If you testify in a court of law in the United States, you have to swear an oath to be honest. This oath requires you to tell the truth, the whole truth, and nothing but the truth. If you violate this oath in sworn testimony, you’re guilty of the crime of perjury: deliberately giving false or misleading testimony under oath.
Noel Sheppard, in his latest entry on the “Newsbusters” site, reproduces a graph prepared by Joe D’Aleo, whose work we have already examined. If Sheppard’s post or D’Aleo’s graph were sworn testimony in a court of law, I’d charge them with perjury.
The “evidence” provided in D’Aleo’s graph, which is the basis of Sheppard’s article, consists of surface temperature estimates from HadCRU (the HadCRUT3v data set), lower-troposphere satellite (MSU) temperature data (from UAH), and CO2 concentration (measured at the Mauna Loa atmospheric observatory):
When I plot the same data for the same time span, I get a very similar graph:
This graph is an example of the deceptive trick known as cherry-picking. Let me show you all the data in the HadCRUT3v data set, all the data in the UAH satellite lower-troposphere data set, and all the CO2 data from Mauna Loa; this is the whole truth:
Now let me show you testimony of Sheppard and D’Aleo; this is not the whole truth:
How astoundingly weak is their case, how deceptive are they, when their “testimony” omits over 90% of the data?
UPDATE:
A graph of 10-year averages of the HadCRUT3v data:
UPDATE #2:
A graph (from Brohan et al. 2006) showing the HadCRUT3 data, together with colored regions indicating the size of the individual components of the uncerainty, for the globe:
203 responses so far ↓
Hank Roberts // April 16, 2008 at 2:34 am
It’s puffery.
Raven // April 16, 2008 at 2:37 am
Tamino,
The climate of the earth has always changed an always will change this means that picking any starting point represents an exercise in cherry picking. In your case, you cherry picked a start point that makes the rise look as dramatic as possible.
In fact, I would say you choice of start point is more misleading than D’Aleo’s because you try to imply that the entire rise is due to humans. You also imply that the historical record is accurate and that temperatures really rose 0.8 degC of that period. A better presentation would include realistic error bars on historical temps.
[Response: Hahahahahaha!!! What are you, new?
For these graphs, I plotted all the available data. That's the whole truth.
I often start my trend analysis with 1975 because completely objective mathematical analysis indicates that that's a "turning point" in temperature data. In other words, I don't "pick" a starting point at all, I let the numbers do the talking. Unlike D'Aleo & co., I don't tell the data what to say -- I listen to what they're telling us.
As for claims that the historical record isn't accurate, that's the last refuge of scoundrels: when the data don't say what you want them to, start slinging mud on the data.]
DrCarbon // April 16, 2008 at 2:44 am
Very impressive. Now, were such cherry picking to be accompanied by a mechanism that negated 100 years of physics then we’d be on to something!
Raven // April 16, 2008 at 3:36 am
Tamino,
You did not plot all of the available data. We have temperature data going back to the 1700s in parts of Europe (which is comparable to today) and we have various sorts of proxy data going 1000s of years. Plot a trend from the Holocene Optimum and you would see a cooling trend. Call it what ver you want but it is still cherry picking to make the point that you wish make.
I also said that failing to plot the data *with appropriate error bars* is misleading because you leave the impression that the historical record is more accurate than it is. Your assertion that we know the global temperature in 19o0 to 1/10th of degree is quite ridiculous.
Your claim that 1975 represents some sort of turning point is amusing because it shows that you acknowledge that trends can change. It is pretty clear that 2002 could be another turning point but we won’t know for certain for another 5-10 years.
Hey. I don’t have any problem with you presenting your graphs and claiming that your cherry picking is better than D’Aleo’s. However, throwing around terms like ‘perjury’ is completely uncalled for given the flaws in your own analysis.
[Response: I didn't pick the data sets. It was D'Aleo who chose HadCRUT3v, UAH-MSU-TLT, and Mauna Loa CO2. But he left out almost all of it. I didn't.
It's typical denialist fare to say "2002 could be another turning point but we won't know for certain for another 5-10 years." Last Thursday could be another turning point -- when will you start on that one? I won't talk about more turning points until there's some *evidence*.
The claim that temperatures back to the 1700s in Europe are "comparable to today" is just more perjury, rather like most of your claims. As for error bars -- go tell Noel Sheppard and Joe D'Aleo to plot 'em.
It's amazing how desperate liars get when they're caught.]
trrll // April 16, 2008 at 3:46 am
Do you really need error bars? Error bars are valuable when you are showing a limited number of data points, and you have other information that tells you how much error there is likely to be in those values. But with this amount of data, anybody with any kind of statistical experience can pretty well “eyeball” the data and get a good idea of how much high frequency noise there is. It is easy to see, for example, that the 1998 peak is a spike of high frequency “noise” superimposed on a much slower upward trend. Similarly, it is easy to see that the level of noise is such that it would be hard to accurately judge the overall trend for a time period shorter than about 20 years or so.
[Response: Bingo!]
cce // April 16, 2008 at 4:19 am
The HadCRU analysis has error bars that are close to ~0.4 tall in 1850.
http://www.metoffice.gov.uk/research/hadleycentre/CR_data/Monthly/HadCRUGNS_3plots.gif
The NRC panel concluded that “most but not all” proxy indicators show that modern warmth has not been exceeded in 2000 years.
The Central England temperature series can be seen here.
http://hadobs.metoffice.com/hadcet/graphs/HadCET_graph_ylybars_uptodate.gif
I’ll leave it to those of us who posess the gift of sight to decide when temperatures were the highest in those records.
nanny_govt_sucks // April 16, 2008 at 4:20 am
I don’t think I understand your choice for the CO2 Y-axis. It looks like you’ve specifically chosen range values to get the CO2 to line up exactly with the temperature chart. Isn’t that a bit misleading? Can you show the CO2 as an anomaly just as you’ve done with temperature?
Dano // April 16, 2008 at 4:21 am
Raven is new, give him a break. He isn’t aware of the long list of denialist rhetoric that has been sifted thru here.
He may be, however, aware of this list, which he is a part of, that details the non-distracting, mischaracterizing or hand-waving excuses so far offered by denialists for their not being able to provide evidence to back their claim. Wanna be sweet 16, Rave?
Best,
D
Dano // April 16, 2008 at 4:22 am
Hmmm. Forgot about the snap preview. I’ll forego TinyUrl in the future. Apologies.
D
nanny_govt_sucks // April 16, 2008 at 4:27 am
Or use the same scale that D’Aleo used.
Simon D // April 16, 2008 at 4:41 am
The irony (to me) is that D’Aleo and others criticize climate models, but their work shows just why we need those models. What they have done is develop a model that is far too simple - monthly Temp ~ monthly CO2 - to explain the system. It is only by considering other processes, whether in theoretical models or actual computational models, that we can learn how the climate system actually works.
John N-G // April 16, 2008 at 5:22 am
Another not-so-subtle subtlety: the overlay of CO2 and T on one graph requires choosing a relative scale. In D’Aleo’s plot, 5 ppb = 0.2 C. With this scaling, the CO2 and T curves would be expected to line up if the climate sensitivity is about 12 C per CO2 doubling, with no lag in the climate system!
T/HB’s scaling on the 1500 yr plot is much closer to the more widely accepted values for CO2 climate sensitivity, including a time lag. Apply that scaling to D’Aleo’s chart, and the CO2 curve only goes up from the equivalent of 0 C in 1998 to 0.17 C in 2008. Voila, the slopes of T and CO2 look much more similar.
Most of the deception in D’Aleo’s plot is from applying an absurd relative scaling of T and CO2.
Zeke // April 16, 2008 at 5:56 am
I notice that both Noel’s graph and your reconstruction of it appear to plot UAH and HadCRU temperature series on different baselines (1979-1998 for UAH and 1961-1990 for HadCRU). Adjusting for that should give you a much greater difference in 1998 and a greater overlap over the rest of the period, and would look like this:
http://i81.photobucket.com/albums/j237/hausfath/UAH-HadCRU.jpg
Its nitpicking (and may have been purposeful on your part to recreate the original graph), but should be pointed out.
fred // April 16, 2008 at 6:26 am
This is an example of the rhetorical silliness that poisons debate on AGW. How exactly is showing a subset of the available data telling an untruth? I do not know this D’Aleo, but what he seems to have done is show a set of correct data, and then probably drawn some mistaken conclusions from it. Perhaps his conclusions were mistaken because he showed too little data? Happens.
It would be far more productive to abandon all the silly rhetoric and just point out that you would not expect short term CO2 fluctuations to correspond to short term temperature fluctuations.
That is rather obvious, but its unfortunately a one liner.
sdw // April 16, 2008 at 6:40 am
The irony is that the ‘complete’ HadCRUTv3 data show how ‘normal’ the recent warming trend appears. The rate of change is almost equivalent to the first part of the century. Low pass filter the data and look at the derivative to convince yourselves.
regards, sdw
Pete // April 16, 2008 at 8:04 am
I don’t see how this is different to those who publish studies of the Sun and then trumpet the claim “no sun link to climate change” (ie the F&L study from 1985 to 2005). 20 Years, are you serious!
Cherry Picking is rife, so it would seem. Just as you showed in your ‘Cycles’ thread that it is bad science from all parties, then so is this example from all sides as well.
I hope you apply your apparently higher standards than everyone else on the Planet to your future work in here. Lead by example!
Tuukka Simonen // April 16, 2008 at 8:12 am
Dude, you’re being so mean to those poor people! I love these short very well argumented posts.
As a whole, I think your blog is the best climate blog on the net; easy to understand and very interesting. Also, poor argument bashing is something which you never get bored. However most of the bashing need more effort, this was an easy one. :)
I keep a similar climate blog (with much less knowledge) in Finnish and I get great material from your blog. Thank you for this post and for your blog as a whole!
Julian Flood // April 16, 2008 at 8:44 am
Re Response to Raven:
quote For these graphs, I plotted all the available data. That’s the whole truth.
I often start my trend analysis with 1975 because completely objective mathematical analysis indicates that that’s a “turning point” in temperature data. In other words, I don’t “pick” a starting point at all, I let the numbers do the talking. Unlike D’Aleo & co., I don’t tell the data what to say — I listen to what they’re telling us. unquote
Have you carried out your — difficult to know what to call it — ‘listening exercise’ on other data? I’d be interested to hear what you… err… hear from Hadcrut SSTs without the Folland and Parker correction. I thought you’d have to allow large error bars at the beginning — the graph starts in 1850 — but, if trrll is correct then that is not a problem.
Eyeballing — or, as some would have it, cherry-picking — I see* a marked change in 1910. By your argument above that would indicate a good time to start measuring warming. An inexorable .14 deg/decade with an upward excursion in the 40s. Is that what you hear?
Perhaps you could you give some indication of the mathematical process you use — in general layman’s terms, nothing too rigorous — or is the maths too complex to summarise?
JF
*Sorry, ‘hear’.
Adam // April 16, 2008 at 8:49 am
Just as an aside, it’s worth noting that the graph of the Had/CRU data (at least) is plotted with error ranges on their website. Hard to see how you’d be able to use them to get rid of the long-term rise in temps.
http://www.metoffice.gov.uk/research/hadleycentre/obsdata/HadCRUT3.html
Tom Woods // April 16, 2008 at 9:00 am
Tamino said:
“As for claims that the historical record isn’t accurate, that’s the last refuge of scoundrels: when the data don’t say what you want them to, start slinging mud on the data.”
—–
Should that not be “doesn’t say what you want them to?
You’re a liar, Tamino, and your grammar is horrible! (/satire)
Okay, back to reality.
Nice repsonse and nice post. You are the path of enlightenment, my friend.
I must also ask you if you have come across anything to sea-level/temperature feeback. I still contest that 15-20% of the global temperature rise from ice age to hypsythermal is due to the feedback of rising sea-levels. Perhaps your mathematical wizardry could write up a quick post in a follow up to your feedback post. Sea-level plus 400 feet would be equal to elevation minus 400 feet (in addition to exposed continental shelf now moderated by the sea) .
I still have yet to see this aspect of feedbacks explored.
The Tuatara // April 16, 2008 at 11:49 am
Thanks, HB. d’Aleo’s graph crops in submissions made by our Kiwi cranks to hearings on Emissions Trading Scheme legislation in the NZ Parliament. Useful to be able to demonstrate the clear cherry pick.
Ta.
dean_1230 // April 16, 2008 at 11:59 am
Tamino,
I doubt you’d ever get a perjury charge to stick in this case, mainly because you’ve wrongly defined perjury. Is the chart in error? are the data points misplotted and more importantly, misplotted on purpose?
If the data is correct, it’s up to the lawyers to delve into the meaning. For D’Aleo to be charged with perjury, you’d have to show that he willfully manipulated the data he presented. You didn’t do that here. He CANNOT be charged with perjury for data he didn’t present.
[Response: The stated definition of perjury comes straight from the dictionary. It includes not just "false" but also "misleading," and I'd say they're being deliberately misleading.]
wolf // April 16, 2008 at 12:18 pm
The AGW theory requires that temperatures rise as a result of increasing levels of CO2 in the atmosphere. The earth receives energy from the sun, some of which is reflected into space and some retained by atmospheric CO2.
Despite increasing amounts of CO2, the temperature has remained more or less stable in the last 10 years. In order for the theory to be true - there must be some explanation of where the energy went. It is not in the land temperatures, ocean temperatures or the atmosphere. Where did it go?
[Response: When you people are pinned down you'll try anything to wiggle out of admitting the truth. Speaking of wiggles, I guess you're one of those who denies the existence of noise in the climate system. You can't defend one denial of truth by resorting to another.]
Alan Woods // April 16, 2008 at 12:28 pm
cce - Tamino’s post is about misleading graphs of cherry picked and truncated data. The CET record you linked to is similar - it leaves out the previous 100 years of the record and the warm 1730’s.
[Response: The 1730s weren't even close to as warm in the CET record as modern times. Must I graph them all?]
dean_1230 // April 16, 2008 at 12:41 pm
No, they are presenting data… It’s up to the lawyers to ask about the holes in the data. For example, D’Aleo would state under oath that “the chart shows that since 1998 there’s been no warming”. That’s not a lie. there’s been no warming. But it is up to the cross-examination to ask about whether 10 years is significant… or whether the data is cherry-picked.
If D’Aleo is asked whether 10 years is considered “climate” as opposed to “weather” and he says “yes”, then THAT’S wrong. If he does so in an effort to mislead, then that’s perjury.
Showing a chart and leaving it up to the lawyers and jury to spot the hole in it is NOT perjury.
wolf // April 16, 2008 at 1:09 pm
The laws of physics don’t take 10 years off. You made no attempt to answer the question.
[Response: Noise in the climate system doesn't take 10 years off either. Noise from denialists doesn't take 10 minutes off. Wiggle wiggle wiggle.
We've seen wiggling of global temperature ever since we started measuring it. It has always wiggled, it always will. Don't you know that energy is constantly exchanged between land, air, and sea, including not only the upper ocean layer but the deep ocean as well? That energy is required for the melting of ice and snow? Come to think of it, questions about "where does the energy go?" usually come from those who don't know beans about the energy budget.]
Tom C // April 16, 2008 at 1:12 pm
What a lot of wasted words. Look, the last ten years of surface data don’t amount to anything conclusive. But, they are suggestive. If the current trend continues for another 5 years we will have a lot to argue about. Another 10 years and AGW alarmism is dead.
It should also be pointed out that most skeptics, such as myself, fully expect modest continued warming from CO2 for the next 50 years, but don’t think it is a cause for alarm. Even if the curve heads north again, it doesn’t prove the cause for alarmism.
climatewonk // April 16, 2008 at 1:38 pm
My interest in this is the effect the debate over climate has on public policy, public opinion and policy makers so this kind of analysis is very signficant. When people like D’Aleo and others present cherry-picked data and base their analysis on it, claiming that global warming “stopped” in this or that year, or is negated, etc., many laypeople are unable to do a sound critique of the evidence presented to them and evaluate the conclusions. They may be swayed by bad science or pseudescience — graphs look professional, statistics convincing — unless you understand them, that is.
It is necessary to show that the work is flawed and why so that people can learn how to judge data.
So, no, not just a one liner.
Leif Svalgaard // April 16, 2008 at 1:40 pm
HB, there is some cherry-picking in your graph as well. Namely the zero-point and scale-factor for CO2. To compare with D’Aleo you should have used his choices, not rescale to make the CO2 curve magically match the Temperature. THAT is cherry picking too. May not rise to the level of perjury, but is intended to be ‘leading’ the eye and mind at least.
[Response: Baloney. The match isn't "magical."]
dhogaza // April 16, 2008 at 1:41 pm
Yes, they suggest three things:
1. The system is noisy. The AGW hypothesis does not predict that climate will suddenly become a noise-free system.
2. El Niño is real. The AGW hypothesis does not predict the end to El Niño.
3. La Niña is real. The AGW hypothesis does not predict the end to La Niña.
Goalpost move.
HB’s post doesn’t attempt to answer the question “will anthropogenic warming be harmful over the next 50 years”.
HB’s post is answering the question, “are these two people lying sacks of organic brown matter”, and the answer is, “yes”, as is anyone who starts an analysis beginning with the El Niño year of 1998.
I also question the accuracy of your statement. The flood of “global cooling has started …” and “global warming ended in the year …” crap we see from the denialsphere would argue against your claim that “most skeptics … fully expect modest continued warming from CO2″.
Leif Svalgaard // April 16, 2008 at 1:48 pm
HB: “The match isn’t “magical.”" Agreed, poor choice of words on my part. I should have said ‘carefully picked’ or, perhaps, ‘cherry picked’.
[Response: Baloney! D'Aleo tries to show that the rise in CO2 doesn't correspond to a rise in temperature -- by displaying such a short time span that the trend in temperature is eradicated. I showed that it *does* -- by displaying all the available data.]
George // April 16, 2008 at 1:54 pm
There is one more thing on that D’Aleo graph that is inaccurate/misleading.
It shows the atmospheric CO2 concentration (in ppm) rather than the radiative forcing due to the CO2 increase.
The temperature change is directly proportional to radiative forcing change, not CO2 concentration change and forcing increases less steeply with time (has a lower slope) than the CO2 concentration.
So by using concentration rather than forcing in their graphic, D’Aleo has accentuated the difference in the slopes.
[Response: If you want to do a realistic comparison, you'd also have to account for the lag in temperature response due to the thermal inertia of the oceans, and you'd want to estimate *all* forcings, not just CO2. But that would be dangerously close to actual *science*.]
Phil. // April 16, 2008 at 2:07 pm
“HB, there is some cherry-picking in your graph as well. Namely the zero-point and scale-factor for CO2. To compare with D’Aleo you should have used his choices, not rescale to make the CO2 curve magically match the Temperature. ”
Tamino, logically the correlation with CO2 should be with log(CO2), how does that plot look for this data?
wolf // April 16, 2008 at 2:08 pm
Still no answer to the physics question. 10 years is not significant in terms of overall climate variability. But it is significant in terms of the AGW theory.
[Response: What part of "deep ocean and cryosphere" don't you get? What part of "10 years is too short to reveal trends" don't you get?
If you're convinced that 10 years is significant in terms of the AGW theory, try looking at a graph of 10-year averages of global temperature. But that would make your premise look ridiculous.]
J // April 16, 2008 at 2:14 pm
There are two issues here. One, which Tamino focuses on, is the cherry-picking of dates. But there are about a zillion examples of this, so the novelty value is minimal.
The more interesting one is John N-G’s point (about the Y-axis scaling). Obviously, Mr D’Aleo has scaled the Y-axis inappropriately, but it’s not always obvious what is the best way to visualize two variables with different units.
To do this right, you really need to first log-transform the CO2 data, because of the logarithmic relationship between CO2 and temperature. (This is why climate sensitivity is derived for a doubling of CO2….)
Just for fun, I took the Mauna Loa CO2 data and the HADCRUT temperature data, log-transformed the CO2 data (base 2), and did a simple linear regression. This assumes that there’s no time lag, which is obviously false, but whatever. Over the whole time period of the ML CO2 data (1958-2008) it shows a temperature sensitivity to log CO2 of 2.24 C/doubling. For D’Aleo’s cherry-picked period, it’s much smaller (0.15 C/doubling). For 1975-present, it’s 2.75 C/doubling.
Again, these are obviously underestimates, because they don’t account for a time lag in the temperature response.
J // April 16, 2008 at 2:18 pm
Oh well, it looks like a several other people made the same points while I was composing that. I always was a little slow … sorry!
Joel Shore // April 16, 2008 at 2:19 pm
Leif: As John N-G pointed out, the scaling used by D’Aleo is deceptive in that the temperature would only track CO2 on such a plot if the transient climate sensitivity were somewhere north of 10 C. (He says 12 C; I estimated 10.something C…Since really the dependence of forcing on CO2 is expected to be logarithmic, the exact answer is not precisely defined but is unambiguously in that generally neighborhood.)
So, in fact, the choice that D’Aleo made for the relative scales of those graphs for temperature and CO2 is very deceptive (although that part may more be due to ignorance rather than purposeful deception). Tamino’s choice of scaling, where the graphs would be expected to approximately align if the transient climate sensitivity is about 2 C, is much more in line with the estimates of what the transient climate sensitivity actually is.
Joel Shore // April 16, 2008 at 2:25 pm
wolf: If you looked at plots from individual runs of climate models under a constantly increasing CO2 forcing, you would see the same sort of ups and downs that you see in the actual climate record. (See, for example, Stott et al., Science, Vol 290, op. 2133, Figure 1.) Thus, your claim that the AGW theory predicts that the temperature be steadily monotonically increasing under a monotonic increase in the CO2 forcing is provably false.
Mike B // April 16, 2008 at 2:35 pm
Atmoz did an intersting correlation of CO2 vs HadCRUT temperatures a couple of months ago, got an R value of 0.725 since 1958.
Also compared NINO, PDO, and NAO indexes against temperature again after standardizing all of them. Not surprisingly none of them correlated well on the longer term.
http://atmoz.org/blog/2008/02/10/on-the-correlation-between-temperature-and-climate-indices/
Leif Svalgaard // April 16, 2008 at 2:36 pm
HB, repeating ‘baloney’ is committing the fallacy of repetition. Don’t get me wrong, I also think that the D’Aleo graph is junk, but we [you] should not commit cherry picking either when you don’t have to. And no matter how many times you say ‘baloney’ the match of CO2 and temperature on your graph did not happen by accident, its visual impact was certainly enhanced [to bring the point across, I guess] by scaling CO2 to match T. This is also cherry picking for effect as it assumes [which may or may not be true] that ALL the change in T is due to CO2.
[Response: I didn't do any cherry picking at all and your repeating it doesn't make it true; the fallacy of repetition is yours. Wikipedia defines cherry picking as "the act of pointing at individual cases or data that seem to confirm a particular position, while ignoring a significant portion of related cases or data that may contradict that position." I specifically showed ALL the data for all three of D'Aleo's selected variables.
As for choosing the scale so that the graphs match, if we correlate one variable with another *that's what happens*.
Your entire argument seems to me to be extremely disingenuous.]
Mike B // April 16, 2008 at 2:43 pm
Leif,
Is the choice of scaling not more or less arbitrary? One could scale CO2 in such a way that the line appeared flat in comparison to temperature.
Tamino’s choice for scaling makes more physical sense than D’Aleo’s. And regardless of scale there is long term (50 year) correlation with CO2 and temperature.
cce // April 16, 2008 at 2:46 pm
There have been many words written about the AGW signal being “about 0.20 degres per decade.” Stated another way, it is about 0.02 degrees per year. AGW is buried under noise with short time periods, but that didn’t stop temperatures from rising 0.5 degrees over the past 3 decades.
The reason that the 1998 record hasn’t been broken in most analyses is because it was an absurdly warm year — well beyond anything predicted by AGW alone. 2005 is either the warmest year, or second warmest year. Every year since 2001 is warmer than all years prior to 1998.
As for Central England, anyone can download the data and plot it for themselves. 2006 is the warmest year on record, and recent years are warmer than any comparable time in the 1700s (apparently, the “Little Warm Period.”)
http://hadobs.metoffice.com/hadcet/cetml1659on.dat
wolf // April 16, 2008 at 3:08 pm
What part of “deep ocean and cryosphere” don’t you get?
Where is the supporting data?
[Response: The cryosphere is shrinking, rapidly; the evidence is everywhere, if you dispute it you're only fooling yourself. As for the deep ocean, we don't have data that I'm aware of, but riddle me this: do you really believe the deep ocean doesn't warm in response to an increase of energy entering the planet?
As for surface temperature, when the same wiggles are observed over the last decade that we've seen ever since we started measuring, where do *you* think the energy comes and goes?]
Neil B. // April 16, 2008 at 3:08 pm
Raven, and like minded skeptics: Your complaints about “you too” cherry picking from advocates of AGW ring hollow because you are ignoring the theoretical reasons why recent AGW should logically be correlated to increased average warming. (Not only that, but the relevant region to pick from would be the combination of significant CO2 rise and good data, which we have for the last few decades and which was shown here. There is noise too, and ten years isn’t really enough to iron that out, but thirty or so is much more convincing.) We already knew that the Earth went through various cycles and has been colder and warmer in the past, the point is that we know that CO2 absorbs IR and should stimulate warming (would you deny that it and other gases are “greenhouse gases” even if you don’t think they are sufficiently efficacious? At least Lumo for example will admit that much.) Hence, it isn’t just a matter of wondering if one trend correlates with another one, in a causal theoretical vacuum. That would be pure statistical analysis, and that’s not the only available tool here.
The first significant deep treatment of how increasing CO2 should warm the Earth was written way back in 1896 by future Nobel winner Svante Arrhenius. Al Gore wasn’t around to press him into writing it.
Hank Roberts // April 16, 2008 at 3:09 pm
Good list there, Dano.
Pity about what’s happened to dot.earth.
P. Lewis // April 16, 2008 at 3:13 pm
The annual average CETs for the warm decade 1730 to 1739 are 10.04, 9.85, 9.69, 10.47, 9.8, 9.54, 10.3, 9.92, 9.81, 9.2°C respectively (decade av. = 9.862).
For comparison, the 1990 to 1999 annual average CETs were 9.52, 9.86, 9.49, 10.24, 10.52, 9.2, 10.53, 10.34, 10.63°C respectively (decade av. = 10.096).
The decades either side of 1730-1739 were much cooler. For decades ending 1719, 1729, 1749, 1759, 1769 the decadal averages were 9.153, 9.307, 8.854, 9.042, 9.058°C respectively.
For “decades” ending 1969, 1979, 1989, 2008 (latter, 8 years only), the “decadal” averages are 9.277, 9.539, 9.518, 10.4448°C (latter, 8 years only) respectively.
Leif Svalgaard // April 16, 2008 at 3:19 pm
Mike B and HB: You would have gained the high ground by simply extending D’Aleos’s graph to include ALL the data [HB, do that] without changing anything else. Also changing the scale [to one that make the two curves match] goes beyond that.
[Response: I certainly included all the data -- that's the point -- and if I'd kept the same ridiculous scale as used by D'Aleo the range of the temperature axis would be so large that it would compress the temperature time series, giving a very *false* impression of the temperature trend -- very much like the misleading tactic which is at the root of D'Aleo's presentation.
Your insistence that I apply a misleading relative scale between the variables is further evidence that you're being extremely disingenuous.]
P. Lewis // April 16, 2008 at 3:32 pm
Danforth, in Arthur Miller’s The Crucible, says:
And from the ‘Lectric Law Library’s Lexicon on perjury:
The question then becomes whether the likes of the aforementioned in the OP would wish it to be known that, if found hypothetically guilty, they were hypothetically perjurers or were just incompetents if a case came before some court or body in which an oath is taken and they were found wanting. Such an outcome would seem to me not to be particularly beneficial to their advocacy (though I’m fairly sure that that would not affect how the vast majority of their acolytes would view their probity and knowledge one iota!).
Of course, unless one ends up in court (say, because of libel or slander) or presumably in some Congressional testimony (?), then this situation, re perjury, is unlikely to be tested.
wolf // April 16, 2008 at 3:36 pm
Joel - ok, agreed - poor articulation on my part.
What I was trying to say - is that if the temperatures are stable - that AGW requires the energy to stay within the earth. There should be a signature - either by increased land and ocean surface temperatures, increased temperatures within the ocean, or increased atmospheric temperatures.
Where is the signature in the last 10 years?
[Response: Where's the signature in the last 10 days?
Your entire argument is nothing more that a thinly veiled claim that there's no noise in surface temperature, so it has to show a "signature" no matter how short a time span is chosen. The global temperature time series has been noisy since we started measuring it, it was noisy before we started measuring, and it will remain so for all time.
As for the last decade, that *starts* at the peak of a giant el Nino -- when heat was transferred to the atmosphere *from the ocean*. It wasn't a sudden increase of total energy in the earth as a whole, it was a sudden exchange of energy between earth subsystems, with energy being deposited in the part which we measure as surface temperature.]
Bob G // April 16, 2008 at 3:54 pm
Tamino,
You mention the importance of accounting for the lag in temperatures due to the thermal inertia of the oceans. I have never seen this calculated, only grossly estimated. Looking at the GISS ocean/land temp graph provided in the handy links at the top of your home page (climate data links/NASA GISS/graphs/page down to land-ocean), there is no evidence in these non-statistically trained eyes of any lag in ocean temperatures when compared to land temperatures. The ocean temperatures are muted, with rates of warming and cooling at roughly half of the land values. But both land and ocean temperatures were at a low point around 1910, rose to a high point in the mid 40s, dropped a bit to the mid-late 70s and rose thereafter. Where is the evidence of a lag? Should it be apparent in this 127 year record?
[Response: The lag due to thermal inertia of the oceans doesn't mean that the oceans will warm later than the atmosphere. It means that the full response of the system is delayed because it takes so long for the oceans to warm compared to the atmosphere. The evidence that this is actually happening *is* that the rate of ocean warming (and cooling) is less than the rate of land/atmosphere warming.]
P. Lewis // April 16, 2008 at 4:07 pm
Added to the legal situation above, there also is the issue of professional misconduct.
Should anyone affiliated to a professional body or academic institution transgress their professional codes of conduct or ethics, especially when dealing with the public, then they can be held answerable to that body.
For example, “Article XII. Guidelines for Professional Conduct” of the American Meteorological Society’s (AMS’s) Constitution (20 January 2008) states the following (any emphasis is mine):
And this set of AMS guidelines is quite wishy-washy compared with my former institute’s guidelines on professional misconduct and bringing the name of the institute into disrepute. In fact, my former institute expressly expects its members to notify “of any significant breach of the Rules for Professional Conduct by another member and should seriously consider reporting another professional with whom they may be professionally connected, to their professional body if he or she appears to be breaching the general principles of ethical practice and who is likely to bring the profession into disrepute”.
What would you do if you saw a brother/sister member acting in a fashion so as to sully the name of your body/institute? Report or not?
I’m surprised this has not (to my knowledge) surfaced in this age of the blogs (other than in some companies not looking favourably on some employees’ blog comments on their employers leading to employee dismissals) — ah, perhaps I can think of one possible case.
Now, innocent until proven guilty in such cases of course, but I know of one AMS fellow and council(l)or who publishes in the public domain who might conceivably be considered to have fallen foul of articles 3A, 3B, 3D and 3E in some respects. Whether such an actual transgression of rules of professional conduct (or ethics) is considered to have taken place, though, is wholly the responsibility of the particular governing professional body. What would I do?
DrCarbon // April 16, 2008 at 4:08 pm
Hey Tamino,
A minor quibble. Could you start labeling your axes? It’s good practice - I’ve noticed a lot of failure to label esp. y-axes for the crowd that advocates post 1998 cooling. By failing to label they can draw attention to perceived trends rather than absolutes. Be nice to show them how it’s done correctly.
David B. Benson // April 16, 2008 at 4:36 pm
cce wrote “There have been many words written about the AGW signal being ‘about 0.20 degres per decade.’” I contrast this with the warming/cooling during the Eem of 0.23 per 50 years, i.e., 0.046 per decade.
I posted regarding the Eem and ‘natural climate variability’ on the Open Thread yesterday, but nobody seems to have any comments. :-(
Oh yes. I chose to start with the Eem since there was no anthropogenic influence (worth mentioning) then…
Leif Svalgaard // April 16, 2008 at 4:52 pm
HB: D’Aleo’s temperature scale differs from yours by a factor of three [relative to the CO2]. Maybe he [as many others of his ilk] has a different opinion on the sensitivity of T to changes in CO2. That is certainly allowed, or maybe the whole issue is that he isn’t allowed to disagree? I do, obviously, agree with you that one should show ALL data and that he is being dishonest by not doing so, but I do not agree that he is not allowed to select another sensitivity without being accused of giving a “false” impression. And that is my point.
[Response: I never claimed, never even hinted, that he was dishonest for his choice of relative scale between temperature and CO2 axes, only that he was dishonest for leaving out all but a tiny fraction of the data so he could avoid the part that contradicted his claim (which is the *definition* of cherry-picking). *You* claimed that *I* was dishonest for *my* choice of relative scale between the axes. That ain't right.]
Leif Svalgaard // April 16, 2008 at 5:17 pm
HB: I never claimed that you were dishonest just that you cherry picked a sensitivity that was your favorite and not his. He is allowed to pick a sensitivity that he likes without being accused of given a *false* impression.
trrll // April 16, 2008 at 5:34 pm
Apologies from previous fragmentary post.
They say that there is no such thing as a stupid question, but this comes close. Virtually all of the energy received from the sun is radiated back into space. This has to be the case, or the temperature would increase. The fact that there is a long-term upward trend in temperature indicates that a very, very tiny fraction of the total is being retained. As the average temperature increases, the average amount of energy radiated to space will increase, until ultimately a new steady-state is established at a higher temperature.
Any kind of energy accounting has to consider not just temperature, but all forms of energy, including the kinetic energy in winds and ocean currents and the chemical potential energy of all the forms of matter on earth. For example, if you put energy into a mixture of ice and water, you will affect the ratio of ice to water, but you won’t affect the temperature at all. You may have noticed that we have ice and water here on earth.
Moreover, there is no reason to expect incoming solar energy to have a constant impact on temperature. Light energy that is reflected from clouds will have much less impact on temperature than light that is absorbed by earth and eventually re-radiated as lower-energy infrared. So such things as cloud cover are going to produce short term (”weather”) fluctuations in temperature even if the amount of solar energy falling upon the earth is absolutely constant.
J // April 16, 2008 at 5:38 pm
Leif writes: “He is allowed to pick a sensitivity that he likes without being accused of given a *false* impression.”
A little consistency would be good. If Mr D’Aleo really believes that climate sensitivity is on the order of 10-12 C per 2XCO2, it would be nice if he would say so explicitly. I doubt that any more than a minuscule fraction of his readership understands the implications of his graph.
Joel Shore // April 16, 2008 at 5:44 pm
Leif: Now that’s just silly…It is very disingenuous of him to implicitly choose a transient climate sensitivity of about 10 or 12 C by his relative choice of axes. If he honestly believes that this is a reasonable estimate of climate sensitivity, then he belongs in the “super-alarmist” camp not the “skeptical camp”…and when the climate then fails to follow his expectations all that it shows is that his expectations were whacked, not that CO2 doesn’t have the affect on climate that most people think it does. So, I would expect Aleo to be hopping mad about the abuse of his graph to claim something with no relation to what we might charitably say he was trying to show.
Sometimes, when you are in hole, it is best to stop digging by just admitting you are wrong.
Joel Shore // April 16, 2008 at 5:55 pm
By the way, it is important to note how the poor choice of the relative scalings of the CO2 and temperature axes is vital to the visual effect of Aleo plot. To see this, just mentally picture what his plot would look like if the CO2 line, instead of going between about 0 and 0.8 in reference to the temperature axis instead went from about 0 to 0.015. Then, it would be immediately visually apparent that the noise in the temperature plot over this time period is too large to tell whether the trend in the temperature plot disagrees with the trend in the CO2!
Lee // April 16, 2008 at 5:58 pm
“They say that there is no such thing as a stupid question, but this comes close”
I think I’ve said this before here - the professor who taught me to teach well used to say that every good teacher tells his students one lie - that there are no stupid questions.
L Miller // April 16, 2008 at 5:59 pm
Putting aside the fact he cherry picked a limited data set, the problem with what you are saying is that D’aleo choice of CO2 sensitivity is exceptionally high. He is setting up a strawman and showing that warming is below that exceptionally high CO2 sensitivity therefore “proving” CO2 isn’t causing the warming.
Even if he hadn’t cherry picked the data set and accounted for lag all he could have show with that scale choice is that CO2 sensitivity is below 10-12 deg. This is clearly a strawman since no one seriously believes sensitivity is that high. Under those circumstances it’s rather ridiculous to expect HB to stick to that choice of scale.
L Miller // April 16, 2008 at 6:07 pm
Sea level rise due to thermal expansion is directly proportional to the total energy in the ocean and has continued to increase in the last decade. Since the oceans hold over 90% of all the energy in the earths climate system this is a very strong indication that energy has continued to increase.
Leif Svalgaard // April 16, 2008 at 6:10 pm
Joel: “By the way, it is important to note how the poor choice of the relative scalings of the CO2 and temperature axes is vital to the visual effect of Aleo plot.”
But the very same is true of the Tamino plot. I don’t know what scaling D’Aleo likes, and I don’t really care, my only point was that if D’Aleo had plotted ALL the data even with his screwy scaling he could not have been guilty of perjury, no matter how bad his scaling was. But changing the scale because otherwise the Figure would give a *false” impression, that “ain’t right”. If the D’Aleo plot with the D’Aleo scaling, but with ALL the data, would look ridiculous, so more the better for people that disagree with him.
P. Lewis // April 16, 2008 at 6:23 pm
PS. There appears to be an HTML gremlin abroad. That was meant to be 2008 followed by a closing parenthesis, not a smiley.
dhogaza // April 16, 2008 at 6:36 pm
If his claim is that his graph disproves IPPC projections based on a roughly 3C forcing per CO2 doubling then, yes, it’s perjury, because his graph shows no such thing, because his scale reflects a 10-12C forcing. That’s not a figure that enters into the scientific argument.
It’s no different than saying “the IPCC is wrong because global temps haven’t risen by 30C this past year”, or “the IPCC is wrong because the greenland icecap hasn’t melted this year”.
dhogaza // April 16, 2008 at 6:41 pm
Leif, consider that one could cause that line to be virtually flat no matter what level of correlation exist by choosing an unrealistic forcing level.
IPCC predicts 3C per doubling, and long term trends track that? You can prove they’re “wrong” by choosing a y-scale for CO2 concentrations that corresponds with a forcing of (say) 100C.
And you’d say that presenting this as “proof” that CO2 doesn’t cause warming isn’t dishonest? Isn’t a lie? Wouldn’t be perjury in a court of law?
If you have any doubt as to whether or not this is being done, go read the post linked to by HB above.
Thomas Huxley // April 16, 2008 at 6:42 pm
Can we be really sure Aleo understood the implications ? :-)
.
P. Lewis // April 16, 2008 at 6:51 pm
Leif
You appear to be guilty of not reading Tamino’s piece closely enough.
You wrote:
Of course he wouldn’t be! Plotting the data doesn’t make one guilty of perjury, and your quoted point is saying something Tamino never actually claimed.
What Tamino actually wrote was:
It’s the swearing of something as true under oath when it is known to be false that would be perjury.
——————-
PS to all. My CET 2008 should, of course, have read 2007 … oops!
David B. Benson // April 16, 2008 at 6:56 pm
P. Lewis // April 16, 2008 at 4:07 pm — I encourage reporting percieved profession misconduct to whomever in the professional organization is charged with receiving such reports. If it is not obvious who that individual is, then send the report to the president of the professional organization.
Professional ethics are not matters to be lightly discarded.
Leif Svalgaard // April 16, 2008 at 7:20 pm
All Tamino had to do to show how ridiculous D’Aleo’s graph is was just to plot ALL the data with the scaling chosen by D’Aleo. D’Aleo is his own worst enemy. But, it is not correct to say that D’Aleo creates the *false* impression that CO2 has no effect by his choice of scaling. Choosing Tamino’s scaling does not make things true as there is debate [not with D'Aleo] on what the scaling is. You can say the scaling is this or that, which would then be your contribution to the debate, but this does not close the debate.
[Response: You're not convincing anybody, and you don't flatter yourself by trying so hard. The scaling wasn't even an issue until you made it one. And your earlier statement that "I never claimed that you were dishonest just that you cherry picked" is an insult to our intelligence.]
Leif Svalgaard // April 16, 2008 at 8:03 pm
HB: “You’re not convincing anybody, and you don’t flatter yourself by trying so hard.”
In these polarized and polarizing blogs, nobody ever convinces anybody of anything. I was just voicing my opinion.
Raven // April 16, 2008 at 8:12 pm
Good grief.
Question: Would a -1 degC/decade trend over 10 years with no volcanic events demonstrate that the IPCC predictions were wrong?
Answer: Most definately. (if someone disagrees then you are basically saying that the IPCC is peddling unfalsifiable psuedoscience).
Question: Would a flat trend over 10 years with no volcanics events demonstrate that the IPCC predictions were wrong?
Answer: It depends. The IPCC did not clearly state its expectations for weather noise.
Since the IPCC did not clearly state what its expectations were that means that anyone analyzing the trends is free to make reasonable assumptions. That means D’Aleo’s claim that the last 10 years of data contradicts the IPCC claims is perfectly reasonable.
Now others can disagree and make other assumptions about what the IPCC expectations for weather noise were. However, you can’t say D’Aleo misrepresented anything. The data prior to 1998 is irrelevant if one is making the claim that the last 10 years of data is not consistent with the IPCC predictions.
[Response: The data prior to 1998 is NOT irrelevant. And EVEN IF we completely ignore it, the last 10 years of data IS consistent with IPCC projections.]
Mike B // April 16, 2008 at 8:30 pm
Leif said:
Tamino never even mentions the scaling of the CO2 in the original post, it wasn’t part of his criticism. If Tamino was guilty of anything it’s rescaling the CO2 plot at a more intelligent level and not mentioning it.
Raven // April 16, 2008 at 8:35 pm
Tamino says:
“And EVEN IF we completely ignore it, the last 10 years of data IS consistent with IPCC projections.”
Only if someone accepts your assumptions about what the IPCC meant. Others have made reasonable and scientifically justifiable assumptions an use those to show the last few years are not consistent with the IPCC projections.
[Response: Baloney. You obviously have no idea of the *probable error* in the trend over the last 10 years.]
Leif Svalgaard // April 16, 2008 at 8:58 pm
Mike B: “Tamino never even mentions the scaling of the CO2 in the original post, it wasn’t part of his criticism. If Tamino was guilty of anything it’s rescaling the CO2 plot at a more intelligent level and not mentioning it.”
It is about the truth and the whole truth, …
Furthermore, later he said “creating the *false* impression”. The more ‘intelligent’ rescaling is using ‘intelligent’ as in ‘intelligent design’, namely as the meaning that fits your own a priory opinion. Hasn’t enough been said about this already?
[Response: Too much has been said. By you. You're the one who *created* the issue of scaling by accusing me of cherry-picking, which *is* accusing me of dishonesty even though you deny it. Shame on you! And plotting the temperature time series but making the range of the temperature axis ridiculously large *will* create a false impression.
You made a foolish objection, it's been shown over and over again to be wrong, but you're too attached to your ego to let it go.]
dhogaza // April 16, 2008 at 9:24 pm
Your task, if you choose to pursue it, is to point us to where the IPCC has claimed that the anthropogenic global warming hypothesis has predicted that El Niño and La Niña events would stop prior to 1998 …
Rattus Norvegicus // April 16, 2008 at 9:38 pm
Raven, you might want to read this post and come back and tell us what you learned.
Leif Svalgaard // April 16, 2008 at 10:28 pm
HB: It is not about ego at all. And I do not take ‘foolish’, etc as worthy from you [true or not]. All I asked you to do was to tell the whole truth but in a way that was compatible with D’Aleo [what he should have done]. I have simply extended his graph to earlier times [I had data handy back to 1880, so it is not the whole truth, but close enough]. The graph is here http://www.leif.org/research/DAleo.png
I don’t know of a way for me to show an image, so maybe I could ask to to do my the favor and show the figure for me, if at all possible? My graph shows the folly of D’Aleo’s graph from the view point of AGW. Now, I don’t think D’Aleo even thought of this issue. He probably just let the graphing software pick the scaling. And people saying that I was advocating a 12C for 2xCO2 just got this backwards. The D’Aleo scaling implies a much smaller sensitivity as is evident from the graph.
TCO // April 16, 2008 at 10:35 pm
Tammy:
A. I agree with you on scaling and using all the data.
B. picking at 1975 can cut both ways. Yes it IS a turn of the data. And we see it visually. And can describe it mathematically. But the point is, is it better to show all the data (including pre-1975) to show the “changeability of data) or is there some reason like aerosols to cut at that point. I think givent the issue is in contention itself. debate, it is not best form to cut there. This is a very mild, minor comment.
C. It would be helpful if you could find a case of offense by “your side” on the perjury issue so that we can disaggregate sidedness of the debate with “what is perjury”. This would NOT prevent a further examination that showed the Rethuglicans (like me) more at fault in frequency offence, but would help in definition of the failing itself.
D. My impression is that you have NOT done a detailed thoughtful investigation of what are the legal perjury findings with respect to expert testimony (looking at law articles, posting on Volokh.com, case example research, etc.) but are trying to use the CONNOTATION of the word or the SPECTRE of legal proceedings to score some points against the idiots against the other side.
E. None of the above changes my wanting to take idiots from my side and hold them firmly head under arm and give nuggies to the extent of pain. Then slappity slappity slap.
P. Lewis // April 16, 2008 at 11:18 pm
David B. Benson said
Yes, of course David. I hope you realise my question was rhetorical on behalf of all, which I grant is not easy to appreciate in a written comment.
Eli Rabett // April 16, 2008 at 11:26 pm
FWIW pushing the CET back later than 1730 is a risky business, first because a bunch of the measurements were not made in central england as I recall, but are fill ins from the Netherlands and then because some of the instruments and methods were very funky. 1730 is a pretty safe start although later might have been better. The origin of this urban legend is M&M, it was originally raised by them in their endearing clueless way. However, if you RTFR, particularly
Parker, D. E., T. P. Legg, and C. K. Folland, 1992: A new daily Central England Temperature Series, 1772-1991. Int J Climatol, 12, 317-342.
The reason that data before 1730 is chancy is clearly explained in the first paragraph of the 1992 Parker, Legg and Folland paper:
“Manley1953) published a time series of monthly mean temperatures representative of central England for 1698-1952, followed (Manley 1974) by an extended and revised series for 1659-1973. Up to 1814 his data are based mainly on overlapping sequences of observations from a variety of carefully chosen and documented locations. Up to 1722, available instrumental records fail to overlap and Manley needs to use non-instrumental series for Utrecht compiled by Labrijn (1945), in order to mate the monthly central England temperature (CET) series complete. Between 1723 and the 1760s there are no gaps in the composite instrumental record, but the observations generally were taken in unheated rooms rather than with a truly outdoor exposure….”
Which means that the Manley reconstruction is only continuous from 1722 on, but the information upon which it relies from 1723-28 has further difficulties, essentially absolute values were not reliable, and the series was constructed by taking the difference between measurements made by those thermometers and ones thought to be more reliable after 1727, and then repeatedly differenced to get values before 1727.
In the light of this, it is perfectly reasonable to truncate the CET series at 1730.
On balance this illustrates the principal of RTFR and the danger of someone unfamiliar with an area trying to do an “audit”. To use the CET before 1730 would clearly have been a mistake, and I think you could make a good argument that it should have been cut off at 1772.
nanny_govt_sucks // April 16, 2008 at 11:43 pm
Leif, thanks for the graphic. I was curious how this data would look using D’Aleo’s scaling for CO2. Tamino/HB, wouldn’t CO2 charted as an anomaly, and lined up with the temp anomaly be less controversial?
Leif Svalgaard // April 16, 2008 at 11:49 pm
Nanny: and both standardized to std-dev = 1 for their common interval.
Leif Svalgaard // April 17, 2008 at 12:04 am
Nanny, here is the graph with zeroed and normalized T and CO2:
http://www.leif.org/research/DAleo2.png
This is the graph HB could have produced instead of dishing out vitriol.
Leif Svalgaard // April 17, 2008 at 12:16 am
I forgot to mention (the whole truth thing) that I also removed the annual variation of CO2, since D’Aleo did and since the temperatures are anomalies too.
Alan Woods // April 17, 2008 at 1:00 am
Tamino, you say “The 1730s weren’t even close to as warm in the CET record as modern times. Must I graph them all?”.
Well, I wasn’t asking *you* to graph them all. But anyway, according to Jones and Briffa (2006). “UNUSUAL CLIMATE IN NORTHWEST EUROPE DURING THE
PERIOD 1730 TO 1745 …”:
“On an annual basis, mean temperatures for the period 1729–1738 are only 0.3 ◦C below the average for the last ten years (1995–2004)”
J&B also say:
“The rapid warming in the CET record from the 1690s to the 1730s and then the extreme cold year of 1740 are examples of the magnitude of natural changes which can potentially be recorded in long series. Consideration of variability in these records from the early 19th century, therefore, may underestimate the range that is possible.”
That is why it is important to plot the record in its entirety.
[Response: I wonder whether their "only 0.3 deg.C below" applies to the CET record -- because I'm looking at the data right now and the difference between the 1729-1738 decade and the 1995-2004 decade is 0.44 deg.C, not 0.3. But we have even more recent data; the difference between the 1729-1738 decade and the 1997-2006 decade is fully 0.59 deg.C, just about *twice* the quoted figure 0.3. A full 0.59 deg.C warmer than the warmest pre-20th century decade you can find fits my description of "weren't even close."
Did you think I hadn't looked at the data?]
L Miller // April 17, 2008 at 1:54 am
That wasn’t the objection. The objection was that any sensitivity less then 10-12 deg for X2 CO2 would make the temperature plot look flat compared to the CO2. IE it would make it appear as if CO2 were rising sharply with little or no change in temperature. This choice of scale allowed D’Aleo to say “look how big the CO2 change is and yet there is almost no change in temperature” even though the sharpness of the CO2 rise is coming entirely from the choice of scale.
Alan Woods // April 17, 2008 at 2:20 am
Tamino, I don’t know why you wonder if their analysis refers to the CET record. You think I’m pulling your chain? Read the paper.
[Response: Because the numbers don't match. The paper is behind a paywall.
You think I'm pulling *your* chain? Go look at the data.]
Alan Woods // April 17, 2008 at 3:54 am
OK Tamino, I realise my post may have been a bit obtuse when refferring to Jones and Briffa, but the quote directly refers to the CET.
Now, I don’t know why Phil Jones and Keith Briffa’s analysis differs to yours, but I think they might be actually looking at annual means within two decades they are comparing, rather then a decadal mean of annual temperatures.
Regardless, I think that my original point (whi I thought was the same as yours) still stands. i.e. that all the data should be included.
I don’t know where you delineate close/not even close when comparing temperatures (is it 0.3, o.45, or 0.6?) but I would have thought that the warmth observed in the 1730’s quite remarkable given the minimal human influences on climate then when compared with today.
[Response: Silly me, actually looking at the data.]
Alan Woods // April 17, 2008 at 4:28 am
Tamino, don’t be so facetious. If you remember this started because cce linked to a truncated version of CET (for some reason you interpreted it as a call for *you* to include the data). Thus, when I say all data should be included I mean all the data in the CET.
With respect to you “actually looking at the data”, so did I, and so did Phil Jones and Keith Briffa. But, if you’re going to use a non-physical and subective term like “not even close” to describe it then obviously we need to determine what you actually mean by it. Not to mention the fact that you’re ignoring the obvious context of the LIA and minimal human impact on climate at that time.
[Response: You objected to the graph linked to by cce because it left out the "very warm 1730s." I pointed out that the 1730s weren't nearly as warm as modern times. You refer to a paper by Jones & Briffa, giving a quote which clearly implies there's a decade "only" 0.3 deg.C cooler than the most recent decade. I had *already* looked at the data before saying "weren't nearly as warm," and 0.3 deg.C decadal comparison ain't so. Now you say "I think they might..." and "you're ignoring the obvious context of the LIA..." And you're telling *me* not to be facetious?
You want to know what I mean by "not even close"? OK, I'll post that too.]
cce // April 17, 2008 at 5:29 am
Here is the full CET data with 5 year moving average:
http://cce.890m.com/cet.jpg
The warmest year is 2006, 0.35 degrees warmer than anything in the 1700s.
The warmest 5 year period is 2002 through 2006, 0.562 degrees warmer than anything in the 1700s.
The warmest 10 year period is 1997 through 2006, 0.589 degrees warmer than anything in the 1700s.
Alan Woods // April 17, 2008 at 7:14 am
I said the warm 1730s, Tamino. I didn’t say ‘very warm’. Warm is a relative term is it not? Relative to the CET record the 1730s were warm. If you don’t think so I suggest you read H.H. Lamb’s “Climate History and the Modern World”.
In the CET record, the 1730s are one of the warmest decades and aren’t really matched until the 1990’s. Considering the LIA was all the rage and there was minimal human impact on climate back then, I find that quite remarkable. Indeed, people at the time found it quite remarkable, which is why they were in for a rude shock when the winter of 1739-40 hit.
Now, why do you have an objection to the CET record being displayed in its entirety?
P. Lewis // April 17, 2008 at 12:37 pm
Leif and n_g_s
Surely it doesn’t matter at all (much?) whether the CO2 data, when plotted on the same graph as the temperature anomaly data, are plotted as an anomaly or as ppm as long as the ppm scale is chosen appropriately in the first place. Why? Because all you have to do to get from any CO2 anomaly to the actual ppm value is to add back in the actual value of your chosen zero level. The relative positions of the two trends will be identical if the ppm scale was chosen correctly; just the actual numbers will change.
And comparing your last linked graph illustrates this nicely it seems. Surely, Tamino’s plot and yours are essentially indistinguishable if you put back in the actual numbers on your chosen anomaly scale. On that basis alone I’d conclude that Tamino chose the correct scale for plotting CO2 on the temperature plot.
And Leif, in the interests of giving the whole truth, it would be helpful to add in what the anomalies in your plot actually mean, since you obviously don’t mean that the global average temperature has gone up ~5°C since 1910 (which might be a normal first assumption when dealing with Hadley temperature anomaly data).
Boris // April 17, 2008 at 1:43 pm
“the 1730s are one of the warmest decades”
CET is Central England Temperature, right? So we’re talking about a very small region, correct? And this is supposed to compare to global temperatures how?
Hansen's Bulldog // April 17, 2008 at 2:50 pm
The discussion of Central England Temperature (CET) was actually spurred by a truly idiotic comment by Raven:
This caused cce to link to a graph (from the Hadley Centre website) of the longest of all temperature records, CET. Then Alan Woods mentioned that the Hadley Centre graph omitted part of the CET record, including the warm 1730s. I replied that the 1730s weren’t even close to as warm in the CET record as modern times. Eli Rabett gave much interesting background on the CET record, pointing out that the Hadley Centre was not being deceptive in omitting part of the CET data because the earlier data were demonstrably less precise, and that there’s plausible reason to argue for cutting off data before 1772. Alan Woods quoted a paper by Jones & Briffa which implies that there’s a decade (very close to the “1730s” ) which is only 0.3 deg.C cooler than the most recent decade (at the time of their publication). I responded that the data contradict that. In addition, data after that publication have made the “most recent decade” even hotter. It’s still not exactly clear what Jones & Briffa were referring to. Finally cce has linked to a plot of the entire CET record. I’ll have more to say about it in an imminent post.
But the whole thing started with Raven’s moronic contribution. CET is not global temperature, “parts of Europe” isn’t either, and those data going back to the 1700s are not “comparable to today.” Raven’s comment was nothing more than an attempt to deflect attention away from the topic of this post: Joe D’Aleo’s blatant cherry-picking (omitting data because they disagree with his claims) and Noel Sheppard’s propagandizing it.
Hank Roberts // April 17, 2008 at 3:19 pm
Well, yeah. There’s either one very active longterm person or a flock of ‘em using the same userid and approach out there on climate blogs.
Leif Svalgaard // April 17, 2008 at 3:44 pm
P. Lewis:And Leif, in the interests of giving the whole truth, it would be helpful to add in what the anomalies in your plot actually mean, since you obviously don’t mean that the global average temperature has gone up ~5°C.
No snide comments please. As I clearly explained the data [both T and CO2] have been ‘zeroed’ [mean subtracted], then divided by the standard deviation of the result. In both cases for the same time interval [1958-2008]. So the Y-scale is obviously not °C. The reason my plot looks just like Tammy’s is not physics or ‘appropriate’ scaling. I guess [for his plot] he did a least squares fit to get the scaling factor. This will force the two curves to agree, regardless what the ‘true’ sensitivity’ is.
David B. Benson // April 17, 2008 at 4:50 pm
Leif Svalgaard // April 17, 2008 at 12:04 am — That is a nicely done, very clear plot. Thank you. Could you do it again using
http://cdiac.ornl.gov/trends/emis/tre_glob.htm
as the basis for the CO2 line?
J // April 17, 2008 at 4:58 pm
Wow, there’s a lot of discord in this thread. I generally enjoy reading what both Tamino and Leif have to say, and the slightly nasty tones back and forth seem a bit uncalled for.
Tamino, your postings are as always great, but sometimes lately your responses to comments seem to escalate to too angry too fast. Leif, I’ve only seen “Tammy” used by people at (or from) CA who are being deliberately obnoxious, and was surprised to see you adopting that.
You should both go sit in the corner now.
:-)
chopbox // April 17, 2008 at 5:32 pm
I’ve found this discussion to be interesting for a couple of reasons.
The first was a good demonstration (for ALL OF US) on how to continue with the main point of a discuss