Open Mind

What if … ?

January 15, 2009 · 141 Comments

NASA GISS has released the year-end temperature figures for 2008. The numbers are bound to change in the near future, but only slightly, because more data is forthcoming; there’s always a delay receiving all the data from stations worldwide. But enough data has arrived to estimate the temperature anomalies for the year.


Here’s the annual average temperature from 1880 through 2008:

globe

There’s clearly a warming pattern present, especially over the last several decades. Just as clearly, there’s a lot of up-and-down jitter from year to year. The overall pattern is the trend, the jitter is the noise. We can reduce the noise level while preserving the main features of the trend, by computing averages over time spans longer than a single year. These data are already 1-year averages; here are 10-year averages (the last of which has only 9 years, not 10, because 2009 isn’t over yet):

10yrave

Although taking averages over longer time spans is the simplest way to reduce the noise level and reveal the trend, we can get a better picture by applying a good smoothing method to the data (in this case, a lowess smooth):

smooth

Although the smoothed version has more detail, it tells essentially the same story as the 10-year averages. In spite of up-and-down jitter from year to year, over the long haul it’s getting hotter. Cleary the trend is upward. But the year 2008 was below the trend, and cooler than 2007, so last year is only the 9th-hottest on record.

We can compute the difference between each yearly value and the smoothed value for the same year. These are the residuals:

resid

It’s evident that the residual for 2008 is negative, but it’s equally evident that the 2008 residual is not at all extraordinary.

Pretty simple, right? Apparently, not to everybody. Realclimate has a post about an episode of “Lou Dobbs Tonight” on CNN, in which Dobbs features an extremely one-sided group of so-called “leading experts.”

Even before the live discussion begins, some comments are reported by videotape. First up to bat is Dennis “Unstoppable Global Warming Every 1500 Years” Avery of the ultraconservative Hudson Institute, and the first words we hear from him are these:


The earth’s temperatures have dropped an average of .6 Celsius in the last two years.

In the Realclimate post, NASA’s Gavin Schmidt says rather plainly:


As for his great cherry pick (0.6º C in two years - we’re doomed!), this appears to simply be made up. Even putting aside the nonsense of concluding anything from a two year trend, if you take monthly values and start at the peak value at the height of the last El Niño event of January 2007 and do no actual trend analysis, I can find no data set that gives a drop of 0.6ºC. Even UAH MSU-LT gives only 0.4ºC. The issue being not that it hasn’t been cooler this year than last, but why make up numbers? This is purely rhetorical of course, they make up numbers because they don’t care about whether what they say is true or not.

I don’t think Gavin gives Avery enough credit. Credit for obfuscation, that is. Why, I can get more than 0.7ºC cooling in the last two years, with GISS data! Of course, I’d have to use monthly rather than annual data to emphasize the noise over the trend as much as possible, but then we already know that two years doesn’t really tell us anything useful about the trend. And of course, it can’t really be “the last two years” because if you end with December 2008 it doesn’t happen. But if you go from January 2007 to January 2008 — from a big el Nino to a big la Nina — then you’ll get the required 0.7ºC cooling:

drop6

Could this be the real trend? It uses an one whole year of data!!!

No.

Whether Avery just “made up” his .6 Celsius, or got it from the kind of wouldn’t-know-the-trend-from-the-noise-if-it-bit-you-on-the-ass cherry-picking I’ve just illustrated, I don’t know. But I do know that his “statistic” is misleading — deliberately misleading — it’s exactly the kind of statement that prompted Benjamin Disrael to say, and Mark Twain to popularize, a famous quote about lies and damned lies.

Now to the title of this post: What if … I did the exact same thing? Suppose I said that over the last 20 years, global temperature has risen by 0.96ºC, and the even more important fact is that the rate of warming is 6.7 ºC/century. We’re sure to be drowning in molten lava by the end of the century!

bigrise

Seriously: what would be the reaction in the blogosphere?

Categories: Global Warming
Tagged:

141 responses so far ↓

  • Kevin McKinney // January 15, 2009 at 2:47 am

    Why, they’d go ballistic!

    Though on reflection, could anyone tell?

    BTW–and forgive a newbie on this site if the status quo represents a well-reasoned policy–but I notice that typically the data source used is not specified. How ’bout adding a little note somewhere that it’s GISS or SR05 or whatever?

  • Malcolm // January 15, 2009 at 2:53 am

    Hi, I predicted a value of 0.43 back in July at:

    http://hot-topic.co.nz/when-gray-turns-to-blueflung-a-dummy/#comment-1823

    I’m not aware of any other forecasts for 2008. I mention this to indicate that, although I am not a climate scientist, I don’t think it unreasonable for me to examine the area and ask some skeptical questions. Thanks to those who engaged with my questions earlier.

    Please take skeptics like me seriously; if some of us can forecast Global Temperature, don’t you think we might have something to say?

  • Malcolm // January 15, 2009 at 2:54 am

    Oops - should probably make it clear that the GISS measurement for 2008 is 0.44, so my forecast was very close indeed.

  • swade016 // January 15, 2009 at 2:55 am

    Wailing and gnashing of teeth!

  • b_sharp // January 15, 2009 at 3:10 am

    That January 2008 is .71 degrees cooler than January 2007 is definite proof that global warming turned into global cooling at that time, however since December of 2008 is only .04 degrees cooler than December 2007 it is equally obvious that global warming is back on because the temperature has risen .67 degrees.

    Whew, for a minute there I was afraid the interglacial was suddenly and abruptly over and about to buy fur stocks. Now I can relax and buy sun screen.

  • b_sharp // January 15, 2009 at 3:12 am

    Malcolm, even a stopped clock is right twice a day.

  • malcolm // January 15, 2009 at 3:21 am

    b_sharp: What were the other clocks forecasting?

  • lee // January 15, 2009 at 3:28 am

    Malcolm, you didn’t forecast global temperature. What you did is, you took a wild-ass guess based on existing trends that show some persistence for SIX FREAKING MONTHS!!! from July, and it happened that those usually-persistent trends (ie, La Nina weakening….) came to be close to what happened.

    The fact that you think that this demonstrates some credible expertise in climate ‘forecasting,’ actually means in reality that you have NO such credible expertise. The act of making the claim discredits the thing you are claiming.

  • b_sharp // January 15, 2009 at 3:52 am

    They were forecasting different values, but they too will be right in due time.

    The point is, that making a successful prediction once is more probably a coincidence than a correct analysis of data and does not give us any indication of how seriously to take your opinion.

    You may very well understand climate extremely well and be able to derive accurate predictions from an analysis of the data, but asking someone to accept that from one correct, possibly coincidental, prediction is expecting too much. Making a few more accurate predictions, enough to show that it isn’t just chance, will change that.

    That said, it certainly is reasonable for you to examine the area, learn as much as possible and ask questions.

  • jyyh // January 15, 2009 at 3:58 am

    Tamino cherry-picking the other way. any news of the apparent 2.8 year cycle (of some ocean indices?) you posted last autumn?

    off-topic, but:
    the sad thing in the arctic is the Greenlands location. as the sea ice melts more and more, the Greenland ice becomes the dominant cold pool in the arctic. this in turn will affect the near-by areas so they won’t warm up as much as some other areas. this means eastern areas of Canada and US, and the western coast of Europe. it is likely the citizens of these areas will not be easily persuaded of the science, as they see less warm effects of the climate change. the cold surges in the summer from Greenland will come more often as the arctic high will be located there in the summer in the future. i’m just hoping the Saharan air will move over the oceans and mediterranean and stay there, so southern and southeastern Europe won’t turn into an infertile steppe style Darfur.

    the Siberians are probably only happy about the warming there.

  • malcolm // January 15, 2009 at 5:06 am

    Lee - actually you don’t know what I did to make the forecast. But if predicting correctly counts for nothing, then you have a different definition of science to me.

    b_sharp. Ah, but unlike the stopped clocks I had a time specific forecast - global temperature anomaly for 2008. I also forecast that the Arctic sea-ice minimum would be less this year than last year. I placed a bet on that in June 2008. Here’s the link with the bet, and another link with my opponent graciously conceding.

    http://hot-topic.co.nz/like-takin-candy-from-a-baby/

    http://hot-topic.co.nz/im-possiblyprobably-a-loser/

    I haven’t made any other forecasts on this topic that I recall.

    For the record, I think the extreme melt this year does tend to support the AGW hypothesis, even though much of it seems to have been due to weather.

  • Simon D // January 15, 2009 at 5:26 am

    Nice post Tamino. After several years of steady improvement in the mainstream media coverage of climate change, there seems to be an uptick in stories and columns on Avery’s sort of data-starved climate skepticism over the past year. Perhaps it is in fact just noise, not a sign of a long-term trend.

  • Gareth // January 15, 2009 at 5:55 am

    Care for another bet, malcolm?

  • Willie // January 15, 2009 at 9:25 am

    Why do satellite data look so different from this data?

    [Response: They don't.]

  • mauri pelto // January 15, 2009 at 12:54 pm

    The NASA GISS page that I am sure you consulted does once again explain the various forcings, noting the importance of La Nina http://data.giss.nasa.gov/gistemp/2008/. And as you have noted before the last time we had a prolonged La Nina 1999 also was a cool year. So yes 2008 is cooler globally, but this was expected. Just wait for the next El Nino, which may occur later this year.

  • Ray Ladbury // January 15, 2009 at 1:46 pm

    Malcolm, Actually the fact that you might have come close to predicting temperature this year is less telling to me than the fact that you are impressed by this feat. It suggests that perhaps you do not have a very deep understanding of climate.
    The correlation in the time series says that you will do better than average by betting the near future will be similar to the near past. However, there are lots of little things that can come up and spoil that correlation. I am much more impressed, for example, with Hansen’s long-term prediction of warming. It is truly climate and it has more important long-term consequences.

  • sod // January 15, 2009 at 1:59 pm

    what to say? brilliant post. as always.

  • Sekerob // January 15, 2009 at 2:14 pm

    Willie, it’s different base lines. The further back you run the reference the higher the anomaly, so GISS will show 0.44 and UAH will show 0.18 for December on baseline 1979-1998 (yes latter, that hottest of hottest El Nino year so surreptitiously chosen, my opinion), where GISS uses 51-80 mean as their base. The profiles for the different compilers are pretty much identical but some things the satellites seem not to or over react to.

  • Barton Paul Levenson // January 15, 2009 at 2:16 pm

    I’ll bet Siberian loggers aren’t happy over global warming.

  • JCH // January 15, 2009 at 2:24 pm

    On La Nina, for those interested, OND (October-November-December) ONI is now charted: -.3.

    http://www.cpc.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml

  • Richard Steckis // January 15, 2009 at 2:37 pm

    Malcom,

    I have to agree with the others. An accurate guess is still a guess.

  • Richard Steckis // January 15, 2009 at 2:42 pm

    And my guess??

    That crackpot Easterbrook (god knows how many papers and about 5 books) is right and this is the beginning of the 30 year cooling!

  • Zeke Hausfather // January 15, 2009 at 2:54 pm

    Willie:

    They are nearly identical when you normalize them to the same base period:
    http://i81.photobucket.com/albums/j237/hausfath/Picture9.png

  • TCOisbanned? // January 15, 2009 at 3:11 pm

    I would be more impressed with “Hansen’s feat” if he had given more of an equation based prediction (for instance, this level of volcano activity, plus this level of CO2 gives this 2 decade temp trend.) So that I can actually plug in the actuals for all three and see how they compared to the temp results.

    I think some of the curves in his results are based on him dropping a volcano in there and also on El Nino predictions. But these are not really predictable, so those wiggles should not be used to give support OR against AGW theory. Tammy has taught us how silly it is to cite the last 10 years of temp plateau without looking at the longer term trend.

    Would like an equation or maybe some sort of nomograph or the like. I think this is how we all intuitively think about the problem…but the conversation becomes confused when we bake in the 3 different scenarios, etc.

  • TCOisbanned? // January 15, 2009 at 3:13 pm

    For that matter, can I just take Hansen’s old model and plug in the actual history (so even the timing of forcings would be inserted) and just see how the overall trend (or even the wiggles if you care) compared to modelled? Did he archive a copy of the code?

  • Hank Roberts // January 15, 2009 at 4:03 pm

    Wasn’t his responsibility to, as junior author.
    You’re referring to “Et Al. and Hansen” from the strawman attack dummy collection again.

  • Lazar // January 15, 2009 at 4:22 pm

    TCO,

    Here for source code, or binaries for the EdGCM front-end. There are differences with the 1988 model, though they appear to be minor, you may find the pristine 1988 version in the development section but that seems to require authorization, you could ask or check the forums.

  • ErikS // January 15, 2009 at 5:00 pm

    Bah! 6.7 ºC/century? Ridiculus.

    I´ve proved that the data from the UAH shows that the globe is currently heating at a rate of 8.5 ºC/decade!

    [Response: Omygod!!! that's 85 deg.C per century!!! I guess climate sensitivity is higher than we thought.]

  • Malcolm // January 15, 2009 at 6:25 pm

    Ray, Richard, thanks for your comments. I just make a few observations:

    - If I were simply forecasting based on persistence, my forecast would have been .34, not .43. So the claim I was merely assuming persistence isn’t supported by the facts.

    - I acknowledge there is a strong element of luck of involved, and I am not claiming to have an accurate climate forecasting system. But I don’t think it’s pure luck. I got the comparative arctic sea ice minimum right as well, as noted above.

    All I am really saying is that if skeptics like me can make such accurate forecasts, we shouldn’t be written off at “flat-earthers” as many tend to do, including on this blog.

    Similarly, having people suggest that being right simply proves me to be ignorant is not, I think, a good look for science. Particularly when they haven’t stopped to query me about my methods or my other forecasts.

    Also, the argument against predictions such as mine, commonly found in the financial markets, is that there is a distribution of predictions and a selection bias in reporting the accurate ones. I don’t think that argument applies in this case.

    There is a selection bias here - I wouldn’t have commented if I was out by a lot! But I can’t find any other predictions of Global Temperature Anomaly for 2008. There is no distribution of predictions, so the Wild-Ass-Guess argument runs into some trouble on that basis.

    If you wish to write my prediction off as a Wild-Ass-Guess, your argument would be much stronger if you could show either lots of similar predictions that were wrong, or lots of accurate predictions using climate models.

    Bear in mind my predictions were public, quantifiable, time-bounded, clearly falsifiable and made well in advance. If it’s so easy to make these predictions, where are all the other ones?

    Gareth - sure: Want to bet on the GISS TEMP anomaly for 2009?

  • TCOisbanned? // January 15, 2009 at 6:28 pm

    Lazar:

    I don’t want to actually do work. Just direct and recieve.

    ;)

  • t_p_hamilton // January 15, 2009 at 6:41 pm

    “Bah! 6.7 ºC/century? Ridiculus.

    I´ve proved that the data from the UAH shows that the globe is currently heating at a rate of 8.5 ºC/decade!”

    So, Willie was right about the satellite data not looking like the GISS data after all, it predicts even more warming!!!One!!

  • gmo // January 15, 2009 at 7:23 pm

    Malcolm,

    I query you about your methods then. Here is one I just thought up:

    X + TREND + SECRET (+/- 2*STD)

    where
    X = previous year value from the end of the lowess smoothed curve posted above
    TREND = about 0.02
    STD = about 0.09
    SECRET = my adjustments like for persistence from a cold or warm start to the year, ENSO events, volcanoes, etc

    Then I could throw away the +/- to come up with a single number. How about 0.61C?

    It is a fun little thing like picking sporting event winners, but it is not really interesting or “climate forecasting”. I would say that is why you cannot find many such predictions out there.

    The more relevant and meaningful questions involve e.g. whether the 2050s will be about 1.0C warmer than the 2000s. Compared to the GISS TEMP anomaly for a given year that is much less likely to be influenced by “weather” events and more likely to be much more directly relevant to the topic of climate change.

  • Lee // January 15, 2009 at 7:26 pm

    Malcolm:
    “All I am really saying is that if skeptics like me can make such accurate forecasts, we shouldn’t be written off at “flat-earthers” as many tend to do, including on this blog.

    Similarly, having people suggest that being right simply proves me to be ignorant is not, I think, a good look for science. Particularly when they haven’t stopped to query me about my methods or my other forecasts.”

    and Malcolm, earlier:
    “Please take skeptics like me seriously; if some of us can forecast Global Temperature, don’t you think we might have something to say?”

    Malcolm, you managed to come close on the 12-month anomaly for 1008, on a ‘prediction/forecast/wild-ass guess made when the fricking year was already substantially more than half over!!

    Being “right” doesn’t “prove you to be ignorant.” It doesn’t, in fact, prove anything at all. But claiming that “being right” on that number somehow validates your credibility on climate change issues - THAT proves you to be ignorant.

    I could be watching a basket ball game, and 5 minutes into the 3rd quarter make a prediction the wining margin of a basketball game. I could be very, very close to right. That would not in any way entitle me to claim that I have any expertise in basketball coaching and strategy. Your claim is equivalent.

  • Lee // January 15, 2009 at 7:30 pm

    Also Malcolm:

    “If you wish to write my prediction off as a Wild-Ass-Guess, your argument would be much stronger if you could show either lots of similar predictions that were wrong, or lots of accurate predictions using climate models.”

    Malcolm, climate models don’t predict temperatures for specific years. When you try to defend yourself by saying ‘wild-ass’ things like that, further illustrating your deep confusion about the difference between weather and climate, between trend and noise, you just dig your credibility hole even deeper.

  • Gareth // January 15, 2009 at 7:35 pm

    Gareth - sure: Want to bet on the GISS TEMP anomaly for 2009?

    Nope, but I’ll take the warm side of the sea ice bet again…

    If you want to test your forecast you could trade on iPredict’s “prediction market”, discussed here…

    Re ENSO. Worth reading Klaus Wolter’s Multivariate ENSO Index page for the current report. He suggests the best historical analogue for the current situation is the mid-70s, when there two successive strong La Niñas…

  • adder // January 15, 2009 at 8:03 pm

    malcolm // January 15, 2009 at 5:06 am:

    “Lee - actually you don’t know what I did to make the forecast. But if predicting correctly counts for nothing, then you have a different definition of science to me.”

    IOW I have a secret theory… and made a correct prediction using it! I’m a Scientist! Give me the Nobel Price!!!!

  • Malcolm // January 15, 2009 at 8:53 pm

    Lee - I don’t care to respond to your ad hominen attacks.

    GMO - similarly, I see no percentage in exposing my methods to more attacks like Lee’s. If you post an email address, I’ll send you a description though.

    Adder - That’s a straw man. All I’m saying is don’t treat me like a flat-earther.

    So much vitriol simply for being right. I am a moderate, and I try to follow the arguments and evidence. AGW proponents should be able to persuade a guy like me. But the abuse/argument ratio is just too high.

    Gareth - you’re on. It’s a bet. Same terms?

  • Gareth // January 15, 2009 at 9:43 pm

    Same terms are fine. Could you confirm in a comment at HT, as per last time? There’s nothing specifically Arctic on the front page at the moment, but there will be soon… ;-)

  • Steve Bloom // January 15, 2009 at 10:06 pm

    Malcolm, eyeballing Tamino’s monthly plot above, a guess in June based on the idea that with a fading La Nina temps would recover to something like the recent mean would have yielded a result similar to yours. It’s a perfectly reasonable approach and probably has the best chance of being right, but it will also be substantially wrong far more frequently than it’s right. While it’s an entertaining exercise, the scientific value is zero and the polemical value not much more than that.

    If you want to establish credibility based on climate predictions, you’ll need to do it with something the climate scientists say is amenable to prediction. Guesses about 6-month global temp anomalies and next year’s Arctic sea ice level are not among your options, I’m afraid.

  • Steve Bloom // January 15, 2009 at 10:16 pm

    Tamino, I suggest as topics some of the most recent idiocy over at the cold weather and astrology blog, in particular the volcano and El Nino postsof the last few days. Both appear to rely heavily on shaky calculations.

  • gmo // January 15, 2009 at 10:39 pm

    Malcolm seems to take two broad seasonal projections that were successful to a particular degree as demonstrative of some fundamental understanding of the climate system. Those projections appear to have been from relatively simple statistical models. There seems to be some info out there about these projections but I am not digging for it right now, and Malcolm is not being forthcoming with his methodology. The predictions have minimal relevance to the issue of climate change, though the apparent weight given them lookss similar to the weight some people give to the drop in temperatures to the early 2008 local minimum.

    Malcolm makes a point to tell how he is scientific and thoughtful but skeptical and being abused. It is not clear what he even wants to discuss. How great he is for his accurate predictions? Perhaps if he either did not seem to think that his predictions were so meaningful as relates to AGW, tried to explain why he thinks his predictions are so meaningful as relates to AGW when repeatedly told they are, or even just gave some sign he was open to arguments and evidence he supposedly tries to follow then the responses to him would not be so… pointed.

  • malcolm // January 16, 2009 at 2:06 am

    gmo: All I’m saying is please don’t write me off as a flat-earther (or non-climate scientist) because I ask some sceptical questions.

    On being amenable to evidence - please note that in my comment above I acknowledge the extreme arctics melt as tending to support AGW. I also acknowledged comments on a previous thread as helpful in allaying some concerns I had about short term climate forecasting. So I am amenable to evidence.

    To answer your question on what I want to discuss, there are several things I would appreciate informed comment on, and if possible references.

    - Could the carbon forcing effect be overestimated? I seem to recall Dr Spencer has a paper to this effect, and I know from my own area that complex models almost always over-fit non-experimental data, making them prone to biased parameter estimates.

    - Will the recent drop in solar irradiance shown at overwhelm carbon forcing effects?

    - Will there be a substantial cooling effect coming from the Pacific Decadal oscillation?

    - Much is made of the noise in the climate system, but there seems to be three sorts of “noise”. Measurement error (fairly minimal). Chaotic effects over 1-5 years (probably substantial). And long term trend effects such as PDO and Irradiance (also probably substantial). Has any work been done to decompose the last two sources of noise to improve the long-term climate forecasts. In other words, some scientists are saying we are in for 10 to 20 years of cooling, notwithstanding an underlying carbon-warming trend. Are they right?

    Thanks for prompting me to ask these questions; you make a fair point.

  • malcolm // January 16, 2009 at 2:07 am

    Sorry, the site ate my URL. I was referencing the solar irradiance shown on the GISS data page.

  • Philippe Chantreau // January 16, 2009 at 4:50 am

    Malcolm 01/15 @08.53.

    Actually Adder does have a point in the following sense: being right counts for nothing unless it can be determined with reasonable certainty that you are right because of your work. To see that you were truly right, you need to show your work so that all can see whether you were right because your method was right or for other reasons, or any combination thereof. Even if your method was right, its domain of application can still be very limited. Noone can tell without knowing what it is. You rightfully acknowledged that gmo made a fair point (several of them , in fact).

    As for the vitriol, get over it. This is a blog, and rather better than most. You don’t even use your full name, no reason to take things so personally.

  • michel // January 16, 2009 at 7:43 am

    being right counts for nothing unless it can be determined with reasonable certainty that you are right because of your work. To see that you were truly right, you need to show your work so that all can see whether you were right because your method was right or for other reasons, or any combination thereof. Even if your method was right, its domain of application can still be very limited. Noone can tell without knowing what it is.

    Oh how true! Yes, this is as clear and succinct a statement as you’ll want of why making all one’s underlying data and algorithms available is so important. No exceptions.

  • malcolm // January 16, 2009 at 9:11 am

    Philippe, what about being wrong. Does that count for anything?

  • deech56 // January 16, 2009 at 10:24 am

    Malcolm for your predictions, have you gone back to other years to see if they hold true? Models are often tested by hindcasting. In referring to Spencer’s paper, do you mean his 2007 paper in GRL? There are a few posts about this paper in the “Spenser’s Folly” thread, starting on July 29, 2008 at 5:16 am.

  • Richard Steckis // January 16, 2009 at 1:08 pm

    Malcolm:

    :Much is made of the noise in the climate system, but there seems to be three sorts of “noise”:

    Much is made of the term noise in this blog. Tamino is fond of outlining noise and even random noise.

    I for one think that calling the noise (aka weather) random is a crock. Weather is caused by forcing agents and those agents are not random. If they were then one would expect snow in Phoenix in mid summer at least some of the time, and 45C days in Lhasa at least some of the time. It just does not happen that way. Randomness in the climate system is a human construct and not an actual occurrence.

    When a climatologist mentions random noise in the climate system they really mean weather that THEY can’t forecast or predict.

  • Nick Zervos // January 16, 2009 at 1:56 pm

    This post by Tamino deflects attention from the relation of the recent trend in global temperature with the recent trend in solar activity. New visitors to this site should also visit http://www.woodfortrees.org to see the data for themselves (I assume the veterans here are already familiar with that site). Click on “Interactive” to view any of the several sources of temperature data, including GIS land-ocean global mean, UAH tropospheric global mean, HADCRUT, and others. That site also provides data for total solar irradiance (TSI). TSI is one of many parameters that provide an indication solar activity. Use the site’s interactive data plotting feature to see the trend since the TSI satellite measurements began in about 1978. TSI has been dropping sharply in the last couple of years.

    I suppose AGW alarmists are saying that the drop in global temperature that is accompanying the current drop in solar activity is just a coincidence. They’ve also said that the 20th century rise in global temperatures that accompanied the 20th century pronounced rise in solar activity was also a coincidence. A very graphic depiction of the 20th century trend of solar activity was presented by Tamino in his July 24, 2007 post (attributed to the reconstruction that Lean published in 2000). The solar activity in the 20th century rose to values higher than it’s been in over 8000 years. The hockey stick TSI trend very much resembles the hockey stick trend in global temperature.

    Now that the solar activity is dropping to levels comparable to mid 20th century, it’s important to see what’s emerging in the temperature response. Of course two years is still too short a time on which to base any statistically definitive conclusions. But we should be vigilant of this emerging trend and not try to mask it in long time-scale plots and averages.

    [Response: Very revealing that you say "still too short a time" and "this emerging trend" in consecutive sentences.

    That seems to be a nearly universal behavior pattern among advocates of "it's the sun!" It's also the kind of uncritical, self-indulgent argument I'm battling against.]

  • Richard Steckis // January 16, 2009 at 2:12 pm

    Weather is chaotic, not random. It is a fully deterministic system in which we do not know all the initialization conditions. Therefore it gives the appearance of randomness.

    Because we don’t know the initialization conditions of weather, climate modellers tend to disregard weather as random. That is one of the fundamental flaws of climate models (in my opinion).

  • sod // January 16, 2009 at 2:31 pm

    Weather is chaotic, not random. It is a fully deterministic system in which we do not know all the initialization conditions. Therefore it gives the appearance of randomness.

    Because we don’t know the initialization conditions of weather, climate modellers tend to disregard weather as random. That is one of the fundamental flaws of climate models (in my opinion).

    the fundamental problem is your lack of knowledge.

    dice aren t random either!
    it is a fully deterministic system, we do not know all the initialization conditions. Therefore it gives the appearance of randomness.

    [Response: I was going to remark that the "not random" argument is a truly desperate attempt to deny the difference between trend and noise -- but you beat me to the punch.]

  • Hank Roberts // January 16, 2009 at 3:27 pm

    For those who’ve paid the AGU twenty dollars:

    Physics Today / Volume 62 / Issue 1 / OPINION
    Physics Today - January 2009
    Solar variability does not explain late-20th-century warming

    Philip B. Duffy, Palo Alto, California
    Benjamin D. Santer, Lawrence Livermore
    Tom M. L. Wigley, NCAR

  • Barton Paul Levenson // January 16, 2009 at 3:32 pm

    Nick Zervos,

    So you think it’s the sun, do you?

    I have annual data for temperature anomalies and TSI for 1880-2007 (Svalgaard’s TSI, not very variable), and 1880-2000 (Lean’s TSI, very variable indeed). You know what? When I regress anomaly on ln CO2 and either TSI reconstruction, CO2 accounts for 76% of the variance and TSI is statistically insignificant.

    Maybe I’m using the wrong analysis? Let me know if you want the raw data. This is your chance to prove all the scientists wrong!

    Want the numbers?

  • Bob North // January 16, 2009 at 3:52 pm

    Tamino, Sod, Richard - I think the last couple of post illustrate a typical problem with communication, particularly with respect to climate change. Richard appears to be using the term “random” with the definition of “without a specific cause, or purely by chance”, which is a perfectly legitmate definition and use of the word. However, Tamino and Sod appear to be using the term random with the statistical definition along the lines of “relating to a type of circumstance or event that is described by a probability distribution.” Again, a perfectly legitmate definition and use of the term (and perhaps somewhat more appropriate for this site).

    In a sense, all of you are right. Although we have not yet identified all of the minor factors that contribute to the interannual variability (aka “noise”) in weather, it is certainly not acausal. Just as certainly, we can describe the weather by statistics, even if we cannot deterministically predict what the weather will be next year.

    Feel free to correct me if I have misinterpreted anyone’s use of the word random, but it just seems to me that some of the “heated discussions” that are had here would be toned down if everyone realised that many words have multiple definitions or uses and may mean something different to the layperson than the scientist, or even may often be used differently between scientific disciplines.

    [Response: I think Richard Steckis knows perfectly well what I'm referring to when I say "noise" and "random." But he pushes the idea that "calling the noise (aka weather) random is a crock" because he dislikes the results, not because he's trying to contribute a useful insight from considerations of causality and chaos. He's spreading his own kind of chaos, and it's a crock.]

  • Hank Roberts // January 16, 2009 at 5:18 pm

    Richard S. claims to be speaking as a scientist. It seems improbable that he’d pick the wrong definition of the word in a discussion about statistics. I can’t say exactly how improbable.

  • Nick Zervos // January 16, 2009 at 5:24 pm

    Barton Paul Levenson. the data you’re referring to may show a strong correlation between temperature and CO2. That doesn’t surprise me at all. But does it allow cause and effect to be differentiated? That is, does the CO2 trend clearly lead the temperature trend? If that’s the case I would very much like to see the data. All the temperature-CO2 corellations I’ve seen so far indicate that CO2 lags temperature.

    With regard to the TSI-temperature trends, is your analysis guilty of the sins Tamino ascribes to AGW skeptics: are the time intervals too small? There certainly is no denying the obvious long-term correllation throughout most of the 20th century of the TSI and temperature hockey stick curves.

  • Lowlander // January 16, 2009 at 5:49 pm

    He [Dick S.] is probably just being obtuse. Randomness is well explained in the wikipedia article where it explains as well the different definitions. He probably stumbled across the article and decided to regurgitate the most convenient parts here.

  • Boris // January 16, 2009 at 6:00 pm

    malcolm says:

    I’m not aware of any other forecasts for 2008.

    http://www.metoffice.gov.uk/corporate/pressoffice/2008/pr20080103.html

    So the MET office predicted a 2008 anomaly of .37 and the actual anomaly was .31.

  • luminous beauty // January 16, 2009 at 6:27 pm

    Nick,

    CO2 concentration began rising in about 1750. Temperature signal began to emerge from noise about 1975.

    That would suggest CO2 proceeding temperature in the modern period, don’t you think?

  • luminous beauty // January 16, 2009 at 6:30 pm

    preceding, I mean.

  • gmo // January 16, 2009 at 6:41 pm

    Malcolm,

    Those are reasonable questions, though they give off an air of formulation from seeing a lot of faulty information on “skeptic” web sites. Many good answers to them can be found even just on this site. The solar irradiance drop may be a good place to start - it is a popular hand-wave these days that the sun is going into some sort of “minimum” and that change will mean large cooling relative to any warming CO2 may cause. That simple argument breaks down with a little scrutiny. See what this supposed drop actually is in terms of W/m^2, consider the fraction of that change that is observed at the earth’s surface because of geometry and aldebo, and compare that to the W/m^2 radiative forcing from CO2 and other anthropogenic forcings (even without considering feedbacks).

    I think maybe the coincidence of the 2007-to-2008 temperature drop like described in the post above and the current solar cycle minimum has led some without any consideration of numbers to vastly overestimate either the magnitude of change recorded in solar forcing or the climate response to that particular forcing.

    As you have seen, if you can be interpreted as just another somebody who says he skeptical, cites some personal supposed experience/expertise that may or may not have relevance to climate, and appears to act as though their line of thinking is right and the consensus opinions of the many experts who have worked so long in the field is wrong, then places like this are not going to be very welcoming. Such is the blogosphere.

    It can be a tough line to tread, but if you want to be seen as truly skeptical you have to not appear dismissive or ignorant of the established views. I think you gave the impression not of being curious but of feeling your answers for anything would be at least as good as those of any experts. That looks like being a “flat-earther” or someone who thinks the Singer cooling stance above is on par with the consensus of continuing long-term warming, and that will get you dismissed here and other places.

  • dhogaza // January 16, 2009 at 7:07 pm

    But does it allow cause and effect to be differentiated? That is, does the CO2 trend clearly lead the temperature trend? If that’s the case I would very much like to see the data. All the temperature-CO2 corellations I’ve seen so far indicate that CO2 lags temperature.

    So you’re suggesting that natural warming was the cause of the industrial revolution, leading to the ever-increasing burning of fossil fuels?

    Interesting “theory” …

  • Hank Roberts // January 16, 2009 at 7:20 pm

    Nick, here ya go:

    http://www.newscientist.com/article/dn11659

    PS, if you wanted to find this, here’s how: paste the words you used into the Google search box.

  • t_p_hamilton // January 16, 2009 at 7:23 pm

    Nick Zervos:”Barton Paul Levenson. the data you’re referring to may show a strong correlation between temperature and CO2. That doesn’t surprise me at all. But does it allow cause and effect to be differentiated?”

    Physcis does.

    ” That is, does the CO2 trend clearly lead the temperature trend? If that’s the case I would very much like to see the data.”

    Compare data from Mauna Loa, to the temperature. CO2 has steadily increased. Temperature has increased over the same period steadily. Other factors known to affect global temperature have remained fairly constant over the past 50 years.

    ” All the temperature-CO2 corellations I’ve seen so far indicate that CO2 lags temperature.”

    Congratulations! You have put your finger on what is different in the modern era. *WE* have put the CO2 into the atmosphere this time, which causes warming.

    In prehistoric times other events caused warming, which caused release of CO2, which caused more warming.

  • Bob North // January 16, 2009 at 7:57 pm

    Tamino/Hank - I do not know Richard Steckis or his motivation for using a fairly common definition of the word random, so I won’t comment on his motivation or the probability that he intentionally used the “wrong” definition in a discussion of weather and climate (not statistics, Hank). Lowlander, I doubt Richard needed to go to Wiki to come up with the word random or one of its common usages. Though I don’t have anything but my own experience, I would be willing to wager that if you went out and surveyed 100 people on what random means, even scientists, Richard’s usage would be the more common answer.

    All of you seemed to miss the point though, which was more about effective communication and realizing that many words have multiple meanings and you can’t just assume that others will necessarily use the word or read your usage of it in the same way. Here are a couple of other examples I have seen here:

    Trend - Usage here seems to suggest that to use the word “trend” one must demonstrate statistical significance. A few days ago, one poster bluntly stated “It is not a trend since it is not statistically significant” or something to that effect. However, there is nothing in the definition or common usage of the word “trend” that demands statistical significance.

    “Warming” - This is another one I brought up on RC a little bit ago. Many, many laypeople are going to interpret this as an active verb rather than a noun referring to the process of AGW. So rather than calling them stupid when they say “it hasn’t warmed since whenever”, first explain how you were using the word (e.g., “Yes, you are absolutely right that there has been little temperature change since (whenver), but here is what is meant by the term” Of course, if they persist, then you can start on the namecalling.

    The point of my post is not about arguing over AGW, but about effective communication.

  • Luis Dias // January 16, 2009 at 9:13 pm

    Just to say that just by looking at the last graph, you could have easily have made a better trend by taking the first point as being in that down point somewhere in 2004. I bet you could have achieved more than 10ºC / Century.

  • Hank Roberts // January 16, 2009 at 9:16 pm

    > not statistics
    What do you come here for, if not statistics?

  • David B. Benson // January 16, 2009 at 9:44 pm

    Nick Zervos // January 16, 2009 at 5:24 pm — During the Little Ice Age, I think you’ll find that CO2 drop led the temperatures. Of course, in the modern era, 1850 CE on, CO2 leads tmperatures. In both situations, the lead time is quite short.

    Regarding early human-caused climate changes, please do read W.F. Ruddiman’s popular “Plows, Plagues and Petroleum”.

  • gmo // January 16, 2009 at 10:00 pm

    Bob North,

    Clarity should be a friend. I agree that communication could be better in many instances. How many people understand that reference to a trend that is not statistically significant as not a trend at all is because the lack of statistical significance means there is not sufficient evidence to the so-stater that there is a non-zero trend? How many people can easily make sense of that previous sentence?

    I think it should always be a goal to make points that will be received by a lay audience readily accessible to most of such an audience. However I say the greater problem is the cases like from Avery above (I think I mistakenly said “Singer” in a previous post) where very simple and clear language is being used but with content that is at very best extremely misleading.

    A lot of the “heated discussions” involving climate change I think greatly involve people trying to take advantage of the uncertainty and vagueness in word usage and definitions to raise doubt that AGW is occurring/bad/continuing/etc. Statements like “the earth’s temperatures have dropped an average of .6 Celsius in the last two years” are much more of a problem than saying a non-stat-sig trend is not a trend.

  • Nick Zervos // January 16, 2009 at 11:56 pm

    Hank Roberts, this is from the article you linked:

    “To repeat, the evidence that CO2 is a greenhouse gas depends mainly on physics, not on the correlation with past temperature, which tells us nothing about cause and effect. ”

    Not exactly a smoking gun. I’d like to see a plot of temperature and CO2 concentration where the latter reasonably consistently leads the former.

    luminous beauty, have you got a reference showing CO2 concentration starting to rise in 1750? Data I’ve seen show the CO2 not starting to rise until well after the temperatures did at end of the Little Ice Age.

    [Response: Here. Some data are here, there are other data sets if you want to track them down, they tell the same story.]

  • Arthur Smith // January 17, 2009 at 1:14 am

    Since Malcolm is so pleased with his July ‘08 prediction turning out pretty close, I guess I ought to inform him about my February 15 2008 prediction for GISS 2008 full-year number: 0.41. See comment # 193 here for proof:

    http://www.climateaudit.org/?p=2698

    (and guesses at follow-on numbers for subsequent years - yeah, that was a little random…)

    Other estimates for 2008 average GISS anomaly from the climateaudit crowd were: (see comments #88, 101, 105, 133):

    Lucia (Lumpy): 0.70
    Geoff: 0.38
    Andrew: 0.34
    Larry: 0.31
    Jae: 0.30
    Earle: .26
    Mosher: 0.23
    PaulM: 0.21
    Sam U: 0.18
    Greg F: 0.15

    Aside from Lucia (who posted her guess before January’s low GISS numbers came out), every one of the regulars there was below my guess, which was by a factor of 2 the closest of them all.

    Let the mocking resume :-)

    By the way, a part of my guess and I’m sure some of the others includes an estimation that the transient response to solar cycle changes is a bit on the high side - but that also means the transient response to all forcings will be high, which makes things *worse* (hotter) for future temperature projections, as GHG’s continue to go up!

  • Philippe Chantreau // January 17, 2009 at 4:16 am

    Michel says:”Yes, this is as clear and succinct a statement as you’ll want of why making all one’s underlying data and algorithms available is so important. ”

    Well, that was done, and GISS went as far as making the codes available. Turned out to be kinda like giving a telescope to a 5 years old.

    With denialists it’s always damn if you do damn if you don’t. You don’t disclose and they accuse you of fraud. You disclose and they’ll torture the data and methods long enough to find any kind of flaw whose significance can be inflated beyond reason and sold to the hoi polloi as proof that scientists are incompetent. Whatever.

  • Richard Steckis // January 17, 2009 at 4:25 am

    Tamino:

    “I was going to remark that the “not random” argument is a truly desperate attempt to deny the difference between trend and noise — but you beat me to the punch.:”

    I was not attempting to deny the difference between trend and noise. Certainly not in the mathematical sense. In the real world the noise is weather and it is not random. It is deterministic. Otherwise weather forecasting would not be possible.

    Lubos Motl has a good post on this issue. Like him or hate him you should read it:

    http://motls.blogspot.com/2008/01/weather-climate-and-noise.html

    Again you cheapen debate by resorting to the word “deny” to link it to denialism. Bob North hit the nail on the head. You look at it in the mathematical sense and I look at it from the perspective of “without cause”. Perhaps we are at cross purposes here.

    [Response: I suggest readers avoid Motl's misleading essay. He makes some patently false claims like "When they talk about noise, they clearly imagine some kind of non-autocorrelated noise ..." I happen to know first-hand that the RealClimate people don't do this, and I certainly don't. He also makes a lot of straw man arguments like "The actual "long-term" climate is not composed of one effect, as they seem to incorrectly assume..." No, they don't; they've been champions of educating people that there are multiple causes at work in climate change, but denialists love to make the false claim that genuine climate scientists ignore important factors other than greenhouse gases (like the ludicrous claim that they ignore the influence of the sun).

    As for denialism ... if the shoe fits, wear it.]

  • Richard Steckis // January 17, 2009 at 4:39 am

    and Hank.

    This blog is not just about statistics. It’s central theme is climate change. The blog looks at climate change from the perspective of statistics because that is Tamino’s area of expertise. If some of us non-statisticians stumble on areas of statistical definition, that is to be expected.

  • Richard Steckis // January 17, 2009 at 4:50 am

    Phillipe Chantreau:

    “Well, that was done, and GISS went as far as making the codes available. Turned out to be kinda like giving a telescope to a 5 years old.”

    When? How does one get the code?

  • malcolm // January 17, 2009 at 5:24 am

    gmo - thanks for that. Food for thought. On a related point, I’m kind of intrigued by some recent minor jags in the Maua Loa CO2 record - it could be explained by temperature changes leading CO2 changes (although I’m pretty dubious about that argument, as I don’t think it is consistent with recent observations). Could it also be explained by the recent reduced solar irradience? I’d be interested in any expert views on this. Intriguingly, of course, it could also be explained by reduced economic activity and thus reduced CO2 emissions, which would be consistent with AGW. I’d love to see a study on global GDP growth and global atmospheric CO2 growth.

    [Response: Perhaps it's wise to see what the experts have to say *before* speculating too much on this. The theories you suggest (temperature change leading CO2 change, solar activity) are regular entries from a number of those who wallow in ignorance (Anthony Watts). I'm pretty ignorant of the subject too (I try not to wallow), but my impression is that it (the "minor jags") has more to do with the biosphere than the thermometer. You might find this interesting.]

    Boris. Arthur Smith. Excellent! Pleased to see the other forecasts. I think the correct response is initially: “Nyah, Nyah, mine was better” :-), but I have to acknowledge the longer forecasting horizon from the start of 2008 is much more challenging. Mock away! I’ll try to get with the program. :-)

    But it does leave me a little confused on the point of whether the short term anomaly can be forecasted. Some here are saying NO!, that’s not part of climate science. Yet the Met is doing it. It seems as if there are still divergent views on short term climate forecasting, even among climate scientists.

    BTW, I notice quite a lot of interest in the statistical significance of trends. I’d just comment that statistical significance has nothing to say about persistence. The significance of a trend line is irrelevant to forecasting. Just ask the financial markets!

    [Response: Perhaps I don't emphasize often enough, or strongly enough, that the forecasts of long-term temperature increase are in no way based on statistical forecasting. They're based on application of the laws of physics, which is pretty solid stuff.

    My concern with statistical significance is mainly to refute such ludicrous claims as "global warming stopped in 1998" and "sea ice is recovering." There are a lot of false claims out there which utterly fall apart when looked at under the microscope.]

  • Hank Roberts // January 17, 2009 at 5:42 am

    Look at the folks still trying to prove CFCs aren’t a problem. Same thing. Look at the folks trying to prove evolution can’t happen beyond microbes. Same thing. Always they pick some early work, something that got people looking at a new area, and try to make hay out of the notion that because the first work wasn’t perfect, nothing since is good.

    All they make is straw.

    They may win. Science isn’t guaranteed to survive — look at the past 30,000 years of human cultures. If there were people thinking scientifically — and there must have been, it’s not like people changed in the past few centuries — there weren’t enough of them, they weren’t in touch with one another, and they got shouted down by those who want things not to change.

    And as a result the world changed rather drastically — without precautions, without thought in advance.

  • Chris O'Neill // January 17, 2009 at 10:41 am

    Malcolm:

    If I were simply forecasting based on persistence, my forecast would have been .34, not .43. So the claim I was merely assuming persistence isn’t supported by the facts.

    Depends on the type of persistence you choose. Instead of choosing just one month we could choose something far less affected by noise, e.g. the 10 year average anomaly for 1998-2007. In this case the average anomaly was 0.504 giving a 2008 forecast of 0.447. My trivial forecast would have been at least as accurate as yours so I am not impressed by your forecast.

  • Frank O'Dwyer // January 17, 2009 at 11:01 am

    I´ve proved that the data from the UAH shows that the globe is currently heating at a rate of 8.5 ºC/decade!

    This is nothing. Here in Woodley, Berkshire the temperature went from -5C to +7C in the space of about 24 hours. It seems that the IPCC has drastically underestimated the warming trend, as this is a rate of about 438,000 C per century.

    By my calculations Berkshire will be hot enough to melt tungsten before 2009 is over. I am considering moving, although then the postman will not know where to deliver my Nobel prize.

    A corollary of my results is that Winter stopped in 2009.

  • Nick Zervos // January 17, 2009 at 5:04 pm

    Tamino, the links you provided do in fact show the CO2 concentration starting to rise in the mid 1700s. Where I was wrong was my recollection of when temperatures started to climb out of the Little Ice Age. As this link shows, that was around 1700.

    http://en.wikipedia.org/wiki/File:2000_Year_Temperature_Comparison.png

    Again, the data show CO2 lagging temperature. And you’re right, any other data that I can track down tell the same story.

    So on what basis does your first linked article claim that “the industrial revolution has caused a dramatic rise in CO2?” The CO2 concentration is a response to, not a cause of, the rise in temperature.

    Considering that the most important tenet of AGW advocates is that anthropogenic CO2 is causing temperature change, I’m dismayed at the lack of scientific rigor in their standards of empirical evidence.

    [Response: You're more dismaying than dismayed. That anthropogenic emissions are the cause of CO2 increase is beyond dispute. It's a fact. If you don't believe it, either you're ignorant of the evidence or you've crossed way beyond the stupid threshold.

    In fact, the origin of CO2 increase is a genuine litmus test. Those who proclaim that it's not due to human activity aren't just mistaken, they're in denial, they're not skeptics they're denialists.

    So here's your test: go read this post. Then come back and tell us whether you still doubt that CO2 increase is due to human activity.]

  • Barton Paul Levenson // January 17, 2009 at 6:13 pm

    Nick Zervos writes:

    All the temperature-CO2 corellations I’ve seen so far indicate that CO2 lags temperature.

    “Correlations.” In a natural deglaciation, CO2 lags temperature because the solubility of CO2 decreases as temperatures rise and CO2 diffuses out of the ocean. That is not what is happening now. The present increase in CO2 is coming primarily from burning fossil fuels, as we can tell from the radioisotope signature, and it is not lagging temperature.

    But I’ll run a Sergeant’s partial-F test for Granger causality on the data and let you know what I come up with.

    With regard to the TSI-temperature trends, is your analysis guilty of the sins Tamino ascribes to AGW skeptics: are the time intervals too small?

    Is 128 years too small to find a climate trend? No, not even remotely. 30 years is usually considered enough.

    There certainly is no denying the obvious long-term correllation throughout most of the 20th century of the TSI and temperature hockey stick curves.

    What part of “I analyzed it and there was no statistically significant relationship” did you not understand?

  • malcolm // January 17, 2009 at 6:48 pm

    Frank - but I don’t think yours was a forecast, was it?

    Tamino - thanks for the CO2 links.

    Nick - while there may well be CO2 lags for temperature changes in the past, that doesn’t mean that is the only mechanism by which atmospheric CO2 changes. Also, I don’t think the CO2 lagging temperature theory can explain the current changes, not just because of the link Tamino gives, but because it is just too early to see a change, according to the lag theory.

  • t_p_hamilton // January 17, 2009 at 8:01 pm

    Since denial of CO2 from industrial activity has been mentioned, some scientists noticed not only is there an atmospheric record from ice, people have been putting samples in sealed bottles for centuries, many of which are still around. We call them wine bottles.

  • David B. Benson // January 17, 2009 at 8:31 pm

    malcolm // January 17, 2009 at 6:48 pm — Please, now, take the time to read W.F. Ruddiman’s popular “Plows, Plagues and Petroleum”

    Come back after that, if you wish, but so far you are simply embarassing yourself.

    IMO.

  • Richard Steckis // January 18, 2009 at 4:16 am

    Tamino:

    “Response: I suggest readers avoid Motl’s misleading essay.”

    My suggestion to readers is that if they wish to remain ignorant then they should avoid Motl’s essay. To be truly informed, one must analyse all arguments in a debate. I am sure the readers here are intelligent enought to make up their own minds.

    [Response: I suggest that reading, and recommending, Motl's essay is one of the reasons for your ignorance.]

  • Richard Steckis // January 18, 2009 at 4:24 am

    Tamino:

    “So here’s your test: go read this post. Then come back and tell us whether you still doubt that CO2 increase is due to human activity.”

    I have started looking at the delta13C data described in the Real Climate post. The relationship between co2 and d13C whilst significant is not convincing as the R^2 is 0.75 meaning there is a large proportion of the change in d13C not accounted for by co2 change. Also the slopes of the two variables over time are significantly different.

    I am not saying that the majority of co2 increase is not anthropogenically caused, just the don’t rely on the d13C data as absolute confirmation of it.

    I regard an R^2 of 0.75 as being too low to be a reliable predictive model.

    [Response: The change of isotopic composition of extra atmospheric CO2 is diluted by exchange with the biosphere and oceans; so of course it doesn't show the same slope as CO2 increase.

    Those who have actually done the math find that the numbers add up and the source of the CO2 increase is depleted in *both* 13C and 14C and is in fact effectively devoid of 14C. That's fossil fuels.

    The *simple* fact that d13C (and d14C) were stable for 10,000 years *until* the onset of massive fossil fuel burning, but start changing at the precise time we start pumping all that CO2 into the air and atmospheric concentration starts its increase -- you have to be massively in denial not to realize that's a dead-sure lock.

    But unless every graph you look at paints a childishly simple picture with 100% correlation, I guess it's beyond your comprehension. You failed the test.]

  • Richard Steckis // January 18, 2009 at 12:55 pm

    Well. If the slopes are different, for what ever reason then d13C cannot be reliable for estimating the level of anthropogenic co2 in the atmosphere.

    Now, obviously you are too quick to the pen and did not read what I said. To wit:

    “I am not saying that the majority of co2 increase is not anthropogenically caused, just the don’t rely on the d13C data as absolute confirmation of it.”

    I stand by that statement. There are numerous external forces to a climate proxy that can influence stable isotopic composition. The Real Climate post talked of using d13C data from tree rings. That would be the one of the last proxies I would use for generating a stable isotope signature reflecting atmospheric co2, for the reasons of changes in the metabolic rate of the plants with time due to stress, disease etc that can affect isotopic composition. Ice cores would be the second last proxy I would use.

    I would be more comfortable with speleoclimatic proxies. I am sure most of the current co2 increase is due to anthropogenic causes but d13C data is not the smoking gun.

  • Richard Steckis // January 18, 2009 at 1:00 pm

    Of course you believe that Motl’s essay is the reason for my ignorance. After all he does not agree with your polemic and therefore is bad. Much like Easterbrook the crackpot scientist (according to Tamino) who has about 150 peer reviewed papers and more than five books to his name.

    Ignore these people and then you will remain ignorant. Just as I will remain ignorant if I ignore your contributions to this debate.

  • Ray Ladbury // January 18, 2009 at 1:01 pm

    The thing I’ve noticed about Motl is that he relies a lot on straw men. I wonder if this is becuase he is being disingenuous or if perhaps his approach to physics is so intuitive that it fails utterly when he is faced with an unfamiliar problem that in some sense defies intuition. The problem with guys like Motl is that they are smart, and while their arguments are based on a misunderstanding of the physics, they tend to be above the understanding of the layman. So, while it’s hard for me to believe that he is sufficiently deluded to make such elementary errors, I can’t rule out that he really does think that compared to string theory, everything should be simple. This would make him an honest fool–albeit a smart one–rather than a fraud.

    Richard, My question to you is why you would waste your time on transparent twaddle like that from Motl. Why not learn the science from the best sources possible? As to the tempest over the word random, I’d point out that randomness is actually a slippery concept–to date there is still no mathematically satisfactory definition of the concept. It’s one of the things that tripped up Kolmogorov in his attempts to put probability onto a firm foundation in mathematics.

  • Richard Steckis // January 18, 2009 at 1:18 pm

    Of course you are on much stronger ground with the 14C data.

  • Richard Steckis // January 18, 2009 at 1:53 pm

    I have just found out that the concentration of 13C when decoupled from 12C is increasing.

    http://wattsupwiththat.com/2008/01/28/spencer-pt2-more-co2-peculiarities-the-c13c12-isotope-ratio/

    I know. It is the evil Watts site and the author of the post is the devil incarnate (Roy Spencer) but the data look sound.

    Trying to find Refs.

  • Barton Paul Levenson // January 18, 2009 at 2:08 pm

    Okay, I promised Sargent’s (not “Seargents,” as I had it) partial-F test for Granger causality.

    I took NASA GISS temperature anomalies (”Anom”) and the natural log of Mauna Loa/Law Dome CO2 readings (”ln CO2″) for 1880-2007 and ran information criteria searches for 1-10 years lag for the period 1890-2007. Appropriate lags were 1 year for Anom and 8 years for ln CO2.

    The partial-F test stat for ln CO2 => Anom was F = 45.42 (p = 0.00001526). The partial-F test state for Anom => CO2 was F = 2.49 (p = 0.02). So CO2 significantly affects temp and temp significantly affects CO2.

    But the effect of CO2 on temp is 1,300 times more significant than the other way around. So I think I can say with pretty high confidence that CO2 is leading temperature this time, not the other way around. Q.E.D.

  • Barton Paul Levenson // January 18, 2009 at 2:09 pm

    P.S. I didn’t account for stationarity or come up with a model for the residuals. Tamino can perhaps tell me where I was off. But I’ll bet a bundle my conclusion stands.

  • Richard Steckis // January 18, 2009 at 2:35 pm

    If there is an increase in overall concentration of 13C (as opposed to d13C) then there must also be a non-anthropogenic component to the co2 increase. The most logical source is inorganic 13C from the oceans.

  • Richard Steckis // January 18, 2009 at 2:43 pm

    Ray Ladbury:

    “The problem with guys like Motl is that they are smart, and while their arguments are based on a misunderstanding of the physics, they tend to be above the understanding of the layman.”

    I don’t know that Motl misunderstands the physics. After all he is a theoretical physicist.

    My personal opinion is that you have to glean information from both sides Ray. Tamino sometimes talks bullshit but most times he doesn’t. Same with Motl.

  • t_p_hamilton // January 18, 2009 at 2:49 pm

    Richard Steckis, it is indeed possible for the ratio to decrease but the absolute number to increase. Quite easy as a matter of fact. What is relevant in kinetic effects is the ratio.

  • Nick Zervos // January 18, 2009 at 3:07 pm

    Tamino, with your reference to the CO2 isotope ratio you’re trying to deflect attention from my main point: there’s no empirical evidence to support the claim that the recent increase in atmospheric CO2 (from whatever sources) is the cause of the 20th Century rise in global temperature. AGW advocates are relying entirely on unvalidated climate models to conclude that CO2 is the cause.

    There’s an analogous situation in the field of heat transfer. Take, for example, heat transfer from a turbulent fluid flowing in a tube to another fluid flowing outside the tube. This is an almost trivially simple phenomenon compared to the complex and chaotic attributes of global climate. Yet any scientist claiming to have a computer model that could predict such heat transfer behavior without empirical data to validate those predictions would be laughed out of town.

    AGW advocates are apparently more willing to rely on faith-based systems of explanation.

    Getting back to the atmospheric CO2 isotope ratio: The only thing that tells us is that the relative fluxes of CO2 from biomass and fossil fuel combustion vs. the non-anthropogenic sources has changed. Duh. Given the fact that humans have started burning fossil fuels, and burn-clearing land, and causing forest fires, etc., how could it not? What that DOESN’T prove is that the current atmospheric CO2 concentration would be lower at the current global climate conditions were it not for anthropogenic sources.

    If the global temperature were to become cooler, the CO2 concentration would drop without any reduction in human emissions. Keep an eye on that Mauna Loa data.

  • dhogaza // January 18, 2009 at 3:07 pm

    If the slopes are different, for what ever reason then d13C cannot be reliable for estimating the level of anthropogenic co2 in the atmosphere.

    Here’s an opportunity for you to gain fame by publishing your first paper related to climate science.

    Get back to us when you’ve succeeded in overturning this bedrock bit of climate science, dude.

  • Hank Roberts // January 18, 2009 at 4:14 pm

    Steckis, the religious framework you see around things is attatched to your point of view, not to the things you’re describing. You’re projecting it.

    You attribute to supernatural diabolical intervention wattsup’s site, which readily can be explained by innumeracy.

  • dhogaza // January 18, 2009 at 5:37 pm

    AGW advocates are relying entirely on unvalidated climate models to conclude that CO2 is the cause.

    This isn’t true. Why don’t you come back when you understand that it isn’t true and why it isn’t true.

    Meanwhile, if you’re going to base your arguments on strawmen you build and knock down, my guess is people here will simply ignore you.

  • David B. Benson // January 18, 2009 at 7:12 pm

    Nick Zervos // January 18, 2009 at 3:07 pm — We rely on the fact established about 150 years ago that carbon dioxide is a global warming (so-called greenhouse) gas. More CO2, more warming. Concentrations of CO2 have been increasing since the start of the industrial revolution. Ego …

    No climate models required.

    You could learn some of the physics by reading “The Discovery of Global Warming” by Spencer Weart:

    http://www.aip.org/history/climate/index.html

    Review of above:

    http://query.nytimes.com/gst/fullpage.html?res=9F04E7DF153DF936A35753C1A9659C8B63

    Please do so before coming back for further cmments.

    Thank you.

  • Malcolm // January 18, 2009 at 7:17 pm

    David B. Benson: every time you engage in an ad hominen attack, an Angel becomes a climate sceptic. :-).

  • luminous beauty // January 18, 2009 at 7:59 pm

    Nick,

    Did you know that the first numeric general circulation model of the atmosphere was written by Ed Lorenz?

    Did you know Ed Lorenz pioneered chaos theory?

    More than forty years ago?

    Do you really think modern day climate modelers are unaware of the constraints of chaos theory?

    Really?

  • Hank Roberts // January 18, 2009 at 8:07 pm

    Malcolm, David has been suggesting your comments would make more sense if you’d have read some of the relevant science before writing them. That is not an ‘ad hominem’ attack.

    If you have read the material and claim to feel attacked because people say that is not apparent in what you write, then cite and quote the particular text on which you’re basing your beliefs, and we can talk about it and how it’s understood.

    If you aren’t interested in reading the science, or believe you don’t need to read it, or believe you understand it well enough to lecture on it without giving any sources, and want us to rely on your authority, that’s something else again.

    Clarify please. You aren’t being attacked, you’re being asked how you support what you say.

  • Hank Roberts // January 18, 2009 at 8:08 pm

    PS: this may help:
    http://moregrumbinescience.blogspot.com/2008/10/ad-hominem.html

  • t_p_hamilton // January 18, 2009 at 8:28 pm

    Nick Zervos said:”Getting back to the atmospheric CO2 isotope ratio: The only thing that tells us is that the relative fluxes of CO2 from biomass and fossil fuel combustion vs. the non-anthropogenic sources has changed. Duh. Given the fact that humans have started burning fossil fuels, and burn-clearing land, and causing forest fires, etc., how could it not? What that DOESN’T prove is that the current atmospheric CO2 concentration would be lower at the current global climate conditions were it not for anthropogenic sources.”

    leChatelier’s principle shows that Nick Zervos’ assertion can’t be true. If one places a stress on a system in equilibrium, then the system responds to reduce that stress. Stress: human adding more CO2. Response: System absorbs some CO2, but not all. Hence the amount of CO2 is higher than in the unstressed system. Therefore the current climate would have a lower CO2 amount if it were not for human activity.

    Other evidence: the record shows 280ppm max for hundreds of thousands of years.

    “If the global temperature were to become cooler, the CO2 concentration would drop without any reduction in human emissions. Keep an eye on that Mauna Loa data.”

    Two things make this statement irrelevant. The first is the “if”, for anything that would last more than two years. The second is that CO2 would drop from a temperature decrease only after the human emissions rate came into equilibrium with the other parts of the carbon cycle. That has a time scale scale of thousands of years.

  • David B. Benson // January 18, 2009 at 9:41 pm

    Malcolm // January 18, 2009 at 7:17 pm — Besides Hank Roberts link, see

    http://en.wikipedia.org/wiki/Ad_hominem

    I fail to see how suggested you acquire some knowledge of the subject matter can be so construed.

  • vibenna // January 18, 2009 at 11:06 pm

    Hank - which comments? Was it my request for opinions on the Pacific Decadal Oscillation? My question about the effects of the reduced solar irradiance being reported by NASA? My concern that carbon forcing may have been overestimated? Was it the question about whether decomposing “noise” into short-term chaotic elements and long term trends would improve forecasting? Was it something else?

    Asserting that I am ignorant and you are knowledgeable is just, well, an assertion.

    David - claiming I am embarrasing myself is an emotional claim against the man. Not a scientific claim. It’s purpose is clearly to put me off, rather than to address particular questions. I’d call that an ad hominen attack, particularly as no specific argument or data was included in your comment, just a non-specific reference.

    (BTW, the reference you suggest is out of stock at Amazon.)

  • malcolm // January 18, 2009 at 11:10 pm

    Rats - on the wrong login. That last comment was supposed to be from me.

  • David B. Benson // January 19, 2009 at 1:16 am

    malcolm // January 18, 2009 at 11:10 pm — Not at all. I simply made an (correct) assertion that your display of ignorance was embrassing yourself. That this was in my view goes without saying.

    I don’t see that as ad hominem, since I don’t consider ignorance to be anything but corrigible through study.

    Anyway, which book is out of stock? I’ll suggest another instead, but first try another vendor.

  • luminous beauty // January 19, 2009 at 2:12 am

    malcolm,

    Clickee on the linkee. It’s free online.

    http://www.aip.org/history/climate/index.html

    Don’t be so sensitive. There is no crime in being ignorant. It’s the default position for learning. David and Hank are just trying to help you find clear and credible sources of information.

  • P. Lewis // January 19, 2009 at 2:28 am

    I think he’s referring to Ruddiman’s Plows….

    Amazon UK have it in stock; Amazon USA temp out, B&N have it in stock, …

    Excuses, excuses.

  • Richard Steckis // January 19, 2009 at 5:19 am

    t_p_hamilton:

    “Richard Steckis, it is indeed possible for the ratio to decrease but the absolute number to increase. Quite easy as a matter of fact. What is relevant in kinetic effects is the ratio.”

    t_p. You obviously do not understand the science behind d13C and anthropogenic co2. If there is any increase in 13C at all, then that increase can only come from non-anthropogenic inputs. This is because all anthropogenic co2 is biogenic in origin and therefore depleted in 13C. Biogenic carbon has a d13c ratio of about -25 to -27 ppt (PDB). Some biogenic carbon has substantially greater depletion than that. Therefore if all co2 increase in the atmosphere was biogenic in origin, there would be no increase in overall 13C. Any increase in 13C must be from non-biogenic and therefore non-anthropogenic sources.

    [Response: You're wrong, and your can't even see it when it's pointed out. Strike one.]

  • Richard Steckis // January 19, 2009 at 5:28 am

    just about all fossil carbon is biogenic in origin. There is an hypothesis that crude oil is abiogenic in origin.

    [Response: Strike two.]

  • malcolm // January 19, 2009 at 5:58 am

    Thanks for the link - that’s what I’m here for - constructive debate, clarification and information.

    Still waiting to hear what specific thing I’m supposed to be ignorant about, though.

  • Barton Paul Levenson // January 19, 2009 at 6:26 pm

    Richard Steckis writes:

    The most logical source is inorganic 13C from the oceans.

    Richard, the oceans are a net SINK for CO2, not a SOURCE. They emit 90 gigatons of carbon a year, but take in 92. You could have looked this up.

  • Barton Paul Levenson // January 19, 2009 at 6:31 pm

    Nick Zervos, who seems impervious to logic, writes:

    there’s no empirical evidence to support the claim that the recent increase in atmospheric CO2 (from whatever sources) is the cause of the 20th Century rise in global temperature.

    “I’m not familiar with the evidence” is not the same as “there is no evidence.”

    Carbon dioxide is a greenhouse gas.
    There’s more of it.
    We’re producing it.
    The temperature is rising.
    The temperature rise correlates with the CO2 rise.

    What more would any sane person want? How much evidence do you need?

    AGW advocates are relying entirely on unvalidated climate models to conclude that CO2 is the cause.

    No, we’re relying on radiation physics. The theory of global warming (1896) long predates global climate models (1955).

    AGW advocates are apparently more willing to rely on faith-based systems of explanation.

    Nobody is interested in your religious prejudices. Stick to the subject or shut up.

    Getting back to the atmospheric CO2 isotope ratio: The only thing that tells us is that the relative fluxes of CO2 from biomass and fossil fuel combustion vs. the non-anthropogenic sources has changed. Duh. Given the fact that humans have started burning fossil fuels, and burn-clearing land, and causing forest fires, etc., how could it not? What that DOESN’T prove is that the current atmospheric CO2 concentration would be lower at the current global climate conditions were it not for anthropogenic sources.

    No, the volume of the relative sources and sinks proves that. Which you could look up. Start with the IPCC AR4. They have a pretty nice diagram of the carbon cycle.

    If the global temperature were to become cooler, the CO2 concentration would drop without any reduction in human emissions.

    You mean if it were to drop to subzero temperatures and the Earth were to glaciate? You could be right.

  • David B. Benson // January 19, 2009 at 7:29 pm

    malcolm // January 19, 2009 at 5:58 am — It doewsn’t matter. Reading Weart’s book will bring you up to speed to the point of asking less naive questions.

    By all means ask such. There is considerable assistance on this site.

  • Hank Roberts // January 19, 2009 at 8:13 pm

    Malcolm, one important difference between debate and science: debaters use the facts they prefer.
    If you’re here for debate you may have come to the wrong place. You understand the difference between debate and science?

  • wishing (ihacan) // January 19, 2009 at 10:47 pm

    Fictional story:

    I once had a very small amount of money in bills and coins that I gave to my father to keep track of for me.

    Then I got a job baby-sitting.

    I was always paid in cash, usually in one dollar bills. Every week I would deposit this money with my father.

    After a few months I had given my father a total of a hundred dollars (from baby-sitting and prior savings), and I wanted to buy something with that money. So I went to him and asked for the full 100.

    Imagine my surprise when I got that money back as 80 one-dollar bills, 2 five-dollar bills and 10 one-dollar coins. Surprise, because I knew for a fact that I hadn’t added any coins at all from baby-sitting, and I had only given him 8 coins before I started baby-sitting. There was only one conclusion: my father must have added some coins!

    Question: What was (were) the source(s) of my savings?

    –wishing (ihacan)

  • Richard Steckis // January 20, 2009 at 2:22 am

    Tamino:

    “Response: Strike two”

    Are you saying that fossil carbon is not biogenic in origin? The it’s a strike for you.

    [Response: You said "There is an hypothesis that crude oil is abiogenic in origin." I don't care whether there's a theory of that or not -- crude oil is from fossilized plants, and that's biogenic. You said that because you're trying to be contrary.]

  • Richard Steckis // January 20, 2009 at 2:24 am

    Tamino:

    “You’re wrong, and your can’t even see it when it’s pointed out. Strike one.”

    Clarify where I am wrong. Or are you getting confused between 13C and d13C?

    [Response: You said "If there is any increase in 13C at all, then that increase can only come from non-anthropogenic inputs." Do you not understand that *depleted* in 13C doesn't mean *devoid* of 13C?

    Strike three. You're out.]

  • Jim Eager // January 20, 2009 at 2:41 pm

    Maybe Steckis would listen to Spencer on this?

    “First, the growth rate of atmospheric CO2 is, on average, only about 50% of what mankind emits. This means that Mother Nature takes out about 50% of the ‘excess’ CO2 that we pump into the atmosphere every year.”

    Thanks to Raven, of all people.

  • Hank Roberts // January 20, 2009 at 3:26 pm

    Just for reference, one of the scientists whose work was misread by the abiogenic oil proponent was inspired to blog about that. See several threads before the one excerpted here:

    http://wah-realitycheck.blogspot.com/2008/11/no-peak-of-abiogenic-oil.html

    “… abiogenic, deep origin, and therefore present in vast quantities that we only need to tap. Does this argument stand up to a reality check?

    Hardly. The foremost western proponent of the abiogenic oil theory is Jack Kenney, who indeed posts several anti-peak-oil articles on his website. I have already written about my weird experience with Kenney, based on which I certainly don’t trust him. But his theory is also thoroughly refuted by many experts in the field….”

  • Raven // January 20, 2009 at 5:01 pm

    Any skeptic who doubts that the CO2 rise is caused by humans should look at this:
    http://www.ferdinand-engelbeen.be/klimaat/co2_measurements.html

    Ferdinand is also an AGW skeptic but does not dispute the source of the CO2 rise.

  • Gary Moran // January 21, 2009 at 2:16 pm

    While Dennis Avery’s conclusions are obviously rubbish, there does appear to be some trends in the data that could be relevant to the static nature of GMST in the 21C (I look at Hadcrut3, but the differences with GISS will not be significant):

    In addition to an underlying warming trend, there also seems to be long term (approx 60 years) cycles of warming and cooling (internal variability?). So we have peaks (approx): 1880, 1940, 2000; and troughs: 1910, 1970. In the 21C GMST does appear to be statistically static, and HADCRUT3 shows the last 3 years getting consecutively colder, such that we are now at the southern edge of the 95% normal distribution (.322 - .315)! If the IPCC hypothesis (good a term as any) is correct then the current short term trend needs to start heading north soon. The last 30 years, which may coincide with a warming phase of internal variability, is currently trending at .14′C/D, obviously below the IPCC estimate for the beginning of the 21C; and the trend is reducing from 1.8′C/D 5 years ago. At this moment in time the empirical data suggests it is not unreasonable to assess the GCM’s predictions are a tad high, and perhaps not so catastrophic after all.

    [Response: I don't know where you get your information, but your "conclusions" are every bit as obviously rubbish as Avery's.

    1) "there also seems to be long term (approx 60 years) cycles of warming and cooling"

    No there doesn't. Proper period analysis shows that any claims about a 60-year cycle utterly fail even to come close to statistical significance.

    2) "In the 21C GMST does appear to be statistically static, and HADCRUT3 shows the last 3 years getting consecutively colder"

    The trend since 2001 is -0.1 deg.C/decade plus or minus 0.3. That's perfectly consistent with the prevailing trend since 1975. As for "the last 3 years" -- could you be more naive?

    3) "The last 30 years, which may coincide with a warming phase of internal variability, is currently trending at .14'C/D, obviously below the IPCC estimate for the beginning of the 21C"

    Wrong and wrong. The last 30 years is trending at .16 deg.C/decade, and the "IPCC estimate" is for the entire 21st century, not the "beginning" of it. The actual model simulations used for IPCC projections indicate early-21st century warming at about .165 deg.C/decade.

    4) "the trend is reducing from 1.8'C/D 5 years ago."

    The present trend (in HadCRU) is 0.16 +/- 0.04; the trend five years ago was 0.18 +/- 0.04. They're well within each other's error limits.

    Are you that seriously misinformed? Or are you one of the misinformers? If the former, you've come to the right place. If the latter, you're definitely in the wrong place.]

  • Gary Moran // January 22, 2009 at 8:48 am

    A few points:

    underlying warming plus 60 year warming cooling cycle – it eyeballs that way to me, and I’m pretty sure I could get a reasonable fit with some work – however I haven’t done the legwork.

    I’m surprised you seem to be trying to undermine my claim that hadcrut3 is showing GMST to be static for the last 10 years (1999 – 2008), as this point is admitted by Hadley themselves (press release Autumn 2008 referencing 1998 - 2007). Whether the short term trend is significant, obviously hasn’t been demonstrated yet, but I didn’t say otherwise. However 2008 GMST was so close to being in the outer 5% of normal distribution (1975 – 2007 as suggested in Open Mind – You Bet 2007), that you must surely have been intrigued? Where will 2009 go? Another year like 2008 would not look good for current GCM projections.

    My figures are: (60yr trend 1949 - .10; 30yr 1979 - .14, 10yr 1999 .07) I have 2008 as .322 and December 2008 at .307. I think they are correct, but I’m rubbish at stats – but I thought you might be using figures to 2007 instead?

    As for my “conclusions”, I’m saying current figures are consistent with the idea that GCM projections are overstated. As such they are patently not rubbish, and in fact you didn’t prove they were, you tried to prove something else, that the IPCC model is not inconsistent either; but I didn’t say it was! It is arguable that a pattern of divergence is appearing, but I’m afraid my maths aren’t up to that. But you are overstating the certainty of the IPCC model!

    However you really missed the point of my post, which was to reference your 2007 You Bet article. 2008 came within a whisker of being a strike against the warming camp using Hadcrut3. How close was it according to GISS?

    [Response: Here's the bottom line: you need to understand that you (and millions upon millions) suffer from a number of very common statistical misconceptions, which is why denialists are so successful misleading people.

    1) "it eyeballs that way to me, and I’m pretty sure I could get a reasonable fit with some work"

    One of the recurrent themes of this blog, one of its reasons for being, is to repeat until it sinks in that such an approach doesn't work. "Eyeball" is a great way to get good ideas. It's also a great way to get wrong ideas. As for "get a reasonable fit with some work," I'm pretty sure I could do that myself -- fitting temperature increase to the decline of Pirates since the 19th century or the shape of Angelina Jolie's legs.

    2) "Whether the short term trend is significant, obviously hasn’t been demonstrated yet"

    On the contrary, it has been demonstrated that it's not significant. Not. That too has been repeated here multiple times. So talking about "static for the last 10 years" is just like "static since last Thursday" -- meaningless. That last 10 years is not meaningfully different from expectation for continued global warming.

    3) "However 2008 GMST was so close to being in the outer 5% of normal distribution (1975 – 2007 as suggested in Open Mind – You Bet 2007), that you must surely have been intrigued?"

    Here's your chance for an epiphany: what would intrigue me is if global temperature did not flirt with the edges occasionally. Such circumstance aren't just possible, they're inevitable, and if they didn't happen I'd know that something strange was going on. Just like if I flip a coin 100 times, getting 5 "heads" in a row is no surprise at all, but not seeing that happen -- then I'd be intrigued.

    4) "I’m saying current figures are consistent with the idea that GCM projections are overstated. As such they are patently not rubbish"

    Maybe you'll understand if I put it this way: "no global warming since last Thursday, so GCM projections are overstated" -- do you see that that is rubbish? Your statements are absolutely no different -- and they're just as much rubbish. When you say "GCM projections are overstated" without any valid evidence (and you're not even close) it'll be called rubbish. Because it is.

    Every one of your misconceptions is a perfectly natural consequence of the sometimes counterintuitive nature of statistical behavior. That's one of the reasons denialists are so successful.]

  • Gary Moran // January 22, 2009 at 9:13 am

    Oops, sorry you’re right: 60yr trend 1949 - .11; 30yr 1979 - .16, 10yr 1999 .08

  • Ray Ladbury // January 22, 2009 at 2:21 pm

    Gary Moran, Hallelujah! Brother! Can I get a “Never mind!”

  • David B. Benson // January 22, 2009 at 7:26 pm

    Gary Moran // January 22, 2009 at 9:13 am — I did an fft power spectrum for the temperature record from central Greenland for the enitre Holocene; there is nothing there but red noise down to the minimum period detectable of 25 yers or so.

    So then I used a method based on the Lomb periodiogram for detecting quasiperiodic variations. At the limit of detectablity, I could just pick out something around 35–90 year periods, nothing shorter and nothing longer out to several hundred years.

    I don’t claim statistical significance for this method (I don’t know how to do that), but I’ll offer the suggestion that the various ocean oscillations are showing up in the termperture records, hardly a surprising conjecture.

    This, of course, has nothing to do with the current warming; the data ends in 1857 CE.

  • t_p_hamilton // January 22, 2009 at 8:25 pm

    Tamino said:”Just like if I flip a coin 100 times, getting 5 “heads” in a row is no surprise at all, but not seeing that happen — then I’d be intrigued.”

    I wonder if a tendency to read signal into noise could be illustrated by the flip side. Ask several people to write a string of random heads and tails for N=100 - I bet the levels of noise away from 50-50 is vastly underestimated by most people. I be you would rarely see someone write five heads in a row.

    I have a subject right here, a graduate student in the sciences. He started off with 5 heads, and had a string of 7 tails.

  • David B. Benson // January 22, 2009 at 8:26 pm

    I shouldn’t have written “red noise”. Power law noise.

    Anyway, no signal(s) stood out of the background.

  • Gary Moran // January 23, 2009 at 9:03 am

    Surely a 10 year trend has more significance, in comparison to a 30 year period than does 7 days! Regardless - so what you are saying is:

    There is no internal variability of the earth climate system; only noise and forcing signals (aerosols, CO2, solar irradiance, stratospheric volcanic eruptions). That looking at the trend from 1975, this is the signal from anthropogenic CO2 and the “static” 21C is just a product of noise?

    How certain of that are you? If you have adhered to an incorrect model that for a time was sufficiently supported by empirical data, then it will take time for that to be demonstrated to be statistically unlikely. In that transition evidence will appear for an alternative model, but that will not be inconsistent with your model.

    I’m not saying this is the case; I’m just interested in the possibilities. And remember before you dismiss this out of hand, a lot of my interest in this was piqued by your “You Bet” article.

    Thanks for your patience in this matter, all of your responses have been interesting.

    [Response: It's not that there's no internal variability, there is. But it's unpredictable (on anything other than the briefest possible time scales) either through randomness or chaos, or through our incomplete knowledge. Because of that, it behaves exactly like random noise, and from the point of view of trend analysis it should be treated like noise: it tells you nothing about the actual trend that's present, in fact it confuses estimates of the trend, making it take longer to determine with confidence.

    Perhaps a good analogy is that trend is like climate, noise is like weather. Weather is natural variability, and it's certainly not random but it is chaotic, so if we're analyzing data to look for a trend we have to treat the weather like it's noise in the data -- not signal.

    The data since 1975 indicate an upward trend plus noise. Statistics tells us little about the cause of the upward trend, although physics tells us a lot (greenhouse gases). As for certainty, I am 100% sure that the data give no evidence of a change in the trend. None. That does not mean there's no change in the trend, it just means there's no evidence of it.

    And no, the recent stasis is not evidence of a trend change. It's really typical of what we expect to see from time to time, in fact what we will see from time to time, given the size of the trend and the nature of the noise. Maybe a close reading of this post will make that point clearer.

    Am I certain that the trend hasn't changed? Of course not. But I am certain that there's no evidence of that. None. Those who point to the last decade as evidence, are selling you a bill'o'goods.]

  • gary moran // January 24, 2009 at 7:40 pm

    I’ve just read the post, don’t get fooled again. And I found it very interesting; and I understand why you get so irritated by the 10 year flat trend proves global warming has ended.

    However, it also showed to me that the points I made in my first post were valid. That’s not to say correct, but causes of reasonable doubt - because it all comes down to assumptions and models (a linear trend is a model, which can be used for prediction).

    The GMST as recorded in the hadcrut3 data set demonstrates a great deal of variability that is certainly not just a linear trend plus weather noise. I talked about an approximate 60 year cycle of warming and cooling and I talk about that as possibly demonstrating internal variability (not weather noise, thats on top again). Given the earth climate system is chaotic, there is little point debating to what extent that pattern is real or fixed.

    It is therefore my contention that the pattern in hadcrut3 represents a small underlying warming trend plus internal variability. To be clear this internal variability is not weather noise, weather noise comes on top of this. There will also be other forcings acting on this such as stratospheric volcanic eruptions, perhaps aerosols will impact, perhaps even small increases in the concentration of IR active trace gasses.

    The IPCC hypothesis differs in that the dips in the record are blamed on aerosols. This is obviously not proven. The IPCC earth climate system (ECS) only has limited internal variability (ENSO) which can be ignored on longer timescales. This is an unproven hypothesis, which may be true or false. According to the IPCC the ECS is chaotic but can be predicted because it behaves stochastically.

    So when “you” talk about the trend, you are taking the trend from the 1970s onwards and treating that as the anthropogenic signal - the basis on which that is done is that even though the ECS is chaotic, it is in a predictable fashion, and that the previous not so linear trend can be ignored because we “understand” why it wasn’t linear. It was the sum of all the forcings (TS1, Aerosols, CO2).

    The problem is that the assumption is too big. It seems highly likely to me that the ECS is truly chaotic, and will shift around all over the place for no apparent reason. GMST appears to demonstrate scale invariance, and self similarity; characteristics of truly chaotic systems. If that is the case, then it is climate scientists who fooled themselves - looking for an anthropogenic signal in climate in the 60’s and 70’s, either a warming or cooling signal, and when they found it - which of course they inevitably would do - and explaining past discrepancies by curve fitting with fudge factors like aerosols. …

    And again, I’m not arguing my viewpoint is correct; I’m just saying it is a reasonable position.

    Finally I thought the comparison between the chances of GMST dipping into the 5% area of normal distribution to coin flipping, on reflection, to not be reasonable.

    [Response: All of your "causes of reasonable doubt" are invalid. A prime example is your repeating the claim of an "approximate 60 year cycle of warming and cooling." There's zero evidence of such a thing: zero. It's no more valid, statistically, than "10 year flat trend proves global warming has ended."

    You fail to grasp that events which appear, to human intuition, to be too unlikely to believe -- cause for reasonable doubt -- not only can happen, they must happen. This applies to the "60 year cycle" mistake every bit as much as it does to the "10 year cooling" falsehood. If such things didn't happen, then I'd know the system was behaving in an unusual way. But when the system behaves exactly as expected, you consider that cause for doubt. That's nothing more than statistical naivete.

    You also "contend" that you understand the pattern of temperature variability. But clearly you don't; you had preconceived notions which are provably wrong, you've even realized that some of your preconceptions are mistaken.

    Your calling aerosols a "fudge factor" isn't just mistaken, it's insulting. They exist, they rose dramatically in prominence in the post-war period (so much so we had to pass laws to limit emissions), and they cause cooling. These aren't opinions, they're facts. Your offensive characterization indicates an ignorance of physics to match your ignorance of statistics.

    Here's my point of view: you don't have hypotheses, you just have guesses and impressions. They're not based on evidence, just on intuition which isn't founded in knowledge of climate science, it's based on the quirky behavior of statistics which fools a lot of people. You really need to add some humility to your skepticism.

    Instead of repeating "I have doubts" you should be saying "I have a lot to learn before I can comment intelligently." If you really want to leave ignorance behind, read Spencer Weart's "Discover of Global Warming." There's a link on the right-hand sidebar labelled "History of Global Warming (Spencer Weart)."]

  • Hank Roberts // January 24, 2009 at 7:55 pm

    http://www.agu.org/pubs/crossref/2009/2008GL036350.shtml

    “… Mainland Europe’s temperature rise of about 1°C since the 1980s is considerably larger than expected from anthropogenic greenhouse warming. Here we analyse shortwave and longwave surface forcings measured in Switzerland and Northern Germany and relate them to humidity- and temperature increases through the radiation- and energy budget. Shortwave climate forcing from direct aerosol effects is found to be much larger than indirect aerosol cloud forcing, and the total shortwave forcing, that is related to the observed 60% aerosol decline, is two to three times larger than the longwave forcing from rising anthropogenic greenhouse gases….”

  • gary moran // January 25, 2009 at 6:41 pm

    You’ve missed the point.

    [edit]

    [Response: You're the one who's missed the point. You continue to speculate from a position of ignorance, and frankly I'm not interested in hosting your speculations.

    I've already answered a number of your "causes for doubt" in great detail. You seem able to learn -- but not willing.]

  • David B. Benson // January 25, 2009 at 9:10 pm

    gary moran // January 25, 2009 at 6:41 pm — First learn what is known and maybe then ask questions. I second Tamino’s advice to you in a reply to a prious comment of yours..

  • Hank Roberts // January 26, 2009 at 1:26 am

    Tamino, is there a way to make this a banner at the top of the main page? It deserves prominence:
    Instead of repeating “I have doubts” you should be saying “I have a lot to learn before I can comment intelligently.” If you really want to leave ignorance behind, read … a link on the right-hand sidebar labelled “History of Global Warming (Spencer Weart).”

  • Hank Roberts // January 26, 2009 at 2:34 am

    Mr. Moran, you’ve posted your opinions many places and gotten a lot of encouragement. You can look yourself up:
    http://www.google.com/search?q=“Gary+Moran”+%2Bclimate

    Be careful whose encouragement you take. “Just because you’re on their side doesn’t mean they’re on your side” is good advice.

    Learning the basic routine of checking everything people tell you against the actual source — via Google Scholar or your local library’s Reference Desk — is the first step.

    What happens when you do this — with anything anyone tells you in our society — is, frankly, somewhere between discouraging and terrifying. But it’s also liberating because it _is_ possible to check what people say.

    Do you know Sturgeon’s Law? It certainly applies to public media and news. Most of it is someone’s opinion, spun to bolster some point of view. Getting past that is hard.

    It is, however, doable.
    David Brin writes about this a lot. I recommend simply testing everything everyone tells you around these issues against published science journal articles, and among those, look for ones with ample “cited by” references from people who found them useful.

    That’s if you think science has a chance of providing accurate information. I don’t know where else one could hope to look. Do you?

  • Hank Roberts // January 26, 2009 at 2:39 am

    One for Dave A, from the current New Scientist:

    ———
    … Before the findings were published in October in The Astrophysical Journal Letters (vol 686, p L49), Kashlinsky knew how heretical his idea would seem. “We sat on this for over a year checking everything,” he says. “It’s not what we expected or even wanted to find, so we were sceptical for a long time. But ultimately it’s what’s in the data.”
    ———–

  • Hank Roberts // January 30, 2009 at 12:43 am

    http://www.calculatedriskblog.com/2009/01/unprecedented-and-shocking-decline-in.html

Leave a Comment