Betting

El Nino Pool

28 Comments

Rumor has it that El Nino is weakening. But, as far as I can tell, NOAA is predicting a continuuing El Nino. So, I think this means it’s time to bet on the value NOAA will publish for the NINO 3.4 3 month average for ASO. That is, we will bet the number that will appear in the circled box below:
ElNinoBet
This image comes from the pdf with the title:“Weekly ENSO Evolution, Status, and Prediction Presentation (PDF) Adobe PDF Reader”.

NOTE: If anyone knows where I can obtain weekly values, I’ll modify the rules and calculate the value based on the average of 12 weeks. (This is a silly bet, and so I would like to have at least three significant figures, even if they don’t “mean” anything.)
Update: Joe B pointed me to this page: Weekly SST data starts week centered on 3Jan1990. I will be using the average over all weeks with any days inside the appropriate period. Partial weeks will be weighted the number of days in the week that fall inside the betting period.)

When betting, bear in mind that the August and most of September data are have already been measured. So, the uncertainty in in October’s values. Also, exact ties go to the first entrant. (The reason I’d like another significant figure is to help spread apart the bets.)

Some betters might be interested in the image showing the most recent data and forecast for the ONI:
Forecast

You can enter until sometime late on Sunday October 4.
[sockulator(../musings/wp-content/uploads/2009/09/UAHBets3.php?Metric=El Nino?Units=C?cutOffMonth=10?cutOffDay=5?cutOffYear=2009?DateMetric=October, 2009? )sockulator]


(Europeans, please use dots ‘.’ not commas ‘,’ to indicate decimal places. MySQL gets really upset if you use commas.)

28 thoughts on “El Nino Pool”

  2. I am feeling pretty confident of my wager. How many quatloos for the thrall Shahna and the spirited newcomers?

  3. George–
    We will have to ask The Providers the price of a Thrall. Shahna and the newcomers appeared to be very skillful, and so must be worth many Quatloos.

  5. Thanks Joel! I’ll use that. It takes care of the whole “everyone is going to tie” issue that happens with only 1-2 significant figures.

  6. I have it on good authority that the next addition of OED will have a picture of this El Nino next to the definition of “dead on arrival”!

    😆

  7. Andrew–
    Don’t count it out. I’ve been looking at monthly data from the ECHAM climate model ,and let me tell you, you never know what next month’s temperature are going to look like.

  8. True. What I really want to know is: which models have had the best past performance predicting El Nino conditions? The spaghetti never seems very useful.

  9. Lucia: You’ll need to qualify the period better. The OI.v2 SST data, on which the ONI data is based, is weekly data centered on Wednesdays. There will be a total of 13 weekly NINO3.4 SST anomaly readings for Aug-Sept-Oct, not 12. If you’re going to use 12 weeks, you need to identify which 12 weeks, Aug 5 to Oct 21 or Aug 12 to Oct 28.

  10. Bob–
    I know. I made up rules before I knew where the weekly data was or how it would be organized. So, what I’m going to do is this:

    The best is based on months.

    The NOAA weekly data goes from Sunday-Saturday (centered on Wed.)

    Aug 1 is a Saturday. So, it’s in “week 1” which I will weight be 1/7th.
    The next bunch of months will be given their whole weight.
    Then October is a Saturday, so that works out ok.

    But, basically, dealing with Aug 1 actually uses the average from July 26-Aug 1 as the Aug 1 value.

  12. Yes. Plus 1/7th from the last week in September. (This can only possibly matter to those of you who can actually predict El Nino! Heh.)

  13. Lucia, of related interest is whether this “El Nino” will meet NOAA’s definition of a “full-fledged El Nino episode”. For this to occur the 3 month average .5C “threshold must be exceeded for a period of at least 5 consecutive overlapping 3-month seasons”.

    I suggest they will not. Anomalies have decreased recently in the Nino 3.4, Nino 3 & Nino 1&2 regions & a positive anomaly in the Nino 4 region, on which these numbers are calculated, is beginning to look very precarious. I prefer to look at the satellite series e.g. Unisys, rather than the ERSST.v3b because I think they have better resolution over this Pacific region.

    Furthermore both the 30 day & 90 day SOI (Southern Oscillation Index) index have gone positive (12.66 & 2.48 respectively), over the past 8 days or more. These are not El Nino numbers.

    http://www.longpaddock.qld.gov.au/SeasonalClimateOutlook/SouthernOscillationIndex/30DaySOIValues/

    BTW Sydney is enveloped with thick red dust from inland Australia this morning; looks like a scene from Mars, absolutely incredible. We’re had these storms streaming in from the Great Australian Bight, with strong winds wiping up dust inland (e.g. around Cobar & Broken Hill) and blowing them east wards over coastal NSW & apparently, as far north as southern Queensland.

  14. Geoff- You may be right. But I don’t know how to incorporate “Will/Won’t be a full fledged El Nino” into the betting pool. (Or, to be more specific, incorporating it would require me to edit the script!)

  15. Lucia. I’m not suggesting you amend the bet; the answer to the question whether this will be a “full fledged El Nino episode” is either “yes” or “no” & may not be known till early Dec.
    BTW there is a error in the 1 month running average for the SOI that I typed above, It should be 3.57; 12..66 is the last days figure.

    Here’s a link to photos of the red dust episode in Sydney at the moment.

    http://www.dailytelegraph.com.au/news/gallery/gallery-e6frewxi-1225778417848?page=17

  16. Geoff–
    There are oddities about internet disucussions. I didn’t actually think you were suggesting I modify the script. But… when I read comments on betting, my first thought is “Could I set up a bet on that?” (Because it could be fun.)

    So, even if you weren’t thinking that, the notion pops into my mind! Then… I type a comment. Maybe I should hold up on responding. (But I don’t, and I probably won’t because I have foot in mouth disease.).

    But, I think people do like betting. Even if it’s only quatloos. I think they like the idea that they recorded something in advance and they’d like to see how that pans out relative to other people’s guesses. It’s mostly just fun! So, if someone suggests something we might predict, I’ll generally immediatlely think: “Could we bet on that?!” 🙂

  19. The models do have el ninos, but their timing isn’t the same as in the observational data. If you look at multiple 20c3m runs from the same model, you see them appearing at different times because each run starts from a different perturbed state so it leads to a unique realization.

    Just thought I’d let anyone who cares know that I’ve finished processing the TLT, TMT, and TLS climate model temperature data. Just need to complete the final validation step tomorrow. I’d appreciate anyone’s feedback. Especially ideas for comparing to RSS/UAH in a way unique to Santer’s tests.

  21. What is the formal definition of a “full-fledged” El Nino? Are there gradations like “Kinda/Sorta” up to “Flaming /Hansenesque”?

  22. George–
    El Nino “is present” when the index reaches some level. So, it could be “present” for 1 day. Then, the index might drop.

    In contrast a “full fledged” El Nino involves some index being high enough for something like 3 months.

    NOAA reports both current conditions (i.e. are do El Nino conditions exist now.) and whether or not a full fledged El Nino formed.

    The reports include a table showing the strength of all historic full-fledged El Ninos.

    Obviously, one can’t know if the current El Nino will be full fledged unit either it a) vanishes or b) last long enough to be “full fledged”.

  24. Bob Tisdale:

    Thanks for the link. Somebody did some really nice work. Very detailed.

    I found this definition set:

    “El Niño: characterized by a positiveONI greater than or equal to +0.5°C.
    “La Niña: characterized by a negativeONI less than or equal to -0.5°C.
    “By historical standards, to be classified as a full-fledged El Niño or La Niña episode,these thresholds must be exceeded for a period of at least 5 consecutive overlapping 3-month seasons.
    “CPC considers El Niño or La Niña conditions to occur when the monthly Niño3.4 SST departures meet or exceed +/-0.5°C along with consistent atmospheric features. These anomalies must also be forecasted to persist for 3 consecutive months.”

    Thank you.

  27. I gave up on prognosticatin’ after getting whupped month after month on Stockwell’s Niche Modeling blog with RSS guesses, but I notice the weekly update has Nino 1+2 looking a little peaked. The models seem to still be predicting a strenghtening though.

    Does anyone know off-hand if any Nino region typically “leads” the rest?

  28. Pingback: The Blackboard » El Nino Betting Pool: ASO NINO 3.4 Average.

Comments are closed.