Finally Some Winners!

The PIOMAS team is betting on a record low. Apparently they believe their own data.

About stevengoddard

Just having fun
This entry was posted in Uncategorized. Bookmark the permalink.

35 Responses to Finally Some Winners!

  1. Peter Ellis says:

    What do you mean, “finally”? As I’ve pointed out to you repeatedly, several of the June SEARCH forecasts predicted a new record minimum. You’re a month and a half late with your “news”.

    The PIOMAS forecast has in fact gone up slightly since June.

    • Last year Lindsey made two wildly different forecasts. This year they are more consistent.

      • Peter Ellis says:

        The Linsday/Zhang forecast is not the PIOMAS model forecast. Everyone has been telling you this for over a year.

        Lindsay does not use the PIOMAS model itself, his is a statistical approach based on a linear regression of various parameters based on a single day’s ice thickness field from PIOMAS. It’s much like your own PIPS-based forecast of last year: presumably you do not claim to actually represent the PIPS modellers.

        The PIOMAS forecast in the SEARCH outlook is the one labelled “Zhang”.

      • Peter Ellis says:

        I’m not sure which “he” you’re referring to – Lindsey or Zhang? Let’s clear this up, noting that the NSIDC minima for 2007 and 2010 were 4.28 and 4.90 respectively:

        Lindsey/Zhang predict 4.9 +/- 0.4 (not a new record) <– Statistical prediction, not PIOMAS
        Zhang predicts 4.0 +/- 0.6 (new record unless weather is bad) <– PIOMAS prediction

        Lindsey/Zhang predict 4.1 +/- 0.4 (new record unless weather is bad) <– Statistical prediction, not PIOMAS
        Zhang predicts 4.3 +/- 0.5 (50:50 chance of new record depending on weather) <– PIOMAS prediction

        The error bars are narrowing slightly as the season progresses, as one would expected. Nevertheless, for at least six weeks now several "alarmists" have been predicting new records, which is why I'm surprised to see you writing "finally" in the title of this post. Did you perhaps miss the June outlook?

        Finally, I'll note once again that when you take the error bars into consideration, 3/4 of the above are essentially the same as the bet I offered you: a likely new record, but if the weather is bad then this year's minimum could yet be similar to last year's figure. Are you willing to take that bet? If so, how much?

      • Peter Ellis says:

        Well, none of the models are predicting that yet – maybe the error bars will narrow more by August and we can come to an agreement. I’m surprised by your lack of confidence in your own analyses though. The most you’re willing to stake cash on is that it won’t be the worst ever recorded? That’s alarmist talk, my friend.

      • Peter Ellis says:

        Would you be willing to take a bet on that basis? That is, I say that total MYI will go down, you say it will go up? If so, how much, and whose data do we use to determine who wins? I presume you mean the end-September MYI fraction as shown by the NSIDC here for last year:

        Are you betting on an increase in ice >2 years (i.e. the dark green fraction), or are you including the 1-2 year ice as well (i.e. the cyan bit too)? I’d be happy to bet on either. I’m a private individual without vast amounts of cash to throw round – something around £50 UK?

      • Tony Duncan says:


        As you say he HAS made that prediction. Seems reasonable to me, but I still think it makes MUCH more sense for him to stick to his guns and bet the 2006 minimum, since that is what all his posts of the conditions in the arctic imply (with huge odds, so he could buy a car and not have to ride his bike all the time)

  2. Peter Ellis says:

    Note also the error bars on the PIOMAS forecast. You know, some actual science. The prediction is for 4.3 +/- 0.5 million km^2. This means that they believe a new minimum is likely, but that if the weather is bad, it could be nearly as high as the September 2010 minimum – but no higher.

    This is exactly the same bet I’ve offered you! Are you in or out? Time for the non-alarmists to put their money where their mouth is, perhaps?

    • suyts says:

      funny stuff……using my predictive prowess, I’ve come up with 5.0 +/- 1.0 .

      • Peter Ellis says:

        Excellent, then we should be able to strike a deal. I’ll bet on the PIOMAS forecast, you bet on your forecast.

        If it’s below 4.0, then you’re conclusively wrong, so I win the bet.
        If it’s above 4.8, (i.e. 4.3 + 0.5 error bar), then I’m conclusively wrong and you win the bet.
        If it’s between 4.0 and 4.8, it’s a tie – we both “win” and no money changes hands.

        How much do you want to put on it?

        • The bet I am offering is record/no record. With all the claims about record low ice thickness, rottenness, record low extent and Arctic heat, that should be a slam dunk for you.

      • Peter Ellis says:

        I wasn’t talking to you, I was talking to suyts, as should be obvious from the threading, and the fact that I refer to his prediction of 5.0 +/- 1.0

      • suyts says:

        lol, Peter, why would I bet against something I wish to occur? Peter, I am very desirous of the ice cap entirely melting. And, the sooner the better! I don’t believe it will, but I wish it to be so. But, as far as your little game goes, obviously, the odds are with you and it would be a fool’s bet to engage in such. And, what of your 4.8 number? If it does end up at 4.8, then I would have been off by only 0.2 but you and PIOMAS would have been off 0.5 but it would be a tie? How do you reconcile that? Get a little closer to proper and I’ll lay $50 down. Or $100. whichever.

      • Scott says:

        But James, what if it ended at 4.01?


      • suyts says:

        lol, then I should be the winner!!! Obviously. 😉

      • Peter Ellis says:

        Um, I don’t see why you say the odds are with me. If you believe your own forecast, then surely you think the odds are with you. That’s what “forecast” means – what you believe will happen!

        In my proposed bet, the “tie” region is far more favourable to you than to me. If the minimum this year is 4.0, then you’d be wrong by a whole million, I’d be wrong by only 0.3 million, we’d both be overestimating it, and yet it would still be called a tie!

        If you don’t like having a “tie” region, then I could take the PIOMAS central value (4.3) and you take yours (5.0), and we just see who’s closest. That means that if it’s below 4.65, I win, and if it’s above 4.65, you win. Would you go for that? It doesn’t really make sense though, since we’ve both given error bars, and there’s a considerable region of overlap in our forecasts.

        What would you see as appropriate boundaries for me winning / you winning / nobody winning, given that my mid-range prediction is 4.3, and your mid-range prediction is 5.0?

      • suyts says:

        Peter, you’re taking this much more seriously than I. And that’s why I say it would be foolish for me to engage in a wager with you. As I stated earlier, I’m one of those skeptics that wish the most dire of scenarios to occur…….A total icecap melt.

        I use the 5.0 +/- 1 as a humorous poke at these predictions. It’s hilarious. My number is pulled out of …… well, thin air based on nothing but general past observation and intuition. And, I see all of these scholarly attempts to divine some sense of how much ice is going to be left on the top of the world for one particular day in Sept. Its stupendously silly. But, again, I say you have me at a disadvantage! Obviously you and PIOMAS have refined your predictive prowess to much more precision than myself. So, I should get deferential treatment!

        Or, Peter, we can do as I’ve often advised, wait until August and then we can start playing this game with more earnest.

        I should note, too, that it is only humorous in an odd way, it is particularly maddening when one realizes that I’ve actually bought and paid for many of these scholarly predictions. Many in the U.S. are speaking of raising taxes because of our shortfall in meeting our debts. Really? We can’t think of anyplace to trim our budgets? I can scroll up and see many are playing this game on my dime.

  3. pdjakow says:

    “This is exactly the same bet I’ve offered you! Are you in or out? Time for the non-alarmists to put their money where their mouth is, perhaps?”

    Maybe he will bet on 5.5 * 10^6 km^2 😉

  4. omnologos says:

    what’s the point Peter? If the sea ice extent were going for a record maximum, alarmists would be running around blaming that on global warming. And if it were around average, alarmists would be running around blaming that on global warming too. Just look at Antarctica.

    Like every other observation, sea ice extent anywhere cannot tell us anything about global warming because each one of its possible states will be blamed on global warming.

  5. Brian G Valentine says:

    “man-made global warming” is supposed to be something evident on a decade, nay century basis – and here we are, fretting on a monthly basis over it.

    Now don’t tell me ..
    I’ve nothin’ to do …

  6. Scott says:

    JAXA hasn’t updated since July 11, but using a combination of the most up-to-date data from CT and JAXA, my spreadsheet gives 4717915 km^2 for the JAXA single-day minimum (effectively a tie with 2008). There are several improvements I’d like to implement to the spreadsheet…maybe I can do that this weekend.


    • suyts says:

      Scott, out of curiosity, what are the variables that goes into your spread sheet?

      • Scott says:

        Right now it’s very simple. I have the entire CT area metric (1979-present) and the JAXA data (mid-2002-present) in my spreadsheet. I’ve formatted into day of year so each day is side-by-side with those of other years (i.e. – the same rows). This allows very quick/easy linear regression with respect to the minimum (which is easily obtained as its own row using the “min” function in Excel) to obtain slope/intercept/R^2 and minimum for the year. Because most people work in extent, I can convert the CT area metric to JAXA extent using an additional correlation/conversion. The predicted numbers I’m showing here are the average of the predictions derived from CT and from JAXA itself. Currently, I’m using the entire CT dataset, but I think changing the analysis to only include 2002-present for CT might be more valid (since I’m working with a JAXA prediction).

        Note that I also have the NSIDC extent and area monthly data in a monthly analysis worksheet in the Excel file, but I haven’t updated that since last winter. Also, at some point I should get the Bremen data included in my file.

        Of particular interest is that 2011 is the lowest extent in JAXA by a fair margin right now, and area is just a smidge above 2007. Yet my analysis still predicts a 2008-level minimum. This statistical result indicates how extreme the 2007 loss was after this date. However, with 2011 tracking 2007 for longer and longer as time goes on, we’re looking more and more like we might end up in that region.

        So that’s it…I just use current extent and area to estimate the final value. Last year and even some this year I’ve dabbled in using other approaches with the above data, but I don’t have the time/desire to try those methods right now.


      • suyts says:

        I hear you on the time/desire thing. I would, cut the data sheet down to 2002, also. And, yes, pretty straightforward. Thanks.

    • Peter Ellis says:

      JAXA has updated the underlying data table, just not the graph.


      Four century breaks on the trot.

  7. Brian G Valentine says:

    Well, we can take advantage of the situation and explore for some oil up there!

    Like washing your car in the wintertime when there is a thaw. Use the situation to our advantage!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s