Did Napoleon Use Hansen's Temperature Data?

It’s colder in Russia in October than in September, as Napoleon found out to his cost in 1812.

Sitting in the ashes of a ruined city without having received the Russian capitulation, and facing a Russian maneuver forcing him out of Moscow, Napoleon started his long retreat by the middle of October.

Flash forward almost 200 years later. NASA has just reported record warmth in October throughout Russia, with many sites experiencing similar temperatures in October as in September – perhaps the sort of situation that Napoleon had hoped for (not similar as anomalies, but similar in actual temperatures in deg C.)

Actually, many stations didn’t just experience similar absolute monthly temperatures. Many stations had exactly the same monthly temperatures in October as in September. Here are the last three years for the Russian station, Olenek, showing NASA GISS monthly temperatures (in deg C) bolding Sept and Oct 2008. October 2007 had an average temperature of -9 deg C, as compared to 3.1 deg C in Sept 2007. October 2008 had the identical temperature as September 2008.

YEAR JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
2006 -34.0 -29.9 -23.5 -18.1 1.6 10.6 16.9 11.5 4.4 -14.6 -27.7 -29.1
2007 -27.9 -41.5 -21.6 -4.0 0.1 12.4 13.5 11.3 3.1 -9.0 -24.8 999.9
2008 -30.0 -29.4 -19.6 -13.4 1.3 12.0 13.1 12.1 3.1 3.1 999.9 999.9

This exact match of October 2008 to September 2008 was repeated at many other Russian stations. A CA reader notified me of this phenomenon earlier today and I’ve confirmed for myself that the information is accurate. Based on what he described as a “Cursory” look, he sent me the following list currently “updated” stations that exactly replicate the Sept data: Almaty, Omsk, Salehard, Semipalatinsk, Turuhansk, Tobol’sk, Verhojansk, Viljujsk, Vilnius, Vologda. I can add Hatanga, Suntora, GMO ImEKF. Not all stations were affected – Dzerszan, Ostrov Kotal, Jakutsk, Cokurdah appear to have correct results.

Let’s consider the opposite situation. Suppose that March temperatures had been inadvertently carried forward into April, yielding a massive cold anomaly in Russia. One feels that Hansen would have been all over the opposite error like a dog on a bone – he would have been his own bulldog.

In any event, we here at Climate Audit are always eager to assist NASA. On earlier occasions, we helped identify the lost city of Wellington, New Zealand, where NASA has been unable to locate climate records for nearly 30 years.

Today, we are able to provide NASA with up-to-date weather reports confirming that October in Russia is colder than September. Verhojansk temperatures are conveniently online here and temperatures are currently a nippy -18 deg C.

You’re welcome, Jim.

156 Comments

  1. DeWitt Payne
    Posted Nov 10, 2008 at 10:25 PM | Permalink

    Just look at the October 1 and October 31, 2008 Arctic images from the CT site. Essentially no snow cover in Siberia on October 1 and full snow cover on October 31. Somehow that doesn’t exactly jibe with the same temperature in October as September. While the Arctic Oscillation may have kept as much cold air from moving south as usual, it still was cold enough to get covered with snow.

  2. Gerald Machnee
    Posted Nov 10, 2008 at 11:10 PM | Permalink

    The climate has become UnReal.

  3. Steve McIntyre
    Posted Nov 11, 2008 at 12:14 AM | Permalink

    This issue is being covered at Anthony Watts.

    The Irish data is screwed up in the same way as the Russian data.

  4. Posted Nov 11, 2008 at 12:25 AM | Permalink

    And over here in Blighty the official Central England Temperature Record records an average for October 1951-80 as 10.52 degrees and Oct 2008 as 9.7 which according to the map is a degree or more warmer…

  5. Patrick Henry
    Posted Nov 11, 2008 at 12:44 AM | Permalink

    Weather Underground shows the mean October temperature for Olenek at -7C, and 2C for September. Hansen missed by 9C.

    http://wxex.wunderground.com/history/station/24125/2008/9/1/MonthlyHistory.html

    http://wxex.wunderground.com/history/station/24125/2008/10/1/MonthlyHistory.html

  6. GeneII
    Posted Nov 11, 2008 at 12:48 AM | Permalink

    that “guy from Canada”, “he whose name must not be mentioned” is back on it again!

  7. Hank
    Posted Nov 11, 2008 at 12:52 AM | Permalink

    Presumably we have an error here we are talking about. This highlights why it is so wrongheaded to become defensive and play “dog in the manger” with one’s methods and data. It is understood that one takes a stand on scientific questions. It is why the double blind study was developed. Researchers have a stake in the “game” and it is only natural that certain results put a gleam it the eye of those conducting researches. Just as a clinician conducting a medical trial will tend to inadvertently advance certain outcomes, so too a researcher handling data will find the errors in his data that displease him and overlook those errors that do not help his hypothesis.
    Jim Hansen of all people should understand that data belongs in usufruct to others who can use it in a fruitful way.

    http://www.columbia.edu/~jeh1/mailings/20070816_realdeal.pdf

  8. Posted Nov 11, 2008 at 12:57 AM | Permalink

    My take on the whole sorry mess is here.

    Cheers,

    Simon
    Australian Climate Madness

  9. Jean S
    Posted Nov 11, 2008 at 1:36 AM | Permalink

    This is apparently the same error some Finnish stations had in the September revision (July/August temperatures) as I reported here. Those seems to have been corrected now. Unless those were corrected by hand (and no credit given to CA), I think it is plausible to think that the error lies somewhere in GHCN data collection algorithm.

  10. Nylo
    Posted Nov 11, 2008 at 2:27 AM | Permalink

    If I was in charge of managing and publishing these data and saw a coloured graph like the one at the begining of Steve’s post, with that huge dark red area, my first reaction would be to verify if something went wrong there. Who is verifying anything at NASA? Don’t they wait until about the 10th of each month precisely to be able to organise the information and correct posible errors for the month before? When they see many stations recording monthly temperatures which are more than 10ºC hotter than the average for that month, shouldn’t they ask themselves why they didn’t hear any of that in the news? Don’t they raise an eyebrow when the anomalous temperature happens to be exactly the same as the month before? Or do they just get so carried away with the red colour that they go out to celebrate? O_O

  11. Ubique of Perth
    Posted Nov 11, 2008 at 2:43 AM | Permalink

    What makes people think it was a screw-up?

    Hansen’s just as likely to have pulled this stunt deliberately, knowing the MSM would fall for it hook, line and sinker; and that any protests would be dismissed as coming from pyjama-clad bloggers in the pay of big oil and GW Bush.

    • Jean S
      Posted Nov 11, 2008 at 3:56 AM | Permalink

      Re: Ubique of Perth (#12),
      The fact that this happened earlier suggests that this is likely an error in some algorithm (not a one time cut&paste problem). Also, I’d like to point out that the error is in NOAA’s GHCN data, so I think it is unfair to blame Hansen (or GISS).

  12. Richard Steckis
    Posted Nov 11, 2008 at 2:52 AM | Permalink

    You can add Kaunas, Archangelsk, Vytegra, Vononez, Sortavala, Verlojamsk and believe it or not MOSCOW.

  13. Posted Nov 11, 2008 at 3:00 AM | Permalink

    The current temperature in Olenek is -28 deg C, about 31 degrees above the “October” temperature. Dozens of other observations, links, and jokes are summarized in my text about this topic:

    http://motls.blogspot.com/2008/11/rss-msu-0013-deg-c-month-on-month.html

  14. Posted Nov 11, 2008 at 3:00 AM | Permalink

    Sorry, I mean “below”, not “above”.

  15. widmerpool
    Posted Nov 11, 2008 at 3:19 AM | Permalink

    As someone in Moscow I don’t find it hard to believe. The temperature here has been about the same for the last three months. If you didn’t know you would never believe it’s November.

  16. John Finn
    Posted Nov 11, 2008 at 3:49 AM | Permalink

    Re: #4

    And over here in Blighty the official Central England Temperature Record records an average for October 1951-80 as 10.52 degrees and Oct 2008 as 9.7 which according to the map is a degree or more warmer…

    Most UK stations are screwed in the same way, i.e oct=sep. However at least one station is just wrong (warmer) with no obvious reason why.

  17. Vincent Guerrini Jr
    Posted Nov 11, 2008 at 4:09 AM | Permalink

    #19 Well it hasn’t been changed. I wouldn’t be surprised if they just left it at +.86C. GISS is not a reliable global temp data resource as has been demonstrated by numerous previous adjustments to data ALWAYS favouring AGW up and up…

    • Jean S
      Posted Nov 11, 2008 at 4:28 AM | Permalink

      Re: Vincent Guerrini Jr (#20),
      Of course it has not been changed, it was spotted less than 24h ago. The previous appearance of the error that was spotted (August = July in some Finnish stations) has been fixed.

      The key question here is if the GHCN algorithm will automatically correct for the error later. For instance, this might be something stupid like that the earlier value is copied if the data is not available yet, but when data is available, it will be assigned to the correct value. If it not automaticly corrected, then we have a serious problem (GHCN data is likely corrupt). So let’s wait until we’ll see where the problem lies before making any far-reaching conclusions. In any case, this does not seem to be a mistake by GISS or Hansen, so they shouldn’t be blamed for more than not detecting the problem.

  18. Urederra
    Posted Nov 11, 2008 at 4:20 AM | Permalink

    Maybe is a distraction maneuver from NOAA with the hope that we will not notice that the North American subcontinent is cooler. (Slightly cooler, I would say, but I am eyeballing)

  19. Harry Snape
    Posted Nov 11, 2008 at 5:24 AM | Permalink

    “In any case, this does not seem to be a mistake by GISS or Hansen, so they shouldn’t be blamed for more than not detecting the problem”

    Give me a break! If I was about to publish a record breaking global temperature and the main reason was because of a 12 degree anomaly in a large region I’d do a hand check of the data. It’s what a sensible person would do. That a research institute of Goddard’s supposed standing in the climate science world can make such an egregious (I’m sure Santer will agree ;-)) mistake is laughable. I guess “conservative and rigorous” methods were used – just like in the past. Could it be that Climate Science might actually learn a lesson about the value of external review of data and methods? Somehow I doubt it.

  20. Phillip Bratby
    Posted Nov 11, 2008 at 6:00 AM | Permalink

    “In any case, this does not seem to be a mistake by GISS or Hansen, so they shouldn’t be blamed for more than not detecting the problem”

    Anybody who publishes data without doing a thorough check of it has made a mistake and is to blame, whether by negligence or by not having good procedures to prevent errors (and even worse for repeat errors). Publish and be damned!

  21. Posted Nov 11, 2008 at 6:05 AM | Permalink

    Always, without exception, Verify the calculations.

  22. Posted Nov 11, 2008 at 6:33 AM | Permalink

    I wonder if their practice is, in the absence of data, to populate the blanks with the prior month’s values. If that’s the case, then it’d seem that their subsequent corrections would trend in one direction in boreal spring and in the opposite direction in the fall.

    As has been said, the remarkable thing is that Hansen’s folks released this with no one questioning such a record warm month. Had it been record cold, then I suspect, as someone said, the report would have been delayed until the blanks were properly filled.

  23. BarryW
    Posted Nov 11, 2008 at 6:57 AM | Permalink

    Are the Antarctic hot spots at Russian sites?

  24. John Goetz
    Posted Nov 11, 2008 at 7:18 AM | Permalink

    The NOAA error seems to be with the processing of the .dly files. I did a spot check of a couple Russian sites that have the September / October twins at GISS. The NOAA .dly files show a clear difference in temperature. The resulting NOAA GHCN v2 file, however, contains the twins.

  25. Posted Nov 11, 2008 at 7:22 AM | Permalink

    The reported average 2008 october anomaly over Russia/Sibiria according to GISS was about +12° C (According to the map legend: 4° – 13.7° for the giant dark red blob there)

    Almost as conspicuous is the look of the 2008 sepmtember anomaly for the antarctic region. There, the giant red blob says 4° to 9.6° over the better part of antarctica.

    Can this really be true? The zonal mean in september shows high anomalies for both polar regions.

    • Jim Pfefferle
      Posted Nov 11, 2008 at 11:04 AM | Permalink

      Re: Jonas N (#29),
      Along the lines of Jonas N’s comment, a 12 degree anomaly for a few days in a specific area is newsworthly, but expected to regularly occur. But 12 degrees in that place for an entire month approaches incredible. Put it on the scale of a continent sized area for that period, the word ‘implausible’ should be applied.

      This leads me to the anomaly color scale that appears for those pretty NASA pictures (WTF, Oct/08 is no longer available). There should be some range (perhaps plus/minus 4 or 6C) where any data becomes highly suspect and should be automatically ruled an outlier until further verification in performed.

      Please ignore my climate ignorance, but I wouldn’t want to based a structural design on data that looked like this.

  26. Flashman
    Posted Nov 11, 2008 at 7:31 AM | Permalink

    snip

    The faulty data is eyeballed by quite a few people before being published, yet nobody cared to double-check the obviously absurd content. The reason is simple, but serious: at GISS as well as NOAA, any action or statement that can possibly be intepreted as “heretic” would probably not be a good career move. Ignoring blatantly obvious errors is not a problem, however…

    It’s tragic, and is anathema to the basic tenets of proper scientific inquiry, but also quite amusing to watch, as the level of absurdity increases.

  27. Posted Nov 11, 2008 at 7:54 AM | Permalink

    From NOAA website:

    Contacts
    For questions/assistance with GHCN-Monthly data, please contact Russell.Vose@noaa.gov

    For assistance with the GHCN-Monthly web site, please contact Jon.Burroughs@noaa.gov

  28. Stuart Harmon
    Posted Nov 11, 2008 at 8:11 AM | Permalink

    Dear Steve

    The following web site is run by Professor Ole Humlum of Oslo University. It provides nuch historical information, news reports etc. It takes no position on climate change.

    I hope it is of use to you or your readers.

    http://www.climate4you.com/

  29. Håkan B
    Posted Nov 11, 2008 at 8:15 AM | Permalink

    Vilnius and Kaunas are located in Lithuania, so it seems not to be a russian only error.

  30. Steve Geiger
    Posted Nov 11, 2008 at 8:22 AM | Permalink

    maybe this was covered, but was the data released along with some sort of press release making the typical declaration s of such ‘alarming’ warming? This, IMO, would make the lack of even basic due diligence in checking the anomalous regions even more, uh, egregious (there’s that new word again!)

  31. Posted Nov 11, 2008 at 8:40 AM | Permalink

    TIME TO ALERT THE MEDIA!
    Last time Steve found an error in GISS (Sept 07), the media did quite a good job of reporting it. It would make quite a good follow-on story that the same Canadian blogger has found another significant mistake in the same dataset.

    By the way, the list of stations where this happened (in Steve’s first post and the comments) is nowhere near complete. It’s the same story at
    Enisejsk, Krasnojarsk, Minusinsk….

    The fact that it also applies to some UK stations (Eskdalemuir, Leeming, Bournemouth…) indicates that the error is with NASA or GHCN (they can’t blame it on some Russian official sending the wrong data).

    Is anyone compiling a complete list, and saving a copy of the dodgy data?

  32. Basil
    Posted Nov 11, 2008 at 8:45 AM | Permalink

    #32

    And that web site has a great graphic of the screw-up:

  33. M. Jeff
    Posted Nov 11, 2008 at 8:49 AM | Permalink

    I informed a PhD physicist about the findings of this thread. His reply:

    The blogosphere is really doing a service by checking these data. For a mistake like this to get through indicates a serious breakdown in the management of these scientific regulatory agencies. I am stunned.

  34. Steve McIntyre
    Posted Nov 11, 2008 at 9:07 AM | Permalink

    I sent Hansen an email this morning notifying him of the error.

  35. tolkein
    Posted Nov 11, 2008 at 9:14 AM | Permalink

    In my day running a price making function for a mutual fund business (unit trusts in the UK) wew daily made prices for our funds. If we got them wrong and the fund or customers lost we had to make good any losses, on top of regulatory punishment. We had checks for anomalous price moves or non price moves and comparisons to indices and markets. In order to ensure that this was taken seriously, all price errors were reported to the Board and explanations, together with steps to ensure they did not happen again had to be given to the Board. It is incredible and a professional dereliction of duty that there are not comparable checks on data before publication, particularly in view of the sensitivity of climate data. We all know that people make mistakes but there should be processes in place to pick them up before publication.

  36. Jaye Bass
    Posted Nov 11, 2008 at 9:14 AM | Permalink

    Don’t think for a moment that this is about science. Its strictly politics with obvious incentives to be more so in the coming months.

  37. RW
    Posted Nov 11, 2008 at 9:35 AM | Permalink

    many other Russian stations…Almaty, Omsk, Salehard, Semipalatinsk, Turuhansk, Tobol’sk, Verhojansk, Viljujsk, Vilnius, Vologda – Almaty and Semipalatinsk are in Kazakhstan. Vilnius is in Lithuania.

    One feels that Hansen would have been all over the opposite error like a dog on a bone – he would have been his own bulldog. – pointless, meaningless, baseless speculation.

    • PhilH
      Posted Nov 11, 2008 at 10:02 AM | Permalink

      Re: RW (#42), This “baseless speculation” by Steve is based on years of absolutely frustrating experience with these individuals. He’s been Hansen’s “bone’ for a long time.

    • RomanM
      Posted Nov 11, 2008 at 10:29 AM | Permalink

      Re: RW (#42),

      Of course, such a conjecture would just be “pointless, meaningless, baseless speculation“. No reputable climate science group would react like that…

  38. Håkan B
    Posted Nov 11, 2008 at 9:39 AM | Permalink

    The sad thing is that those guys are not alone. Earlier this year our Swedish national bureau of statistics, SCB, issued an erratic level of inflation. The error was found to be in the prices for shoes, but before the error was discovered the Swedish national bank raised there basic interest which cost Swedish house owners hundred of millions SEK! Think of it, shoe prices! And this is a true story. One wonders where we are heading?

  39. Posted Nov 11, 2008 at 9:55 AM | Permalink

    GISS has now deleted the duff station data.
    But the incorrect 0.88 anomaly is still there in Fig.C.txt, as is the ‘red Russia’ graph.
    I wonder whether they will acknowledge the error, or try to pretend that it never happened?

  40. Demesure
    Posted Nov 11, 2008 at 10:34 AM | Permalink

    Must be “pointless, meaningless, baseless speculation” like 6 m sea level rise or hundred million “climatic refugees”.

  41. Steve McIntyre
    Posted Nov 11, 2008 at 10:38 AM | Permalink

    NASA:

    2008-11-11: Most data posted yesterday were replaced by the data posted last month since it looks like some mishap might have occurred when NOAA updated their GHCN data. We will postpone updating this web site until we get confirmation from NOAA that their updating programs worked properly. Because today is a Federal Holiday, some pages are still showing yesterday’s data.

  42. Posted Nov 11, 2008 at 10:50 AM | Permalink

    Stupid Jesters!

  43. Peter
    Posted Nov 11, 2008 at 11:11 AM | Permalink

    Is it just me, or is the same error repeated for Ireland?

    For an idea what October was like here – temperature-wise – try this on for size:

    http://news.bbc.co.uk/1/hi/northern_ireland/7696286.stm

    BTW – today is absolutely bitter!!!

  44. Chris
    Posted Nov 11, 2008 at 11:19 AM | Permalink

    #48 I had a quick look into Jan 2007 earlier.

    GISS was more consistent with other temperature records for that month i think.

    But nonetheless, I did find a station which appeared at odds with Wunderground:

    Station = Krasnojarsk
    GISS mean for Jan 2007 = -7.7C

    http://data.giss.nasa.gov/work/gistemp/STATIONS//tmp.222295700006.1.1/station.txt

    Wunderground mean for Jan 2007 = -10C

    http://www.wunderground.com/history/airport/UNKL/2007/1/11/MonthlyHistory.html?req_city=NA&req_state=NA&req_statename=NA

    Maybe someone else would know how to find out if they use the same station/location?

    In any event, urban heat effects are also very likely a factor here in the longer term:

  45. Posted Nov 11, 2008 at 11:23 AM | Permalink

    John V should install some flags in his GMST code and send it to GISS. Maybe we should pass the hat for a new laptop?

  46. Pierre Gosselin
    Posted Nov 11, 2008 at 11:25 AM | Permalink

    Like money, so managea the government data.

  47. Posted Nov 11, 2008 at 11:37 AM | Permalink

    http://data.giss.nasa.gov/gistemp/

    Latest News

    2008-11-11: Most data posted yesterday were replaced by the data posted last month since it looks like some mishap might have occurred when NOAA updated their GHCN data. We will postpone updating this web site until we get confirmation from NOAA that their updating programs worked properly. Because today is a Federal Holiday, some pages are still showing yesterday’s data.

  48. Posted Nov 11, 2008 at 11:44 AM | Permalink

    The graph with Siberia colored in Maroon is still up. Yes…. I took a screen shot. :)

  49. Pierre Gosselin
    Posted Nov 11, 2008 at 11:45 AM | Permalink

    Every Sunday evening I download a colour-coded temperature chart of North Asia and North America. I’ve just compared the September maps to the October map and, not surprsingly, the October maps are much bluer (colder) than the September maps, which are predominantly green (no frost). Based on these maps, it’s impossible for October to have been simmilarly warm as September.
    Furthermore. one only has too take a look at the Arctic sea ice growth in October and compare it to September. It had to be much colder in October in Northern Russia. Indeed Arctic sea ice recovered very quickly during October.

    Surely the results presented for October should have signaled something was terribly amiss and should have led to a serious double-check before releasing them.

  50. Posted Nov 11, 2008 at 11:45 AM | Permalink

    Oh… shoot. I should have read steve’s comment. (Oh well…. I guess if GISS can repeat September, I can repeat a quote.)

  51. Pofarmer
    Posted Nov 11, 2008 at 11:52 AM | Permalink

    I assume you guys have seen this paper on why tree ring proxies basically suck?

    The Divergence Problem and the Failure of Tree Rings for Reconstructing Past Climate

    I found it on World Climate Report

  52. Bob B
    Posted Nov 11, 2008 at 12:11 PM | Permalink

    What a joke, Gavin tries to just shake off the errors and tries to take a swipe at Skeptics–what a loser!

    http://www.realclimate.org/index.php/archives/2008/11/mountains-and-molehills/

  53. Janama
    Posted Nov 11, 2008 at 12:51 PM | Permalink

    he obviously hasn’t read the replies to his attack on Michael Duffy in the Sydney Morning Herald.

    http://www.smh.com.au/letters/?page=fullpage#contentSwap1

  54. Bruce
    Posted Nov 11, 2008 at 12:56 PM | Permalink

    Has anyone noted that a station like Vitim has about 30% entires of 999.9 since 1990.

    http://data.giss.nasa.gov/work/gistemp/STATIONS//tmp.222300540001.1.1/station.txt

    2003 and 2004 have only 9 valid months data out of 24.
    Is that for real?
    Were the values just dropped by another error moving the data from GHCN to GISS?

    • Demesure
      Posted Nov 11, 2008 at 3:01 PM | Permalink

      Re: Bruce (#65), The GISS database is marred with such “anomalies”: for example, there is no data (9999) for Jan & Feb & Mar 1991 for ALL French stations and most Spanish stations. Come on, their temperature is crap.

  55. Posted Nov 11, 2008 at 1:22 PM | Permalink

    As I see it, the problem is that this error was very easy to spot. The British Islands colored in red on the GISS map? Come on. Even if the CET cool anomaly for October hadn’t been published a while ago, those of us weathernuts checking the weather around the world knew that the Britons are experiencing an abnormally cold Autumn.

    If a layman can immediately see that an anomaly reported by GISS must be wrong but they can’t realize it before publishing it, what trust can one have in their diligence? How many less obvious errors have slipped though? As noted above, it may not even be a case of simple carelessness: who can believe that GISS would have let this happen if the error had rendered a big cold anomaly?

    BTW, even with the glaring errors over much of Eurasia, the reported October anomaly didn’t get to be as large as those of Jan ’07 or Feb ’98: http://data.giss.nasa.gov/gistemp/tabledata/GLB.Ts+dSST.txt

  56. Sam Urbinto
    Posted Nov 11, 2008 at 1:25 PM | Permalink

    Ah, yes. What a very scientific essay.

    Me, I’m exuberant any time there’s a transcription error.

    It’s clearly true that the more eyes there are looking, the faster errors get noticed and fixed. The cottage industry that has sprung up to examine the daily sea ice numbers or the monthly analyses of surface and satellite temperatures, has certainly increased the number of eyes and that is generally for the good.

    You’re welcome.

    As we discussed last year, there is a strong yearning among some to want to wake up tomorrow and find that the globe hasn’t been warming, that the sea ice hasn’t melted, that the glaciers have not receded and that indeed, CO2 is not a greenhouse gas.

    Who says the global mean temperature anomaly isn’t trending up, that ice in the sea doesn’t melt, that some glaciers don’t recede, or that carbon dioxide doesn’t react to outgoing long-wave infrared? Anyone know somebody that yearns strongly to wake up the next day and find out physical reality has suddenly shifted into some other dimension or what have you?

    This will undoubtedly be disappointing to many, but they should comfort themselves with the thought that the chances of this error happening again has now been diminished. Which is good, right?

    Probably more like wondering why a possibility of such an error hadn’t been fixed long ago, or if the people doing the work would have ever noticed (and then corrected it in a timely manner) such an error.

    But is it good? Sure, just like “The cottage industry that has sprung up to examine the daily sea ice numbers or the monthly analyses of surface and satellite temperatures”

    Again, you’re welcome.

    It is clear that many of the temperature watchers are doing so in order to show that the IPCC-class models are wrong in their projections.

    Why bother? It’s fairly clear already that the models are basically individually meaningless/unskillful/unphysical, and that in the whole can cover just about any answer. I think everyone is aware the anomaly trend isn’t over the 1 degree resolution the measurements are recorded. Why would anyone care if a range of models is similar in appearance?

    But I’m glad he thinks it clear the goal of the game is showing projections are wrong. I wonder what game he’s playing, it doesn’t look like the same one I’m watching. Maybe it’s easier to see the player running when you’re in the stands rather than when you’ve just been knocked out of bounds. Or that the ball cleared the net when you’re on the side of the net rather than hitting the ball.

    Instead, the anti-model crowd focuses on the minor issues that crop up every now and again in real-time data processing hoping that, by proxy, they’ll find a problem with the models.

    This assumes the real-time data is comparable to models. Wait. Anti-model crowd? Is there a club one can join if one is anti-model, or do anti-modelers just wing it as they go along?

  57. Vincent Guerrini Jr
    Posted Nov 11, 2008 at 1:25 PM | Permalink

    Looks like USA GISS/ or is it NDCD? data subjected to major fiddle as well

    http://icecap.us/images/uploads/Year_Thus_Far_Cold.pdf

    One would think this has to hit the major news centers?

  58. fred
    Posted Nov 11, 2008 at 2:11 PM | Permalink

    Evidently for GISS there is available:

    1) the station details
    2) raw temperature readings by station
    3) adjustment algorithms by which raw station readings are changed
    4) averaging algorithms by which the global anomaly is generated fro the set of adjusted station readings.

    Is this a fair summary? If so does anyone know how much of this is available for the UEA material? Do we have a list of stations, a file of raw readings, and the algorithms by which the UEA global anomalies are generated?

  59. Vincent Guerrini Jr
    Posted Nov 11, 2008 at 2:14 PM | Permalink

    Amazing that real climate actually posted anything on this and allowed considerable AGW dissent!. Maybe there is a shift occurring even there? (as they run for imminent cover required LOL)

    http://www.realclimate.org

  60. Steve McIntyre
    Posted Nov 11, 2008 at 2:27 PM | Permalink

    CRU (UEA) has refused to provide any data. So it would have been impossible to pin down an equivalent error by them. Nor is there information on their algorithm.

    The GISS algorithm only became available as a result of the Y2 K embarrassment. Prior to that, it was a dark secret as well. We’re still trying to figure out their code, from time to time.

  61. Richard Hill
    Posted Nov 11, 2008 at 3:27 PM | Permalink

    Has anyone else noticed that the Mauna Loa CO2 graph showed
    a strong up-tick for October.
    Is there any connection?

    • hswiseman
      Posted Nov 12, 2008 at 10:12 AM | Permalink

      Re: Richard Hill (#73),

      The Mauna Loa site clearly indicates that their data is subject to review and revision (presumably following qc review). I have not seen any reason to think that the MLO is an agitprop organization, that they are not using appropriate instruments or calibrating them properly. MLO is a single data point on CO2 and should be reviewed in the context of all the reliable reporters. This data can be used to analyse important questions lijke ‘how well mixed is CO2?’, ‘what is the interrelationship between SST and CO2?’ etc.

      One cannot ignore the monotonic increases in atmospheric CO2 levels over the past century, and the questions it raises about environmental and climate impact. Those questions can only be answered with good science untainted by political concerns. Good science requires first and foremost good raw data. Long term climate data is highly compromised. Most of it was collected, and continues to be collected, as weather data without concern for global station-to-station consistency. Which leads us to calibration through statistics, and remote sensing. These are the areas that need the most fundemental scrutiny before you get anywhere near a GCM. What is valid raw data, how do you perform homogenization and adjustment and what are the valid statistical methods for determining trend/noise/SD. Jumping over these steps insures that the rest is a complete waste of time.

  62. M. Villeger
    Posted Nov 11, 2008 at 3:43 PM | Permalink

    This post is a gem!

  63. Ruth Dixon
    Posted Nov 11, 2008 at 3:52 PM | Permalink

    A bit off-topic, but here is a link to a depiction of Napoleon’s Retreat described as ‘perhaps the best statistical graphic ever drawn’.

  64. Pofarmer
    Posted Nov 11, 2008 at 3:56 PM | Permalink

    Craig Loehle.

    Thanks for the work.

  65. Francois Ouellette
    Posted Nov 11, 2008 at 4:02 PM | Permalink

    #67

    Gavin is right, there are people who would like to see that temperatures are not increasing, and who would love it if human emissions of GHG’s had no consequences. I, for one, would be quite happy if that were the case.

    But that’s not the point. The point is that there are also people would are adamant at blaming everything on GHG’s. There are people who are deeply convinced that humans are destroying the earth, and that using gasoline in our cars and coal in our power stations is a sin against nature. And some of these people find themselves among the scientific community.

    So it’s a good thing that if there are ideologically motivated people on one side, there are some on the other side as well. That’s just politics: a good democracy requires a good opposition.

    If AGW had no incidence on public policies, or the price of gasoline, nobody would care. But it’s a policy debate. That one side wants to use the “scientific truth” or the “scientific consensus” argument to brush aside criticism is their right. The only problem is that both history and human nature tell us that we should not buy such an argument. They can keep trying, but there will always be people to watch them at every step. It’s annoying, but that’s the price you have to pay if you pretend to save the planet.

  66. cookie
    Posted Nov 11, 2008 at 4:07 PM | Permalink

    I enjoyed Gavin’s response to a commentor at his blog.

    “I mostly agree, but you fall into a trap when you think that science has to be put into a box that is labelled ‘settled’ or ‘unsettled’, …”
    The science isn’t settled? Shurely shome mistake, Dr Schmidt? (One for UK readers)

    He continues with the priceless: ‘…or that scientists must be either ‘AGW-ers’ or skeptics – these are completely false dilemmas (sic) and lead people into making all sorts of mistakes.’
    Well, he got that right!

    But sadly then lapses with: ‘There is instead a spectrum of confidence that the IPCC tries hard to quantify (moderately successfully).’

    But finally he (partially) redeems himself with: ‘And all good scientists are sceptics (sic, assuming consistency in American spelling) – but that isn’t the point at issue at all.’

  67. Bruce
    Posted Nov 11, 2008 at 4:18 PM | Permalink

    Francois,

    there are people who would like to see that temperatures are not increasing

    I think UHI comprises all of the land based warming.

    The Satellite record essentially says it warmed a tiny bit coming out of the late 1970’s, but we are just in a long period of minor flucuations up and down (except for 1998 which is just a replay of 1934).

    It isn’t any warmer than 1934. And I doubt it is any warmer than than the Medievil Warm Period.

  68. Richard deSousa
    Posted Nov 11, 2008 at 4:20 PM | Permalink

    “But finally he (partially) redeems himself with: ‘And all good scientists are sceptics (sic, assuming consistency in American spelling) – but that isn’t the point at issue at all.'” Gavin Schmidt must be smoking something if he believes there are any “good” scientists in the AGW camp.

  69. Gerald Machnee
    Posted Nov 11, 2008 at 5:08 PM | Permalink

    Of course Lynnie at RC comes up with a real winner. As if the errors are not too high as shown by Mckittrick et al.

    Leaves one wondering about errors that lead to underestimating global warming, and whether they might be significant. That would be the real concern, not slightly underestimating it in non-sig ways, or even in sig. ways.

    • Posted Nov 11, 2008 at 5:56 PM | Permalink

      Re: Gerald Machnee (#81),
      Lynnie is half right. The potential for errors in either direction should be a matter of concern. That models might be woefully skill-less or biased should be a matter of concern. That’s true whether they are biased high or low.

      The fact that there are dangers in incorrect projections means it’s important to a) compare observations to projections focusing on observations that arrives after projections are made and b) scrutinize the the data products for problems.

  70. Steve McIntyre
    Posted Nov 11, 2008 at 5:43 PM | Permalink

    #81. Actually, I think that that sort of question is fair.

    • Gerald Machnee
      Posted Nov 11, 2008 at 5:58 PM | Permalink

      Re: Steve McIntyre (#82),
      I agree that it is fair. However, when you know what side of the fence the person is on, you can predict the response. The open minded attitude is best as you follow when doing an audit. The pieces fall where they may, either above or below. Similarly we are not sure which way the temperatures will go next year, but a certain group insists on upward. And we just saw one upward trend come crashing down with the typical response from RC.

    • Mike B
      Posted Nov 11, 2008 at 6:41 PM | Permalink

      Re: Steve McIntyre (#82),

      The statistical question is whether or not the models are biased; the sign doesn’t matter, just the magnitude.

      For the scientific questions, the magnitude and sign both matter, because they tell us whether we’ve got the feedbacks and forcings right.

      On the political/policy questions, sign is the most important, with magnitude only of secondary importance, especially so long as “significance” can be claimed.

      If only statistical and scientific questions matter, Santer would welcome all elves to help in his workshop.

    • Alan Wilkinson
      Posted Nov 11, 2008 at 7:49 PM | Permalink

      Re: Steve McIntyre (#82),

      Above post was meant to link to #82

  71. Alan Wilkinson
    Posted Nov 11, 2008 at 7:48 PM | Permalink

    Steve, global warming isn’t something you estimate, it is something you measure. If the temperature isn’t going up, then under-estimating global warming is not happening. If we don’t understand why it is not going up then we are in no position to predict what will happen next.

  72. Bruce Foutch
    Posted Nov 11, 2008 at 8:06 PM | Permalink

    Just to provide a break in your discourse and since you mentioned Napoleon, here is a link to a redraw of the famous Minard graphic of Napoleon’s march on Moscow. This graphic provides several layers of data including two that fit into GW discussions – troop numbers and temperature. Although battle losses were staggering, losses from cold weather alone were in the tens of thousands. Should be quite illuminating to those who worry the world might be getting a little warmer…

    http://www.napoleonic-literature.com/1812/1812.htm

  73. Bruce Foutch
    Posted Nov 11, 2008 at 8:19 PM | Permalink

    Sorry Ruth. Didn’t see your post about Minard’s Graphic before I hit the submit button. Don’t mean to step on your toes. Hope the one I provided shows the temps better. They were quite bitter.

    May also be pertinent. Ruth’s link was to Edward Tufte, who is an interesting guy, and who has written some excellent books on how to usefully and ethically display quantitative data. Would recommend his books and seminars to anyone (including NASA folks;-) who who use displays of quantitative information to communicate.

    • Ruth Dixon
      Posted Nov 12, 2008 at 2:00 AM | Permalink

      Re: Bruce Foutch (#89), the more mentions of Tufte the better :-) And thanks for mentioning the temperature – note that this is in degrees Reaumur.

  74. Posted Nov 11, 2008 at 9:45 PM | Permalink

    To summarize: — And PLEASE check my summary!

    Many stations/cities in Russia (and a few of its East Block neighbors) are duplicate entries, copying Sept data into October.
    Ireland repeats as well.
    Denmark numbers are duplicated as well.
    Some, but not all, of UK’s Sept values are duplicated.
    Finland’s July and August data were duplicated (since corrected.)

    So, what else should I include when I write my congressman asking that he call Hansen in to a hearing into his scientific rigor and fundamental scientific/political processing?

  75. Mark_T
    Posted Nov 11, 2008 at 10:38 PM | Permalink

    I’m curious. Hava any of these scientific sources of temperature ever made a mistake where the temperature was reported lower than it actually was?

    Seems like every time I read about an error in temperature reporting, it’s always in the positive direction. Am I missing something?

  76. anna v
    Posted Nov 12, 2008 at 12:39 AM | Permalink

    In my view either:

    a) This was deliberate to give a push to Obama’s new administration towards implementing cap and trade, with a sub clause that it could be one researcher with his/her hands on the data and strong ideology.

    or

    b) A bright eyed and bushy tailed graduate student was let loose on the data and was trusted because the results were so sexy for AGW

    or

    c) it is the usual mentality that happens in civil service when one is just warming the chair. i.e. indifference to working too much and if liking the outcome, just push it along to the politicos . It can be also the effect of too many people stirring the broth.

    Here is a lovely story from student days: In the mechanics lab students were set up in groups of four around pendulums and told to find its constants. So they started the pendulum swinging, the one with the stop watch started it, and stopped it at the agreed time. OK, he said, how many swings? Nobody had been counting the swings, trusting that the other three were doing so !

  77. crosspatch
    Posted Nov 12, 2008 at 1:09 AM | Permalink

    “So, what else should I include when I write my congressman asking that he call Hansen in to a hearing into his scientific rigor and fundamental scientific/political processing?”

    Well, you could start working your way back through the record to see how many other examples of this you can find. There might be some months with only a few scattered duplicated values. Keep track of your time and bill NASA and/or NOAA for your trouble.

  78. braddles
    Posted Nov 12, 2008 at 2:01 AM | Permalink

    I suppose it has come up in past threads, but it seems strange that the “state of the art” global analysis map manages to include just one temperature station from Australia (Forrest in Western Australia, it would appear, which closed down for a while in the 1990s).

    • Rich
      Posted Nov 12, 2008 at 5:04 AM | Permalink

      Re: braddles (#95),
      Last time I looked Hadcrut3 had 42% of gridcells that had no data for any year, none at all. How does that get turned into a “global” temperature? True, many are near the poles where the relative area is smaller but still….

  79. Len van Burgel
    Posted Nov 12, 2008 at 2:32 AM | Permalink

    braddles notes:

    (Forrest in Western Australia, it would appear, which closed down for a while in the 1990s).

    Forrest Aerodrome was closed in 1995. From 1993 readings were taken at the Forrest township about 1km to the southwest. Don’t think there is a heat island effect. The town has just 6 houses. It is an official Bureau site. Average temperatures are .7C higher at the new site, but that only has 16 years of record compared to the older site which has 49 years of record.

  80. widmerpool
    Posted Nov 12, 2008 at 2:36 AM | Permalink

    Is there any actual evidence so far? Again, as someone actually living in Moscow, it was anomalously warm until last Thursday. Every time I caught the TV weather in October they were talking about it being 10 degrees above average. Right now it is 8 degrees, which is 6 degrees above the norm.

  81. Sven
    Posted Nov 12, 2008 at 3:03 AM | Permalink

    What’s up? The data that was withdrawn from the GISS website last night until the mess was to be sorted out seems to be back and the same as before (station 0,88 and land-ocean 0.78C)?! The explanation at the beginning of the page that there are problems with data is removed and the graphs bear a text: “(Last modified: 2008-11-10)”. No, it’s not in my cache…

    Sorry, if it’s going to be a duplicate post with my comment at wattsupwiththat but a spam filter seems to catch my comment, so I’m removing any links and trying it here

  82. Pierre Gosselin
    Posted Nov 12, 2008 at 3:10 AM | Permalink

    Widmerpool,
    Perhaps so in Moscow.
    But the Arctic Ocean above Siberia would not have frozen as fast as it did in October if the temperature had stayed at September levels. And all the temperature charts I looked at clearly show it was colder in October. Hansen had to have seen that his temperature results were way out of the ballpark, and should have suspected a gross error. Yet, he didn’t bother to doublecheck!
    Every 6th grader learns that when you solve a science problem, you have to step back and ask yourself if the answer is realistic.

  83. Chris
    Posted Nov 12, 2008 at 3:39 AM | Permalink

    widmerpool – i justed posted a reply to you which was caught by the spam filter. But basically it used data from Wunderground cross-referenced with GISS data, to show that for the nearest two stations to Moscow (data for Moscow was unavailable), Oct 08 was indeed several degrees cooler than Sep 08, and was also cooler than numerous years in the historical record including earlier in the 20th century.

  84. Sven
    Posted Nov 12, 2008 at 4:02 AM | Permalink

    Re my post #98 – oops, the data and maps are gone again! Though the explanation about errors in source data that was there yesterday is still missing?!

  85. Vincent Guerrini Jr
    Posted Nov 12, 2008 at 5:35 AM | Permalink

    A conclusion from all posts above is that Satellite data is the most accurate. In any case discussion about changes in temps being significant for the last 100 years in climate terms is ludicrous (even satellite data!)

  86. Mike Bryant
    Posted Nov 12, 2008 at 5:55 AM | Permalink

    Did NASA GISS shut down their website?

  87. Mike Bryant
    Posted Nov 12, 2008 at 5:58 AM | Permalink

    From a post at Watt’s website, was the October 2007 Siberian anomaly similar to this year? The Siberian heatwave October 2007 was concurrent with the historic Arctic sea ice comeback.

    • Rejean Gagnon
      Posted Nov 12, 2008 at 6:38 AM | Permalink

      Re: Mike Bryant (#105),
      I think you mean October 2008

      Re: Ruth Dixon (#94),
      Keep in mind that this did not include windchill either! (a 20th C invention). Those troops really went through hell.

  88. Posted Nov 12, 2008 at 7:07 AM | Permalink

    As someone in Moscow I don’t find it hard to believe.

    • Jared
      Posted Nov 12, 2008 at 12:15 PM | Permalink

      Re: kurtlar vadisi (#108),

      Yes, Moscow was very warm last month. The satellites and wunderground stations show that. But Moscow is just a miniscule part of Russia, the largest nation geographically on earth. Observations and satellite measurements clearly show that October 2008, while being quite warm in parts of Russia, was still much cooler throughout than September. Even Moscow was 3C cooler in October compared to September.

  89. stephen richards
    Posted Nov 12, 2008 at 7:16 AM | Permalink

    REALCLIMATE announce that the data has been pulled in under 24 hours.

    You are making a mountain out of a cock up

  90. Mike Bryant
    Posted Nov 12, 2008 at 7:20 AM | Permalink

    Rejean, No the 2007 ice comeback was also historic during a time that GISS had a heatwave in Siberia.

  91. Steve McIntyre
    Posted Nov 12, 2008 at 7:39 AM | Permalink

    #109. They pulled the data AFTER I sent them an email notifying them of the error (which had been pointed out to me by a CA reader and which I had confirmed). They did not identify the error on their own. Students are obliged to properly credit their sources at risk of plagiarism. In this case, by not crediting or acknowledging his sources, Hansen has apparently caused you to think that they identified this error on their own within 24 hours. A student would get into trouble for that sort of trick.

  92. Posted Nov 12, 2008 at 7:53 AM | Permalink

    Some thoughts about this strange GISS/NASA verification methodology are here. I kind of like a closing paragraph:

    Now GISS/NASA has estimated the costs for trying to apply a little quality control to the temperature calculations. What’s so expensive about actually thinking about the numbers for which you are responsible? Looking at the fancy multi-colored plot and saying, Hmmm … that’s strange, it’s hotter than two Hells in Siberia in October; not to mention a couple of other places. That’s not expensive. And actually thinking about the numbers for which you are responsible is part of the job in the first place; the most fundamental part as far as I’m concerned. It’s a job requirement in Engineering.

    I’ve also added a comment (3.) to this post about the coding in ModelE.

  93. drawp
    Posted Nov 12, 2008 at 8:35 AM | Permalink

    Steve #111,
    Just who is the “CA reader” who pointed out the error? Shouldn’t you also be “obliged to properly credit sources”?

    Steve:
    In my post, I clearly stated: “A CA reader notified me of this phenomenon earlier today and I’ve confirmed for myself that the information is accurate.” I did not pretend that I had discovered this. The reader in question did not identify himself other than through a nickname; it was “John S”. I suggested to him that he do up his own post on the matter, offering him space here, but he said that he was too busy and suggested that I or John Goetz do it. On the other hand, Hansen obviously left some observers of the incident with the false impression that NASA identified this problem in their regular course of business and through their own quality control.

  94. Dotto
    Posted Nov 12, 2008 at 8:51 AM | Permalink

    In the blog http://wattsupwiththat.com/2008/11/10/giss-releases-october-2008-data/#comments
    I pointed out that there were heat waves in October 2007, 2005, 2000, 1999 and 1997, i.e. temperatures 4-8 degrees above normal temperatures, similar to the data for October 2008. I urged someone wiht the knowledge to check if these data. Shortly after the GISS site was closed down. And so it remains to this moment when I am writing this post.

  95. Steve McIntyre
    Posted Nov 12, 2008 at 8:58 AM | Permalink

    HEre is a complete list of GHCN sites with 2008 October equal to 2008 September. Some of the tropical sites may be corret. Most of the problem areas have been picked up by various poeple – Russia obviously, some former Soviet countries, Ireland, UK and Finland.

    id site country
    212 11763402000 JIMMA ETHIOPIA
    216 11763471000 DIRE DAWA ETHIOPIA
    261 12263723000 GARISSA KENYA
    587 15260725000 JENDOUBA TUNISIA
    588 15260735000 KAIROUAN TUNISIA
    590 15260745000 GAFSA TUNISIA
    593 15260765000 GABES TUNISIA
    1182 20742933000 AKOLA INDIA
    1437 21135394000 KARAGANDA KAZAKHSTAN
    1443 21135700000 GUR’EV KAZAKHSTAN
    1445 21135796000 BALHASH KAZAKHSTAN
    1448 21136177000 SEMIPALATINSK KAZAKHSTAN
    1454 21136870000 ALMATY KAZAKHSTAN
    1688 22220292000 GMO IM.E.K. F RUSSIAN FEDERATION (ASIAN SECT
    1690 22220674000 OSTROV DIKSON RUSSIAN FEDERATION (ASIAN SECT
    1692 22220891000 HATANGA RUSSIAN FEDERATION (ASIAN SECT
    1701 22223205000 NAR’JAN-MAR RUSSIAN FEDERATION (ASIAN SECT
    1703 22223330000 SALEHARD RUSSIAN FEDERATION (ASIAN SECT
    1706 22223472000 TURUHANSK RUSSIAN FEDERATION (ASIAN SECT
    1707 22223552000 TARKO-SALE RUSSIAN FEDERATION (ASIAN SECT
    1710 22223724000 NJAKSIMVOL’ RUSSIAN FEDERATION (ASIAN SECT
    1713 22223884000 BOR RUSSIAN FEDERATION (ASIAN SECT
    1717 22223933000 HANTY-MANSIJS RUSSIAN FEDERATION (ASIAN SECT
    1720 22224125000 OLENEK RUSSIAN FEDERATION (ASIAN SECT
    1722 22224266000 VERHOJANSK RUSSIAN FEDERATION (ASIAN SECT
    1725 22224507000 TURA RUSSIAN FEDERATION (ASIAN SECT
    1726 22224641000 VILJUJSK RUSSIAN FEDERATION (ASIAN SECT
    1728 22224738000 SUNTAR RUSSIAN FEDERATION (ASIAN SECT
    1735 22225173000 MYS SMIDTA RUSSIAN FEDERATION (ASIAN SECT
    1738 22225399000 MYS UELEN RUSSIAN FEDERATION (ASIAN SECT
    1755 22228275000 TOBOL’SK RUSSIAN FEDERATION (ASIAN SECT
    1763 22228698000 OMSK RUSSIAN FEDERATION (ASIAN SECT
    1766 22229231000 KOLPASEVO RUSSIAN FEDERATION (ASIAN SECT
    1767 22229263000 ENISEJSK RUSSIAN FEDERATION (ASIAN SECT
    1770 22229570000 KRASNOJARSK RUSSIAN FEDERATION (ASIAN SECT
    1775 22229866000 MINUSINSK RUSSIAN FEDERATION (ASIAN SECT
    1776 22230054000 VITIM RUSSIAN FEDERATION (ASIAN SECT
    1785 22230554000 BAGDARIN RUSSIAN FEDERATION (ASIAN SECT
    1800 22231004000 ALDAN RUSSIAN FEDERATION (ASIAN SECT
    1804 22231253000 BOMNAK RUSSIAN FEDERATION (ASIAN SECT
    1900 22938687000 CARDZOU TURKMENISTAN
    1903 22938880000 ASHGABAT TURKMENISTAN
    1908 23138262000 CIMBAJ UZBEKISTAN
    1909 23138413000 TAMDY UZBEKISTAN
    1910 23138457000 TASHKENT UZBEKISTAN
    2041 30382106000 SAO GABRIEL D BRAZIL
    2057 30382704000 CRUZEIRO DO S BRAZIL
    2219 30984628000 LIMA-CALLAO/A PERU
    5996 50548657000 KUANTAN MALAYSIA
    5997 50548665000 MALACCA MALAYSIA
    6068 50998444000 LEGASPI PHILIPPINES
    6076 50998550000 TACLOBAN PHILIPPINES
    6147 53191366000 KWAJALEIN/BUC MARSHALL ISLANDS
    6257 61402836000 SODANKYLA FINLAND
    6260 61402869000 KUUSAMO FINLAND
    6262 61402875000 OULU FINLAND
    6263 61402897000 KAJAANI FINLAND
    6267 61402935000 JYVASKYLA FINLAND
    6269 61402958000 LAPPEENRANTA FINLAND
    6270 61402963000 JOKIOINEN FINLAND
    6272 61402972000 TURKU FINLAND
    6432 62103955000 CORK AIRPORT IRELAND
    6435 62103962000 SHANNON AIRPO IRELAND
    6438 62103967000 CASEMENT AERO IRELAND
    6439 62103969000 DUBLIN AIRPOR IRELAND
    6443 62103976000 BELMULLET IRELAND
    6444 62103980000 MALIN HEAD IRELAND
    6570 62826629000 KAUNAS LITHUANIA
    6571 62826730000 VILNIUS LITHUANIA
    6721 63822550000 ARHANGEL’SK RUSSIAN FEDERATION (EUROPEAN S
    6725 63822802000 SORTAVALA RUSSIAN FEDERATION (EUROPEAN S
    6727 63822837000 VYTEGRA RUSSIAN FEDERATION (EUROPEAN S
    6730 63826063000 ST.PETERBURG RUSSIAN FEDERATION (EUROPEAN S
    6742 63827037000 VOLOGDA RUSSIAN FEDERATION (EUROPEAN S
    6749 63827612000 MOSKVA RUSSIAN FEDERATION (EUROPEAN S
    6755 63834123000 VORONEZ RUSSIAN FEDERATION (EUROPEAN S
    6771 63834880000 ASTRAHAN’ RUSSIAN FEDERATION (EUROPEAN S
    7145 65033345000 KYIV UKRAINE
    7147 65033393000 L’VIV UKRAINE
    7155 65033837000 ODESA UKRAINE
    7162 65033946000 SIMFEROPOL USSR UKRAINE
    7167 65034300000 KHARKIV UKRAINE
    7172 65103026000 STORNOWAY UNITED KINGDOM
    7178 65103091000 ABERDEEN/DYCE UNITED KINGDOM
    7179 65103100000 TIREE UNITED KINGDOM
    7183 65103162000 ESKDALEMUIR UNITED KINGDOM
    7187 65103257000 LEEMING UNITED KINGDOM
    7189 65103302000 VALLEY UNITED KINGDOM
    7224 65103862000 BOURNEMOUTH A UNITED KINGDOM
    7228 65103917000 BELFAST/ALDER UNITED KINGDOM

    • Dotto
      Posted Nov 12, 2008 at 9:11 AM | Permalink

      Re: Steve McIntyre (#115), What about September-October 2007, 2005, 2000, 1999 and 1997? Can you check for these years too?

  96. Dotto
    Posted Nov 12, 2008 at 9:10 AM | Permalink

    What about September-October 2007, 2005, 2000, 1999 and 1997? Can you check for these years too?

  97. Kenneth Fritsch
    Posted Nov 12, 2008 at 9:14 AM | Permalink

    I am addressing this post to my fellow jesters and other interested acolytes of the executives of the fossil energy industry. In it I am committing the cardinal sin of jestering, as it was taught to me at an undisclosed site of an unnamed fossil fuel interest, i.e. interpreting an act of jestering.

    What can we, as jesters, take away from this thread? Please pardon my serious tone here, but I think all of you true practitioners of the art of jestering know that we, like our cousin clowns, are required to think seriously about our craft.

    Great men do not make mistakes. They get bad advice and incorrect data, but they are too busy themselves thinking important thoughts to get involved at the levels where all mistakes are made.

    One could readily blame Hansen and Napoleon for their apparent mistakes in Russia, but look more closely and we see that they were handed bad information. Therefore, I implore you not to go a bridge too far.

    Do not go a bridge too far with the clever, by too far, juxtaposition of Napoleon and Hansen. That issue is a complex one not intended for banter on a blog.

    We can take advice that we have received on the limits of the MWP and apply it here and again avoid going the bridge too far. The phenomenon is local and at best regional and surely short in duration, the Missed Weather Period that is.

    The bottom line jester lesson here should be obvious to you jesters and that is that adjustments of data while necessary can be vulnerable to a biased application of them. Again, my fellow jesters out there, I implore you not to go a bridge too far and heed what Steve M states in a reply above where he notes that errors on temperature change could go in either direction, be they measured or predicted.

    So I say to my fellow jesters. Jester on, but avoid that bridge too far – unless, of course, you are very, very clever.

  98. Julie
    Posted Nov 12, 2008 at 9:32 AM | Permalink

    re #115 – Aberdeen/Dyce was definitely not the same temps for Oct & Sept. We were between 0.65 & 1.2 degreeC below average here in October at 8.5 or 7.7 (depends on what data you use), whereas we were slightly above the average for Sept (12.5 or 12).

    The former numbers come from Philip Edens Climate-UK website, the latter were calculated from the averaged Wundergound daily values.

  99. gjg
    Posted Nov 12, 2008 at 9:39 AM | Permalink

    In comment about Urban Heat Island Effect.
    Here is a comparison between urban and rural sites that are within 100 km of each other across the US. This covers 111 years of data.

  100. crosspatch
    Posted Nov 12, 2008 at 9:50 AM | Permalink

    “I pointed out that there were heat waves in October 2007, 2005, 2000, 1999 and 1997, i.e. temperatures 4-8 degrees above normal temperatures, similar to the data for October 2008. I urged someone wiht the knowledge to check if these data.”

    Well a simple check for those months would be to see if the GISS temperature is way out of line with RSS and UAH data for the same months. I believe that was the tip-off in this case. GISS was so grossly out of line with the others that someone decided to have a look. But there could be more subtle problems if it happened in the past with only a few stations retaining a previous month’s value and it would be very difficult to spot. Nevertheless, months where GISS is significantly different from UAH and RSS would be where I would look for past occurrences of this sort of error.

    • Nylo
      Posted Nov 12, 2008 at 12:30 PM | Permalink

      Re: crosspatch (#123),

      Nevertheless, months where GISS is significantly different from UAH and RSS would be where I would look for past occurrences of this sort of error

      I run a quick comparison of global monthly averages as registered in RSS and GISS. There is an average difference of 0.3 degrees (logical, as they use a different time period of reference too). Appart from that, although in past decades there were sometimes big discrepancies at some given months, this only happened twice in the last 8 years, one in March 2002 and the other in March 2008. Of these two, I am only suspicious about the later one, because the GISS anomaly for that month was also very different from the GISS anomaly in the surrounding months. But it was in March, so it is very unlikely that the same kind of mistake happened (if they had copied February data in some stations in the NH, the resulting wrong anomaly should be colder, not hotter). Of course, the wrong reporting could have happened in the SH…

  101. crosspatch
    Posted Nov 12, 2008 at 10:17 AM | Permalink

    Actually, come to think if it, it might not be too hard to see such an error in stations at higher latitudes, but it would depend on the month. Spring and Fall being easy to spot. A computer program running through the station data should be able to spot a value that is out of line with other values for the same month and check that value against the previous month’s value for that year. Still, someone would need to have the time to code it.

    It really wouldn’t surprise me in the least if we learned that such an error has happened many times in the past though I would think that the error would be just as likely to go either way (just as likely to happen in April/May as September/October).

  102. crosspatch
    Posted Nov 12, 2008 at 10:24 AM | Permalink

    “One cannot ignore the monotonic increases in atmospheric CO2 levels over the past century”

    Sure one can. You are assuming that the CO2 levels are a cause of something rather than a result of something. You also seem to assume that climbing CO2 levels are somehow bad for the environment. Maybe the opposite is true. Maybe atmospheric CO2 levels were getting too low and having an adverse impact on plant life.

    CO2 levels have been this high in the past. They have been much higher, 2x as high, 5x as high … and temperatures didn’t “run away” into greenhouse oblivion.

    • hswiseman
      Posted Nov 12, 2008 at 6:52 PM | Permalink

      Re: crosspatch (#126), I am not assuming anything of the sort. I am saying perform the science without confirmation bias, with validated data and methods. And presumptive contempt for the work of research scientists making real observations in the field without any evidence of gamesmanship or manipulation can only chill reasonable debate and discussion.

  103. Jaye
    Posted Nov 12, 2008 at 10:58 AM | Permalink

    You are making a mountain out of a cock up

    And you are simply parroting the words of Schmidt.

  104. tetris
    Posted Nov 12, 2008 at 11:09 AM | Permalink

    Re: 114 and 116
    Dotto and Lucia,
    As of time of posting [0906 PST] the GISS site was still unaccessible.
    Re: 115 and 118
    Steve,
    It seems to me that Dotto’s suggestion is an interesting. Do you have access to those data series?

  105. Sean J
    Posted Nov 12, 2008 at 11:12 AM | Permalink

    Based on Dotto’s graphs over at Watts, I am willing to bet that the GISS problem is a bit more systematic than this one off.

    Figuring out how to save face when GISS revises its temp trends down has got to be the reason for the site being down.

    Maybe John G or V can shed some light on this.

  106. Mike Bryant
    Posted Nov 12, 2008 at 11:35 AM | Permalink

    Perhaps GISS will issue a correction that covers all the September/October “surprises” for the last 20 years.
    Nahhhhhh.

  107. tetris
    Posted Nov 12, 2008 at 11:40 AM | Permalink

    Re: 109
    Steven Richards,
    Have you noted that whatever the “cock up” at GISS it’s always in the same direction: up? Have you also noted that at RC any GISS “cock up” is always explained away with a lot of hand waiving? For perspective, imaging if the sign on the “cock up” were reversed..

  108. Mike Bryant
    Posted Nov 12, 2008 at 11:53 AM | Permalink

    Imaginary GISS press Release:

    Accidental Release of Data Quickly Corrected

    An intern with GISS accidentally released data that had not been approved for release. Even though it was quickly removed from the website, numerous right-wing bloggers are calling for the removal of Dr. James Hansen. Hansen said, “It’s really just a tempest in a teapot, I had left specific instructions to wait until after the holiday to post this, since I realized that something was not quite right. I also removed the corrupt files as soon as I noticed the mistake. Now that we have released the correct October anomaly, you can see that it hasn’t really changed that much. There were higher anomalies in other places that the bloggers did NOT catch. The bloggers obviously have confirmation bias. The intern will be reprimanded and the procedures will be followed.”

  109. Dotto
    Posted Nov 12, 2008 at 12:01 PM | Permalink

    I managed to get copies of the temperature anomalies for October 2007, 2005 and 1999 before http://data.giss.nasa.gov/ was shut down.
    If the administrator sends me an email I will be glad to share them.

  110. fred
    Posted Nov 12, 2008 at 12:10 PM | Permalink

    #71 Steve M

    You say “CRU (UEA) has refused to provide any data. So it would have been impossible to pin down an equivalent error by them. Nor is there information on their algorithm.”

    How odd. It is probably not widely known, but there has been a fairly vigorous debate in East Anglia, the UK, lately about alleged rising sea levels and consequently about AGW. CRU is based in Norwich, the largest city in the East Anglia region. The question is, what the Government’s committment is and should be to maintaining sea defences along the Norfolk and Suffolk coasts. In particular, one Government agency has been floating a trial balloon of a policy on coastal defense which would see the Norfolk Broads waterlands, villages and farmland gradually vanish into the North Sea. The argument is, naturally, that the data shows that warming is happening, that sea level rises are therefore inevitable, and that resistance is futile.

    One wonders whether, in the circumstances, the local MP might take an interest in CRU secrecy. It really does not seem reasonable that a body whose conclusions figure prominently in a significant national political debate should be able totally to avoid any examination of the basis for them.

    I think you are saying this, correct? That we do not know the stations used, we do not have the raw readings from them, and we do not finally have the algoithms used to get from the readings to the global temperature trends.

    How very extraordinary. In no other field, from finance to farming to medecine, would one get away with simple assertions in this way.

  111. Urederra
    Posted Nov 12, 2008 at 12:14 PM | Permalink

    This site http://www.dailytech.com/Deja+Vu+All+Over+Again+Blogger+Again+Finds+Error+in+NASA+Climate+Data/article13410.htm covers the story aknowledging “Steve and his gang” for finding the error (once again).

    There is a paragraph that worries me:

    The error not only affected October data, but due to the complex algorithm GISS uses to convert actual temperature readings into their output results, altered the previously published values for several other months as well. The values for August 2008, for instance, changed by 0.11C and the global anomaly as far back as 2005 increased by a hundredth of a degree.

    Has anyone noticed that the values for August 2008 changed by 0.11C?

  112. DG
    Posted Nov 12, 2008 at 12:17 PM | Permalink

    Why are errors at the grocery store nearly always in favor of the store?

    One wonders what was changed in their model code to cause these errors, and the date it was done. Was there such a difference from Sept to Oct 2008 that made this stick out and is an isolated case? If there is no revision history with date stamps, this could be a bigger mess than meets the eye.

  113. Dotto
    Posted Nov 12, 2008 at 12:27 PM | Permalink

    Very interesting! If you look at the temperature anomlies maps you see that the largest anomalies always occur in Siberia!

    http://www.climate4you.com/GlobalTemperatures.htm#Monthly%20surface%20air%20temperature%20anomalies%20versus%20average%201998-2006%20in%20areas%20between%20%2072oN%20and%2060oN

  114. Patrick Henry
    Posted Nov 12, 2008 at 12:30 PM | Permalink

    The discrepancy between the GISS data and other sources should have raised a red flag. In industry, a major error like that sends people packing.

  115. GP
    Posted Nov 12, 2008 at 12:41 PM | Permalink

    Apart from the lack of a value for Dec 2007 has anyone noticed how erratic the these values are in many months yet how similar in others, month to month and year to year?

    YEAR JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
    2006 -34.0 -29.9 -23.5 -18.1 1.6 10.6 16.9 11.5 4.4 -14.6 -27.7 -29.1
    2007 -27.9 -41.5 -21.6 -4.0 0.1 12.4 13.5 11.3 3.1 -9.0 -24.8 999.9
    2008 -30.0 -29.4 -19.6 -13.4 1.3 12.0 13.1 12.1 3.1 3.1 999.9 999.9

    I would accept that the greater annual differential in temp might offer more scope for larger differences then one might see in region so f the globe woth more temperate or otherwise consistent values, but some of these seem rather odd.

    For example 2006 , other than for October, looks quite reasonable. October 2006 looks possibly correct but somewhat lower then previous and subsequent months.

    But look at Feb and April 2007 which seems respectively much colder and much warmer whlst MAr 2007 in between then seems to be right in the middle of Mar 2006 and Mar 2008 showing March warming, year on year by about 2 degrees. (Right or wrong who knows?)

    May through September are all relatively consistent – or at least more so that some, but not all, of the winter months.

    What are we seeing here? Fact or interpolation?

  116. DG
    Posted Nov 12, 2008 at 12:51 PM | Permalink

    Given past temperatures in GIStemp data constantly change [erratically], how is it even possible to backtrack past years for these errors? It would appear only a thorough software going-over can solve the problem.

  117. John S.
    Posted Nov 12, 2008 at 1:21 PM | Permalink

    Congratulations, Steve Mc, you did a great job with my tip, far better than I could have managed on a busy schedule. And John Goetz deserves kudos for tracing the problem to the GHCN data files at NOAA. It was never a question of erroneous reports from the stations themselves!

    Being deeply continental, Siberian average temperatures have some of the highest variability year-to-year on the globe, with sigmas in excess of 1C at many stations. And brief heat waves are not uncommon throughout Western Russia in autumn. The results posted by GISS, however, were so ridiculously out of range, that anybody with a modicum of climate sense could have spotted the error.

  118. Posted Nov 12, 2008 at 1:59 PM | Permalink

    Steve M

    As Urederra already has asked: Did this abnormal ocotber anomaly also have ramifications backwards in time? Så that it made august 2008 quite a bit warmer globally?

    Are these ‘corrections’ known and publicly available. It sounds quite astonishing to be, that an (admitteldly excessively) warm october also could heat up the preceeding august!

    Maybe you have a link/post on that topic?

  119. Urederra
    Posted Nov 12, 2008 at 2:00 PM | Permalink

    Congratulations, John S.

    BTW. NASA GISS website is up again.

  120. Urederra
    Posted Nov 12, 2008 at 2:06 PM | Permalink

    GISS website is up again but It cannot generate a Global map from GHCN October data.

  121. Mike Bryant
    Posted Nov 12, 2008 at 2:55 PM | Permalink

    Jonas M,
    The temperatures are tele, and temporally connected in such a way that any new anomaly, up or down, always affects every temperature in the historic database. Normally the effect is that very old temperatures become colder, as more recent temperatures rise. This effect was discovered by Dr. James Hansen. I expect he is in the running for the Nobel Prize. I think it has something to do with relativity. It may seem confusing now, but eventually everyone will accept the new paradigm.

  122. crosspatch
    Posted Nov 12, 2008 at 3:15 PM | Permalink

    I was able to generate a map (from September’s data). I find it interesting, though, how they always seem to show a huge positive anomaly in the polar regions even when the areas just below the polar regions show a negative anomaly.

  123. Mike Bryant
    Posted Nov 12, 2008 at 3:37 PM | Permalink

    Crosspatch,
    This is really quite easy to explain. If a tree falls in a forest and there is no one there to hear it fall, could the temperature have been a few degrees warmer? It is a corollary of the Hansen Effect that the warmest temperatures are always in the most sparsely populated parts of the globe.

  124. Len van Burgel
    Posted Nov 12, 2008 at 9:29 PM | Permalink

    GISS amended october global temperature anomalies now on their page.

  125. Soronel Haetir
    Posted Nov 13, 2008 at 6:34 AM | Permalink

    crosspatch,

    A positive artic reading seems like it would stem from the comparison being against a known cold period. Isn’t the 0 line the average from 1950 to 1979 or some similar range and not the average of all years?

  126. Posted Nov 13, 2008 at 9:10 AM | Permalink

    From the Gistemp website:
    “2008-11-12: It seems that one of the sources sent September data rather than October data. ”

    So, which ‘one source’ sends them the data for Russia, Denmark, Britain, Ireland, Australia…..

  127. Posted Nov 13, 2008 at 10:31 AM | Permalink

    You can get a measure of just how cold it was on the retreat from Moscow from “The Diary of a Napoleonic Foot Soldier” by Jakob Walter. You needed a horse to get back, and you needed to tie yourself to the horse at night to discourage theft. But suppose someone cut the rope and stole your horse? Then you’d go and steal someone else’s horse, cutting the rope, etc…

  128. Robinedwards
    Posted Nov 13, 2008 at 2:22 PM | Permalink

    Regarding #147 (UC), I have been studying historical temperature records (data) from the NW Atlantic recently. For several stations it gradually dawned on me that during the depth of winter missing values were quite common. It soon became evident that the recorded temperatures /never/ went below -19.9 C. Dec, Jan and Feb are the months when the high teens (negative) occurred regularly. I would have expected -20 to be exceeded now and again, given the values for surrounding months.

    I concluded that it must have been deemed too cold for anyone to venture out to read the thermometers – and I don’t blame them!

    Might the same sort of filtering still be going on?

  129. ba
    Posted Nov 16, 2008 at 9:53 AM | Permalink

    Such high temperatures make one wonder if there is a thermistor being read remotely, under 2, 5 or 10 ft of insulating snow, potentially making ground temperatures and the electronic instrumentation factors…

    • John F. Pittman
      Posted Nov 16, 2008 at 9:56 AM | Permalink

      Re: ba (#164), That is a good comment. Thanks.

  130. henry
    Posted Nov 16, 2008 at 4:53 PM | Permalink

    Seems to me that this “carry-over” of data would be very easy to check for:

    Simply take the list of a station’s data, and subtract it from a one-month shifted list (done twice, both shifted forward and back one month).

    Any zero result is immediately suspect…

  131. James P
    Posted Jan 1, 2009 at 10:05 PM | Permalink

    If it is any consolation, both the international weather service and 16 different countries cannot agree on a single attribute that should define what is “normal” in normal temperature range during a single season, much less on what is abnormal temperature variations and acceptable measurements.

18 Trackbacks

  1. [...] di anomalia che per il NOAA risulta essere di +0,91° rispetto ai +0,66° di settembre. Su Climateaudit c’e’ la segnalazione dell’errore [...]

  2. By Desde el Exilio on Nov 11, 2008 at 3:36 AM

    Escándalo: los datos de temperatura para Octubre del GISS falsificados!…

    Ha saltado la noticia y ya corre como la pólvora: los datos validados por la NASA (NOAA) sobre los que basa el Goddard Institute of Space Studies (GISS) sus tablas y notas de prensa están corruptos! El anuncio del Octubre más clálido en la historia…

  3. [...] the “delayed winter in GISS records” problem is rather widespread. It may also include England. Steve McIntyre and John S. both looked at the GISS record in more details, and many Russian records [...]

  4. By Cooking the Books at NASA « The Liberty Boys on Nov 12, 2008 at 6:30 AM

    [...] Did Napoleon Use Hansen’s Temperature Data? [...]

  5. [...] My my, at waterloo napoleon did surrender Oh yeah, and I have met my destiny in quite a similar way… [...]

  6. [...] Daraufhin entbrannten wilde Spekulationen. Siehe: Did Napoleon Use Hansen’s Temperature Data? [...]

  7. [...] http://www.climateaudit.org/?p=4318 [...]

  8. [...] Following a blunder at NASA’s Goddard Institute for Space Studies, Steve McIntyre, at the Climate Audit blog, has reminded James Hansen, from NASA, that its colder in Russia in October than in September, as Napoleon found out to his cost in 1812.  Read more here . [...]

  9. [...] http://www.climateaudit.org/?p=4318 [...]

  10. [...] Following a blunder at NASA’s Goddard Institute for Space Studies, Steve McIntyre, at the Climate Audit blog, has reminded James Hansen, from NASA, that its colder in Russia in October than in September, as Napoleon found out to his cost in 1812.  Read more here . [...]

  11. [...] I decided to put Google Earth to work to see what I could see. One of the stations mentioned in a recent post at Climate Audit cited the station of Verhojansk, Russia, which has  temperatures conveniently online here at [...]

  12. [...] http://www.climateaudit.org/?p=4318 [...]

  13. [...] GISS data again, as it claimed there was historic warming in Russia, which turned out to really be stale September data left in the October [...]

  14. [...] for the perspective from Climate Audit: CA link to initial wonderment about the Russian anomalies. CA link to NASA blaming NOAA. CA link to discussion on corrected data and a “warm” [...]

  15. [...] The editorial is based on Steve McIntyre’s research, reported on his site Climate Audit in the post Did Napoleon Use Hansen’s Temperature Data? [...]

  16. [...] a bit of an embarrassment for NOAA and GISS, that was pretty short lived when two blogger meteorologists went through the data and found a number of anomalies.  They discovered at least 10 Russian stations that oddly enough reported the exact same [...]

  17. [...] Today, we are able to provide NASA with up-to-date weather reports confirming that October in Russia is colder than September. Verhojansk temperatures are conveniently online here and temperatures are currently a nippy -18 deg C….(Full Story) [...]

  18. [...] question is asked by Stephen McIntyre (www.climateaudit.org/?p=4318), who called attention to a recent error by NASA Goddard Institute for Space Studies (GISS). James [...]

Follow

Get every new post delivered to your Inbox.

Join 3,196 other followers

%d bloggers like this: