Thursday
May122011
by 
Bishop Hill
Bishop Hill A climate bet
May 12, 2011 Climate: Surface Toward the end of the Cambridge conference, delegates were offered the chance to put some hard cash behind their opinions on AGW in the shape of a bet on temperature trends. The wager was put forward by Dr Chris Hope of the Judge Business School in Cambridge. This is what is on offer:
If the global mean temperature in 2015 is more than 0.1 deg C below the global mean temperature in 2008, I will pay £1000. If not, the other party will pay me £1000. The global mean temperaure to be determined by the NASA GISS data set.
If anyone is interested in taking part, drop Chris an email at chris dot hope and the domain is jbs dot cam dot ac dot uk.
Reader Comments (91)
Dear Angus,
You're right, the graph in the paper uses HadCRUT. So they are 2-0 down, and needing snookers in the third frame in a best of seven match.
Akasofu talks about a +-0.2C 60 year cycle superimposed on a 0.5C/100 year linear trend
If you take 2008 as being the high end of the 60 year cycle then a 2015 prediction would be of the order 0.11C less,
...........but we are talking in absolute terms: short term variabilty + 60 cycle + linear trend = 14.44C.
Comparing that directly to 2007 and 2009 higlights that very point about short term variability.
Lets take an average of these three years - 14.53C
A more realistic bet based on Akasofu's analysis would be to say,
If you checked the odds with your local bookmaker that could be worth a punt.
Now if you included years 2005 thru to 2010 to provide an average for this bet well I think Chris Hope would looking to offset the money as quick as he could.
2015! Love it! The doomsday of 2100 is too far, but the IPCC 2050 is already far, far beyond 2011. The start-point for disaster used to be 1980, now seems to have slithered to 2000. Doesn't matter. 2015 is 30% of the way to 2050. If we can't tell what is going on by then, it ain't happening or the required rebound events are so far out of the IPCC models that the models are ... wrong.
The trouble with the 0.1C by 2015 is that the number is within the normal +/-0.2C of the global temperature changes. The SST varies by +/-0.15, approximately. What is important is the other areas of the world that show multiples of the global warming events. The land data specifically - Meteorological Station Data, I think GISSTemp calls it, shows about a 2.1X change relative to the global average. The Arctic, about 3.5X.
This regionally excessive response is a key component of the IPCC global warming scare: global is not global, and regional dwarfs global. The skeptics can use this weirdness to more easily point out the errors.
The "allowable" 2C planetary warming is bizarre in this light. The continental USA will go up by about 5C under their math; very significant, and for crops, very bad (wheat crop yield very sensitive to 35C temperatures). So you would think an acceptable rise would be, well, acceptable.
This regional vs global disconnect has recently become more obviously missed by the main groups on both sides when Archibald's predictions of 1.5C in Europe and 2.0C at the US/Canadian border were discussed on WUWT. He looks to Sunspot Cycle 24 bringing these effects, but his comments refer to the land masses. What is the multiple to be used when correcting to find the global drop? Perhaps my earlier estimate was excessive (Ithought a translation to -0.3C globally would be right), but a 3X multiple is probably not bad. So a regional drop of -2.0C in the northern states shows up as a -.05C on the global stage.
For this wager we need a startpoint as well as a non-2011 endpoint. The difference needs to be outside the 5-year variability. We also need a start reference point the warmist would agree to.
Suggestions?
The idea behind the proposed bet, that people with more money than sense would create an index-type market by betting on things that none of us can predict, is not a new one and does seem to work quite well except in cricket where it has attracted some fairly corrupt individuals. But you need a start point that is nearer the median view of those who consider themselves informed, rather than an outlier, and you need a data set that commands wider respect.
We don't know what the effect of UHI is on global temperature measures, nor do we know how much the GISS treatment of the sparse Arctic data has affected their overall trend, so rather than get into the specifics of those issues why don't you go with UAH and the 5 year average from 2006-2010?
The bet should then be that the 5 year average for 2011-2015 is either more or less than .075ºC above that for 2006-2010, ie 1.5ºC per century warming trend, giving believers the whole of the IPCC range and sceptics anything below. Any interest Dr Hope?
This gets interesting. Lets take an average for 2006 to 2010 to counter natural variability around the bet year 2008.
Average temp = 14.56C
Now take that as the temp for the Akasofu cyclic high point in 2008.
Then the determined Akasofu temp for 2015 would be 14.45C.
So there is very little chance that Chris Hope would have lost his original bet.
He must have known that !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Is this another trick?
If I wanted to bet against warming this isn't how I'd do it - I'd rather bet on the individual components embedded in the presumptions that temperature will rise.
For example, what predictions is Chris Hope making about the oil price in 2015? Energy price affects consumption, and consumption affects emissions. Hence energy price affects emissions. What's the consensus assumption for the 2015 oil price? If you can predict temperature changes caused by emissions you must be able to predict the oil price too. A huge salary at Goldman Sachs awaits those whose can.
What predictions do CAGWers make about emissions of CO2, accumulated atmospheric CO2, and what correlation do they predict between those? They seem poorly correlated to me, so I'd be interested in a bet on the level of correlation.
Some of the constituents of warming will take longer to manifest - technology change, land use change, industrialisation / deindustrialisation, third world infant mortality, population growth - so it will be about 2030 or so before CAGWers' current assumptions start to look indisputably wrong. But you can be wrong about the oil price and about CO2 starting right away. What do you say, Chris?
Weather is not climate, we are told. So why bet on a single year versus another single year? Why not, say, a 3-5 year average (like 2913-2017) versus a similar baseline (say 2005-2010)?
Considering that Mac has shown how Chris cherry picked his year I may have been unwise to offer to take up his bet if he was willing to gamble on a 0.1 degree of actual warming, particulalry using Hansen's figures, though he may no longer be employed there in 2015 I will stand by the offer if he goes for it.
http://tvnz.co.nz/close-up/mr-climate-change-new-zealand-s-future-5-48-video-4169480
James Hansen is even in New Zealand spouting his guff. Seeing is believing. I'd heard what Hansen was like but seeing him in the flesh rabbiting on about, inter alia, 75 metre sea-level rises is quite something.
Going by comments I've seen, he seems to have rattled quite a few people - the gullible.
The interviewer was disappointing in that he never asked single question that might've put Hansen on the spot.
I believe this to be quite a good bet to take if you look at the data. Looking at the Hadley data, there appears to be a cyclical pattern. Most who follow the data (or at least have seen presentations by Monckton and Carter) will know that there was a warming period from 1910 to the mid-1940s remarkable similar to that between 1976 to the present in time and magnitude. Will also know that there was a (lesser?) warming period up to 1878.
What is most remarkable is the drop in average temperatures at the end of the warming periods. In the 23 months from Feb 1878 to Dec 1879, the anomaly dropped by 0.782 degrees. In 17 months from Aug 1945 to Dec 1946 the anomaly dropped by 0.736 degrees.. Aug 1945 and Feb 1878 are separated by 810 months. We are now 790 months from Aug 1945. So a significant fall in temperatures might be likely.
Put in context, these two peak to trough falls are about the same as the 20th century warming trend.
FijiDave - if "the science is crystal clear" Hansen is still seriously suggesting 75m sea level rises he needs locked up, for his own good, as much as everyone elses.
David S: "The idea .. does seem to work quite well except in cricket where it has attracted some fairly corrupt individuals."
And the contrast with 'Climate Science' is what exactly?
The average temperature in 2015 (as made up by "Honest Hansen"), like the Summer 2015 Arctic Sea Ice coverage is only fractionally more interesting than would be a model "prediction" of how many angels "might" dance on the head of a pin.
The Climate models are the only "evidence" that there is a problem (despite the awkward fact that they clearly have almost no predictive skill). It is also notable that the much hyped doom predictions (produced by yet more models) are the reason we are hozing Trillions into "low carbon energy" scams which don't really work.
I'd be very interested in placing a substantial bet on how much "renewable" electricity (averaged over the year and based on the published figures) will be produced in 2015.
LoonyHuhne was telling the HoC that we now get 7% of our electricity from "renewables". Since we're spending so much, who will bet on (say) 10%?
Come on Chris Hope! You are one of those blowing up this bubble, entirely careless of the 25% (and increasing) portion of the population who are already in Fuel Poverty.
Fancy a flutter?
Jane
You have got me there. I give up.
Chris - yes - how about how many rolling blackouts there will be in 2020 when we have lost 15-18GW from aging coal and nuclear plants which have only been replaced by unreliable and inefficient windmills, which produce diddly squat in winter cold spells? Or how many old people will die or suffer from malnutrition related illnesses, due to fuel poverty? (it is already happening - http://www.thecourier.co.uk/Community/Health/article/12975/third-world-conditions-in-tayside-linked-to-heating-or-eating-quandary.html ).
Why would anybody trust GISS data?
Divergence Between GISS And HadCRUT:
http://stevengoddard.wordpress.com/2011/04/21/divergence-between-giss-and-hadcrut/
Chris Hope,
I'd suggest you offer a bet similar to this: http://rankexploits.com/musings/2011/bastardis-wager-you-bet/
Its much less subject to natural variability.
Chris, how about "Will 2015 be as warm as it was in the Medieval, Roman or Minoan Warm Periods"?
http://jonova.s3.amazonaws.com/graphs/lappi/gisp-last-10000-new.png
I suspect that the number of "scientists" actively employed, predicting the consequences of catastrophic global warming, will be lower in 2015, than any other year this century.
Dr Chris Hope announced this as a 'business opportunity'. It would seem that the Judge Business School isn't very clued up about business. Del Boy would be proud!
Thanks for all the comments and sorry not to have been able to get back on the site before now to respond.
I came up with the bet at a conference where the majority of the people either did not believe the global temperature is increasing, or believed that if it is, it's nothing to do with our activities and could easily go into reverse at any time - the Akasofu hypothesis. Some people seemed quite angry that their views were not being taken seriously; I spoke with James Delingpole for instance and he seemed very angry. So this seemed a good way of showing that I did take their views seriously by putting up £1000 of my own money and inviting them to do the same.
Unlike many people on here, I don't think that I'm bound to win the bet. As Angus has pointed out, the data for 2011 so far look bad for me, and rather good for anyone who is willing to take the bet. The anomaly for 2011 so far is 0.26 degC vs 0.33 for 2008 (using HadCRUT data). Even using the GISS data it is 0.43 for 2011 so far vs 0.44 for 2008. As I said before, I'm willing to use another data set like HadCRUT rather than GISS to settle the bet. So is there really no one on here who thinks there is a good chance that the temperature anomaly in 2015 will be at least 0.04 degC lower than today, which would still leave the temperature over 0.2 degC higher than the 1961 - 90 average?
Seems to me this really has nothing to do with believe/don't believe AGW. Chris Hope just *hoped* (hehe) to make a ton of money out of a lot of mugs. Can't blame him, academic salaries being what they are these days...
Stephen Richards? Manicbeancounter? Send me an email if you are interested.
Dear Zeke,
Thanks for bringing your bet to my attention. Very nicely stated and reasoned. What is the current status of it?
Chris - thanks for your further thoughts. I think there is a good chance that it will be cooler, but I won't be taking your bet for reasons explained by others above; 2015 could be a warm or cool year due to natural variability - the AGW signal (if there is one) is tiny compared with the noise from the chaotic climate system. There is more heat energy in the top 6 feet of the ocean than there is in the entire 20 miles of atmosphere, and oceans have currents, which move this heat around. A 1% or 2% change in average cloud cover for the year will have a much more significant impact on global temperatures than an extra 10%, 20% or 30% CO2. And the H2O cycle is even more chaotic than ocean currents. Your bet is equivalent to taking a bet on whether it will be raining or sunny in Fort William in two weeks time. May is traditionally the driest month in the Highlands, but Ben Nevis can generate rain clouds almost any day.
Hi Chris,
I'll raise you.
$10,000 says that we won't see 50 million climate refugees by 2010.
$20,000 says we will see snow again in our lifetimes.
$30,000 says there will be ice at the North Pole in 2015.
Time to man up.
If we are using the Akasofu hypothesis as the basis for this type of bet, then you have to account for and discount natural variability. The best way to do that is to construct averages around the 2008 and 2015 bet years.
Taking the average for 2006, 2007, 2008, 2009 and 2010 for GISS temps gives 14.56C in absolute terms.
You could do the same for HadCruT, UAH or RSS. What we are testing and betting on is the Akasofu hypothesis.
Now the Akasofu hypothesis basically states that the we have +-0.2C 60 year cycle superimposed on a 0.5C/100 year linear upward trend, and that 2008 is a high-point inn the 60 year cycle. Calculating the global temp change on this basis for 2015 gives in absolute terms gives a prediction 14.45C, a 0.11C drop in temps.
Again we have to account for a certain amount of natural variation, so I would argue to complete the bet an average has to be taken for 2013, 2014, 2015, 2016 and 2017 to determine who has won the bet.
NOTE: Of interest is the someone like Chris Hope entertains the notion of cyclic climatic events that impact on global temps. There is evidence to suggest that not only do we have a roughly 60 year cycle but also rougly a 1500 year cycle and a smaller 150 year cycle that impact on global temps. It is this much longer 1500 year cycle that may gives rise to this underlying linear trend that we eyeball in the GISS and HadCruT data. As any engineer will testify complex wave-like processes can be broken down into their constituent parts using Fourier analysis, and that you can get quite close to the orginal complex waveform by simply adding some of the main cyclic constituents.
Has anyone used this approach with the various data sets at hand?
Because if you did you could make forecasts and hindcasts of global temps to determine future predictions and the limits of such a method, and if the method proves useful you could further try to understand the nature of these natural cyclic variations. It all adds to the picture.
'I came up with the bet at a conference where the majority of the people either did not believe the global temperature is increasing, or believed that if it is, it's nothing to do with our activities and could easily go into reverse at any time - the Akasofu hypothesis. Some people seemed quite angry that their views were not being taken seriously; I spoke with James Delingpole for instance and he seemed very angry. So this seemed a good way of showing that I did take their views seriously by putting up £1000 of my own money and inviting them to do the same.'
Chris, thanks for your engagement and interesting comments and contributions here, but I think you seriously misunderstand and/or mis-characterise what people do or don't 'understand' or 'believe'. Their beliefs and concerns are very wide ranging and while it is politically useful to place all under a convenient label, it contributes nothing to the advancement of the discussion.
For example, as many have stated here, it is not that they don't 'believe' that temperatures are increasing, but rather that with the current data and knowledge, and by virtue of the overall make-up of the climate system, we have absolutely no way of knowing what will be happening in 2015. As such, setting a lower baseline, and a still lower threshold level for an opponents winning bet, is simply offering a biased sucker bet on random or chaotic and unknowable system behaviour.
Chris has not budged from a bet which he would be far less than 50% likely to lose if Akasofu cooling is correct (because he has cherry picked the coolest year as a base). This is quite remarkable for somebody who is paid to persuade us to spend trillions betting on diminishing an alleged catastrophic degree of warming.
I bet that you can't find a positive integer less than zero. After all, there's an infinity of choices, and as we all know, if something's infinite, it must be there somewhere. Right?
That bet makes about as much sense as this one.
Dear Neil,
At the conference there seemed to be people who had views about climate that would have made the bet a good deal for them. I still think there might be others on here.
Dear Ollie,
Not sure I understand you, but are you saying that the temperature being 0.1 degC lower in 2015 than in 2008 is a logical impossibility?
Interesting that you immediately jump on my (quite deliberate) spurious logic, and yet don't respond to the many valid criticisms of your proposed bet, which have all been up far longer than my post.
Curious.
Dear Ollie,
I have tried to respond to the comments, to the extent that I worry that I am wearying people. I have, for instance, agreed to use other data sets than GISS to settle the bet if people object to GISS. We seemed to have reached the curious position where the many people on this blog who don't think we are affecting the climate are apparently more confident than I am that the temperature in 2015 will be no more than 0.1 degC below its value in 2008!
I note two studies that have studied the cyclic nature of recorded temperatures.
http://www.springerlink.com/content/g28u12g2617j5021/fulltext.pdf
http://www.fao.org/docrep/005/y2787e/y2787e01.htm#TopOfPage
Both these studies note a significant 60 year cycle.
There are also Bond events
http://rivernet.ncsu.edu/courselocker/PaleoClimate/Bond%20et%20al.,%201997%20Millenial%20Scale%20Holocene%20Change.pdf
Basically a 1500 year cycle.
More here;
http://www.ncpa.org/pub/st279/
Also there is Craig Loehle's study that suggests that the impact of a 1500 year cycle and possible a 60 year cycle on global temps.
http://www.ncasi.org/publications/Detail.aspx?id=3025
This all raises the possibility that it maybe possible to reconstruct past global temps based on these cycles going back 2000 years and raises the possibility of projecting ahead for the next 100 years.
All that would be required to do that would be amplitude, periodicty and phase.
Hindcasting would give some determined confidence about the method, the projections for the next 100 years could be used to test AGW.
Maybe a new bet can be constructed on the basis of that approach.
The only thing curious is that you don't allow for year-to-year natural variation. The original bet is heavily biased in your favour on that basis.
I would argue that if you are taking the Akasofu hypothesis seriously then an average for 2006, 07, 08, 09 and 10 should be compared against an average for 2013, 2014, 2015, 2016, 2017.
Initial average temp: 14.56C (Hi 14.63C. Lo 14.44C)
Target average temp: 14.45C (+- 0.1C)
Not much of a bet over such a short period.
Perhaps 2025 would prove better.
Well I wasn't there Chris but I don't get the impression that there were a large number going round saying the Akasofu theory is so optimistic that the probability is the cooling will not only be worsebut will be worse compared to the single coolest year cherry picked by you.
The fact remains that you and indeed everybody else earning their living promoting a catastrophic level of alarmism (often predicting 6 C a century = 0.53 C for an 8 year period) & costing the taxpayer trillions aren't willing to put £1K of their own on the prediction it will be 0.1 C warmer than the cherry picked coldest.
"We seemed to have reached the curious position where the many people on this blog who don't think we are affecting the climate are apparently more confident than I am that the temperature in 2015 will be no more than 0.1 degC below its value in 2008!"
I and many others have clearly explained why few if any people on this blog will take up your bet, i.e. cherry picking 2008, natural variability, and because 2015 isn't far enough away to make any cooling/warming signal discernible from the noise. As for initially suggesting the GISS dataset: http://stevengoddard.files.wordpress.com/2010/10/1998changesannotated.gif?w=500&h=355&h=355
http://www.youtube.com/watch?v=DFbUVBYIPlI&feature=player_embedded
Enjoy the Holocene while it lasts.
"The global mean temperaure to be determined by the NASA GISS data set."
The way the NASA GISS surface data set mutates with time, will make Chris Hope's bet pretty safe.
If the temperature followed a random walk then 0.0 would be the line where even odds were fair (50 % chance of up or down).
These guys tell us the temperature 'walk' is not random but actually biased towards upward steps, and yet he still insists on a line well below 0 (<2008) for his even odds.
Obviously whoever proposed the bet isn't confident the temperature will rise.