Eagle eyed Chip Knappenberger noticed fiery-tounged Joe Romm offered a bet he would win even if the IPCC multi-model mean over-predicts warming. In a post where he sets flame to Andy Revkin Joe makes this offer:
“I will be happy to bet anyone that the 2010s will be the hottest decade in the temperature record, more than 0.15°C hotter than the hottest decade so far using the NASA GISS dataset.
Gosh.. Joe is willing to place the cut-off at only 0.15C warmer than the hottest decade so far? Even though IPCC models predict “about 0.2C/decade” of warming during the first part of the 21st century? And slow warming since 2000 is supposedly due to “weather noise”?
It’s interesting to ponder what two people who believe the IPCC models are on track might consider an even money bet– you know, the the type where they would be equally willing to wager on either side of the bet.
Let’s consider a few ways of finding the “even bet” temperature rise two people who believe IPCC models might reason out:
- The IPCC projects “about 2C/century” of warming during the first three decades. This simplistic method would suggest that the decade from 2010-2019 should be roughly 0.2 C warmer than the decade from 2000-2009. Joe is willing to take bets– at a lower rate of warming.
- Using the SRES and computing using August 1999-2009 vs August 2009 – 2010, the IPCC multi-model mean projects the latter decades will be about 0.19 c warmer than he earlier one. Once again, Joe is willing to take bets– as a lower rate of warming.
- There are some out there who believe the current decade has been cooler than projected due to “weather noise”. If so, then one might suggest that “weather noise” will regress to the mean. Suppose, that the current decade can be explained as falling 1 standard deviation below the multi-model mean due to “weather noise”. The standard deviation of 10 year average temperature over 22 models used in the A1B scenario was ±0.032C. If the weather reverts to the mean, one would expect the temperture rise to be 0.22C. But Joe is willing to bet “at least 0.15C”.
- There are some who might suggest that if this decade experience cool “weather noise”, we might be due for a warm decade. Also, some might suggest that the current cool weather noise is partially due to the minimum in the 11 year solar cycle, which is due to rise from the minimum. If so, those people might consider a rise of 0.26C to be the fair bet. But Joe is willing to bet “at least 0.15C.”
- Finally, the trend since 1980 has been 0.158 C/decade. So, people who base their bets on extrapolation might think a gain of 0.158C represents a fair even money bet. But Joe is willing to bet “at least 0.15C.”
There might be someone out there who is willing to bet Joe “Hell and Highwater” Romm taking the low side of “at least 0.15C”? Maybe.
But given Joe’s foaming at the mouth rants regarding anyone who seems to think warming might fall anywhere below the most extreme upper bound of IPCC projections, his offer to bet money on “over 0.15C” seems a bit, well, lame.
Personally, I’m waiting to see someone who likes to gamble offer to bet Joe the average temperature in from 2010-2019 will be “no more than 0.26C” higher than 2000-2009. If you buy into the IPCC projections, that bet seems as fair as the one Joe offered.
I wonder if Joe would take that bet?
There should be lots of people who would take Joe Romm’s bet since they think AGW is a fraud or climate scientists are stupid. They should be eager to take his money. Jeff Id?
“AGW is a fraud or climate scientists are stupid”
I sort of agree with this. I believe AGW is a fraud, but it’s not the climate scientists that are stupid, if you catch my drift. 😉
Andrew
Boris,
Does Jeff Id think the surface is not warming?
The interesting thing is that Joe won’t place a bet that’s in line with his rhetoric.
“The interesting thing is that Joe won’t place a bet that’s in line with his rhetoric.”
I agree. But this is true of skeptics as well. Lindzen claimed that warming was equally likely as cooling over the next 20 years, but wanted 20-1 to take the bet.
Sure.
Out of curiosity, did Lindzen throw that offer “out there”? Or was it a counter offer in response to someone else’s offer?
To some extent, if Romm publically challenged me to a bet that was not inline with his beliefs and that I knew gave him a huge edge based on his beliefs, I might counter with an offer that is equally ridiculous.
Based on what Romm rhetoric I’ve read, I think Chip’s suggestion that Romm should be offering bets that the decadal will be no more than 0.50C would be fair might represent a ‘even money’ bet if Joe were betting with someone who shared his outlook on warming.
I really hate the thought of betting… but if I got to vote the cool side of 2010-2019 being no more than 0.50C relative to 2000-2009, I’d be pretty tempted to place a bet!
If Pielke Sr. et al are right, 0.15 degrees may be less than the built-in warming bias of most current surface measurements.
And with a little Hansenesque thumb-on-the-scale correction algorithmic adjusting past and present outputs…
Hell, Joe oughta go 0.35 if he really believes…
What skeptic would take that bet based on GISS? I’d trust Nathan Detroit before I’d trust “Big Julie” Hansen.
If he’s talking average for the decade you would expect the value to be less than .2 deg given that the temps are flat right now wouldn’t you?
OTOH, for the IPCC projection to hold what would be the rate that you would have to see to make up for the present rate? In other words .4 deg by 2019.
Give Joe Romm a break. I’m sure he’s read Mojib Latif’s projections that global tmperatures may run stable or slightly cool due to natural variations for up to a decade. Placing a bet lower than the projected average decadenal increase may simply reflect some internal conflict between the models and “natural” variations.
BarryW–
Hansen is applying a published methodology, relies on data collected by others, his code has been “freed”, and it is scrutinized. There is no way he can fiddle with the reported values in any nefarious way and have the fiddling remain undetected over a decade. I know you don’t trust him. But, the thing is, his successfully fiddling the future reported temperatures a lot is Not….even… possible.
Sean– Joe is criticizing Andy Revkin for reporting the truth. Then to emphasize how wrong Andy is, he proposes a bet that aligns with Andy’s story being fair and accurate. How does this make sense?
Didn’t Joe offer 2-1 odds?
Regarding Lindzen’s bet, in the only followup to the story, the reporter messed up. Supposedly he said I would bet on cooling, and then Dr. Annan contacted him, and claimed that Lindzen wanted 50-1 odds.
Reason then followed up and reported that Lindzen said he was misquoted and he only meant to say the odds were 50-50, which sounds like the same thing.
He offered a counter bet along the lines of if warming is less than .2C in 20 years I win, and if it’s more than .4C you win, in between the bet is off. He wanted 2-1 odds, but I’m not sure who gets the odds.
Joe Romm has updated on his site.
UPDATE: Global warming deniers, who are famous for making stuff up and taking things out of context, are trying to spin my proposed wager as a statement of what I actually believe the warming will be in the next decade. It ain’t. I was in fact offering a specific wager to call out the various delayers out there — or to see if Revkin himself would be willing to back up his absurd statement that “global temperatures … may even drop in the next few years.â€
For the record, climate warming is not a linear phenomenon, it is an accelerating phenomenon, in part because of well-known delays in the equilibration of global temperatures with all the exogenous forcings and in part because of positive, amplifying feedbacks in the carbon cycle — the impact of global dimming is also relevant to the recent and near-term trend line. So the planet can warm, say, 0.2°C next decade — just as it warmed 0.2°C this decade, using NASA’s data, which is probably the best — and still warm 5°C this century, if we don’t act quickly to reverse emissions trends. That said, I am expecting the planet to warm more than 0.2°C next decade, particularly if there is not a major volcano.
NASA’s data is “probably best” why again? Because…It shows “.2 degrees over the last decade” or whatever. I think that sounds totally wrong, but I’ll let others work it out.
Romm is now being very silly with this backpedaling.
And Andy ain’t no delayer.
“For the record, climate warming is not a linear phenomenon, it is an accelerating phenomenon, in part because of well-known delays in the equilibration of global temperatures with all the exogenous forcings and in part because of positive, amplifying feedbacks in the carbon cycle”
This is not credible! I don’t get how many times it needs to be said-Chip specifically looked at the twenty year trends in models, and Romm’s rate is way less than the models show. If the model rates aren’t indicating the “accelerating phenomena” (they generally don’t) then they must be missing Romm’s “exogenous forcings” “amplifying feedbacks” etc.
MikeN– I don’t see the 2-1 odds offerred by Joe. The article is long– maybe I’m missing it.
I think the update is hilarious.
Who says that his wager is a statement of what he believes warming will be in the next data? I always assumed his bet was rhetorical posturing. But what’s the message? Even as a rhetorical ploy, he doesn’t pick rate of warming commensurate with his dire warming. Maybe he’s afraid Andy would accept the bet if Joe were willing to wager that the trend would be 0.26C or above? Then Joe would either have to say “Just kidding… I didn’t mean to bet that!” or swallow hard and accept the risk.
Evidently, Joe ain’t willing to risk that, so he picks 0.15C.
You also got to wonder why Joe thinks this statement is absurd
Recent papers suggested negative trends of over 10 years are consistent with models and/or brisk. Is Joe telling us he doesn’t believe that. 🙂
Also, if Joe is trying to rebut that, why doesn’t he offer a bet applying to the next few years? Is a decade a “few” years? I should think 3 to 5 is “a few”. (And in any case, Andy said “may even”, that hardly suggests Andy is predicting temperature are just as likely to drop as rise over the next “few” years. )
I think Joe needs to calm down and breath while he reads Andy’s articles. That will help oxygenate his brain cells and improve his reading comprehension.
I seem to remember that Joe Romm once said at Climate Progress on April 13, 2009 that the “median” forecast was for warming in the US by 2100 of 10-15F, or 5.5-8.3C, and he made it very clear that if he had to pick a single number, it would be the high end of that range.
On average, the 8.3C implies about 0.9C per decade of warming. This might vary slightly by what starting point he intended (he is not very clear in the post) and I understand there is a curve so it will be below average in the early years and above in the later. But it sure is daring for a guy who is constantly in the press at a number around 0.9C per decade to a number 6 times lower when he puts his money where his mouth is.
“Global warming deniers, who are famous for making stuff up and taking things out of context”
As a Global Warming Denier, I’d like to know what stuff/things did my Denier Brethren famously make up and/or take out of context?
Did we do Atlantis? That was a whopper wasn’t it? 😉
And Al Capone’s Vault? That one was gold, Jerry. GOLD.
Andrew
So Boris,
what you are saying is that Warmers, who understand weather could upset the bet, are smart, but deniers, who understand that weather could upset the bet, are stupid.
Does that sum up your statement??
Wow,
So even Joe Romm doesn’t believe Hansen’s “Scenario B”!
“what you are saying is that Warmers, who understand weather could upset the bet, are smart, but deniers, who understand that weather could upset the bet, are stupid.
Does that sum up your statement?? ”
No. What I am saying is that if Romm believes that we should see 0.30C of warming over the next decade and a denier believes that we should see 0.00C of warming then 0.15C should be an equally attractive bet for both of them.
Theoretically.
“As a Global Warming Denier, I’d like to know what stuff/things did my Denier Brethren famously make up and/or take out of context?”
Really?
“No. What I am saying is that if Romm believes that we should see 0.30C of warming over the next decade and a denier believes that we should see 0.00C of warming then 0.15C should be an equally attractive bet for both of them.
Theoretically.”
Why would you assume that? It would depend on the probability distribution each party would hypothesis.
Boris–
Maybe if you found someone who specifically predicts 0C/century for the next decade, the 0.15C might be fair. (That is assuming Je only expects 0.3C/century w hen it sometimes seems he expects 0.9 C/century).
But about an AGW-non-believer who thinks the recent 0.158 C/century is the result of rebounding from the little ice age and that we will continue rebounding? Or due to cosmic rays which may still be doing whatever cosmic rays do to warm? Or who thinks it’s due to the sun and the sun is likely to revive?
Why would they take Joe Romm’s bet?
And anyway, didn’t Joe offer the bet in a post criticizing Andy Revkin? Is there any reason to believe he thinks the best estimate for warming in the next decade is 0C? So what point is Joe’s offer to bet supposed to communicate?
What it does communicate is that Joe offers bets where he wins even if the IPCC has wildly over predicted warming!
Romm turns down a serious offer of a bet of 0.30:
http://climateprogress.org/2009/09/22/new-york-times-andrew-revkin-suckered-by-deniers-to-push-global-cooling-myt/comment-page-1/#comment-125508
Lucia,
Surely there are people who think we will be cooling. I wouldn’t expect people who think cosmic rays are going to cook us to bet, but there is a huge chunk of the denial community who thinks we haven’t warmed and we won’t warm.
So if you are after the highest EV bet, you should offer lowball temp rises because there are people who will (er, should) take them based on what they believe and say.
No idea on Joe Romm and Revkin. I’ve said before I don’t follow Romm and what I’ve seen of his rhetoric I don’t like.
Boris: “Really?”
Andrew_KY: “Really”
Let’s have ’em. Should roll right off the tongue since That’s What We Deniers Are Famous For, right? 😉
Andrew
Lol at James M Taylor offering Romm a terrible bet.
Boris,
If Joe Romm believes we will have .30c why would he bet low at .15c instead of high at .45c???
Same thang for the denier. If he believes in .00c why would he rather bet on .15c, the same as Romm rather than -0.15c???
Of course, in each case it shows their doubt in their base position, .30c and .00c. Or, it could also show they both have a slight idea of probablility and want the best chance to win while seeming to support their position!!
Another excellent example of “Money Talks, BullS#!t Walks!!”
Where is your bet Boris??
Andrew,
Check out my favorite. The worst climate paper ever: Khilyuk and Chilingar Touted by Pat and Chip at World Climate Report, cited by Lindzen. I fear, however, that you may find it convincing.
And don’t you guys find it odd that it’s the Deniers (name implies we are going around denying things) who are famously ‘making things up’? If ‘Deniers’ are famous for ‘making things up’, should our name be ‘Famous Maker-Uppers’ rather than Deniers? lol
Andrew
Lucia,
“But, the thing is, his successfully fiddling the future reported temperatures a lot is Not….even… possible.”
Would you bet a large amount of money without knowing whether or not JH has some friends at NOAA who can add/drop stations??
HAHAHAHAHAHAHAHAHAHAHAHAHAHA
“If Joe Romm believes we will have .30c why would he bet low at .15c instead of high at .45c???
Same thang for the denier. If he believes in .00c why would he rather bet on .15c, the same as Romm rather than -0.15c???”
The point of betting is to win money. Why would people bet in such a way that they will lose substantial amounts of money if they are right?
If I think the Chargers will beat the Dolphins by 3 points, why would I go out and bet that they will win by 6?
I really cannot fathom your question. But you are welcome at my home poker game!
Not sure why you believe “his successfully fiddling the future reported temperatures a lot is Not….even… possible.”
He fiddles with the past temperatures with every “adjustment”, but that won’t affect the future eh…..
“there were quite a few changes propagating back to October 1882.”
“I’m not sure the changes consistently make the past colder. At least sometimes past temperature have risen!
That said, the shifting past temperature tends to give ’some’ the heebee-jeebees.”
I have yet to be convinced that current temps require you to adjust temps back to the 1880’s to keep things “correct”.
Boris,
Ah yes… the famous Khilyuk and Chilingar papers. Those were popular gifts around Christmas a couple of years back.
Maybe our concepts of
Main Entry: fa·mous
Pronunciation: \ˈfÄ-mÉ™s\
Function: adjective
Etymology: Middle English, from Anglo-French, from Latin famosus, from fama fame
Date: 14th century
1 a : widely known b : honored for achievement
aren’t quite in sync.
Andrew
Haha…I’d play word games too if my logical position were as poor as yours.
Boris,
“The point of betting is to win money. Why would people bet in such a way that they will lose substantial amounts of money if they are right?”
So your position is that betting has nothing to do with who believes what, but, only in winning the money.”
Your statement then supports the position that Joe Romm, who has been castigating deniers, and warmers who are not extreme enough for him, really believes that the temps are going up at a rate that is a non-issue and is a total FRAUD!!!
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
Thank you.
Now, where is YOUR bet!!!
“you may find it convincing”
Boris, I can sincerely state that I do not.
Andrew
“Your statement then supports the position that Joe Romm, who has been castigating deniers, and warmers who are not extreme enough for him, really believes that the temps are going up at a rate that is a non-issue and is a total FRAUD!!!”
I shudder at the abuse of logic that leads you to this conclusion. You should never be allowed to be alone with a syllogism.
Seriously
This is the stupidest post I have read in a long time.
What a bunch of morons.
You don’t like his bet don’t take it up. Simple.
Idiots.
What does the Two Box Model say?
——————————————
Boris – you seem like a nice guy. Vigorously misinformed and not afraid to show it, but a good guy all the same. Have a great evening!
“You don’t like his bet don’t take it up”
We don’t and we’re not. Simple.
Andrew
Boris–
Joe offerred a bet he considers a sucker bet and claims that makes some sort of point. James Taylor makes a counter offer sucker bet he considers a sucker bet. Joe won’t take it. Joe gives a complicated reason.
But the likely real reason Joe doesn’t take James counter offer is simple: Joe thinks there is a high probability he would lose.
Ok… maybe I’m wrong and Joe really has some mysterious convoluted reason why he thinks his offer was some sort of lesson, while James is some manifestation of “linear thinking”. But no matter how Joe tries to spin this, the fact is: Joe refuses to take James bet.
People will process this information. Many will decide what they think it means for themselves– the don’t need Joe to tell them! Hah!
But we’re still going to gossip about Romm’s publicity stunt. Wasn’t that what he hoped for? (He just didn’t seem to have a clue what we’d think about it!! )
Boris speaks of “huge chunks” of the “denial community”.
I seem to recall Boris referring to Spencer and Christy as deniers. And Lindzen. And Chip, and Pat, etc. etc.
All of whom are on record saying it has warmed and some on record saying they think that will continue.
So we are clear Boris, is it your view also that 9/11 truthers are representative of the entire anti-war movement?
I predict Boris’ response “No, [insert explanation for his inability to consistently apply logic]”
I’ve done this dance before. As someone once said “You sir, are a moron.”
“But we’re still going to gossip…”
Indeed. 😉
And Lucia, I’m glad you have good-spirited wagering at your site, and not the Bitter Betting that goes on at other nameless sites. 🙂
Andrew
The fix for Joe’s sucker bet isn’t in the trend level undershoot, it’s in his choice of the keepers and the data base. Try a satellite data-based bet and you could probably give flailin’ Joe odds.
mccall–
Based on his answer to James Taylor, I get the impression Joe doesn’t want to negotiate the bet or consider other bets. He has some delusion that his bet would “call out the various deniers out there,”. I have no idea what that means nor do I know if he succeeded in calling anyone out or even making any sort of point.
But yes, if he’d offered to bet the IPCC trend and used satellite data, he almost certainly would have found a taker. It wouldn’t have been me, but it would have been seen as a fairer bet by the people Joe calls “deniers”.
I guess if Joe wanted, he could make some sort of point by progressively lowering his bet threshold. Next week he could offer a threshold of a 0.20C rise. The next week 0.25C and so on. Eventually, someone would take his bet. He might “prove” something that way– but he would also be forced to actually bet because at some level he’s going to find a taker. (James Taylor is evidently willing at 0.3C.)
“But we’re still going to gossip about Romm’s publicity stunt.”
What you have here is a pile of garbage.
What you have is your own publicity stunt.
“He has some delusion that his bet would “call out the various deniers out there,â€. I have no idea what that means nor do I know if he succeeded in calling anyone out or even making any sort of point.”
This comment is great. You claim that he has a delusion, but then admit that you don’t know what he means! How can you come to the conclusion he is deluded when you don’t even know what he is talking about?
This blog is a crack up.
I think this sums it up
“This simplistic method would suggest that the decade from 2010-2019 should be roughly 0.2 C warmer than the decade from 2000-2009.”
Is 0.15 for the next decade so far from about 2/century for the first three decades.
Let’s see some Thuper Thtatistics, show us how different they are, and why not include some fancy errors bars in there too.
Hey Nathan,
Seriously
Yours is the stupidest comment I have read in a long time.
Moron.
You don’t like the post then say something constructive. Simple.
Idiot.
0.15 doesn’t equal 0.20. I don’t think we need “fancy” error bars to show that. Romm bitches constantly about how 0.2C/decade is too low an estimate from the IPCC. He had an opportunity to back it up. He didn’t.
What people miss is the terms “denier” and “alarmist” are not short hand for opinions on the science. They are short hand for one’s *political* opinions.
You can believe that we are headed fro 5degC warming this century but if you also believe that CO2 control policies are wrong and doomed to fail then your are a “denier”.
Of course, Joe and his ilk like to create strawmen and pretend that the people who disagree with his politics must disagree on the science but no one but his choir take such things seriously.
Joel
“0.15 doesn’t equal 0.20. I don’t think we need “fancy†error bars to show that. Romm bitches constantly about how 0.2C/decade is too low an estimate from the IPCC. ”
The problem is, Joel, that the 0.2 was actually “about 0.2”, and was over 3 decades. So the question becomes how close is 0.15c/decade to an average of “about” 0.2/decade.
“He had an opportunity to back it up. He didn’t.”
What are you on about?
Nathan, we don’t have to quibble over the “about” 0.2C/decade and what this means for the first 20 years of the 21st century, we can just look at the model runs themselves. As Chip has done in Figure 2: http://masterresource.org/?p=4922
Lucia, I would take an evens bet at 0.20 based on satellite data. I don’t share your confidence that GISS is fiddle proof, but I guess I would accept 0.30 based on GISS.
There’s a problem with almost all of the bets I’ve seen.
It shouldn’t be a fair bet. I wouldn’t want to make a bet that was 50-50 chance of winning.
So Romm betting .15C is OK, but for the skeptic it is too close to a break even point.
For the skeptic, you would want something like less than .3C of warming. So how does one reconcile that?
If Romm’s belief is for higher warming like .5C he should be willing to bet around .25C, a point at which the skeptic should also feel he is getting a good deal. But if Romm believes the likely warming is lower for the first few decades, then I don’t see a fair bet for both sides as possible. There just isn’t enough of a differential between the two sides.
Well there you are Joel, 0.15 seems like a quite reasonable bet, no?
Nathan–
If you think the topic of the blog post is stupid and unworthy of your time, why have you been wasting your time posting 4 comments?
On this
Yes. But the graph in the report shows nearly linear warming during the first few decades, if you download the runs, you see nearly linear warming, and mean for the 3 decades exceeds 0.2C/decade. Joe knows this. He has given people the impression he thinks that mean is too low.
But, when dreaming up bets, he offers 0.15C. You try to infuse whatever meaning you like, but people will notice he is not willing to place an even money bet at the rate of warming he claims we should expect.
As for this
Romm was given the opportunity to take a bet the 2010-2019 temperatures would exceed the 2000-2009 temperatures by more than 0.30C. Based on his rhetoric, it appears he has been suggesting this is more likely than not. In fact, his rhetoric would suggest he thinks a 0.50 C might be likely. But even though 0.30C would give an easy 0.20 C buffer (according to the believe in 0.50C), he won’t take a 0.30C bet.
This means he won’t back up his words with his bet.
Does this matter? Well… normally. No. But in this context (and as you know from Joe’s response to the offer of a bet at 0.30C we should put this in context…) Joe is the one who first implied the notion the act of being willing to bet some how reveals someones confidence in some particular outcome. He did this when he used his first offer to bet as some sort of “proof” that he had confidence in warming. The challenge if not picked up was supposed to give some sort of “proof” that non-betters by simply not taking up his offer somehow don’t have confidence in non-warming.
Joe continues to imply not-picking up a bet implies lack of confidence in a position– he does this claiming his bet has something to do with demonstrating something to deniers. But Joe won’t take a bet on the outcome he seems to be perpetually threatening us with. (That is warming greatly exceeding IPCC projections.)
Of course people are going to point out that whatever it was Joe was claiming his bet proposal demonstrated, we know that he will not bet for warming unless he gets to pick a value well below IPCC projections and even below the 30 year mean trend!
If Joe thinks his bet demonstrated something, I don’t know precisely what that might be. What he says it demonstrates is is both vague and confusing.
But I do know what I think his bet demonstrated. And Nathan, there is absolutely nothing illogical about my saying what I think his bet shows us.
I know very well that the +0.15 bet is what Joe Romm thinks is a safe bet, i.e. in the lower limit of his confidence interval and well under what he considers to be more likely, but still I am more than willing to accept a bet like that. However I was banned long ago from ClimateProgress. I have tried with a different name and address but I think he got my IP. I cannot post a reply. Hey Joe, if you read this and you really think that your terms mean easy money for you, why don’t you accept it here.
Copy of what I seem unable to publish in Climate Progress, under the line.
———————
What are exactly the conditions of this bet? As I understand it:
Joe Romm wins if the average global temperature for the period 2010-2019, according to GISTEMP data, is greater than the average global temperature for the period 2000-2009 plus 0.15C. Joe Romm loses if the difference is equal or lower than 0.15C.
Well, I am willing to take that bet, but 2020 is too late to collect. I propose the posibility of an early collection in 2015. I would collect if the 2010-2014 average global temperature is not greater than the 2000-2009 in any ammount, making it very unlikely that the whole period be 0.15C greater (a sudden step of +0.30 would be needed for all of the remaining years). Joe would collect if the 2010-2014 temperature is already +0.15C compared to 2000-2009, making it very unlikely that the whole period warming be less than that (a full stop of the warming would be needed for me to win at that point).
Is that acceptable? Or do you want to include some caveats in case there is a huge volcanic eruption? I could accept that, but only in case of volcanoes as big as Chichón or Pinatubo.
Just for fun I used the Woodfortrees grapher to come up with the decadal trends since 1880.
I think Romm may have a point. If you look at the previous decades there have been very few 0.1-0.2 slope trends. They’ve mostly either been much steeper or negative. Hence if you do believe global warming is stopping then you should take his bet. Mind you I’d be interested to make the bet as some kind of a spread (0.1-0.3 no bet, x for the (-0.1-0.1 and 0.3-0.5) range, 2x for (-0.3- -0.1 and 0.5-0.7) etc).
image got swllowed in previosu comment – http://www.woodfortrees.org/graph/gistemp/plot/gistemp/from:1880/to:1889.9/trend/plot/gistemp/from:1890/to:1899.9/trend/plot/gistemp/from:1900/to:1909.9/trend/plot/gistemp/from:1910/to:1919.9/trend/plot/gistemp/from:1920/to:1929.9/trend/plot/gistemp/from:1930/to:1939.9/trend/plot/gistemp/from:1940/to:1949.9/trend/plot/gistemp/from:1950/to:1959.9/trend/plot/gistemp/from:1960/to:1969.9/trend/plot/gistemp/from:1970/to:1979.9/trend/plot/gistemp/from:1980/to:1989.9/trend/plot/gistemp/from:1990/to:1999.9/trend/plot/gistemp/from:2000/to:2009.9/trend
Lucia
It’s hilarious that you see this as significant. This is some really cheap point scoring.
And for you all to “know” this is a bad deal, you have to accept the IPCC is correct. By default you are all accepting the IPCC position.
You’ll have to show us some stats that show this is bad from YOUR position. Where are your ‘Lukewarmer’ stats?
“If Joe thinks his bet demonstrated something, I don’t know precisely what that might be. What he says it demonstrates is is both vague and confusing.
But I do know what I think his bet demonstrated. And Nathan, there is absolutely nothing illogical about my saying what I think his bet shows us.”
Great! Now you say you know what it says, when earlier you said you had no idea.
See:
“He has some delusion that his bet would “call out the various deniers out there,â€. I have no idea what that means nor do I know if he succeeded in calling anyone out or even making any sort of point.â€
Now I will leave so that you will stop making suggestions that I value this discussion for anything other than hilarity.
MikeN
I suspect Joe has enough egg on his face. He opened this can of worms by offering a bet. Now he’d like to stifle discussions of what other bets he might consider. Those discussions would provide further evidence he is not willing to place 50/50 bets at the level of warming he claims he believes has a 50/50 chance of occurring.
On the one hand, he’s not required to place the bet. On the other hand… he brought up the subject of betting. How did he think people would respond?
Nathan
Huh? What I previously said was this
Do you have some difficulty understanding the written word? I said
* I don’t know what Joe means by “call out the various deniers out there,â€
* I don’t know what Joe thinks his bet demonstrated and
* I do know what I think Joe’s bet demonstrated.
There is nothing inconsistent with these three things. This is not the first time you have had difficulty logically parsing the meaning meaning of written words. I seem to recall the “You can’t say El Nino is present even though it is present by NOAA’s definition” controversy.
I seriously think might want to take a refresher course in reading comprehension. It would help your game in comments on posts you think aren’t worthy of your time.
So you do accept the IPCC Stats… Great.
Nathan–
Who accepts the IPCC stats? In what sense do you mean by “accept” the stats?
Most people accept the notion that the IPCC has published stats, the are readily available and we can know what those statistics say. Those are simple facts. Why would anyone dispute this?
Nathan–
A luke warmer who wants halfway decent odds might bet this:
The temperature rise for the next decade will fall between + 1 standard deviation and -3 standard deviations of the mulit-model mean of the IPCC projections. If it’s outside that range, I lose.
The betting kind would have to do some computations to firm up the specific numbers. But they aren’t likely to take a mere “high/low” bet when their position is that there is warming, but IPCC projections are biased high. The certainly aren’t going to do a high/low bet at the 0.15C value Joe picked is right in the middle of the ‘lukewarmer’ range.
I cold imagine a “stone-cold the ice ages is imminent” denialist might take accept that bet but depending on their precise position. Or a “hell and high water alarmist” might accept that bet. Or, betters might want to negotiate over the specific intervals. But if we go by the principle that people want to place bets that give then favorable odds, it’s possible that no agreement could be reached when a lukewarmer is negotiating a bet with others.
If we add the principle that when the bet is mostly for bragging rights in a politicized dispute, it’s going to be even more difficult to find a bet. The reason is: If my goal is to make a point, I’m not going to put my upper range at 0.20 which is within 0.05 of a value I considerer highly probable when betting with Joe Romm who seems to constantly suggest he thinks the probable jump is 0.50C. If the goal of the bet is somehow to make a point that Joe and I really believe our claims as opposed to money, and I think the best estimate is 0.15 and Joe thinks it’s 0.50, the break point is close to (0.15+0.50)/2 = 0.325C.
Nathan, I’ll make this simple for you.
Romm has proposed a bet that the next decade will be 0.15C warmer than this decade based on GISS.
These are the approximate GISS trends in C/decade (please correct me if I’m wrong):
From 1979 – 0.16
From 1989 – 0.19
From 1999 – 0.18
Now, what do you think the lukewarmer position would be for the next decade based on the GISS dataset? What do you think Romm’s should be? Lower than 0.15C looks like a horrible bet for a lukewarmer.
Terry,
“Boris – you seem like a nice guy.”
Who’s vigorously misinformed now? 🙂
lucia (Comment#20604) I was just thinking about that. So perhaps Taylor thinks the mean is zero and 0.3 is at least one standard deviation away from the mean (with a skew distribution like a gamma distribution). I am not sure what Joe thinks is the mean is but he must think that if go all the way to 0.15 from the mean you are atleast one standard deviation away.
However, this is by hypothesis, if you look at the past trends the data averaged over a decade may suggest less variance then either of them are hypothesizing.
John —
I also think the data from the past decade (and periods with no volcanic eruptions) suggests less variability in “weather noise” due to the earth’s climate.
But, when betting, there is also uncertainty due to the possiblity of volcanic eruptions, not knowing what will happen with GHG’s, not knowing
the effect of cosmic rayswhat the Leprechauns will do.Whatever theory one might have about the cause of recent warming, it stand to reason that people think there is more uncertainty than the variability due to “weathern” over the past decade.
What happened to cryosphere’s today objection to the idea that global ice was back to normal?
http://arctic.atmos.uiuc.edu/cryosphere/IMAGES/global.daily.ice.area.withtrend.jpg
This guy Joe Rom is a godsend to skeptics and denialists LOL
vg,
0 on that graph is “normal”. Not -1.
Hint: just because the data for a particular point is above the OLS regression line doesn’t mean that the trend has suddenly changed. It might suggest that 2007 was anomalously low relative to the long-term trend, and that projections of an ice-free arctic by 2014 are risible, but we’ve been saying both of those things for awhile now :-p
http://nsidc.org/sotc/images/arc_antarc_1979_2009.gif
According to this
washington post arcticle they’re now saying it’s about 6 deg F in 100 years which seems to be based on this model
BarryW (Comment#20613) September 25th, 2009 at 12:44 pm
I’m curious what kind of climate sensitivity there models assume because when I fit the current growth in CO2 levels to recent history (1960-2010) I estimate that if the trend in CO2 growth continues it would take a climate sensitivity of 20 to get that much warming from CO2 growth alone. A climate sensitivity of 20 for doubling CO2 is much higher then the IPCC predictions.
http://earthcubed.wordpress.com/2009/09/19/logco2-and-scary-graphs/
Well, here’s a lame observation that I make whenever I see the usual rounds of climate betting.
Taking a bet is a function of many, many things. Of course, the point of view of the individual making the bet is very important. But there are two other factors which are equally (or perhaps more) important.
1. The individuals willingness to take risk. This is fairly obvious, and most people expect it, but most people underestimate the magnitude of the effect. It is BIG. Some will only bet when the odds are massively stacked in their favour, others will launch into a bet willingly when the odds are clearly stacked against them. I think I read an anecdote that there was some Italian guy who would happily take evens on bets that were clearly stacked 1000-1 against him.
2. An observation by Arthur Dempster (he of EM fame). The development of Dempster-Shafer evidential reasoning is on the observation that people modify their risk based on the confidence not just of the outcome, but based on how informed they feel. For example: consider two sprinters, who have run many races against each other during the year, and have set similar times and won a similar number of races. Most people would be comfortable with a 50/50 bet, plenty of information, and it is clear the odds are similar. Now consider the two same people sitting down to a game of chess. Most would not take a 50/50 bet, because they have no basis on which to assess the correct odds. The Bayesian prior should be 50/50, just like the sprint race, but people don’t feel comfortable with the bet, and fewer people will take the odds.
How does this affect climate bets? Well, I would suggest that the pro-warming side of the debate have a high degree of confidence that we understand the climate system therefore can be confident in what might happen next. Sceptics, on the other hand, may well feel we do not fully understand the climate system. Therefore, most sceptics will typically require a bigger margin (since they do not believe the odds or the system can be well characterised) than the warmers, all other things being equal.
(Obviously, this does not apply to those sceptics who are convinced there is some other cause of the warming)
Romm’s claiming of linearity should not save him.
To see what the MIT boys (faves of Joe) are projecting for the coming decade, see Figure 8b on page 23 of their report (pdf warning).
Clearly their median rate of warming in the next decade is at least 0.3C, if not a touch more. So Joe ought to be plenty comfortable with that.
BTW, James Tayler has now lowered his offer to a 0.19C rise (see his comments over at MasterResource.
-Chip
Zeke I was thinking more along the lines of “tale of the tape” its a straight line trend NOW really from 1979 to current?
There is a summary of the GISS data in my slide set, along with ice core data and a lot more.
http://www.palisad.com/co2/slides/siframes.html
This is the raw data retrieved from the ISCCP project web site, has not been processed with multi-year averaging and is not presented as anomalies. There’s a clear discontinuity in 11/2001 when NOAA-14 was replaced with NOAA-16, which was never properly accounted for, although it is marginally documented. This shows the error of Hansen’s preferred anomaly analysis, which can’t distinguish between data anomalies and anomalous trends.
George White (aka co2isnotevil)
Ok, Ok I’ll take the bet. Who cares the odds, it’s climate. $1.00 payable if the next decade is the “hottest decade in the temperature record, more than 0.15°C hotter than the hottest decade so far using the NASA GISS dataset.”
Hey wait a minute…those aren’t the same thing, hottest decade in temperature record and NASA GISS record.
Let’s see in 12 years (I get one year for NASA GISS to revise and correct their numbers) $1.00 will be worth 15 cents figuring inflation. This could be a real money maker.
Oh and Nathan, when someone is considered delusional it usually implies most people can’t understand their jabbering.
Lucia,
I have posted the results of the hypothesis tests on the synthetic MSU data. Have a look.
Lucia
You have a habit of making statements like this, completely misrepresenting what people say, then claiming victory. Joe made a specific claim, he made a specific bet in the context. I see numerous claim among the denierarti that we are about to enter an ice age, Joe responds by offering a bet that it will get warmer, not colder. You take this and pillory him for something entirely different of your own creation. It’s the Tamino mathematical model all over again.
VG you keep refering to a flat trend in the “tale of the tape” graph. There is no trend line shown on that graph. The flat line is just an axis drawn at zero. It will stay flat no matter what the trend. The trend on the graph though is down.
What’s going to happen is, they will succeed finally in changing the meaning of ‘denier’. Right now, it carries a pejorative implication that what you are skeptical about is obviously true. Over time, this will change, so that it comes to mean simply someone who denies the truth of some possibly unfounded hypothesis. In short, it will come to mean ‘skeptic’, and will lose all its pejorative overtones.
You generally cannot change things by changing what you call them. But you can change what the things you call them are generally understood to mean.
The UK is full of this. ‘Asylum seekers’ now just means illegal immigrants. ‘Jobseekers allowance’ means unemployment insurance. ‘Care in the community’ now means, throwing people out on the street and good luck to them.
It is really extraordinary that people keep doing this. I guess its the fault of liberal arts educations, confusion between words and things.
Bugs:”You have a habit of making statements like this, completely misrepresenting what people say, then claiming victory… It’s the Tamino mathematical model all over again.”
You are hilarious, Bugs. You have misrepresented Tamino to say, that his model is nothing but a mathematical curve fit, completely ignoring that Tamino furiously banned Lucia from commenting at his blog for suggesting that his model might be _unphysical_. You are obviously doing precisely what you accuse lucia of: misrepresenting what Tamino said and then claiming ‘victory’. You seem obsessed with ‘winning’ or ‘losing’, Bugs (eg your claim that UAH data lost, models won).
Maybe Romm is too and that’s the reason for his sucker bet, but it does convey the message to me and others that he doesn’t really believe what he preaches. Yes, that message is my own ‘creation’ 😉
bugs–
Yes. Joe made some specific claims. He made a specific bet in context. He now has egg on his face.
I’m not seeing how my saying this is claiming a victory for me. But..whatever….
Yes. Joe made some specific claims. He made a specific bet in context. He now has egg on his face.
That it was going to get warmer, not colder. It was in response to all the claims that we are heading for the next ice age.
Bugs–
1) But people can make more than 1 claim. You are ignoring other claims made by Joe to try to make your case. Since we all know you did that, you will fail to convince anyone.
2) Joe’s was absolutely not written in response to any claim we are headed for an ice age. His post was a screed criticizing Andy Revkin’s NYT news article. Andy has never, ever, ever come remotely near claiming there is an ice age, cited no one who claims we are headed to an ice age and made not the slightest intimation we are headed into an ice age.
Lucia (Comment#20540) NASA GISS is continually fiddling the figures for the Rest of the World, that is where the so called warming trend is coming from, see Steve Mc’s where’s Waldo series. The upward trend of Continental US is almost non existent, and will eventually disappear altogether when Anthony Watts corrections for faulty data collection are available.
MarkR,
What will remove the upward trend in the UAH dataset for the continental US?
Is the raw UAH data available? All I can seem to find is processed anomaly data. This isn’t particularly useful, as time averaged anomaly analysis can make data anomalies masquerade as trends. One suspicious feature is the blip in 1998, which is not present in the GISS ISCCP data, which is available in raw form. Other than the UAH peak in 1988, there’s no statistically relevant trend whatsoever. There might be a small upwards trend up until the mid 2000’s, but that seems to have reversed direction. In mid-1998, the NOAA sounder analysis was changed and made a few tenths of a degree difference in temperature calculations based on that data. If this wasn’t properly accounted for, it may be the source of the 1998 blip. It’s possible that the same error was replicated in the RSS data and the land based temps were fudged to fit. Also, the only RSS data I can find is processed anomaly data. In general, you should never trust anomaly data when the raw data is not available.
George
David Gould (Comment#20672)
“What will remove the upward trend in the UAH dataset for the continental US?”
“…here is a chart of global monthly temperature trends in the two satellite-based lower troposphere series (UAH, RSS) and the three surface series (NASA-GISS, HadCRU and NOAA)….Note the strong divergence in UAH monthly trends. Such a strong annual cycle is simply not credible, and casts serious doubts on the validity of this temperature series. For more on the UAH annual cycle see my previous posts on the topic..” Deep Climate http://deepclimate.org/2009/06/05/uah-annual-cycle-continues-in-2009/
Here is the GISS ISCCP data. Note the large seasonal variability in the global average. This peaks are out of phase with perihelion, because at perihelion, the most energy is entering the thermal mass of the system (about 10 W/m^2). At aphelion, the most energy is leaving the system (about -10 W/m^2). The earth is still warming from Jan to April until the sign of dE/dt (E is the energy stored in the Earth’s thermal mass) flips. This is a consequence of the solution to the differential equation Pi = Po + dE/dt, where Pi is the energy entering the system, Po is the energy leaving the system and dE/dt is the rate of change of the energy stored in the Earth’s thermal mass.
http://www.palisad.com/co2/avg_temp.gif
Note the anomaly after 11/2001. This is caused by a calibration change when NOAA-14 was replaced with NOAA-16. When time averaged, it appears as an anomalous warming trend, but is really an anomaly in the data. This is why you can’t trust anomaly analysis unless you can see the raw data. Anomaly analysis uses independent averages for each month in order to make the annual variability go away. This is highly misleading because knowing the month to month variability is crucial for understanding how the climate responds to change.
Here is the same data, spit into the upper 60 degrees of latitude, the lower 60 degrees and the equatorial 60 degree slice.
http://www.palisad.com/co2/avg_temp_3.gif
George
The same people who faked the moon landing did it. Those guys are awesome!
Boris:
Get real. That was 30 years ago, and they are long retired. I’m guessing they have homes in the same retirement neighborhood in Tucson as the guys responsible for the Kennedy assassinations.
David Gould (Comment#20672)- What does that have to do with anything?
MarkR (Comment#20676- I’ll tell you what’s not credible, that link.
See:
http://vortex.nsstc.uah.edu/data/msu/t2lt/readme.18Jul2009
co2isnotevil (Comment#20674)- Assuming you mean data not corrected for orbital drift etc. ….Why would anyone want that?
I’m sorry but you are being pretty paranoid here. The UAH data is a well supported and documented analysis.
Boris (Comment#20681)- Real mature. 🙄
co2isnotevil (Comment#20678),
All the data I’ve seen says the global temperature, whether lower troposphere or surface, peaks in late July. The reason is because there is about twice as much land surface in the NH as the SH and the temperature range of the land surface is much, much larger than the ocean surface. Your graph illustrates that well as the temperature swing in the NH extratropics, which is mostly land, is much larger than the swing in the SH extratropics, which is mostly ocean. So when you sum the two hemispheres, the cycle of the NH dominates and the orbital eccentricity contribution is not visible. The world ocean heat content peaks in March, but that’s because the time constant of the ocean is so long that the phase of the response is shifted 90 degrees or three months from peak insolation in December at the solstice. The NH OHC peaks in September, but because there is more ocean in the SH, the sum of the SH and NH is dominated by the SH. Again, orbital eccentricity contributes, but the effect is not obviously visible.
Dewitt, re #20687
The asymmetry between the hemispheres is actually due more to the ebb and flow if ice and snow. In the N hemisphere, snow gets a chance to build up, while in the S, most snow falls on water and can’t accumulate. This becomes more apparent when you examine the reflectivity differences.
Andrew, re #20685
What you call paranoia, I call questioning the methods by which the data was put together. If it smells like a turd …
I would still like to get a hold of the raw UAH data. As I said anomaly analysis is useless. Surface energy is about 390 W/m^2, the flux in and out of the earth’s thermal mass is on the order of 10 W/m^2, the seasonal variability of about 4C represents a p-p energy variability of about 22 W/m^2, a .1C ‘anomaly’ represents a change in energy of about .5 W/m^2, or about 0.12%, which is only about 2.2% of the yearly surface variability. Even with data averaging, a .1C trend is so far in the noise that it’s meaningless. The main problem is that the data is not all from one source, and that the boundaries between data sets must be held to within a fraction of a percent so that discontinuities in these boundaries don’t masquerade as anomalous trends. Is there any question why people like Hansen, and in fact all AGW’ers, prefer anomaly analysis?
George
Andrew_Fl,
It was a response to MarkR, who seems to think that there is no warming trend over the continental USA. He gave reasons why the GISS data will soon show this, despite the ‘fiddling’. I was wondering if he also had reasons to doubt the UAH data, which also shows strong warming over the continental USA.
co2isnotevil (Comment#20688) September 28th, 2009 at 10:41 am ,
But if it’s albedo (snow and ice in the winter) and the average albedo of the NH is higher than the SH, then the NH should be cooler than the SH. It’s not. It’s about 1 C warmer.
David Gould. No recent unprecedented Temperatures in the Continental US NASA GISS, and no Warming since 1998.
http://data.giss.nasa.gov/gistemp/graphs/
PS. GISS don’t do a recent Monthly Continental US Temp, maybe it shows a downward trend?
I believe we will see a cooling through the next decade. If I had any money at all, I would take Romm’s bet. Romm is simply comical and I enjoy listening to him rant as it offers a little comic relief in an otherwise stressful life. Keep it coming Romm!
LMAO…
DeWitt,
It’s not that simple. Yes, the ratio of land and water makes a difference, but that’s only one influence of many. The hemispheric asymmetries in the dynamic reflectivity is very important. The fact that the average altitude of Antarctica is over 2000 meters is also important. The fact that most SH snow falls on water and can’t accumulate is important. The fact that during half of the year, the Earth’s heat mass is gaining energy and that during the other half it’s loosing energy, at average monthly rates of up to +/- 10 W/m^2 is important. The fact that the Earth’s reponse to change is the solution to differential equations is important. The fact that average solar input varies by over 20 W/m^2 between perihelion and aphelion is important. Understanding how the average climate changes on a monthly basis is most important to understand how the climate behaves. Unfortunately, the tendency to rely on anomaly analysis, which cancels out seasonal variability, obscures the mechanisms behind climate change and hampers it’s understanding.
George
co2isnotevil (Comment#20773) September 29th, 2009 at 10:55 am ,
I base my conclusions on data not hand waving. You should try it.
Daily lower troposphere temperatures for NH, SH and global:
http://vortex.nsstc.uah.edu/data/msu/t2lt/tltdayac7998_5.2
Monthly gridded absolute surface temperatures:
http://www.cru.uea.ac.uk/ftpdata/absolute.zip
TOA insolation for any latitude, longitude and date with orbital eccentricity included:
http://aom.giss.nasa.gov/srlocat.html
Global Mean Monthly Surface Temperature Estimates for the Base Period 1901 to 2000:
http://www.ncdc.noaa.gov/oa/climate/research/anomalies/index.html
And the annual variation in ocean heat content:
ftp://ftp.nodc.noaa.gov/pub/data.nodc/woa/PUBLICATIONS/grlheat04.pdf
Yes it is that simple.
MarkR,
But there has been warming since 1996 …
Yes, I love cherrypicking data too. 😉
David Gould (Comment#20719)- Why is it always “strong warming” and never just “warming”?
Well, I think you miss my point. MarkR seemed to be suggesting that the warming trend in the entire land surface data set for the US might not have an over all warming if corrected (I don’t agree) and UAH could at best confirm the warming since the seventies. It can’t tell us, for instance, whether the year 1998 really was warmer than 1934 in the US or not.
MarkR (Comment#20733)-I’m fairly certain it’s impossible for monthly data to differ that substantial from the annual averages. Nevertheless it annoys the crap out of me that there is no easy comparison to make between monthly GISS US surface temps and monthly from NCDC (which can be extracted, with a little effort, here:
http://lwf.ncdc.noaa.gov/oa/climate/research/cag3/na.html
Because interestingly enough, NOAA’s data warms significantly relative to GISS in the US (and if you are going to do the comparison yourself, remember to convert the Fahrenheit degrees of NOAA to Celsius first):
http://www.climateaudit.org/?p=4852
David Gould (Comment#20810)-I’m also unclear as to why it is cherry picking to observe that the last 144 months at least have seen no warming. It’s literally true, and there is no reason why that should be that I’m aware of…it starts in the beginning of the 1997 El Nino, but the temperature effects hadn’t really gotten going, because temps lag ENSO by about six months IIRC. That’s very similar to the current situation. The sun was all that active at that time-sure, it’s less active now, but I doubt that could totally overwhelm AGW. So the behavior of the temperature records in the last twelve years really is rather odd, especially in the context of expectations of ever faster warming. You could say that it is a short period, but that doesn’t make it any less baffling.
Andrew_FL,
It depends on your purposes. Starting with 1998 an obviously warm year, as it is around two standard deviations above longer term trendline and ending with 2008, an obviously cold year as it is around two standard deviations below the longer term trendline, is pretty obviously cherry picking (as is me starting in 1996, which was significantly colder than the trendline).
I have no problem acknowledging the fact that there has been no warming this decade. My very simple and very simplistic model shows that such periods occur quite regularly with a warming trend of .2 degrees per decade and a SD and autocorrelation that matches the early years (1880 to 1920) of the GISS data – (or better yet, the 1850 to 1920 years of the Hadley data if I am doing Hadley testing. )
However, it does become problematic over periods too much longer than a decade. Thus, my model predicts that it is extremely unlikely that we will see flat temperatures over the next decade. In betting terms, I think that we will see .15 degrees of warming over the next 6 years and a further .1 degrees of warming – at least – from then until 2020. If something close to this does not happen, the notion that .2 degrees of warming per decade is occurring is gone from my perspective.
In less complicated terms, if 6 of the next 12 years (2009 to 2020) are colder than 2008, I will say that the world is not warming fullstop, let alone at .2 degrees per decade. (My model shows this as being effectively impossible if the world is warming significantly).
David–
You can find years with zero trend before the 70’s when the expected warming was slow. You can also find them after 1970 during periods when the models predicted low or negative trends due to eruptions of volcanos like Fuego, El Chichon or Pinatubo.
It’s not quite right to explain the current slow warming with variability induced by volcanic eruptions like Pinatubo. There’s been no eruption recently.
That said: The evidence doesn’t say “no warming”. What it suggests is “not as high as 2C/century right now”.
Lucia,
Regarding volcanos, if I exclude the major ones as listed on wikipedia (http://en.wikipedia.org/wiki/Volcanic_Explosivity_Index)
when I am working out the SD, it looks as though I would need to use the hadley dataset and work from 1850 to 1880 and 1890 to 1900 – something like that, at any rate.
In advance, I am wondering how big a difference this is going to make. Could be big; could be not.
Hmmm. Removing the years most likely to have been affected by significant volcano activity seems to have no effect on the SD. So, either I have excluded the wrong years (quite possible, particularly if wikipedia has failed to mention significant volcanos, or if their effects were delayed for some reason) or the impact of volcanos is so short that it is outweighed by all the other weather noise when looked at over multi-decadal periods.
However, if there was a warming or cooling trend over the period, that would tend to increase the sd, however. And that appears to be the case if we use the 1850 to 1880 data and the 1890 to 1900 data. (There is no trend for the 1850 to 1920 data.)
So, correcting for that trend, I find that the SD does drop, from .129 to .123
I will have to see how significant that is in regards to my model.
However, too many howevers, however … 😉
David Gould (Comment#20816) “Starting with 1998 an obviously warm year” I didn’t do that-the last 144 months starts in 1997. “ending with 2008, an obviously cold year” I didn’t do that, I ended in August 2009. “I have no problem acknowledging the fact that there has been no warming this decade.” I’m curious as to whether you feel the same about 1.2 decades. Since 144/12=12.
lucia (Comment#20817)-“The evidence doesn’t say “no warmingâ€. What it suggests is “not as high as 2C/century right nowâ€.” I think it would be more accurate to say that the 144 month trend is not significantly different from a slightly positive trend (even though the trend itself isn’t) but it is significantly different from .2 degrees per decade. If evidence of no warming means a negative trend:
http://i23.photobucket.com/albums/b370/gatemaster99/hadsincesep97.png
The last 144 months (September 1997-August 2009) of Hadley data, with trend.
http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/monthly
David Gould (Comment#20819) Oi, I should have commented on this. The index linked to doesn’t do a good job of indicating the climatic effects of volcanic eruptions (which, BTW last a few years). A little better is this:
http://data.giss.nasa.gov/modelforce/strataer/
Now, note that a major issue you’ll run into is that volcanic effects pervade the record during periods of warming, thus making an analysis like yours especially difficult.
Andrew_FL (Comment#20814) Thanks for the links. I was thinking of a Continental USA version of: http://data.giss.nasa.gov/gistemp/graphs/Fig.C.lrg.gif
If most of the warming is coming from the non Continental USA as Steve McIntyre suggests, then the Cont USA should show less warming according to GISS now, and if Watts eventual reductions are applied, a definite cooling will be shown. It’s only a matter of time.
PS CO2isnotevil, I agree with you. No one has shown the exact mechanism, and no one has done an experiment to show it. I bet that CO2 is ultimately shown to have zero effect on Temp. I consider Atmospheric Temp is a fuction of Pressure or Density, and that Gravitational Field and Radiation from the Sun are determining factors as to the average atmospheric temperature. Differences at the surface are determined by cloud layer, thermodynmics etc.
By the way, who here believes in Push Gravity rather than Pull Gravity? ie Gravity is the result of random motion of Cosmic Matter, and Gravity is the result of Shielding by Planetary Bodies. See http://en.wikipedia.org/wiki/Le_Sage%27s_theory_of_gravitation
Andrew_Fl,
The slope of those 144 months seems heavily influenced by two years data, that of the first 12 months and the 12 la nina months in 2007-08. Removing them, I get a trend of .01 per year. If we also consider high sun/low sun, I do not think that it is that dire yet. But a few more flat years and, yes, the .2 per decade looks bad. Needless to say, I do not think that that will happen. 2009 is already pushing to be relatively warm. 2010, by my reading of the tarot cards, looks like it will be a record or very close to it.
Andrew_Fl,
I do not understand the GISS link that you posted re volcanos, unfortunately. Can you tell me which years you think that I should exclude? It looked to me as though excluding the 1880s and the 1900s was the best bet to get rid of volcano effects.
I understand it better now – I have printed out a table with the global AOD per year.
Excluding the 1880s looks right to me and after 1900. I probably should also exclude 1856 and 1857, and possibly 1858, 1896 and 1897. I will do that now.
Removing those years changes the SD upwards, to .125. So I think that my original pick was likely okay.
And thanks for that table – it is very interesting. Pity it does not have the 1990s.
And re the 144 months, yes, I know that removing data from a dataset is not a good idea – it is the very definition of cherrypicking. But testing the effect of removing certain values gives an indication of the ‘weight’ of those values on the trend or the average (depending on what you are looking at). Values that are ‘heavier’ should be looked at carefully. This is why medians rather than means are used in some analyses of certain datasets, and I imagine that there is something similar regarding trends (although I have never heard of it, so am just speculating here, I suppose).
I wouldn’t exclude the eruption years per se, but perhaps find a way to account for the effect. Hm, well, I have to say that I was planing on examining the issue of the Pinatubo effect, but unfortunately people seem to have no interest in such an analysis, because whenever I ask for help…crickets. 🙁
Anyway, yes, the 1998 El Nino and the 2008 La Nina probably effects that trend by being near the end points. It is still a very unusual twelve year period (models don’t do that terribly often, many not at all). But accusations of cherry picking often point to the problem of going straight for the ’98 peak to start your trend. I could have done that, but why when I can show a truly odd, even longer period?
Removing outliers I think is not typical for something we assume to be a real physical part of the history. If we thought it was due to error, maybe, but we seem to know the origin of the outliers.
Andrew_FL,
Excluding the volcano years, though, removes any amplfying effect that they can have on the SD. In other words, it gives us the *minimum* SD for all other non-volcanic weather/climate effects. Given that we are looking at the last 12 years, during which there has *not* been a volcano, comparing that period to other periods during which there have been no volcano effects is probably pretty reasonable.
As to excluding the outliers, the issue is to do with reality and variability. If we get an unusually warm year at the start of a period and an unusually cold year at the end of that same period, how reasonable is it to use that period as a period over which to examine the trend? If the weights are not very heavy – for example, they are within 1 SD of the apparant longer-term trend – then it is probably fine to use them. If they *both* are 2 SD outside that overall trend, then we can come to one of two conclusions: these are outliers or the longer-term trend is *wrong*.
If we have reasons to believe that these are in fact outliers – and 1998 was a very strong El Nino year, and there was a relatively strong sun then, while 2008 was a middle of the range La Nina year with a very weak sun – then we are perhaps (perhaps) justified in discarding them for the purposes of examining the trend *over that particular period*.
For longer-term analysis, the weight of these two years would reduce as more data was added, so they can be ‘safely’ included.
David–
You can’t just remove the specific year with the volcano. The volcanic aerosols hang around and affect the temperature for a while. Plus, there is the recovery after they clear up.
It’s not clear exactly how to remove volcanos, but just taking out one year isn’t quite right.
David
The AOGCMs show strong effects from El Chicon and Fuego. The effects last more than 1 year. So, you need to take out some of the 70s and 80s, and also take out anything as big as either of those.
lucia,
I have not simply taken single years out. First, I removed whole periods based on my guess as to how long the volcanic effects lasted. Thus, I removed the 1880s and anything in the 1900s.
Then Andrew_Fl pointed me in the direction of a sheet that estimates the Aerosol Optical Depths in any given year from 1850 to 1990.
Thie confirmed that removed the 1880s was a very good idea, and that remoed the 1990s was also a good idea.
However, a couple of years were high in the rest of the data: 1856 to 1858 and 1896 and 1897. These are also not single years.
Lucia,
The 1870s show very tiny effects on the chart – a 6 on the AOD scale is the highest 1870 year, as opposed to 125 for 1885, for example. If you are saying that I can only use periods in which there are 0s, that pretty much excludes all of the data. There are only 26 scattered years in the 1850 to 1990 dataset that have zero.
David, you could remove the effects of volcanoes from your data using a method similar to this:
http://bobtisdale.blogspot.com/2009/02/impact-of-north-atlantic-and-volcanic.html
See Figure 4.
http://pubs.giss.nasa.gov/docs/1993/1993_Sato_etal.pdf
This publication has the chart that I am referring to. It is page 4. It is unfortunately sideways and I cannot rotate it, so I had to print it out to use it.
David–
If you are trying to discover the sd that would exist when volanic aerosols are constant, that technically, you need to find periods when aerosols are constant. So, technically… yes… you need to find periods when the aerosols are nearly constant, which leaves something like the 30s -50s. Then that period has a problem because it’s been decided there are possibly serious problems with sea temperature measurements around WWII.
This leaven almost zero time to do the test.
In principle you would want to relax the criterial. But, if your goal is to find the effect on SD, you need to know the volcanoes didn’t affect the value. The fact that Fuego and El Chicon are smaller eruptions than Pinatobu isn’t sufficient reason to ignore them. You need to think they are insufficient to affect global temperatures.
But then you have a problem– the multi model mean from the AR4 says they affected the temperature.
In fact, if you compute the sd of 8 year trends for the multi-model mean (which has very little weather noise) you’ll see variations around those two eruptions.
So, if the models are right, those eruptions added significantly to the variability of 8 year trends. If you keep the periods around those eruption in, part of the sd is due to those eruptions. (Or at least you can’t be confident it’s not.)
Joel,
Thanks. Yes, that looks good – once I work out precisely what it means, of course!
In degrees C, from Bob Tisdale’s graph it looks as though 50 AOD equals .1 degrees of effect. So Krakatoa would have cooled the world by .25 degrees at its peak effect. Interesting.
Lucia,
But ‘constant’ can have error bars!
As an example, if I exclude the years that I am talking about and simply run from 1850 to 1882 (for example) the AOD ranges from 0 to 14.
This of course means that my SD would also have error bars. But the question is: how big?
If removing periods in which the AOD varied from 0 to 125 affects the SD by .006, I think that I can pretty safely conclude that including years with a AOD range of 0 to 14 will affect the SD by significantly less than .006.
And if removing those years *increases* the SD (which, as it happens, it does), then I can safely resort back to the *lower* SD as the lower bound.
If I then remove the years with greater than 9 AOD, I get an SD of .123 – the same as my SD for the period I tested previously.
So, I think that my lower bound SD is getting pretty robust here.
But Bob Tisdale’s method might enlighten me further and show that I am wrong.
David–
I’m confused by what you are saying. Why would you remove the years with low optical depth (0-14)? To find the SD during years of low optical depth, you want to include those and exclude the others.
I tried this exercise a while ago. Since I don’t believe in “correcting” for volcanos, ended up computed for years from after Kamai to an eruption in the 50s which appears on some volcano charts but not others. Then… right when I was done, a group published a paper explaining that the huge source of SD around WWII was experimental error! If we took that out using a correction suggested by the paper, the SD dropped.
Or if we just didn’t correct the “buckets-to-jet inlet” issue in the data, that something like 2 decades to find standard deviations. The SD was quite a bit lower over that period than the rest of the time. It’s maybe 20% higher than we would estimate using “red noise” now. But… the period is short!
BTW–
I’ve been silent because I’ve been looking at the SD’s predicted by models for the 20th century. Oddly, the statistical test show that AR(1) kinda-sort predicts the correct standard deviation over multiple runs for the models taken as a collection. (More details later. I want to see what happens with the A2 models and the 20th century. But for the A1B models, that’s how it’s looking for the group as an ensemble. )
Okay: using Bob Tisdale’s method, and correcting for the slope, I get a SD that is significantly less than the one that I calculated above. It is .116. So it looks as though the method of exclusion fails, as Lucia said it would.
lucia,
I had excluded the high ones. The ones 0 to 14 were the ones that I had left. Then I was excluding the highest of the ones I had left to see what happened when I did that.
David G,
Another way to drop out anomalous years is to use an underlying trend method like is used for the CPI in Aust. Line up all your years by AOD and exclude the bottom and top thirds by rank. That way you exclude unusually low and unusually high AOD. If you only using one third of the data is dropping too much then drop the outer quarters. Yes this doesn’t drop years by how far they are from the mean but it does assume an underlying value which is what you are looking for. This non-parametric approach works well when your data can’t be expected to be normally distributed which I would guess is the case with AOD. Using SD assumes normal data — have you tested the AOD data for normality?
Andrew23,
You are right. It is not normally distributed – it has a hard wall at zero, where most of the values are clustered, and then it has higher values than that. So I guess it might be one half of a normal distribution. I am not sure of the implications of that.
David–
In table 2 in Sato, 1857 has values in the 60s, not the 10s.
lucia,
That is covered by ‘exclude the years that I am talking about’ – 1856 to 1858.
Editing comments does not seem to work for me: what I wanted to add was that I could have been clearer – the comments about excluding 1856 to 1858 were in a post higher up.
Andrew23–
That sounds fine from a statistics view, but it is not useful from a phenomenological point of view.
With respect to standard deviations in trends or residuals, the problem is that the change in the level of the AOD affects the amount of heat entering the climate system. The earth’s temperature responds, but the time constants of the earth matter.
So, if you applied a sudden step change like this:
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,100,100, 100,100,100,100,100,100,100,100,100,100,100,100,100,100,100,100,,0,0,0,0,0,0,0,0,0,0,0,0,
The big changes in temperature happen during the few months of years after the AOD’s switch from 0 to 100 and then again when during the years when AOD switches from 100 back to zero. I colored those in blue.
If the levels are constant for a while, you expect the variability to go back to the loser baseline level.
So, it isn’t the high years per se that are variable.
Lucia,
I agree about the diff between a stats view and a phenomenological point of view. Especially if what you are trying to do is relate one phenomeon (in this case AOD) to another like GMTemp and the issue of lag. But if all you are doing is is using one measurement to remove samples from the data set before analysing a trend or average then I just wanted to give David G a better outlier test. We expect that AOD effects temp but we don’t know by how much and even if it is linear so using a normal model to test for outliers is bad.
Probably pulling the stats view with the phenomenological point of view and the lack of an understanding of the magnitude and linearity of volcanic effects together means we are stuck with using the whole data set and the consequence of a largish Error.
Somebody else in this thread said that you should only reject data when you know why it is an outlier and that is true. Include it all and let Student sort it out (sorry sort of a stats plus religion joke).
Andrew23
But that’s not quite what David wants to do. He wants to estimate the magnitude of the standard deviation of “N” year trends. (I think.)
In David’s case, incorrect correction could cause problems that sort of wash out if you are just trying to estimate the bias in a 30 year trend the way Bob did.
If David is trying to take out the effect of volcanoes on variability he actually has to remove the effect from the years that the volcano affected. So, the fact that it’s the change in the aerosols that matters is important for him.
Lucia,
GIven that even the mercurial Steve MacIntyre does not yet confess to fully understand GISS methodology, I suspect you are too forgiving of GISTEMP.
Hansen has significantly revised past trends. I would not like to be relying on the effifacy of future trends.
Geckko
I know Steve says he doesn’t fully understand the methodology yet. I anticipate someone who is suspicious of GISS will understand GISSTemp within 10 years.
“I anticipate someone who is suspicious of GISS will understand GISSTemp within 10 years.”
I take to things from the above statement:
1) Lucia doesn’t understand GISSTemp either.
2) Lucia thinks it’s good that people are suspicious of GISS.
Or am I wrong about these, too? 😉
Andrew
Andrew_KY
Sort of. I have read the paper so I have a basic understanding of the theory. I have not delved into the code to see how the details play out when the method is applied “in the wild”.
(Note based on past discussion: This is not a yes/no response. As with most things, ‘yes/no’ is insufficient to convey any useful information.)
Once again… sort of. I think that “trust provisionally but hope someone external and unconnected to the creator of the product verifies” is the best policy toward all research products (and even commercial products).
As it happens, the only people likely to be motivated to check are those who are at least a bit suspicious in some way. (Their level of suspicion could range from thinking GISS has perverted the record to merely thinking ‘people make mistakes’. But, only those somewhat suspicious would be motivated to go to the effort.)
Lucia,
Here are the yes/no questions and answers:
Does Lucia understand GISS in a practical, real-world sense? No
Does Lucia think it’s good to be suspicious of GISS, as such suspicion would motivate necessary objective checking? Yes
😉
Andrew
As SteveMc has shown with tree rings, you don’t need to change the code just cherrypick the data set. Same can be done with temp sites. Remember sites seem to “disappear” from the official data even though others can find them.
But Andrew_KY–
No. Those answers don’t apply to the questions you asked.
Had I simply answered the questions you asked, I would call my answer to “is it good to be suspicious of GISS” “no”. First, you didn’t say suspicious of what? I don’t think it’s good to be suspicious of GISS per se, to be suspicious of their motives or suspicious of general incompetence.
I think ideally all important results are verified independently of their creators and if it’s not done one trusts results but provisionally.
Second, if I’d just read your question without the addition of “practial real world sense” you now add, I would have answered “yes”.
So, as you can see, if I’d treated your questions as “yes/no”, I would have given the opposite answers from the ones you think I would have provided.
However, based on past experience, I know that your “yes/no” questions are generally badly worded and the “yes/no” answer to your question would not provide you the information you are trying to elicit.
Lucia,
The questions that have yes or no as answers are always good, because they lead to more clarifying questions. See how they are extracting info from you? Now we least know what your position is on the last two questions. You can’t say maybe about them and retain any consistency. We both understood the questions and the answers. That’s the beginning of a knowedgebase about you.
I didn’t say (on this occasion) be suspicious of anyones motives. I asked about being suspicious as it relates to checking what is produced.
Anyway, has your practical, real world understanding (I have not delved into the code to see how the details play out when the method is applied “in the wildâ€) of GISS produced anything valuable?
Andrew
Andrew_KY–
The only reason you got information is that I did not answer “yes/no”. If I’d answered yes/no, to those questions, I would have given the opposite answer from the ones you list in your summary (Comment#20890).
Based on the wording of your question, how could the person being asked have known this?
Yes.
Lucia,
Because I wrote in the question “as such suspicion would motivate necessary objective checking?” The objective checking is what is relevant, and it is included in the question.
“Anyway, has your practical, real world understanding (I have not delved into the code to see how the details play out when the method is applied “in the wildâ€) of GISS produced anything valuable?
Yes.”
What? 😉
Andrew
Andrew_KY–
Go back and read what you asked in Andrew_KY (Comment#20888). Only a psychic would have known how you would reword them after I answered the original questions.
Many things.
Lucia,
I reworded the questions based on your own responses, so ideally you would come down on them one side or the other as a yes or a no, provided you were engaged in a serious discussion. Clearly you aren’t interested in that.
What you are engaged in here elusion. And for what reason? I haven’t the foggiest notion.
Andrew
Andrew_KY–
You reworded questions after I answered them. Then, when I said the questions were vague before I answered them, you said they weren’t– pointing to the rewording after I answered them.
I am interested in having honest discussions. But today you started by putting words in my mouth in (Comment#20888), followed by what would appear to be a yes/no question that was not a yes/no question.
I gave a detailed explanation.
Then, you reworded the questions, and gave answers to your reworded questions calling them my answer. The “yes/no” answers you put in my mouth were the opposite of what I would answer to questions worded as in (Comment#20890).
I don’t know what you think you are doing, but your method of eliciting information from me is very wasteful of my time.
If you have notions about GISSTemp that you want to advance, then just explain your notion or theory. Don’t try to advance them by writing comments telling people what you think I think and mangling my position.
Andrew KY-elusion is not a word.
Et tu Andrew_FL , my esteemed compatriot?
http://education.yahoo.com/reference/dictionary/entry/elusion
😉
Andrew
FWIW Andrew_KY, I haven’t the foggiest idea what you think I am trying to elude. I’ve answered your questions.
I’ll grant you it comes from a word, but it’s still not proper.
lucia,
Yes, ultimately. And I can only do that knowing the SD of a set of years. However, (and I know that I am not explaining anything here, except to myself) if I use the SD from a set of years that includes volcanos then using that to model against a set of years that does not include volcanos (eg, 2001 to now) is … problematic.
David
Yes. This is a notion I have advanced in quite a few blog posts– mostly before your time here.
Unfortunately, it’s not clear we have enough volcano-eruption-free years to estimate the SD without fairly big volcanic eruptions. You can try correcting, but that’s fraught with danger too. What to do? Dunno.
Unfortunately, it is a fact that sometimes the answer is “You don’t have enough data to discover what you’d like to discover with any degree of certainty or support any claim as well as you’d like.”
Chiefio has been looking into Gistemp. He concludes it is the addition of extra stations that is causing the warming.
MikeN Good point.
From Chifio blog:
“I’ve put up a simple intro to the “issues†with the AGW thesis:
http://chiefio.wordpress.com/2009/07/30/agw-basics-of-whats-wrong/
Not exactly GIStemp, but in “characterizing the data†for the GHCN input, I’ve found that all the “warming signal†is carried in the winter months. The summer months do not warm. That can not be caused by CO2.
Further, the “warming signal†arrives coincident with the arrival of large numbers of new thermometers. When you look at the longest lived cohort, those over about 100 years lifetime, there is no warming signal present in the data to speak of. When you look at the much shorter lived cohorts, you find a very strong warming signal, especially in the winter months. On further inspection of the data it looks like a lot of thermometers “arrived†at places with low latitudes AND at airports (newly built as the “jet age†arrived).
http://chiefio.wordpress.com/2009/08/05/agw-is-a-thermometer-count-artifact/
http://chiefio.wordpress.com/2009/08/10/well-theres-your-global-warming-problem/
http://chiefio.wordpress.com/2009/08/13/gistemp-quartiles-of-age-bolus-of-heat/
My “working thesis†at this point is that GIStemp is a “filter†that tries to remove the impact of this bolus of thermometers arriving in a spike of time and space (using zones, boxes, grids, et. al.) , but is just not adequate to the task. GIStemp is just not a perfect filter. Looking at the impact of the temperature steps (up to zones), they act as a mild amplifier, so some of the “filter†effect in the later steps will only be removing what was added in the early steps.”
Now here is a test that can be proved statistically.
David and Lucia — I know David is trying to uncover the pattern in the data that would exist without volcanoes and I can see that that may cast some light on both the underlying trend and the likelyhood of a decade of below trend results but … don’t volcanoes really happen? Shouldn’t we consider their impact as part of the white noise? Does a trend found without volcanes have any relevence to the actual earth’s climate? Surely GCMs should be able to cope with the real world and hence include the noise of volcanoes along with El Nino and sunspots etc. If we accept volcanoes as white noise then we don’t need to worry about the lag but yes Standard Error will be large and the power of tests reduced but that is the real world.
Andrew–
It’s a matter of context.
Volcanoes really happen. But when trying to gauge whether the current flat trend is consistent with projections, we also know that there has been no big eruption since Pinatubo. So, we know the cause of this slow warming is not a volcanic eruption. In contrast, dips after Pinatubo were caused by Pinatubo.
With respect to predicting the future we don’t know when the next major volcano occurred. But with respect to explaining the past, we do know whether a volcano erupted. So the eruptions switch form random to deterministic when explaining what has already occurred.
I’ve written on Joe Romm’s blog offering to take him up on his bet, to the tune of USD 1000. I could very well have offered a much larger amount, this bet is a no-brainer since we have entered a cooling trend which will last for at least 20 years more. Probably 30. But there are much better ways to double ones capital in 10 years than this bet.
At the present we are in a solar minimum, the next solar cycle is numbered 24. Even solar cycles are MUCH more liable to produce cooling, and solar cycle 24 will last at least until 2019. OK, solar cycles no 10 and 22 did produce a warming of at least 0.15 degrees Celsius, but solar cycles 12, 14, 16, 18 and 20 did not. See diagram at http://energiminnesfonden.se/index.php?option=com_content&view=article&id=74:klimatet-och-solcyklerna&catid=36:ovrigt&Itemid=59 . So it’s a good bet that solar cycle 24 will produce cooling and not warming.
Svempa– Wow! I didn’t think anyone would take his bet as offerred! Did he accept your bet?
Well, in all honesty he hasn’t had many minutes to answer yet. Hopefully he’ll respond within a day or two.
The whole thing about ten years or twenty years, so it’s not a reasonable investment makes no sense, and it didn’t when it popped up on RealClimate a few years ago. Then someone was willing to give 25-1 odds to James Annan with regards to either methane or CO2 concentrations, betting $25 against $1. Annan wanted to make it $2500 vs $100, and he said no cause it’s a bad investment.
The bet is the bet. You are not putting the money in escrow.
So if it is $100 being bet in ten years time, then it’s really a $50 bet now.
MikeN, I am indeed planning to put the money in escrow, and I have asked Joe if he has an escrow account set up for this bet. One of the reasons being my age (68) and some health issues, which means I am not certain I will be around to collect my winnings in ten years time, even if I of course hope to do so. The other reason is that I am not sure Joe is being honest about this bet, and with an escrow account I can be sure that he has to pay when (I’m not saying IF) he loses this bet. I have suggested we should use an arbiter, I would prefer Anthony Watts if he is willing.
So the “bad investment” still stands, even if doubling of my money in ten years comes down to 7% interest per year on my “investment”.
No answer from Joe Romm yet. I do not even know for certain that my posts (I wrote in two discussion threads on his blog) have been publshed, I can see them myself but can others? Take a look at http://climateprogress.org/2009/09/27/time-magazine-names-joe-romm-one-of-the-heroes-of-the-environment-2009/ , should be the newest post and it reads like this:
“Hi Joe! I found a reference to your bet on another blog, and I would like to take it. In order not to impoverish you (and possibly your children) I will be satisfied to bet only USD 1000, which also should mean that I leave room for other takers.
Do you have an escrow account set up? Due to my relatively advanced age (68) and some health issues I would also like to name an arbiter or caretaker, perhaps Anthony Watts of WUWT would be OK? So if I’m not here by 2019 I could be sure that my winnings go to whatever person or charity I will name.”
Svempa – I see your message there but no takers yet 😉
EW – thanks! At least that is one questionmark erased.
Svempa–
I think you’ll probably lose your bet. But… what is more important is: Joe posted the bet. Will he even take it? Joe hasn’t accepted by tomorrow, this may need a post. After all: Joe picked a number that is well below what he tells people is certain. If he won’t even take that after offering… well…. hmmm….
Thanks Lucia! But I wonder what makes you think I will lose my bet? I recently posted this little exercise on another climate blog, using my alias Greyshark:
“But I look at things in a much simpler way, and I think the information we have about temperature changes and forcings in the last 100 years can give us a more reliable result with simpler calculations.
I agree to the idea of temperatures depending on forcings in a linear manner. But as we know, CO2 forcings depend on CO2 concentrations in a logarithmic manner. So let me repeat what I think we are in agreement on:
1) CO2 has increased from 280 ppm to 385 ppm since pre-industrial times.
2) Net forcing from this increase is 1,7 W/m2 (according to IPCC, check my numbers)
3) Total forcing for a doubling of CO2 is 3.7 W/m2 (according to IPCC, check my numbers)
4) Temperature increase during this period is 0.6 C .
So there is another 2 W/m2 and 175 ppm CO2 to go until we arrive at a CO2 doubling. The time needed for this is probably 80-90 years, if we do manage to burn enough fossil fuels.
Let us now assume that the entire temperature increase of 0.6 C is due to the 100 ppm of CO2 increase (I of course think otherwise, but for the sake of argument), giving a forcing of 1.7 W/m2. Since temperatures depend linearly on forcing, this (using regula de tri) means that another 2 W/m2 will cause a temperature increase of 0.7 C. Total is 1.3 C, not far from the theoretical value of 1.1 C for a doubling of CO2.
Now IPCC says that we need to introduce a postive feedback that is quite large, and which will cause temperatures to rise even more. This is obviously wrong, since the first increase of 0.6 C is INCLUSIVE of all feedbacks in the system. Therefore the prognosticated 0.7 C increase also is INCLUSIVE of all feedbacks. This is simply a reality check showing that no positve feedbacks can exist!
The above reasoning depends omn two things:
1) We do not reach any “tipping points” along the way, and this seems to be true.
2) There are no hidden “heat deposits” that will come into play, that is I am assuming that all feedbacks for the temperature increase of 0.6 C have played out. And this seems to be a very reasonable assumption.
So my conclusion is that the “IPCC scenario” at most can mean a temperature increase of 0.7 C over what we already have got. Which does not cause me any alarm. And in REALITY, even this will not happen – because the increase of 0.6 C was “most likely” NOT caused by CO2 alone.”
So which side of the argument are you on :-)?
Svempa–
I think there is a long term warming.
I don’t think it’s as strong as the IPCC projected…. I guess I don’t know if you’ll lose– but I think it’s going to be darn close!
Still, a bet is a bet!
Close to the middle– but on the lower end. Anyway, you might win. Or not.
I think you are wise to insist on escrow. That’s the thing that makes bets real. You could insist the money be put in a back account or a 10 year certificate of deposit and whoever wins gets the interest too.
Well, I also believe there is a long term warming trend. But i also feel quite certain that we have entered a short term (if 30 years is short term) cooling trend. The CO2 warming is somewhere in the background, but I calculate it to be quite small, accounting for perhaps a 0.04-0.10 C/decade temperasture rise. And in the foreground there are MUCH stronger forces at play and they will cause a short term cooling. So I trust that history will repeat itself and my belief is that Joe Romm picked the wrong decade for his bet.
Thanks for the 10 year certificate idea, I had been thinking about some way to get interest on the money and that could be the best way.
No way he will let Watts be the arbiter.
He will insist on GISS.
Well he bet against a cooling trend, so there you go.
Ultimately, your calculations amounted to a linear extrapolation of temperature trend. The CO2 levels could go higher faster.
Here is the thread on RealClimate I was talking about. I just realized I confused the bettor with Ammann.
http://www.realclimate.org/index.php/archives/2005/06/betting-on-climate-change/
I think Romm’s bet is actually a bit more favorable to Svempa, and perhaps even a fair bet.
He didn’t say 2019 will be .15C warmer than 2009, but that the decade will be .15C warmer than the 2000s.
Running the numbers, this is actually a bit worse than a .15C bet, because 2000 is so much colder.
If we assume a linear gain for each year from 2010 to 2019, then it would take a .14C gain for the 2010s to be .15C warmer than the 2000s, which has annual numbers of
39 55 67 64 58 76 63 72 54 69(assuming Sept-Dec stay at 71)
MikeN, thanks for the link to the Annan post. But I am not impressed, since almost 20 years back I am an avid student of the stock market (and an investor and/or speculator) and I can tell you that any belief in pefect market forces is grossly misplaced. Also this was from 2005, and I can understand that even someone like Lindzen hesitated to put money on the table. But indeed there WAS a bet made as of August that year, you should taka a look at this: http://julesandjames.blogspot.com/2005/08/bet.html . In this case temperatures in 1998-2003 are to be compared with those in 2012-2017 (6 year average in both cases) and the amount (split on two bettors against the house) is a lot larger, USD 10000. Furthermore, it is based simply on warming or cooling, so I think my own odds should be better :-)!
Well I would take Annan’s side of that bet. These were the numbers he could see then.
1998 0.5971 1999 0.4199 2000 0.3886
2001 0.5173 2002 0.5736 2003 0.5809
2004 0.5409 2005 0.6147
1998-2003 average .5129
He probably thought it was free money.
Now he may be a little bit more nervous.
2006 0.5583 2007 0.5455 2008 0.4792
MikeN, I googled on the bettors’ names, and as expected they turned out to be russian solar scientists. Already two years before the bet they are cited here: http://www.freerepublic.com/focus/fr/981669/posts as having come to the conclusion that a cooling is coming, and might even be in progress at that time (2003).
This is exactly my own reasoning, only they were a number of years ahead of me. Only natural, they are professional and I am only a rank amateur. However, I would like to point out the additional fact that if we have a continued warming over the next solar cycle, that is until about 2020, it will be the first time in recorded temperature history that a warming OR cooling trend has lasted for more than three solar cycles. Which I find highly improbable, even disregarding all other evidence that points to increased CO2 as having only a minor effect on global temperatures.
Some support for this is also found here: http://www.nsstc.uah.edu/atmos/christy/2009_DouglassC_EE.pdf where Douglass and Christy find a CO2 driven trend of 0.07 C per decade. I believe this to be close to the truth, even though I would rather say 0.04 – 0.10 C/decade adding that 0.04 – 0.07 is in my mind more probable than 0.07-0.10 .
Well the breakeven point for your bet is .14C, not .15C, as I explained above. This should be an easy bet for Romm to take. I’ll move further comments to the new thread.