Open Mind

CO2 Blip

November 12, 2008 · 96 Comments

The October 2008 estimate has been posted of CO2 concentration as measured at the Mauna Loa Atmospheric Observatory. The graph over the time span since observations began looks like what is to be expected (I’ve added a smoothed curve to the graph):

mloco2


However, I noticed something unusual. There’s an obvious seasonal signal in the data, so in addition to reporting the monthly average, also given is a deseasonalized version, one for which the seasonal signal has been removed. Here’s the deseasonalized estimate since 2005:

recent

There’s an unusually large increase in the most recent deseasonalized value.

The smoothing method I applied in the first graph also computes residuals, the difference between the actual value and the smoothed value; this effectively “detrends” the data, so those residuals show the annual cycle quite well. I removed the annual cycle from those residuals by fitting a 4th-order Fourier series, leaving these “anomalies” (these aren’t “anomalies” in the usual sense, but that name will do in a pinch, although they can also be called “detrended deseasonalized”). Here they are:

anomaly

The latest monthly value is strongly out of line with the prevailing trend/seasonal pattern. Perhaps it’s correct, in which case it’s an unusual event whose cause deserves investigation. I think it’s more likely to be an error. I sent an email inquiry to the scientist in charge of releasing these data, but I haven’t yet received a reply.

On an entirely different track, I also subjected the detrended deseasonalized data to Fourier analysis. Longer periods (longer than about 6 yr) are removed by the smoothing process, so we won’t see them if they’re present in the original data. But a shorter period, 3.62 yr, is strongly indicated:

p36

I’ve seen the same period in other data, including temperature data and indexes of the el Nino phenomenon. For instance, it’s present (estimated at 3.65 yr) in temperature data so I included it When I made my “just-for-fun forecast” of future temperature. I’ve now seen it in enough time series that I think it deserves closer investigation.

UPDATE

It turns out that the October CO2 estimate from Mauna Loa was in error, as is explained in the change log:


19 November 2008

We corrected an error in the data posted for the previous October on 3 November 2008. The assigned CO2 values for the reference gas mixtures in use at our Pt. Barrow Observatory, Alaska, had mistakenly been used in calculating CO2 mole fraction values for the Mauna Loa data, resulting in a value for October that was ~1.5 ppm too high. The error became apparent because of an abrupt shift in the comparison between measurements of flask air samples taken at Mauna Loa, but analyzed in Boulder, Colorado and the continuous analyzer operating at Mauna Loa. There is always a delay of several weeks between the collection and analysis of the flask air samples.

After the correction, the October 2008 value was revised down from 384.75 to 382.98, and the de-seasonalized value was revised down from 387.80 to 386.25.

Categories: Global Warming

96 responses so far ↓

  • t_p_hamilton // November 12, 2008 at 5:42 pm

    Faux outrage AGW-denier mode: I think if we find that this data was in error, this shows that AGW sientists are a disgrace!

  • Tom G // November 12, 2008 at 5:52 pm

    Curious….
    First the GISS temperature “errors” and now this supposed “error”.
    What are the chances that these are not errors and are true numbers?

    [Response: The GISS figure for Oct. 2008 is due to data for some stations being repeated from Sep. to Oct. by NOAA (or its precedent organization in the data-handling chain); there's really no chance that the high temperature anomaly is correct, and in fact the data files containing the global and hemispheric averages have been removed pending correction.

    As for the CO2 data, it's not impossible that the number is correct, but it's certainly unusual. As I said, I suspect it's in error but I don't know for sure. If it is correct, the cause deserves investigation.]

  • tamino // November 12, 2008 at 6:07 pm

    I received a reply to my inquiry:

    Thank you for your note. We had noticed that October looked high, although it was not unprecedented. We are looking into it, and will make a correction it we find an error. In that case the correction will be described in the Change Log on the web site.

    Thanks again,
    Pieter Tans

  • Ray Ladbury // November 12, 2008 at 6:25 pm

    Well, the denialists are basking in the success of their strategy: If you never contribute anything, you’ll never be in error.

  • Tom G // November 12, 2008 at 7:21 pm

    Thank you for the quick response.
    I hope I didn’t give the impression that I was prematurely jumping to some sort of conclusion,
    but the old saying is ‘timing is everything’ and to have these two blips come out at the same time is kinda attention getting…
    In regards to ‘faux outrage’….the comments at RealClimate about mole-hills gets a little mind numbing after a while. To me it just proves that a fair number of people have nothing to say but are completely willing to say it anyway…..

  • David B. Benson // November 12, 2008 at 9:33 pm

    Puzzeling, that 3.6 year period peak in both CO2 and temperature and some elNino index. This is very close to 1/3 of an average solar cycle; probably just a coincidence.

    Tamino — could you say more about how this arises in elNino indicies?

  • Richard Steckis // November 13, 2008 at 1:16 am

    Tamino,

    If you are seeing this anomaly in three different climate datasets is it possible that it is a real but as yet unknown phenomenon?

    [Response: Definitely possible. It could be a statistical fluke, common to many data sets simply because they're correlated on various time scales. But I consider it more likely (not certain!) that it's a real phenomenon. Whether or not it's been noticed elsewhere, I don't know.]

  • Phil Scadden // November 13, 2008 at 3:35 am

    An ocean overturning effect perhaps linked to El Nino?

  • michel // November 13, 2008 at 10:39 am

    Well, the denialists are basking in the success of their strategy: If you never contribute anything, you’ll never be in error.

    It might be fairer to put it like this.

    There are some people who doubt that GISS is delivering its product to quality standards or managing the production process in a quality controlled way. The product is the monthly global anomaly series.

    They have quite a bit of evidence on their side from the quality failures that have been documented, including the latest one.

    On a totally separate issue, the motive which prompts most of these people is skepticism about the views of Dr Hansen of GISS on AGW. These views are generally supported by the product, and the product is usually publicly associated with him and his views, so skeptics are about equally anxious to discredit the product, Dr Hansen and his views on AGW.

    There is really nothing wrong with this, whether one agrees with their skepticism or not. Any motivation which leads one to scrutinize the data series carefully is probably constructive, and can only benefit accuracy. In a similar vein it is probably constructive to have the official numbers for Iraq civil casualties skeptically examined regardless of motivation of the examination.

    In any event, its clear that GISS has a serious quality management problem, and moving into denial on this, or reacting with hostility when its pointed out, is pointless and just makes the matter worse.

  • Ray Ladbury // November 13, 2008 at 1:20 pm

    Michel,
    Horse puckey! The climate datasets all show similar trends. The differences between GISS and HADCRUT are likely attributable to the different ways they treat the polar regions.
    And what quality failures? The error was discovered and corrected rapidly. Hell, their track record is better than Microshaft–which also relies on users to find problems.

    Michel, science is both collaborative and cooperative. Scientists compete for resources, talent and in the marketplace of ideas. However, they are continually making suggestions to each other–and finding each others mistakes. If you seriously think there is a quality problem at GISS, you haven’t the foggiest notion what you are talking about. However, I’m sure the scientists there would be more than happy if you’d cough up some cash to hire someone to improve quality.

  • Richard Steckis // November 13, 2008 at 2:07 pm

    Ray Ladbury:

    “And what quality failures? The error was discovered and corrected rapidly. Hell, their track record is better than Microshaft–which also relies on users to find problems.”

    Sorry Ray, must disagree. The issue here is due diligence. GISS failed in that due diligence. Such a failure, whether minor or not, is one way to lose the confidence of the public very quickly. It can then call into question the care and quality control they put into their modelling.

  • Ray Ladbury // November 13, 2008 at 2:40 pm

    Richard Steckis, Lose the confidence of the public–you mean the folks watching Dancing with the Stars and playing with their Wii’s? Do you really think they’ve ever heard of GISS? Do you mean the scientists who actually work with data and expect occasional errors in datasets, find them, notify the issuing lab and get back to work? Or do you mean the wannabes over at CA and WUWT who don’t have the foggiest notion of what even constitutes data? Frankly all you guys are doing by harping on this point is showing that you are utterly clueless when it comes to science.

  • thingsbreak // November 13, 2008 at 2:44 pm

    It can then call into question the care and quality control they put into their modelling.

    Seeing as how the products you are comparing are produced by two different groups, that makes as much sense as claiming that the space shuttle Columbia disaster calls into question ModelE qc.

    The reaction to this kerfuffle are quite an easy way to separate the serious from the unserious in all of this. If it wasn’t easy enough already…

  • Former Skeptic // November 13, 2008 at 4:02 pm

    Richard Steckis:

    Due diligence for the data volume you are talking about requires $$ - much more than ~0.25FTE that has been estimated for the “diligence” that I assume you are desiring. Gavin’s put up a number ($500kpa) that would suffice; wanna foot that up?

    The alternative would be to let the knowledgeable public scrutinize the data and note for any errors - which has been the case for a looooong time. This is what science is supposed to be right? Unfortunately, certain folk with biases have tried to take this error and run with it with their prior agendas, instead of taking it for what it’s worth.

    Besides, your conclusions (GISS WRONG! = public loss of confidence = modeling faulty) smack of a classic slippery slope. Do you honestly think, like several commentators in WUWT?, that once the GISS Oct data are corrected that global warming would have stopped? *shock shock horror horror*

    And from your response I take it you’ve been ignoring the interesting debate going on at RC? Methinks Gavin’s answer to John Finn below would apply to your concerns.

    From RC: http://www.realclimate.org/index.php/archives/2008/11/mountains-and-molehills/langswitch_lang/po#comment-102758

    GISS is not the international weather service. GISTEMP is an analysis of data collated by the NWS and NOAA, and it is not independent of their efforts. GISS maintains no stations, employs no on-site monitors, and takes part in no international negotiations on the sharing of weather data. GISTEMP provides that analysis as is and can’t possibly certify the work of all the individual agencies whose data gets used. When problems are noticed (as now), all GISTEMP can do is query the originators of that data. Most often problems are noticed in comparison with other products (and that goes for the other products as well), it is clearly more efficient to release the analyses relatively quickly and have more people looking at the output than spend weeks comparing everything internally before release. If that’s what you want, only look at the data at the end of the year and ignore the monthly releases.

  • michel // November 13, 2008 at 4:21 pm

    Ray, you really do not understand. The GISS global anomaly series are not science. It is not research. They are products. They are the outputs of a defined process, which like any other process that produces products, needs quality control. Now, this particular process permits defective product to come off the line undetected until product arrives in the hands of consumers. This is what is meant by inadequate quality management. This is what happens when you think quality is something you inspect in. Or that it is something which is still delivered if you do rework when the customer sends the thing back.

    It is not like this is the first time there has been a problem. But the response is always the same, to fix it and defend the status quo and impugn the motives of those objecting. Someone on RC suggests this should be handed over to people who do data generation for a living, like D&B lets say. Indeed yes. What is going on now is the equivalent of trying to deliver product from your prototype assembly facility.

    If you cannot see this, you really do not understand about production environments. And this is data so important that it has to be produced from a real serious professional production environment. The future of the planet is at stake, after all.

  • t_p_hamilton // November 13, 2008 at 4:25 pm

    Ray Ladbury:

    “And what quality failures? The error was discovered and corrected rapidly. Hell, their track record is better than Microshaft–which also relies on users to find problems.”

    Richard Steckis:”Sorry Ray, must disagree. The issue here is due diligence. GISS failed in that due diligence. Such a failure, whether minor or not, is one way to lose the confidence of the public very quickly. It can then call into question the care and quality control they put into their modelling.”

    That did not address Ray’s point. Does Microsoft fail in due diligence? Does the public lack confidence in Windows, in spite of the weekly reminders (called updates)? I don’t see it. There are alternatives (Mac and Linux) yet Windows predominates by far still.

    Does the public lack confidence in Intel and AMD chips, in spite of well publicized hardware bugs? Did they fail due diligence? Perhaps we should not buy their chips - what is the alternative? Not have computers, so we won’t be susceptible to making those errors? That would be the denier solution, which of course is directly opposed to the idea of progress.

    The fact is simple and plain: complex tasks will have mistakes, so the criteria for judging success is performance. That is how progress is made, not waiting for perfection, which can NEVER be guaranteed. It would be FAR better to remind the public about that than to mislead them about the reliability of GISS, and maybe even remind the public about GISS’s purpose (which is NOT to verify another agency’s data).

    What is important is whether the GISS model is correct, and whether it matches reality. It is quite clear from the lack of coherent counterargument that GISS is correct.

  • t_p_hamilton // November 13, 2008 at 4:42 pm

    michel:”In any event, its clear that GISS has a serious quality management problem, and moving into denial on this, or reacting with hostility when its pointed out, is pointless and just makes the matter worse.”

    Are you in the habit of asserting serious quality management problems based on a single instance of a supplier goof that is caught quickly? How much worse then the quality must be at Mercedes-Benz since they have recalls on problems that were not caught quickly, for example a 2003 recall for M-class vehicles 1998-2003.

    You remind me of the Monday morning quarterback saying how well they would have done in the Sunday game!

  • Phillip Bratby // November 13, 2008 at 5:02 pm

    Ray ladbury:

    “Or do you mean the wannabes over at CA and WUWT who don’t have the foggiest notion of what even constitutes data? Frankly all you guys are doing by harping on this point is showing that you are utterly clueless when it comes to science.”

    There are a lot of scientists (not just wannabes) at CA and WUWT who have worked with data and are used to quality controls and quality audits.

    Name-calling helps no-one, least of all scientific veracity. I am a retired physicist and I do understand a little bit about data and science.

  • Ray Ladbury // November 13, 2008 at 5:16 pm

    Michel,
    Uh, Dude, first, the “product” that had the error in it was not produced by GISS. Second, your saying that they “should have” caught this error implies that you believe that they should check for every conceivable error–whether it has ever occurred before or not–before releasing the product. Leaving aside the impossibility of doing this with the extremely limited staffing typical of NASA, do you have any idea how much this would delay the release of data? Most SCIENTISTS–you know, the ones who use this data in their day jobs–would prefer to have the data more quickly even if it does contain the rare error, since the errors do get detected and corrected pretty quickly. Your ability to concentrate on trivialities continues to astound!

  • Ray Ladbury // November 13, 2008 at 7:00 pm

    Great. You’re a physicist. Then you should know that all the chest thumping over a trivial error that would, in any case, have been difficult to anticipate, is simply absurd. You should also know that if a scientist has something of value to contribute, he or she does so in a peer-reviewed journal. Until they do that in the field under discussion they are either amateurs learning the field (into which category I fall) or wannabes–or, worst of all, trolls.

  • michel // November 13, 2008 at 8:10 pm

    You’re defending the indefensible. Its easy to write screens for the kind of errors we are talking about. Take a student a half day if that to write, and a couple of seconds to run once written. All it has to do is flag up a case for possible manual verification. But so what if it took a student a week? What’s so hard about that? You’re saying they really cannot do sanity checks once a month on a few hundred numbers because its too hard or would take too long? Its totally nuts.

    As to is it GISS product or not, of course its GISS product. Their brand is on it. They get components in from suppliers. Their task is to set standards for them, and do goods-in inspection to make sure those suppliers meet those standards. They failed in one or both.

    Whenever anyone proposes to improve quality, or put quality management into effect, the reply of the refuseniks is always that it is (a) impossible (b) too expensive. Fact is, its easy, and its a lot cheaper than the alternative. I’ve never been able to understand why it is so resisted. But what is being said here is fairly typical. And typically wrong.

  • Zeroth // November 13, 2008 at 8:21 pm

    Two errors! You’d think that upon being embarrassed once, they’d double, double check the data before re-release. But no, Gavin now reports four more sites as erroneous… a short time after McIntyre posts on his blog about the four stations in question. The comedy continues apace… replete with punctured, overinflated egos. Watts not to enjoy?

  • sod // November 13, 2008 at 8:40 pm

    the only thing typically wrong here, are your claims.

    conspiracy theories are NOT “quality control”.

    i would surely love to see some results of your “quality control” of the denialist “science”…

  • David B. Benson // November 13, 2008 at 9:19 pm

    Phil Scadden on November 13, 2008 at 3:35 am wrote “An ocean overturning effect perhaps linked to El Nino?”

    All comments after that are off-topic for this thread, just discussing a tempest in a teapot. How about taking it to the Off-topic thread, please?

    [Response: Please.]

  • t_p_hamilton // November 13, 2008 at 9:23 pm

    michel said:”As to is it GISS product or not, of course its GISS product. Their brand is on it. They get components in from suppliers. Their task is to set standards for them, and do goods-in inspection to make sure those suppliers meet those standards. They failed in one or both.”

    Actually, GISS’s job is NOT to set standards for NOAA, as these are separate government agencies. Your argument is invalid.

    GISS’s product is a global model of climate. The global temperature anomaly is a criterion used to evaluate their model. The way that GISS evaluates global temperature anomaly is widely accepted, and the procedures thoroughly checked.

    Please put in this space what GISS does for quality control currently, before they release their year report:

    Please put in this space all possible checks they should do for the data to guarantee that there is no error in NOAA’s data (after all, it’s so EASY!):

    [Response: How about taking this to the open thread.]

  • t_p_hamilton // November 13, 2008 at 9:40 pm

    michel:”Whenever anyone proposes to improve quality, or put quality management into effect, the reply of the refuseniks is always that it is (a) impossible (b) too expensive. Fact is, its easy, and its a lot cheaper than the alternative. I’ve never been able to understand why it is so resisted. But what is being said here is fairly typical. And typically wrong.”

    What is wrong is your assertions that people resist attempts at improving data quality. For example, a screen to check for that problem will surely be incorporated, if not done already.

    What we dispute is that all possible errors in data from other researchers should be foreseen, and using that as a reason to doubt the reliability of GISS. particularly from people who have done nothing.

  • Phil. // November 13, 2008 at 11:17 pm

    t_p_hamilton // November 13, 2008 at 9:40 pm

    michel:”Whenever anyone proposes to improve quality, or put quality management into effect, the reply of the refuseniks is always that it is (a) impossible (b) too expensive. Fact is, its easy, and its a lot cheaper than the alternative. I’ve never been able to understand why it is so resisted. But what is being said here is fairly typical. And typically wrong.”

    What is wrong is your assertions that people resist attempts at improving data quality. For example, a screen to check for that problem will surely be incorporated, if not done already.

    In fact according to NOAA such checks are already done (see below), for some reason their screening failed, that’s what should be addressed.

    “Both historical and near-real-time GHCN data undergo rigorous quality assurance reviews. These reviews include preprocessing checks on source data, time series checks that identify spurious changes in the mean and variance, spatial comparisons that verify the accuracy of the climatological mean and the seasonal cycle, and neighbor checks that identify outliers from both a serial and a spatial perspective.”

  • David B. Benson // November 13, 2008 at 11:32 pm

    Phil. // November 13, 2008 at 11:17 pm — This is off-topic here. Carry on over on the Open Thread. That’s what it is for.

    I’d actually like to follow up on the mysterious 3.6 year cycle, but that is hard to do with all the off-topic c**p smeared all over the thread. Thanks.

  • dko // November 13, 2008 at 11:47 pm

    Not to spoil a perfectly good OT discussion, but…

    Tamino, have you heard any more from Tans?

    [Response: No. I've checked the change log, but nothing yet.]

  • Phil Scadden // November 14, 2008 at 1:12 am

    “An ocean overturning effect perhaps linked to El Nino?”

    David, Tamino - the above was suggestion for basis of 3.5yr cycle which I thought was very much on topic - the kind of thing that might induce such a cycle. Confusing me with the “Phil”???

  • David B. Benson // November 14, 2008 at 1:25 am

    Phil Scadden // November 14, 2008 at 1:12 am — Your comment certainly is on-topic and be assured I haven’t confused you with that other Phil. :-)

    If the effect is real in all three datasets, it poses something of a conundrum: there is no other such long-term (nearly) cyclic behavior anywhere in climate, AFAIK. Ocean oscillations are not even pseudoperiodic but changes in interval seem to be correlated with changes in length of day (LOD).

    I suppose it is possible that there actually is a 3.6 year cycle in LOD, since the earth’s angular momentum is affected by the entire solar system. But 3.6?

  • BillC // November 14, 2008 at 1:37 am

    If the original corrupted data had shown that it had been the coldest October in decades then it would not have been released. They would have gone over it with microscopes to find something wrong, (which would not have taken long).
    Since a hot October it suited their case they just pumped it straight out as a press release. If it wasnt for those “denialist blogs” it would have taken a week to correct, not a day. The “correction” might never have been made public either.

    Plenty of “denial” on this thread.

    [Response: On the contrary. If the anomaly had been low, then the denialists would never have scrutinized it -- they would have trumpeted it as proof of global cooling.

    This topic belongs on the open thread, not here.]

  • dko // November 14, 2008 at 2:06 am

    David B. Benson // November 14, 2008 at 1:25 am

    “If the effect is real in all three datasets, it poses something of a conundrum….”

    Thinking out loud: CO2 correlates with surface temperature (especially when surrounded by all that water); and surface temperature correlates to El Nino index (especially when surrounded by all that water).

  • Ray Ladbury // November 14, 2008 at 2:33 am

    Tamino, be reasonable. Do you really expect the short-bus crowd from WUWT to be able to discern what is on and off topic when they can’t even distinguish a trivial error from a catastrophic blunder?

  • Phil Scadden // November 14, 2008 at 2:41 am

    dko - my thoughts too - thats why I wondered about ocean effect. Now is pseudo-periodic or periodic? How far back does the cycle go? If I had more time, I’d like to play with these series too.

  • t_p_hamilton // November 14, 2008 at 4:15 am

    Why does the spectrum show a peak at 1 year if the data have had the annual cycle from those residuals by fitting a 4th-order Fourier series? Residuals from the fit?

    [Response: Because the annual cycle has changed its amplitude over the years (it's gotten bigger). Such "amplitude modulation" can be modeled as a superposition of two sinusoids with nearly equal frequencies. The main component (1 cycle/yr) has been removed, but the "other" frequency (nearly equal to 1 cycle/yr) remains.]

  • Gavin's Pussycat // November 14, 2008 at 4:45 am

    > they just pumped it
    > straight out as a press release
    There was no press release.

  • Gavin's Pussycat // November 14, 2008 at 4:58 am

    Ocean oscillations are not even
    pseudoperiodic but changes in interval seem to be correlated with
    changes in length of day (LOD).

    Yes… and also changes in atmospheric circulation. The reason is simply conservation of angular momentum, of solid Earth + ocean + atmosphere (though some exchange of angular momentum takes place with Sun and Moon through the tides). The angular momentum budget of the Earth is currently pretty well understood.

    I suppose it is possible that there actually is a 3.6 year cycle in LOD,
    since the earth’s angular momentum is affected by the entire solar
    system. But 3.6?

    No, I don’t think the rest of the solar system can torque the Earth enough to produce observable effects (except for the lunisolar tidal torque). One possibility would be the saros period of 18 years, of which 3.6a is one fifth. I don’t buy it.

    dko’s speculation makes a lot of sense.

  • dko // November 14, 2008 at 6:12 am

    I just read what I wrote above and think I should have been more clear. (Or maybe I should stop thinking out loud…)

    Let’s try this. Suppose there really is a ~3.6-year cycle to Pacific sea surface temperature. The surface water cools, the MEI goes negative, and we say we’re in a La Nina phase. Over time, the Pacific absorbs a lot of heat and it begins to show up in your favorite global temperature anomaly. Testable Hypothesis #1: Temperature anomaly follows MEI at some lag.

    Meanwhile, cooler Pacific surface water absorbs a little more CO2, depressing the values reported at MLO. Not much, but enough to produce a signal — a signal that would be missing in data from, say, the South Pole. No lag would be discernible in monthly data. Testable Hypothesis #2: MLO CO2 and MEI are closely correlated; no pattern in polar data.

    Finally, and this is even more speculative, if the blip was real could it have come from warming sea water upwind of MLO giving up some CO2? A sort of burp heralding the return of El Nino?

    OK, I’ll stop talking now. ;-)

  • adder // November 14, 2008 at 7:16 am

    3.6 years: see

    http://www.cosis.net/abstracts/EGU05/00193/EGU05-J-00193.pdf

    [Response: Fascinating.]

  • Pete // November 14, 2008 at 4:05 pm

    See http://cat.inist.fr/?aModele=afficheN&cpsidt=846074

  • t_p_hamilton // November 14, 2008 at 4:32 pm

    Tamino,

    Thanks for answering my question about the peak at around 1 yr. The amazing thing is that I actually understood the answer immediately.

    [Response: Good! I wasn't sure it would make sense (and it probably doesn't for everybody).]

  • mauri pelto // November 14, 2008 at 8:53 pm

    3.6 years very interesting indeed. Two different references to ocean circulation by dko and adder. I recall seeing aspectrum analysis of el nino-3 and sset with a repeat interval of 3.5 years. http://ces.snu.ac.kr/uaw05/ses.5/Tianjun.ppt
    The above powerpoint has this spectrum on slide 21. however, I had seen it before. I am not clear on the s-mode versus t-mode of the el nino referred to by IPCC. I know the two have been identified with different modes of two and five years, is the combination yielding 3.5?
    http://cat.inist.fr/?aModele=afficheN&cpsidt=17538567
    http://www.cosis.net/abstracts/EGU06/01348/EGU06-J-01348.pdf

  • David B. Benson // November 14, 2008 at 9:51 pm

    After the links provided (thank you), the existence of a 3.6 year period cycle in climate is not to be doubted (at least by me). The problem is that the peak is so sharp, at least in the CO2 signal. The internal mechanisms proposed in comments above all seem plausible, but fail to explain why this is actually periodic; most oscillations in climate are quasi-periodic at best.

    So I proposed an astronomical influence, simply because such come much closer to be actually periodic. Gavin’s Pussycat didn’t care for this, but looking more carefully at LOD anomalies may well be warrented. I’ll just state I’ve seen one (unpublished, internet) paper showing LOD microsecond anomalies and stating an influence of the solar system (mostly the sun and Jupiter).

    Another possiblity, suggested by the Mediterrean-Atlantic abstract, is a regular driver for oceanic Rossby waves. An computational experiment discussed awhile back on one of Michael Tobis’s blogs strongly indicates that Rossby waves in the North Pacific shift ENSO phase. The problem for a steady 3.6 years is that I, knowing essentially nothing about what energizes oceanic Rossby waves, can imagine no means to provide the steady beat in that region.

    It is plausible (don’t know how likely) that the Rossby waves observed in the Atlantic side of the Straits of Gibraltar, being energized by the dense water flowing out of the Mediterrean and downwards towards the bottom, might have a 3.6 year period of activity. This, then, would drive the entire climates response at this frequency.

    I’d welcome additional ideas about how a 3.6 year cycle can maintain its metronomic beat.

  • Gareth // November 14, 2008 at 10:50 pm

    No oceanographer, me, but periodic “sloshing” in the ocean could be a function of ocean basin shape, and the driver winds. Fill up a bath, set up and end-to-end “slosh” by paddling with your hand - it will continue for some time without further energy input, but the slosh period will be the same - determined by the size of the bath. To keep the slosh going, you have to do some carefully timed paddling: on a global scale that could be prevailing winds…

    BTW: I’ve always thought that the shape of South America and the Antarctic Peninsula suggests that the Antarctic Circumpolar Current is very powerful… ;-)

  • David B. Benson // November 14, 2008 at 11:18 pm

    To my surprise, it seems that Rossby waves in the Pacific are biennial:

    http://www.highbeam.com/doc/1P3-30283530.html

    http://en.wikipedia.org/wiki/Rossby_wave#Oceanic_waves

  • dko // November 14, 2008 at 11:21 pm

    If there truly is a ~3.6-year in the Pacific _and_ the Atlantic/Med, then the cause must be external — unless someone can show how sloshing in one basin telegraphs to the other.

    Google [+"3.6 year" +cycle] and you get all kinds of returns. The first three concern fire patterns in Alberta, the stock market, and red-backed voles. Either this thing is a *lot* bigger that I would have thought…or maybe FFT is prone to find 3.6-year spikes in monthly data?

  • David B. Benson // November 14, 2008 at 11:47 pm

    Rosby waves don’t ’slosh’ much

    http://www.noc.soton.ac.uk/JRD/SAT/Rossby/Rossbyintro.html

    but are somehow connected to changes in SST (earlier paper) and hence changes in the atmosphere. The atmosphere changes could certainly propagate widely.

  • David B. Benson // November 15, 2008 at 12:25 am

    “Rossby wavelike features observed in the projection mode are order 1000 km long with most periods ranging between 2 and 4 years.”

    Not sharply 3.6 years, but

    “Since the Rossby wavelike features observed in the projection mode have a majority of their periods ranging between 2 and 4 years, their forcing can be attributed to long-period Kelvin waves. Spectral comparisons between the nearshore values in the projection mode and coastal sea level show greater than 90% coherence in the period band 3-4.4 years.”

    which is closer.

    From

    http://adsabs.harvard.edu/abs/1991JGR….96..749S

  • Hank Roberts // November 15, 2008 at 12:31 am

    dko wrote
    > cause must be external —
    > unless someone can show how … one basin
    > telegraphs to the other

    Wrong word (search term). Telegraph is old, telecom is in, and lo!

    http://www.google.com/search?q=teleconnection+Atlantic+Pacific+ocean+basin

  • Boris // November 15, 2008 at 1:05 am

    “Do you really expect the short-bus crowd..”

    Please don’t insult the people who have a thinking handicap by comparing them with people who choose not to think.

  • Philippe Chantreau // November 15, 2008 at 2:30 am

    Now this IS interesting. Whether or not there is anything significant to it, that’s what climate blogs should be about.

  • Gareth // November 15, 2008 at 6:32 am

    I think Hank has shares in Google…

    From his search, this NOAA page is a good intro. It will be through changes in these patterns (and perhaps new ones brought about by sea ice decline) that “climate change” manifests itself.

    See also: UCB ice core analysis.

  • george // November 15, 2008 at 5:08 pm

    Not sure if this has any relevance to the apparent cycle above (may be pure coincidence) , but there is some evidence that the sun may have a cycle of about that length. At the very least, it’s sure to spawn (more) “interesting” theories on some sites (eg, wattswrongwiththis?)

    Large-scale Non-axisymmetric Magnetic Fields on the Sun
    and Cool Stars

    S. V. Berdyugina
    Institute of Astronomy, ETH Zentrum, 8092 Zurich, Switzerland
    Astronomy Division, PO Box 3000, 90014 University of Oulu, Finland

    Major spot activity on the Sun alternates the active longitudes in about 1–3 years, which results in flip-flop cycles of 3.8 and 3.65 years in the Northern
    and Southern hemispheres respectively.

    I belive the following paper was the first suggestion of the cycle
    “Persistent active longitudes on the Sun”
    S. V. BERDYUGINA and I. G. U SOSKIN

    Applying a new sun-as-a-star approach to solar observations for the last 130 years we showed that sunspots are formed preferably in two migrating active longitudes 180 apart, which are persistent through the entire studied period. Their migration in Carrington longitude is determined by changes of the mean latitude of sunspots and the differential rotation.
    The two longitudes periodically alternate the dominant activ-ity with about 3.5-year period similar to the”flip-flop” phe-nomenon known in starspot activity. Our results imply that a large-scale toroidal field regenerated during the solar cycle is essentially non-axisymmetric and oscillates with the three-
    fold solar cycle frequency.

  • sky // November 15, 2008 at 6:16 pm

    By George, you’ve got it. Anyone who thinks our sun has less influence on climate than trace gases is the ultimate denier.

  • Hank Roberts // November 15, 2008 at 7:15 pm

    > our sun has less influence on climate
    > than trace gases …

    Too general a statement. Consider the ozone layer; trace gases made a far greater change in both the upper atmosphere and the ground-level UV measurements than any possible change in the Sun, including the worst solar flares.

    It took high altitude nuclear bomb tests to make a similar change — and that was a brief transient rather than a change that’s lasted decades.

    Try being a bit more specific about whatever it is you can’t believe possible, eh? Maybe you’re right, if you can find the specific cite for it.

  • dhogaza // November 15, 2008 at 8:07 pm

    Anyone who thinks our sun has less influence on climate than trace gases is the ultimate denier.

    No one thinks this. We all know that without the sun, we’d freeze our sweet bippy off. No source of warmth at all other than geological sources.

    However, variability in solar output is another thing, and, no, there’s no evidence that such variability is dominating the climate today.

  • dko // November 15, 2008 at 8:35 pm

    george — Sunspot activity is interesting but I don’t think that is the answer. Just look at TSI since the 1970s and you immediately see the 10.8-year solar cycle. Is there another cycle at 1/3 that? Perhaps, but it would have to be puny. I can’t even get a good correlation between the 10.8-year cycle and temperature anomaly…and that cycle jumps off the screen at you.

    Hank — Thanks for the updated Google search phrase. The AMS link has a plausible connection between oceans, if I can follow the jargon-dense prose. Pretty well dashes the red-backed vole hypothesis I’ve been working on…but I’ll bet Pasteur didn’t get it right the first time, either. :-)

  • Ray Ladbury // November 15, 2008 at 9:09 pm

    Sky, Uh, well, when the Sun isn’t changing…

    Maybe if you actually looked at the data, you might understand this.

  • David B. Benson // November 15, 2008 at 9:16 pm

    george // November 15, 2008 at 5:08 pm — Thanks, but this does not appear to have the precision of the sharp peak at 3.6 years. It is this precision which is unusual; most weather and climate phenomenon are nowhere near this precise unless astronomically driven.

    So somewhere there has to be a very good resonator to produce the metronome like beat seen in a variety of data from various parts of the world.

  • llewelly // November 15, 2008 at 9:27 pm

    I didn’t read most of the thread - since so much of it is off-topic copy-and-paste-ranting about the sept-for-oct GISS temp error that was fixed.
    So if I missed someone else bringing this up - I apologize. ESRL has data from CO2 monitoring stations other than Mauna Loa available, and a global composite as well. Among other things, the character of the annual variations differs somewhat between the NH and the SH.

    Does the unusually large increase in the most recently deseasonalized value show up at the other monitoring sites? What about the 3.62 yr periodicity? I’m guessing the others show but if they’re different in character it could be a clue.

    [Response: I've kept my eye on the global composite numbers, but they haven't posted October's value yet. The others you link to are only updated through the end of 2006.]

  • David B. Benson // November 15, 2008 at 9:34 pm

    “Finally, an important property of oceanic Rossby waves is that they possess a thermodynamic signature in the surface temperature field, by which they may have some climatic impact. The thermodynamic signature of Rossby waves can be caused by several processes, including the advection by Rossby wave velocity field accross mean temperature gradients, and the coupling with the mixed layer.”

    from

    http://earth.esa.int/workshops/venice06/participants/1146/

  • dko // November 15, 2008 at 9:47 pm

    Back to the CO2 blip…is there an oceanographer in the house?

    Is it possible that a large gyre of seawater parked itself upwind of MLO and has been warming and outgassing CO2?

    Still no notice in their changelog, but here is a link to their quality-control process. They throw out nearly half their data:

    http://www.esrl.noaa.gov/gmd/ccgg/about/co2_measurements.html

  • george // November 15, 2008 at 9:55 pm

    Sky: Got what?

    I posted that as a curiosity more than anything.

    I don’t think anyone denies that the sun has an effect on climate.

    The question is “how great?”

    But more to the point of this post: a similar period in a solar cycle may have no relevance whatsoever to the earth, though it is curious that the period seems to be so close in this case.

    Of course, that assumes that both cycles are actually real — not to mention the fact that numerology is actually full of such coincidences. I’d be the first to admit that it may (probably does?) mean nothing.

    The period similarity says nothing at all about a mechanism for a solar link to say nothing of the magnitude of the effect (if there is one).

    I don’t recall Tamino even saying anything about the magnitude of the effect of the 3.6 year cycle in CO2 or temperature (relative to the overall change in both over the past 3 decades).

    Also, unless there were some sort of resonance involved, it’s really hard to see how a 3.6 year cycle would lead to a fairly constant increase in temperature over several decades.

    But it’s not an either/or a situation: either CO2 causes warming or the sun does.

    It is quite possible that both influence climate, but to different degrees (if you pardon the pun).

  • David B. Benson // November 15, 2008 at 10:02 pm

    Aha. From J. Phys. Oceanography:

    Long Rossby Wave Basin-Crossing Time and the Resonance of Low-Frequency Basin Modes FRANCOIS PRIMEAU
    ¸
    Canadian Centre for Climate Modelling and Analysis, Meteorological Service of Canada, University of Victoria, Victoria, British Columbia, Canada

    ABSTRACT
    The ability of long-wave low-frequency basin modes to be resonantly excited depends on the ef?ciency with which energy fluxed onto the western boundary can be transmitted back to the eastern boundary. This ef?ciency is greatly reduced for basins in which the long Rossby wave basin-crossing time is latitude dependent. In the singular case where the basin-crossing time is independent of latitude, the amplitude of resonantly excited long-wave basin modes grows without bound except for the effects of friction. The speed of long Rossby waves is independent of latitude for quasigeostrophic dynamics, and the rectangular basin geometry often used for theoretical studies of the wind-driven ocean circulation is such a singular case for quasigeostrophic dynamics. For more realistic basin geometries, where only a fraction of the energy incident on the western boundary can be transmitted back to the eastern boundary, the modes have a ?nite decay rate that in the limit of weak friction is independent of the choice of frictional parameters. Explicit eigenmode computations for a basin geometry similar to the North Paci?c but closed along the equator yield basin modes suf?ciently weakly damped that they could be resonantly excited.

    Full paper:

    http://kuroshio.ess.uci.edu/download/Primeau02b.pdf

  • dko // November 15, 2008 at 11:05 pm

    http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_update/sstanim.shtml

    For those keeping score at home, it’s Voles, 1; Gyres, 0.

  • hswiseman // November 16, 2008 at 3:30 am

    Not the first outlier of the year.

    Sent to Dr. Tans 4/5/08

    Dear Dr. Tans:

    Looking at a 4 month winter period, December to March, MLO Monthly CO2 Trend Data, for 50 years,
    I identified the following:
    11 Winter Periods showed concentration unchanged or declining.
    WINTER PPM
    _____________
    07-08 -.30
    88-89 -.02
    85-86 -.09
    83-84 -.10
    74-75 -.26
    71-72 -.35
    70-71 -.04
    67-68 -.26
    66-67 -.07
    59-60 -.00
    58-59 -.01
    This year’s decline is the largest since Winter 1971-1972, is the second largest in 50 years, and the first winter period decline in concentration in 19 years. Average 50 Year Concentration Increase .366 PPM. I suspect this is a La Nina effect as well as lower global SST generally. Then again, it might be nothing at all.

    Dr. Tans sent a prompt, thoughtful, polite response. I never post any responses I receive in private correspondence with researchers who are kind enough to answer a layman’s email.

    On 4/11/08, I wrote back
    Dear Dr. Tans:

    Thank you for taking the time to respond. Of course it is highly unlikely that the long term trend in CO2 will be derailed, and in fact it will be interesting to see if the additional dissolved CO2 from this period contributes to a spike when SST rises. The 4 month event in question probably represents a standard deviation in the 3-4 range, and as such is worthy of note. Given the persistent and strong la Nina, I wonder if this is another signal of a definitive move towards a negative PDO. If SST’s remain suppressed, the question of how much additional CO2 is absorbed before some saturation equilibrium is achieved remains unanswered.

    There are very few good continuous longitudinal climate records, so thank you and the ML staff for building a clean observational record.

  • mauri pelto // November 16, 2008 at 12:18 pm

    Looking at dko’s link to Mauna Loa data, it is amazing to see the seasonal variation in the fluctuations of CO2 with altitude at the Wisconsin site. http://www.esrl.noaa.gov/gmd/ccgg/about/co2_measurements.html#variations
    Of more importance the global signature of el nino indicates it is the best choice for the 3.6 year interval, there is no other short term ocean-atmosphere cycle we have observed that is global. The cycle is not consistent in strenght or frequency but the 37-39 cycles since 1865. Do indicate a 3.5 to 3.7 year cycle. http://www.dar.csiro.au/information/soi.html

  • sky // November 16, 2008 at 4:40 pm

    Just a brief heads-up, one that I don’t intend to argue over with true believers in AGW:

    The solar “constant” varies by a few w/m^2, in step with the sunspot cycle (~11yrs). The record is puctuated by frequent downward spikes of several w/m^2 due to to intense sunspot groups. The third harmonic (~3.7yrs) is quite prominent in the solar radiance spectrum. Inasmuch as insolation is what powers the entire climate and bio-system , it should come as no suprise that this frequency should appear widely in many earthly records.

    GHGs produce no power of their own, providing only a minor contribution to the system heat capacitance, which is dominated not by trace gases or even water vapor in the unretentive atmosphere, but by the retentive oceans. Changes in net downward LW radiation are balanced in the climate system by countervailing changes in moist convection and cloud formation. End of story.

    [Response: You know what would be really nice? If you (and other members of the church of bury-your-head-in-the-sand) could get your facts straight.

    The amplitude of the variations in solar output due to the solar cycle is about 1.05 W/m^2, not "a few w/m^2." And that corresponds to a climate forcing of about 0.18 W/m^2.

    The third harmonic is NOT "quite prominent" in the power spectrum of solar irradiance. Its amplitude is no more than 0.12 W/m^2, corresponding to a climate forcing of 0.02 W/m^2.

    As for your ludicrous argument that the main impact of greenhouse gases is a negligible contribution to heat capacity -- you'll need several years education just to get to the point that we could discuss it intelligently.

    End of story.]

  • dko // November 16, 2008 at 5:06 pm

    A sharp-eyed reader caught an error in the October Vole data posted above. The corrected numbers are: Voles, 0; Gyres, 0.

    We regret any confusion (or sudden clarity) as a result of this error.

  • Lazar // November 16, 2008 at 6:14 pm

    sky,

    Changes in net downward LW radiation are balanced

    No, the lapse-rate feedback is small compared to the water vapor feedback, and the cloud feedback is positive. If feedbacks were balancing the change in forcing, we would not have a temperature rise.

  • Ray Ladbury // November 16, 2008 at 6:29 pm

    Gee, Sky, so why is Earth 33 degrees warmer than it would be without an atmosphere given solar iradiance? I eagerly await your latest piece of science fiction.

  • Barton Paul Levenson // November 16, 2008 at 6:36 pm

    sky says:

    Changes in net downward LW radiation are balanced in the climate system by countervailing changes in moist convection and cloud formation. End of story.

    So basically the temperature at the Earth’s surface can’t vary? That’s a relief. Those signs of ice ages must have been misinterpreted.

  • Lazar // November 16, 2008 at 7:15 pm

    sky,

    Heat capacity is a temperature rise per unit of energy added to the system. The GHG effect is the ability of the system to retain energy.

  • dko // November 17, 2008 at 12:34 am

    sky,

    Since late 1978 satellites have provided accurate TSI data over three complete solar cycles. During this time, solar output has trended down slightly while global temperature has risen markedly.

    How do you account for this?

  • sky // November 17, 2008 at 1:24 am

    Thank your very much for helping me win a bet! Your predictably cavalier response, replete with erroneous values of solar radiance variations and climate forcings–quoted to a hundredth of a w/m^2– unmistakably shows the scientific maturity of your views. Small wonder that the importance of moist convection in the actual thermodynamic balance escapes you, or that some your avid readers can’t even distinguish between equilibrating changes in the system and the indisputably salutary “greenhouse” effect itself. That this effect depends upon SW thermalization by the Earth, which would be grossly different in the absence of oceans, seems entirely beyond your collective ken.

    Having survived unauthorized access to my disc subsequent to the last posting, I have no interest in visiting your blog again.

    [Response: Thanks for proving to everyone with even half a brain that not only are you misinformed, you're unwilling to learn.

    As for your promise never to return -- your departure is a welcome relief.]

  • Ray Ladbury // November 17, 2008 at 2:10 am

    Sky, Thanks for the laugh. I always did like slapstick.

  • Phil Scadden // November 17, 2008 at 3:52 am

    Sky represented a worrying phenomenon though. Clearly he/she has some sort solar-lag model for GW and having got it sorted in brain, believes it sufficiently to ventures into the lions den. When this is immediately challenged, then retreats rather than discuss. Sky does not want to know the truth - he/she just wants to believe so I suspect visits disinformation sites where this kind of view gets supported. 7-day creationists are other group with this mindset.

    And then there is the amazing veiled accusation that Tamino or the site admins used profile information to hack his/her hard disk! Paranoid into the bargain.

    Reverse paranoia - is sky a exxon-paid PR employee cunningly trying to poison the blog?

    (and no, I dont really believe that. I’ve dealt with enough people like Sky to think that former is much more likely and frankly a lot more worrying).

    [Response: Or maybe just a troll.]

  • Ray Ladbury // November 17, 2008 at 1:54 pm

    I vote moron.

  • george // November 17, 2008 at 6:49 pm

    Having survived unauthorized access to my disc subsequent to the last posting..”

    I’m sure the Robinson family of “Lost in Space” could commiserate with Sky, having suffered the same fate with the stow-away Dr. Smith.

  • David B. Benson // November 17, 2008 at 7:36 pm

    One more piece of the mystery possibly explained:

    Detection of Rossby Waves in Ocean Colour Data
    Paolo Cipollini, Peter G. Challenor, David Cromwell, Graham D. Quartly and Stefano Raffaglio

    Full paper:

    http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=00858134

    offers as a possible source the bringing of nutrients nearer the surface by the action of Rossby waves. This could be the cause of the 3.6 year period oscillation in the Keeling curve.

  • David B. Benson // November 17, 2008 at 10:34 pm

    Hypothesis: North Pacific Rossby waves cause 3.6 year period signals.

    (1) The North Pacific basin acts as a resonant chamber supporting eastbound Rossby waves and the return westbound Kelvin waves. This can happen at a period of 3.6 years.

    (2) From computational experiments, North Pacfic Rossby waves affect the timing of El Nino. The result is seen in El Nino indicies.

    (3) Both directly and through El Nino effects, the temperature is forced with a 3.6 year period. Via teleconnections this same period appears in Atlantic SST measurements.

    (4) Phytoplankton productiion is effected, possibly by the Rossby waves, certainly by El Nino. The result is that the 3.6 year period is found in the Keeling curve of atmospheric carbon dioxide.

    Note: The csiro link provided by mauri pelto // November 16, 2008 at 12:18 pm justs times-out for me today, but I take the point of his post to be that either the North Pacific Rossby wave in question is not always present or does not always affect El Nino.

  • John L. McCormick // November 18, 2008 at 2:08 pm

    Tamino,

    OH radical reaction with CO produces H and CO2.

    The global increased fossil fuel consumption puts more CO into the atmosphere and the radicals do their job. The product CO2 hangs around to be measured.

    Not an answer; but a part of the puzzle.

    John McCormick

  • John L. McCormick // November 18, 2008 at 2:29 pm

    RE: radical movements

    CH4 (on the increase globally) combining with OH yields CO which reacts with OH to form more CO2.

    John McCormick

  • David B. Benson // November 19, 2008 at 11:04 pm

    Brief followup to my November 17, 2008 at 10:34 pm comment — The csiro link worked; there is a graph of a n El Nino index, beginning in 1866 CE. Visual inspection shows there is almost always an effect which I take as about 3.6 years.

    What is a bit surprising is the precision of this peak in the Keeling curve; perhaps that is an artifact of the processing used to find it?

  • tamino // November 20, 2008 at 1:29 am

    The change log indicates that the October 2008 estimate actually was in error; details are in an update to the post.

  • HankRoberts // November 20, 2008 at 2:19 am


    19 November 2008

    We corrected an error in the data posted for the previous October on 3 November 2008. The assigned CO2 values for the reference gas mixtures in use at our Pt. Barrow Observatory, Alaska, had mistakenly been used in calculating CO2 mole fraction values for the Mauna Loa data, resulting in a value for October that was ~1.5 ppm too high. The error became apparent because of an abrupt shift in the comparison between measurements of flask air samples taken at Mauna Loa, but analyzed in Boulder, Colorado and the continuous analyzer operating at Mauna Loa. There is always a delay of several weeks between the collection and analysis of the flask air samples.

    Stop the presses! … oh, wait …. this is normal.

  • dko // November 20, 2008 at 2:49 am

    Good catch, Tamino.

    Now if we could only figure out why January 1893 was so abnormally cold in the temperature records….

  • Bob North // November 20, 2008 at 5:13 am

    Hank - The error was not normal, but not uncommon nor of particular concern. What is exceptional is the prompt, forthright explanation detailing exactly what the error was and how it happened. No excess verbiage or commentary. Kudos to Dr. Tans and his crew!

  • tamino // November 20, 2008 at 12:46 pm

    What is exceptional is the prompt, forthright explanation detailing exactly what the error was and how it happened. No excess verbiage or commentary…

    This is not exceptional, it’s the norm. It’s all we’d have gotten from GISS concerning the recent error in October temperature averages, if they hadn’t been crucified in the blogosphere for someone else’s mistake which was not of particular concern.

  • Hank Roberts // November 20, 2008 at 4:06 pm

    C’mon, Bob North. Everyone reading in the sciences knows about reading for corrections, both in published data and once the work gets to the journal stage. Indexes and the online sites always give notice when corrections are posted.

    You can find corrections and updates files _everywhere_ in science, both paper and online.

    E.g.
    http://scholar.google.com/scholar?&q=%22data+file%22+FTP+%2Bcorrections+errors&btnG=Search

    Where did you get this notion? Some guy on a blog telling you it’s a tsk!andalous problem?

  • Bob North // November 20, 2008 at 5:30 pm

    Tamino & Hank - Actually my comment could apply to more than just the recent NOAA/GISS kerfuffle. Certainly there are those on the skeptic/contrarian/denier (choose your term) side that belabor the whys and hows of how they made a mistake (if they ever own up to one) trying to point the finger at getting bad info from others.

    On the specific GISS issue, I posted early on at realclimate that I was of the opinion that it was more of a molehill than a mountain, so no, I don’t believe it is tsk!andalous. Going back and re-reading Gavin’s post, the first 2 paragraphs are fine and probably all that were needed other than a brief acknowldegement that the error was caught by a third party (no need to say who) and that GISS would be looking at additional QA/QC measures to help prevent such issues from re-occurring. Everything else in that post and nearly all of the comments were excess verbiage and commentary.

    Tamino - There is no doubt that the original error occurred in NOAA’s processing of the temp data from various parts of the world. However, if I use data from others to develop my own report, it becomes my resposnisbility to review that data for accuracy. Bad data in my report are my error regardless of where I got the data.

  • Bob North // November 20, 2008 at 5:33 pm

    P.S I take responsibility for the obvious typos in the above post.

  • geologyjoe // November 20, 2008 at 6:29 pm

    up and up and up.
    we’d better start doing something beyond pilot testing PDQ.
    Methinks those StadtOil people just might have part of the solution.

  • HankRoberts // November 20, 2008 at 8:18 pm

    Ps, Tamino, has the graph posted at the top of the thread been revised now at the original source?

    IF so might want to add a link to the source page.
    I doubt it’s noticeable!

    [Response: I believe that yes, it has been revised. I was able to notice, but probably only because I've examined the precise issue.]

  • Andrew Dodds // November 24, 2008 at 3:57 pm

    Just out of interest, I had a look at what your Just For Fun prediction for 2008 was.. and it looks spot on (local minima in a rising trend..)

    [Response: I still don't put much stock in the "just-for-fun" prediction. If it's correct, there'll be a warm 2009 and 2010 -- we'll see!]

Leave a Comment