Betting

Bet on March UAH

53 Comments

This month, we are betting at the last minute. Not that that makes any difference to some people’s ability to let what they want temperature to out weigh what temperature are likely to do. Someone will still bet -5C or +5C or something like that. If you’d like to take their quatloos, enter your bet for the value of the March UAH temperature anomaly Roy Spencer will post in early April.

53 thoughts on “Bet on March UAH”

  3. Stephan

    My Channel 5 chart is not a good indicator of UAH performance. It is based on a different satellite and is not used by Roy Spencer in calculating the monthly UAH anomalies.

    Here is a link

    http://img859.imageshack.us/img859/6875/dailyamsuvsactualuahtem.png

    to a comparison of daily UAH and AMSU (Channel 5) by Billis Illis.

    Please notice that there is a significant and variable difference between UAH and Channel 5 – AMSU. You can only use the Channel 5 values to get a rough sense for the situation. UAH, GISS, RSS, Hadley and NOAA are the real indicators, not Channel 5.

  5. Kelly thanks but your link ain’t working. Also I don’t trust ANY data from GISS, Hadley. NOAA who use same data and adjustments anyway and are creators of the “hide the decline” applies to anything these people spew out see Muller’s Video. AS an aside I haven’t noticed any heating/climate change since I was born 58 years ago have you?

  6. Maybe time to delete that UHA channel 5 graph from that site it ain’t showing what the team likes… lets make a bet it does disappear sometime soon….

  7. Kelly–
    I think Stephan discussion may motivate me to write a script to find out whose bets have consistently under or over estimated the temperatures over all time. 🙂

  9. Kelly–
    I don’t know if I should share the database. People might have an expectation of privacy. It’s a script I’ve planned to write– in php.

    If you do know php, maybe I could send you partial information in the database and then you could write the guts of the thing in php to pull out the bets for all matched names and then we could display.

  10. Lucia

    Sorry, I’m just an R and Excel guy. I understand the confidentiality concern, so you could possibly use anonymous ID code for each person so that no-one is explicitly identified.

  12. Stephan,

    Yes, as a matter of fact I have. I’ve lived on a lake since 1978. When I first came here, the ice-out date was consistently in mid-April. For three years in a row in the 80’s, it was on April 14. In the 90’s it began moving back to maybe the 10th or 8th. During the past decade it’s been in the first week of April. Last year it was March 30. It was a cold winter so it’ll be later than that this year, but I doubt it will be as late as the 14th again.

  13. Privacy?

    Looks like I’m already being monitored!

    How else could you know this? “Note: You’ve tried to bet at least 0 times.”

  16. steveta_uk–
    The script hides the number of times you tried to bet in the fresh form it sends you. Then, when you submit, it reads that number. So, if you can’t add… well….

    You can come back and try again. It’s just to slow down the pesky add-enabled ‘bots.

  17. Lucia,

    quite well given that this is completely statistical model, and made 31 months ago:

    http://www.climateaudit.info/data/uc/HADCJan2011.png

    Last month I promised to DeWitt to check if there are too many observations above the prediction mean, still in my to-do list. Anyway, models including observed up-to-date TSI and CO2 etc. as parameters should do much better job in prediction (where are such predictions?).

  18. Kelly thanks but your link ain’t working. Also I don’t trust ANY data from GISS, Hadley. NOAA who use same data and adjustments anyway and are creators of the “hide the decline” applies to anything these people spew out see Muller’s Video. AS an aside I haven’t noticed any heating/climate change since I was born 58 years ago have you?

    This post has it all. Associating “hide the decline” with groups that had nothing to do with it? Check. Complaining about adjustments in the surface record without realizing that adjustments of the satellite record are far more widespread and problematic? Check. Citing a You tube video as scientific evidence? Check. Reliance on personal observation to diagnose global climate? Check.

    96/100

  21. UC –
    “where are such predictions?”
    Most of the relevant inputs to any temperature model have two common features: first, they’re unpredictable over the short term (i.e. monthly), and second, in the long term (i.e., multi-decadal) they net to zero or near zero. I’m thinking here of volcanoes, ocean cycles, solar irradiance, etc. The one relevant input that doesn’t fit that description is long-lived greenhouse gas forcing. So the first question is, why do we need such predictions? And the answer is, in the short term, we don’t (which is good, because predictions can’t be made short term anyway). Only the long-term predictions are needed for climate, and those have been made in the literature. See, e.g., IPCC AR4 WG1, chapters 10 and 11.

  22. Tim W 0.501 5

    Perhaps this gentlemen should Google P. T. Barnum’s most famous quote.

  24. Lucia
    Is there a graph/frequency table of the ‘winners’ on this blog? I want to see the frequent winners and place bets close to their values.
    .
    I saw your post on Cronon at the Collide blog. I agree largely with what you wrote, in the first post there.

  25. On average, we guessed that UAH would be -0.072 for March, by my figuring.

    Excluding the +8 degrees outlier from the calculation, there were 95 of us. We will see if the wisdom of crowds applies to this crowd.

  26. KAP,

    “Most of the relevant inputs to any temperature model have two common features: first, they’re unpredictable over the short term (i.e. monthly), and second, in the long term (i.e., multi-decadal) they net to zero or near zero. ”

    what do you mean by unpredictable? What I meant is that if

    Forecaster 1, pure statistics guy observes past temperatures and predicts future temperatures

    Forecaster 2, observes past temperature _and_ future CO2, TSI, volc, and predicts future temperatures

    then the forecaster 2 should win the forecasting game. Or tie, if CO2, TSI, volc have no impact on temperature. I’m looking for conditional prediction involving the main drivers, CO2, TSI, volc.

  27. Shub–
    I think I posted balances after January. But I have yet to write a script that tells us how close people are to winning over time. I need to do that. It will also tell us who is consistently high or low. But… right now, I’m running R. (I’m doing something woefully inefficient in the hopes of seeing if an idea is a dead end or not. If it’s not, I’ll be modifying to make R do whatever I’m doing efficiently. That will involve… learning more R!)

  28. UC (Comment#72355)

    And how does one “observe” future CO2, TSI, and volcanoes? Aren’t we talking about prediction or projection rather than observation? I agree that if the only input forecaster 1 uses is past temp, he won’t be building a very good model. But I don’t know of anyone who actually works that way. Do you?

    The bottom line for long-term forecast is this: ENSO and ocean cycles have zero or near-zero secular change. The same goes for solar and volcanic effects. Anthropogenic forcing has significant secular change, so if you’re building a long-term model without it, you won’t be building a very good one.

  29. KAP and UC,

    It is possible to forecast out at least a year with the ocean cycles as your main natural variable.

    There is a 3 month lag in the ENSO’s impact. An El Nino is coming in 2011 (soon that is) and it should peak around mid-December as most ENSO events do. The Nino 3.4 index influences the UAH temperatures with a coefficient of 0.07 to 0.11 with a 3 month lag.

    The north Atlantic responds to the ENSO with an 8 month lag about 50% of the time (not always but it has to be your best bet). It has a coefficient of about 0.5 with respect to UAH temperatures.

    Last week, the ENSO increased to -0.82C and looks to get to +1.0C by mid-December. Last week the AMO declined to +0.024C (after being as high as +0.93C in August) and looks to go a little lower in the near future before stabilizing and then starting back up in the fall.

    My forecast for both going out to the end of the year.

    http://img228.imageshack.us/img228/5629/weeklyensoamomar232011.png

    And how does this impact global temperatures, here is just the impact on global SSTs. Going down a little for awhile yet but then going back up in the fall. (Sometimes a little sticky on the downside).

    http://img508.imageshack.us/img508/4264/weeklyensoamoglobalssts.png

    UAH hits its low point for the year in March, is a little warmer than this until July and then warms to +0.15C by December.

  31. DeWitt Payne

    A new Arctic extent and area low based on what?

    The critical issues are SST, wind strength/direction and ice thickness.

    From the current starting point the minimum this year could easily be the highest in the last 6 years, particularly given the fact that the air temp impacts of the present La Nina are delayed in impacting on the Arctic.

    Like last year the new cycle will be an interesting one to follow.

  32. Lucia, is the most people to bet for a one month period?

    Zounds, nothing like gambling to get some traffic to a site eh?

  33. DeNihilist, we’ve even heard rumors of table top dancing (sorry, not exotic). That would definitely increase traffic!

  34. Since we’re all guessing about the future here I’d like to ask if anyone has considered what, if any, impact the magnitude 9 earthquake will have on temperatures? Taking Wikipedia’s word for it on magnitude vs. energy I think the quake was playing with something like 5×10**22 joules, some largish portion of which would have been converted to heat.

    Some comparative context on that magnitude from Pielke Sr.

    bob

    This is an accumulation of heat of 1 * 10**22 Joules per year. We now have data to assess what actually occurred in terms of this metric of global warming up through the end of 2008 (i.e. see the Figure in Pielke (2008), Figure 1 in Willis et al (2008) and personal communication from Josh Willis to extend the data to the end of 2008).

    Perhaps I’m being silly here (wouldn’t be the first time) so please correct me if this is “just wrong”.

  35. DeWitt Payne (Comment#72369)
    March 28th, 2011 at 9:52 pm
    Re: Bill Illis (Mar 28 19:10),

    If you’re correct, bet on a new record low for Arctic Ice. It looks like we’ve seen the maximum for extent and area for the year and it’s in the 2007 range.

    ———————————–

    The AMO has a reasonable correlation with the NH sea ice extent so it shouldn’t be a bad year for the sea ice. There should be some recovery (assuming the forecast works out) although it looks like it takes a few years of warm or cool AMO conditions to leave an impact.

    Monthly NH sea ice extent anomaly and the AMO from 1972 to Feb 2011.

    http://img576.imageshack.us/img576/2598/nhse72anomamofeb11.png

  36. Re Ice. Unfortunately graphs are showing that a LOW max is more related with a HIGH min. Look at the ice graphs. BTW why is global warming only occurring in the NH ice?

  37. Stephan #72379

    The variability has been much greater, especially since 2007. I suspect that has to do with the loss of multi-year ice.

  38. Lucia, is the most people to bet for a one month period?

    Sometimes we bet for longer periods. But then I have to remember that bets were placed!

    Yes. It brings in traffic. We can all chat too. 🙂

  39. Re: Stephan (Mar 29 10:31),

    For minimum area, years before 2007 may not count. Here’s a detrended (linear) residuals plot of the Arctic area minimum (Cryosphere Today data). The horizontal lines are 1sd apart. The process is clearly out of control with 3 of the last four years greater than 2 standard deviations below the trend (1979-2006). Three of those four years also had maxima above the trend.

    A plot of maximum vs minimum has a trend of 1.098 with an R^2 of 0.5 so in fact a low maximum is correlated with low minimum. Considering that the maximum area this year was a record low, it’s not at all unreasonable to think that this year’s low will be a new record as well. All that goes out the window if the AMO goes negative, but Bill Illis thinks it won’t.

    The AMO was near record high and so was the UAH NoPol anomaly. It’s therefore likely that the new ice that formed this year is thinner than usual. Predicting Arctic ice is somewhat of a fool’s game, so I wouldn’t be at all surprised if I’m wrong.

  40. Im not going even try to predict ice this year (but agree that it will be probably much greater than past years that is MInimun). Kay’s Graph here
    http://processtrends.com/images/RClimate_UAH_Ch5_latest.png
    really highlights how stable global temps in fact are. I would myself have expected greater fluctuations. BTW I know UHA is adjusted at end of month but the TREND over the past 9 years is actually completely FLAT. I predict you will not see any significant change in your lifetime. he problem is I probably agree with most extreme global warming fanatics that something has to be done BUT NOT WITH C02 emissions! How about population, urbanization ect? I don’t think climate is the issue at all.

  41. DeWitt Payne (Comment#72388)
    March 29th, 2011 at 1:55 pm
    All that goes out the window if the AMO goes negative, but Bill Illis thinks it won’t.
    ———————

    I should have been a little more clear. My chart above is the inversed AMO.

    The AMO will go negative (starting next week probably) and stay in the negatives until the early winter when it will probably go back up again.

  42. Stephan re comment # 72392

    It is Kelly, not Kay. If you are going to continue linking to my chart, at least use my name correctly!

    You have linked to my chart for the umpteenth time. It is time for you to look at another of my charts: my comparison of UAH and GISS trends.

    http://chartsgraphs.files.wordpress.com/2011/03/gu_move_avg2.png

    This chart shows that UAH is often higher than GISS when both are set to same baseline period, 1981-2010 in this example. Your claim that the ” … TREND over the past 9 years is actually completely FLAT. ” is not valid. Both UAH and GISS include trend and cyclical patterns.

  46. Well there you go Kelly you removed it and replaced it with what you though looks like a warming graph. Thanks, I leave others for them to judge whether you are prone to any bias.
    http://chartsgraphs.wordpress.com/2010/08/19/enhanced-uah-channel-5-temperature-anomaly-trend-chart/
    Actually this graph is even better it shows a cool world from 1980-1997 and a warm world between 2001 and 2007 and cool world again (now), thanks again. Anyone trying to deduce a trend here should have their degrees removed LOL . I think you should remove the graph entirely.. its not helping your cause. just leave the Giss ones etc.

  48. Unfdortunately Zeke, Kelly fell for it. The graph did not show any warming even if it was 10 years and removed it (30 year data was more meaningful). I agree 10 years means nothing in climate terms. He/she still removed it, because on its own does not show warming, which proves you guys/gals are actually biased. Then Zeke, you go on to show your 10 year graph comparing UHA and GISS. This is typical “team” behaviour and that’s why I lost confidence a long time ago. What Kelly should have done is leave the 10 year graph and maybe added the 30 year graph. I’m sorry you guys are biased and that is why your are losing support (ie Judith Curry etc). I have lost confidence in your data and data graph postings. If you did not act in this way you would maintain much more support for your cause. That is all I am saying and I suspect this is what Judith Curry is trying to tell you fellows LOL

  49. BTW I accept the new graph data. In my view it shows no significant warming/cooling for the 30 year period 1980-present. Maybe in your eyes it does. Go for it and good luck.

  50. Kelly and Zeke please help…
    But I am very confused. This new graph posted on the Blackboard purports to show global warming. What amazes me is that from 1980 to 1998 there seems to be significant cooling. I thought according to Hansen it was way above anomaly?
    http://processtrends.com/images/RClimate_UAH_Ch5_latest.png
    There is warming from 1998 to 2007 but then it cools again please help LOL

  53. KAP,

    “And how does one “observe” future CO2, TSI, and volcanoes? Aren’t we talking about prediction or projection rather than observation? ”

    Like I said, I’m looking for conditional prediction. Once the future CO2,TSI,Volc become available, we feed them in to the Forecaster 2 prediction, and verify with observed temperature. No room for ad hoc auxiliary hypotheses or change of definitions.

    “I agree that if the only input forecaster 1 uses is past temp, he won’t be building a very good model. But I don’t know of anyone who actually works that way. Do you?”

    Not a good model, but a good baseline for the scientists to beat.

Comments are closed.