Betting

UAH December: +0.265C. Annual average: +0.236

27 Comments

Roy Spencer announced the UAH temperature anomaly: +0.265C. We now see JCH , denny, MpG took win, place and show. And yes, it looks like once again, the low bias in betting persists. Here are the other bets:

Winnings in Quatloos for UAH TTL December, 2013 Predictions.
Rank Name Prediction (C) Bet Won
Gross Net
— Observed +0.265 (C)
1 JCH 0.259 5 82.397 77.397
2 denny 0.256 5 65.918 60.918
3 MpG 0.25 5 32.685 27.685
4 pdm 0.222 5 0 -5
5 Jefff 0.22 4 0 -4
6 DenisC 0.22 5 0 -5
7 Tamara 0.22 5 0 -5
8 AndrewKennett 0.311 5 0 -5
9 lance 0.216 5 0 -5
10 PavelPanenka 0.202 3 0 -3
11 BobW 0.199 3 0 -3
12 ivp0 0.198 5 0 -5
13 EdForbes 0.191 5 0 -5
14 AMac 0.341 3 0 -3
15 MDR 0.183 3 0 -3
16 DavidJay 0.181 5 0 -5
17 ScottBasinger 0.18 5 0 -5
18 Cassanders 0.175 5 0 -5
19 LarryGoldberg 0.17 5 0 -5
20 KreKristiansen 0.169 4 0 -4
21 Rick 0.16 4 0 -4
22 LesJohnson 0.154 5 0 -5
23 RickA 0.147 5 0 -5
24 RobertLeyland 0.144 4 0 -4
25 edward 0.143 3 0 -3
26 stevet 0.14 4 0 -4
27 mct 0.127 5 0 -5
28 march 0.123 5 0 -5
29 JohanAbrahamsson 0.12 2 0 -2
30 Perfekt 0.119 5 0 -5
31 RiHo08i 0.119 4 0 -4
32 eddieo 0.118 4 0 -4
33 Skeptikal 0.117 4 0 -4
34 mwgrant 0.112 3 0 -3
35 RobertInAz 0.1 1 0 -1
36 LaurieChilds 0.099 5 0 -5
37 angech 0.09 2 0 -2
38 TomBolger 0.09 3 0 -3
39 LukeWarmist 0.087 3 0 -3
40 BobKoss 0.087 5 0 -5
41 NielsANielsen 0.085 5 0 -5
42 Ruth 0.066 5 0 -5
43 GeorgeTobin 0.014 5 0 -5

The net winnings for each member of the ensemble will be added to their accounts.

Good luck with betting in the upcoming year. (It might be fun to bet how many pseudo-scientific expeditions will become sufficiently trapped in ice to require rescue by ice-breakers. But… nah!)

27 thoughts on “UAH December: +0.265C. Annual average: +0.236”

  2. “It might be fun to bet how many pseudo-scientific expeditions will become sufficiently trapped in ice to require rescue by ice-breakers.”
    A few boats here!

  3. Lance – payday loan interest rates might be available.

    I think I won this twice. I could be a satellite.

  4. Wow! One more win like this and I will be able to pay off my debt – maybe 🙂
    This is clear proof that you don’t have to know anything about climate to win in this game.

  5. Nick Stokes,
    My brothers-in-law started discussing breaking out the ice fishing hut. I think it hasn’t been used in 4 or so years. We’ll see. . .

  6. Has there been a volcano recently or something that I missed? The UAH LS temperature is the highest it’s been since the early 1990’s after the eruption of Mt. Pinatubo.

  7. DeWitt

    Has there been a volcano recently or something that I missed?

    Don’t worry. These eruptions are often overlooked. If temperature don’t rise, several years from now I’m sure one will be found to have erupted about now.

  8. DeWitt Payne (Comment #122242)

    lucia (Comment #122243)

    I don’t see an anomaly that seems as large as you are suggesting, so I would suspect there is no major eruption going on. Given ENSO neutral conditions, if there has been one we should see a down spike of global surface temps in 7 to 9 months, although it may still be hard to see in the noise.

    I’m seeing -0.16 for December:

    http://vortex.nsstc.uah.edu/data/msu/t4/uahncdc_ls_5.6.txt

    Which is exactly the same as December of 2010.

  10. Hm, seems that way.

    Something a little fun, I’ve used the GISS AOD data to try and estimate the lower stratopheric warming effects of volcanoes. Did it a couple of different ways, not satistified with anything yet; part of the problem is that GISS AOD data appears to be smoothed somewhat-and oddly enough, Stratospheric warming from Pinatubo appears to have peaked *before* the peak of AOD data. Weeeeeird.

    Anyway, depending on whether I use a higher coefficient or a lower one, there is either a weird warm spike for one month left over (and it is quite large) or what looks like it might be several months of over corrected data:

    http://devoidofnulls.files.wordpress.com/2014/01/uahlsstratvolcano.png

    I’ve also marked the dates of solar minima and maxima, because I think you can vaguely see a solar cycle effect.

  11. Re: Andrew_FL (Jan 4 19:54),

    I’ve also marked the dates of solar minima and maxima, because I think you can vaguely see a solar cycle effect.

    That would be logical if, as I remember, the solar cycle has a larger effect on the UV part of the spectrum compared to the total flux. According to Wikipedia, among other things:

    The 30 hPa atmospheric pressure level has changed height in phase with solar activity during the last 4 solar cycles.

    Have you tried wiggle matching with the sunspot number?

  12. DeWitt Payne (Comment #122253)-Just did a multiple regression model with monthly sunspot number, the AOD data, and a trend that stops in August of 1994, looks like a good fit, and the sunspot coefficient looks like it is definitely greater than zero. Specifically it’s about 0.002 per spot, which sounds small but keep in mind the monthly sunspot number is pretty large, this translates to temperature swings on the order of .3 K.

    Anyway, here’s a plot:

    http://devoidofnulls.files.wordpress.com/2014/01/uahlsmodel.png

    Red is the model, blue is the original data. Not bad for just three variables.

  13. Re: Andrew_FL (Jan 4 21:11),

    I would say more than three. A trend has a slope and an intercept and the cutoff date is a variable as well.

    Of course it’s more complicated than that. You have a cooling effect from increasing CO2 and an effect from chlorofluorocarbons that were increasing and causing cooling and are now, I think, decreasing which should result in warming. It’s possible that CO2 and CFC’s are currently cancelling each other which might explain the lack of trend since 1994.

  14. DeWitt Payne (Comment #122264)-Fair enough. Yes some of the CFC’s have declined a bit recently, I think others are flat, I need to find the time series..es. Anyway, at this level, modeling studies I have seen suggest the CO2 impact is negligible compared to Ozone effects, so I don’t think it could be canceling, it’s just the CFC trend is very close to flat.

  17. Lucia

    “…the low bias in betting persists.”

    Bias? What bias? My models have just has gone south. I think they are now stuck in the ice. Then again maybe a change is afoot.

    Somewhere between bemused and amused.

    Happy New Year.

  18. My entry was supposed to be 0.14 not 0.014 so that I would once again be nearer to The Consensus rather that the actual temp.

  19. ”
    Andrew_FL (Comment #122274)
    January 5th, 2014 at 10:16 am

    Web, get lost with your stupid meaningless fit. The grownups are talking.
    ”

    Oh, that’s what “Quatloos” are about. I thought it was play money made by Hasbro.

  21. “Andrew_FL (Comment #122274)
    January 5th, 2014 at 10:16 am
    Web, get lost with your stupid meaningless fit. The grownups are talking.”

    Oh be fair Andrew, he only has 49 variables in his fit

  23. In the area of *interesting* science:

    Revisting what I did here:

    http://rankexploits.com/musings/2013/indisputably-correct-climate-model-results/#comment-122011

    and here:

    http://rankexploits.com/musings/2013/indisputably-correct-climate-model-results/#comment-122016

    I was interested in how much of what I was looking at was due to the presence of residual solar or volcanic forcing. I’ve done the phase analysis I’ve been doing for solar cycles and volcanic eruptions to residual sunspot cycles, and removed the previously established volcanic effect, in addition to the low frequency component I was already removing. Those look like this:

    http://devoidofnulls.files.wordpress.com/2014/01/hadcrut4andlowfrequencyplussolarandvolcanoes.png

    The new PSD of the residuals looks like this (old in blue for comparison):

    http://devoidofnulls.files.wordpress.com/2014/01/psdhadcrut4highfrequencyminussolarandvolcanic.png

    The general pattern is still present, with some interesting changes.

  24. Re: Andrew_FL (Jan 8 21:01),

    Either there’s a low frequency signal your filtering process doesn’t remove, which seems unlikely, or your it’s being added back by the rest of the process. A decent high pass filter, which is more or less the equivalent of subtracting the low pass filtered signal from the original, should have a cutoff frequency below which you don’t see much.

    Why don’t you generate some time series with random noise, white, pink and red, say, and see what that looks like after running through your process. Then you could add some volcanic eruption signals and try it again.

  25. Model fit to GISTEMP.

    http://imageshack.com/a/img845/6034/7ju6.gif

    James Hansen wrote in 2007:
    “Two noticeable discrepancies with the temporal variation of observed global surface temperature are the absence of strong cooling following the 1883 Krakatau eruption and the lack of a warm peak at about 1940.”

    I have also revealed that the Krakatoa eruption shows at least a 2x reduction from what is expected from the estimated aerosol forcings. I also find a discrepancy during the WWII years, with an anomalous warm peak that is at least partly explained by a temporary change in the way that SST readings were being made on board ships taking the measurements (not wanting to be identified by German U-boats, they pulled in the trailing buckets).

    There are other uncertainties, such as small gaps in 1936 and 1953, and one right before the Pinatubo eruption (almost a 6-month causal anticipation in the cooling), but that is a start.

  26. Re: Andrew_FL (Jan 8 21:01),

    Another possibility is that you’re overestimating the volcanic response and making what amounts to a notch filter. Is the frequency of the dip about the same as the time constant for the volcanic response?

  27. DeWitt Payne (Comment #122427)-I could try a less aggressive filter (I use an iterative process, which I repeat 10 times, so I have the less filtered versions of the time series.) I’m probably leaving in anything with a frequency less than about 30 years.

    I actually think it’s probably more accurate to call what I’m using an “attenuator” than a filter.

    Since the rest of the process is removing volcanic eruptions and a solar cycle, and the basic shape is present whether I do so or not, I would tend to think I can’t be adding much to the thing.

    DeWitt Payne (Comment #122435)

    This seems more likely, except I don’t think you are looking in the right place with the volcanic eruptions.

    For one thing: my estimate of the magnitude appears to be significantly less than others find.

    It looks like the dip occurs around 4 years. This looks more like the origin of the dip is my solar cycle estimate (lag 45 months versus volcanic lag 8 months), which could *easily* be mistimed or overestimated, since I more or less went fishing for it.

    I’ll see about working with some synthetic data.

Comments are closed.