribution from contributors that own or run other sites, it might be enlightening. At the very least, the final results should not be expressed as mean and std deviation. Show the histogram of full results.

  24. kadaka (KD Knoebel) says:

    From Stephen Rasey on June 1, 2012 at 3:20 pm:

    What is the value of such a predictive poll when there is no individual accountability? Some of WUWT biggest detractors can enter into the poll absurdly high or low values with no cost to their own reputations, but to soil the reputation of WUWT as a blog site.
    (…)

    There would be a notable hypocrisy factor if Anthony restricted it to “qualified” WUWT commenters or otherwise restricted participation. Back in March Anthony directed our attention to an essay posted on Daily Kos, “Michael Mann is a Modern Hero and we need to acknowledge that!” Also mentioned was the included unrestricted poll about Mann. While Anthony never told his readers to go there and “freep” the poll, although accused of that, and in fact never told us readers to vote at all, we went and voted. BTW, poll is still there, still open, with 71% that Mann “is distorting evidence to prove his point” and 26% “should be fired from the university”.

    Since we voted in someone else’s unrestricted poll, how can we justify insisting on a restricted poll here?

  25. @Kakada: hypocrisy factor if Anthony restricted it to “qualified” You are right.
    But the word I used was “stratify”. Identify by category, not hide. Nowhere did I use the word “restrict”.

    another distribution from contributors that own or run other sites. Shown, not hidden.

    Anyway, I still think that the full distribution needs to be shown as there is no reason to believe that the distributin will be normal and adequately described by mean and standard deviation. I would not be surprised to find it bi-modal.

  26. kadaka (KD Knoebel) says:

    From Stephen Rasey on June 1, 2012 at 4:35 pm:

    Anyway, I still think that the full distribution needs to be shown as there is no reason to believe that the distributin will be normal and adequately described by mean and standard deviation. I would not be surprised to find it bi-modal.

    There will be two main groups, those with hope that will predict high and risk disappointment, and those who are pragmatic who will predict low and risk being pleasantly surprised.

  27. ferd berple says:

    So what happened to the bet Connelley offered and Smokey accepted for $100? As far as I can see it was a legally binding contract. 2 standard deviations below normal were the terms as I recall.

  28. gopal panicker says:

    did anybody else notice that the UIUC site went offline on april 28th when the arctic ice extent was very close to the average and when it came back after a few weeks it was well below the average….also there is a huge discrepancy with the NORSEX data which is still close to the average…watts up with that?….is somebody cooking the books?

  29. DMarshall says:

    I’m guessing a low of 4.1 mil sq km; and a late finish to the summer melt.

  30. HenryP says:

    G.Panicker
    ?….is somebody cooking the books?

    Henry says
    I have been suspecting this too
    especially in Antarctica where I think it has been cooling more dramatically than on average

    http://www.letterdash.com/henryp/global-cooling-is-here

    but I cannot get any (real) daily data from Antarctica

  31. Brian H says:

    Anthony;
    When and if you fix the 5.3-5.4 duplication, will you combine the votes?

  32. Michael Jennings says:

    I went with 4.125

  33. well it looks like WUWT readers are going to come in above 5.0.

    Anybody want to bet how much higher WUWT predicts than the rest of the world.

  34. HenryP says:

    Steven Mosher says

    well it looks like WUWT readers are going to come in above 5.0.

    Anybody want to bet how much higher WUWT predicts than the rest of the world.

    Henry says
    Ok, Steven, I challenge you.
    I want to know specifically why you would think that it is getting warmer on earth rather than cooler.

    http://www.letterdash.com/henryp/global-cooling-is-here

  35. Anthony Watts says:

    Mosh is such a killjoy lately.

  36. kadaka (KD Knoebel) says:

    Anthony Watts said on June 2, 2012 at 9:54 am:

    Mosh is such a killjoy lately.

    He needs more fiber. And cowbell.

  37. steven mosher says: June 2, 2012 at 9:31 am

    Mosh, the WUWT predictions tend to be optimistically high, but you did notice that they also tend to be closer to the observed numbers than many of the others, including most of the models, whcih tend to be fatalistically low. So what was your point?

  38. Chris Frey says:

    Do I remember right? Wasn’t there a prediction in 2007 that within the next 5 years the Arctic may become ice-free? None of the forecasters here seems to follow this…

    Chris Frey, Germany

  39. HenryP says:

    Is more ice “better”?
    I fear people do not realize what they are wishing for.
    I voted for >6 million
    but I still hope it will be a lot less.

    http://www.letterdash.com/henryp/global-cooling-is-here

  40. As poorly as we guess here at WUWT, an actual Science-ist: Serreze has been wrong every year since his “ice-free 2008″ prediction. I mean, unless he’s apologized and (explicitly) owned up to lying for publicity and I somehow missed that.

    REPLY: Predicting a future state of a nonlinear chaotic system is a bitch, no matter how smart and how many papers you have. – Anthony

  41. Smokey says:

    “Mosh is such a killjoy lately.”

    That’s how some folks react when things don’t go their way. Planet Earth is falsifying all the alarmist crowd’s predictions. That must be hard to accept.

    I know that most men, including those at ease with problems of the greatest complexity, can seldom accept even the simplest and most obvious truth if it be such as would oblige them to admit the falsity of conclusions which they have delighted in explaining to colleagues, which they have proudly taught to others, and which they have woven, thread by thread, into the fabric of their lives.
    ~ Leo Tolstoy

    I like Mosh. He’s a nice guy. But the models are wrong.

  42. Dinostratus says:

    “But as usual, the race to forecast the minimum is on again. Will it be another Serreze death spiral media opportunity? Or will it be ho-hum- nothing to see here, move along?”

    A bad year for icers, a Serreze death spiral media opportunity. I’m guessing less than 2011, 4.6e6 sqkm but am hoping higher than 2007. I’ll go with 4.4e6 sqkm.

    “Anyone can submit a forecast to ARCUS, all you need is a rationale and you have to put your name on it. Even “SWAG” qualifies as a rationale, though there are many who will use models and statistical techniques to try predicting the sea-ice minimum.”

    Right now, it’s just a SWAG based upon the Bering straight going like it did in 2007 and the ice north of Siberia being as thin. My quantitative method is to plot A vs. dA/dt in true strange attractor style, compare the rate of change as the ice goes from 10e6 sqkm to 8e6 sqkm. The faster the rate of change during that period, the lower the minimum extent.

  43. idkfa says:

    Hillary Clinton in Arctic to see impact of climate change

    http://ca.news.yahoo.com/clinton-arctic-see-impact-climate-change-151737949.html

    Clinton told reporters that she learned “many of the predictions about warming in the Arctic are being surpassed by the actual data.”

  44. ppaul says:

    What is the accepted way to prove that all time series curves are not different, running the total data against the individuals and looking at the coefficient confidence intervals? taking the sd at all timeslices? thaks

  45. walterdnes says:

    One question about IARC/JAXA seaice extent data. I thought that it had gone down Oct 3rd, 2011 due to satellite problems. The satellite is operational again. Fine. But where did they get the Oct 4 to Jun 1 data from???

  46. Caleb says:

    I am expecting a lot of pressure ridges will fall apart, as the arctic warms, and mess up our calculations. It was windy over the pole last winter, which creates both open leads (which swiftly freeze over) and pressure ridges, which involve a wide area of ice scrunched together like an accordian.

    My logic is that, if ice an area of ice a mile wide is scrunched together like an accordian to make the jumbled ice of a pressure ridge only half a football field wide (fifty yards,) then, when that pressure ridge turns slushy, it will expand like an acordian and again cover an area a mile wide.

    This will mess with our heads, because our ordinary logic does not expect melting to expand the area floating ice covers, however melting can (according to my theory) cause a pressure ridge fifty yards wide to become a mile wide, or even wider, if it spreads out to cover only 15% of the water.

    I think this is what happened in 2006, and explains why 2006 extend, so low in the winter, graph out as so high by September. My guess is the extent will be like 2006’s. However if the winds blow all the ice south, I’m screwed. O well. Won’t be the first time I’m wrong.

  47. John Whitman says:

    One June 1 I voted for 5.0 to 5.1 Million km2.

    It will be fun to see the actuals later this fall.

    John

Comments are closed.