Lost at Sea: the Search Party

Uninsulated Buckets
A CA reader emailed me, observing that there may be relevant differences in insulated and uninsulated buckets in the post-World War 2 period, which could easily affect adjustment schedules. This makes a lot of sense to me and might reconcile a few puzzles and opening others.

Let’s say that the delta between engine inlet temperatures and uninsulated buckets is ~0.3 deg C (and here we’re just momentarily adopting one of the canonical Folland numbers as this particular number surely deserves to be cross-examined). Insulated buckets would presumably be intermediate. Kent and Kaplan 2006 suggest a number of 0.12-0.18 deg C. So for a first rough approximation to check our bearings on this – let’s suppose that it’s halfway in between. Maybe it’s closer to engine inlets, maybe it’s closer to uninsulated buckets. We’re not trying to express viewpoints on such conundrums here – we’re merely examining what assumptions are latent in the temperature estimates.

We know that 90% of all measurements in 1970 with (supposedly) known provenance were done by buckets (Kent et al 2007), while there was a turnover in proportion to about 90% engine inlet and hull sensor by the 2000s. In my first cut at estimating the effect of unwinding some of the erroneous adjustment assumptions, I posited that the above information implied that the 0.3 deg C adjustment between buckets and engine inlets didn’t disappear merely because of reversion to “business as usual” after WW2. On this information, the only time that the delta could be introduced was between 1970 and 2000. This in turn poses new conundrums, as you’re getting into periods with satellite measurements. So there are issues with pushing the delta entirely into the post-1970 period.

However, let’s suppose that there was a transition from predominantly uninsulated buckets immediately post-WW2 to predominantly insulated buckets as at 1970 or so. Then the 0.3 deg C total adjustment would be spread proportionally between the two periods – with the delta between uninsulated buckets and insulated buckets being allocated to the 1945-1970 period or so (together with other relevant instrumental changes) while the delta between insulated buckets and engine inlets would be allocated to the 1970-2005 period (again together with any other relevant instrumental drifts e.g. changing proportion of hull sensors, buoys, whatever.)

I’m headed away for the weekend, but I’ll redo my rough guess based on these variations in a day or two.

A couple of observations, which readers should bear in mind.

In doing an “audit”, if an auditor identified a situation where the assumptions did not warrant the conclusions (e.g. the IPCC hypothesis of a sudden and permanent changeover to engine inlets at Pearl Harbour), he would not substitute his own assumptions; he would reject the assumptions and throw the problem back at the authors. I’ve been pretty consistent about this in the proxy areas, avoiding the temptation to posit what really happened. I don’t want to go beyond this policy here either.

Now I did a graphic showing the impact of abandoning the Pearl Harbour hypothesis under certain other assumptions. Please construe any such calculation as equivalent to a sensitivity calculation to show the effect of (say) excluding bristlecones, to illustrate the impact of certain assumptions, but not advocacy of a specific alternative. My language at the time may not have been explicit on this, but it’s the way that I do things. There are any number of intricacies in the interpretation of SST buckets. What are the effects of hull sensors? Drifting buoys? Etc. etc. I have no information on such matters at present. Could hull sensors and drifting buoys offset changeover from insulated buckets to engine inlets? Could be. I’m not opining on this. I’d like to see proper expositions of these populations and biases and one of the good outcomes of Thompson et al is that it will almost certainly achieve this goal.

As to my original observations on this: I was in a position to observe that the Pearl Harbour assumption did not hold up and that the 0.3 deg C adjustment did not occur in one bite in 1941, but needed to be spread much later. Right now, as noted above, it looks like some proportion gets spread to the period before 1970 and some proportion after 1970.

I might also make a point here about reviewing. Posts at this blog are not “peer reviewed”, but they are read by a lot of people, who have an opportunity to comment on them. In this particular case, a reader drew my attention to a distinction between insulated and uninsulated buckets that I hadn’t thought about and which makes sense in the context. One could make a case that blogs are actually a very good way of carrying out review of articles. In economics, journals like articles to have been around for a while and to have been reviewed in a variety of venues before being published; the whole idea of embargo-to-publication that has been promulgated by Nature and Science seems like something that needs to be talked about.

The audience is also astonishingly active. Demetris Koutsoyannis wrote to thank me a favorable recent comment; he said that he got 2700 downloads of the article in one day after being recently mentioned and linked at Climate Audit and 5700 since then, with the vast majority coming directly from CA. It’s no accident that we rank so high on any number of Google searches.

Provenance Bias versus Sudden Changeover
A second point that I’d like to quickly mention before leaving for the week-end.

My original criticisms were based on the arm-wavingness of assuming that there was a “sudden but undocumented” change in observing procedures coinciding with Pearl Harbour. This seemed far too much like Briffa’s Cargo Cult explanation – and originated once again from essentially the same crowd. Here is the precise quote from Folland and Parker 1995 (originally cited here in 2005):

Barnett (1984) gave strong evidence that historical marine data are heterogeneous. He found a sudden jump around 1941 in the difference between SST and all-hours air temperatures reported largely by the same ships. Folland et al. (1984) explained this as being mainly a result of a sudden but undocumented change in the methods used to collect sea water to make measurements of SST. The methods were thought to have changed from the predominant use of canvas and other uninsulated buckets to the use of engine intakes. Anecdotal evidence from sea captains in the marine section of the Meteorological Office supported this idea. .

or again:.

The abrupt change in SST in December 1941 coincides with the entry of the USA into World War II and is likely to have resulted from a realization of the dangers of hauling sea buckets onto deck in wartime conditions when a light would have been needed for both hauling and reading the thermometer at night. The change was made possible by the widespread availability of engine inlet thermometers in 1941 (section 4)

Now there is a real discontinuity in World War 2 and it’s related to the above. Thompson et al 2008 provide a far more plausible and convincing explanation of this discontinuity than the one provided by Folland and Parker 1995. It’s partly a difference in nuance, but the difference is important. Here is a graphic from Thompson et al 2008 showing a change in US contribution to the total population of SST measurements. Clearly the discontinuities in the increased proportion of US measurements (engine inlet) to the total population in WW2 is directly connected to discontinuities in the SST history. It’s not that thousands of ships suddenly and overnight changed how they did things; it’s that the data sources changed. In a way, it’s exactly like the Hansen Y2K problem.

From Thompson et al 2008. Bottom panel shows US/UK proportions.

This explanation makes sense, while the Folland and Parker 1995 one doesn’t. If you then look at their graphic with this idea in mind, one wonders what is happening in the 1960s, when US contribution inexplicably goes to nearly zero for a while, which seemingly coincides with a downdip in world SST, with a later sharp increase in US contribution coinciding with an increase.

Wooden Buckets

Left undiscussed by Thompson et al 2008 is the thorny issue of wooden buckets prior to 1940. One of the distinctive innovations of Folland in the 1980s and Jones et al 1991 was the hypothesis that heavy wooden buckets (which behaved like insulated buckets) were used in the 19th century, with a more or less linear changeover to uninsulated canvas buckets.

In the last few days, I’ve been reading some of the older literature on buckets. Brooks (Weather Review 1924) is an early and interesting consideration, reporting on experimental comparisons of canvas bucket and engine inlet measurements in a variety of situations. There’s not a whiff of mention of wooden buckets. One of the discussants of the paper wonders why they use canvas buckets at all, when engine inlet appears superior; and another says that it’s just a holdover from sailing vessels. Another discussant mentions the use of engine inlet measurements by the Canadian Meteorological Service in the NW Pacific, a point that may tie in to an issue in Rayner et al 2006 [note to self].

Folland and Parker 1995 carried out elaborate measurements on heat dissipation from an oak bucket, noting, in passing, that there was no evidence that this type of oak bucket had ever been used to measure SST.

A nineteenth-century oak ships’ bucket covered in iron bands has been studied though there is no indication that it was used for taking sea temperatures

This did not stop the development of an elaborate adjustment methodology based on a transition from oak buckets to canvas buckets, a transition which had the effect of lowering adjusted 19th century temperatures relative to what would have been calculated if canvas buckets had been used.

Surely at some point, someone has to show that these wooden buckets actually were used in the 19th century. If I had to take a stab at it right now, I think that it’s far more likely that canvas buckets were used until engine inlet changeover and that it would be well worthwhile doing a variation on this assumption. Also the assumption of 0.3 deg C isn’t necessarily written in stone. Brooks 1924 observed many values well in excess of this. Saur 1963 (J Appl Met online) is an interesting direct comparison of bucket and engine inlet measurements on the same voyage – the sort of evidence one would like to see more of – and the differences were higher than 0.3 deg C.

Whatever the end result, there are a lot of models being developed on this sort of information and careful cross-examination of the wooden buckets is long overdue.


  1. BarryW
    Posted May 31, 2008 at 8:49 AM | Permalink

    One question would be what is the percentage of US observations that are done by warships vs commercial vessels. Secondly are commercial vessels considered US even if they are chartered under another flag? The sudden drop could be due to the Vietnam war or a change in registration.

  2. jerry bono
    Posted May 31, 2008 at 9:08 AM | Permalink

    Maybe this is in the literature I don’t know since I have not read it but it seems to me all the bucket data came from right on the surface of the ocean, the real interface between air and sea. Now, how deep was the sample taken from the engine inlets. I know it depends on the size of the ship but it should be deep enough to get away from potential wave troughs, maybe say 10 feet? I don’t know. From my scuba diving experiences I can tell you that I have felt some very sever thermal gradients. Most of this was near the cost obviously and I would expect open ocean to to be less variable but there should be a variation.

    the specialists are aware of this sort of issue and there are elaborate discussions.

  3. agesilaus
    Posted May 31, 2008 at 9:36 AM | Permalink

    As someone who spent far too much time in a US Navy engine room I am puzzled by some of this. Condenser inlet temp is an important process control value on steam ships. And until recent times most USN ships were steam driven. And that reading is collected houly on board any USN ship.

    And all those records are probably stored in some warehouse someplace, the Navy not being prone to discard anything once it is commited to paper. So someone could find those records and correlate them with the ship position, also stored somewhere, and generate as much data as they had the time and money to search for. I suspect these records probably go back to at least pre WWI times.

    And a chunk of that data would be from submarines which could add data from waters below the surface layer. Especially in the nuclear submarine era. Nuke boats travel mostly submerged and the depth data would be recorded with the position data. This sounds like an excellent Ph.D thesis project to me.

    So is this just a case of these people not wanting to expend the effort to find real data? And has anyone done a comparison with bucket vs inlet temps?


    Steve: Millions of data points have been collected in the COADS program,

  4. Alan S. Blue
    Posted May 31, 2008 at 9:44 AM | Permalink


    In this particular case, a reader drew my attention to a distinction between insulated and insulated buckets that I hadn’t thought about and which makes sense in the context.

    This sentence from paragraph ten would seem to need a “non-” in there.

  5. Posted May 31, 2008 at 11:05 AM | Permalink

    Steve M., you wrote “the whole idea of embargo-to-publication that has been promulgated by Nature and Science seems like something that needs to be talked about”.

    Open peer review was trialed by Nature, and it failed: almost no comments on submitted papers were posted. For details, see

  6. steven mosher
    Posted May 31, 2008 at 12:06 PM | Permalink

    It’s Gilligan’s Island. Not lost at sea.

  7. Robert Wood
    Posted May 31, 2008 at 12:37 PM | Permalink

    Just how certain can one be of a theory when the error bars and uncertainties in measurements are equal to the effect being studied?

  8. Posted May 31, 2008 at 12:52 PM | Permalink

    The more I read this story the more I find it hard to believe. Are the those temperature graphs so ominously presented to the world in Al Gore’s movie going to be overturned by an argument over what type of materials survey ship buckets were made out of? The general public have no idea of what has gone on in the name of global warming-it would be outraged if it did.

    For the record; I think the buckets I used back in 1981/2 were made out of plastic.

    Sorry- no idea of the insulation properties.

  9. Josh
    Posted May 31, 2008 at 1:36 PM | Permalink

    heavy wooden buckets (which behaved like insulated buckets) were used in the 19th century, with a more or less linear changeover to uninsulated canvas buckets. … a transition which had the effect of lowering adjusted 19th century temperatures relative to what would have been calculated if canvas buckets had been used.

    Ummm… doesn’t the HADCRUT data display a near-linear increase in temperatures from the late 19th century until the war? What is the magnitude of this “heavy wooden bucket adjustment”? If it is the same 0.3C, then it eliminates roughly half of the increase in temperatures in the early part of the century.

  10. Colin Hunter
    Posted May 31, 2008 at 2:34 PM | Permalink

    This article from 1926 reports the use of canvas bags on a British ship in 1925.

    It also reports the measured temperature difference between the buckets and
    engine inlet. Apparently they were alive to this issue in 1926.


  11. jeez
    Posted May 31, 2008 at 2:50 PM | Permalink

    RE 6,

    No Mosh, it’s:

    The Bucket List–things to do before the [insert favorite policy noun(s) here] die.

  12. Kenneth Fritsch
    Posted May 31, 2008 at 3:12 PM | Permalink

    Just saying Steven Mosher, but wooden buckets to canvass around the turn of the century, buckets to inlets around 1945 and perhaps more insulated buckets to inlets later (1980ish?).

    Click to access 100694.pdf

  13. kim
    Posted May 31, 2008 at 4:47 PM | Permalink

    There’s a hole in my bucket, Dear Liza, Dear Liza, there’s a hole in my bucket, Dear Liza, a hole.
    Well, fix it, Dear Jimmie, Dear Jimmie, Dear Jimmie, well fix it, Dear Jimmie, Dear Jimmie, fix it.
    With what shall I fix it, Dear Liza, Dear Liza, with what shall I fix it, Dear Liza, with what?
    With adjustments, Dear Jimmie, Dear Jimmie, Dear Jimmie, with adjustments, Dear Jimmie, Dear Jimmie, what else?

  14. DocMartyn
    Posted May 31, 2008 at 6:37 PM | Permalink

    I suspect that a change in materials when in this way:-
    wooden buckets> enameled buckets> plastic buckets. In the UK, enameled steelware was around in my childhood (born in 1964). You can see a lot of them in restored canal barges (where they are hand painted). I think I have seen the original enameled buckets on HMS Belfast (in the Dentists and in the Officers Cabins).

  15. Geoff Sherrington
    Posted May 31, 2008 at 9:16 PM | Permalink

    It seems futile to argue that the present official SST adjustments are correct. Is it possible to correct them in a beneficial way? Probably. Is it worth while to strive for perfection? No, that would be exhausting and the T of the shallow sea is too noisy.

    If corrected plausibly, to what use can SST be put? There is a well-known problem of layered ocean models and another problem of which layers can or should be used in connection with statements on climate change.

    It seems that the upwards gradient from 1990 to now that appears in a lot of global surface temperature data (whether correct or not) is being used to drive policy. While strongly supporting solving the riddles of that pesky 0.3 degrees, my inclination would be to also concentrate on the last 2 decades. This is easier for the public to relate to. From figure 2f from Kent et al 2007 (see “Lost at Sea” leader), the main instrumental change in this time is graphed as a change from intake to hull sensors, so the problem might be simpler.

    I do not have a good impression from my limited reading, as to whether a global temperature increase of a postulated magnitude, in the atmosphere, will raise a nominated depth of water by a nominated number of degrees, and how rapidly. I see heat transfer calculations as a nice addition to the measurement problem. The phase connection between air, sea and land is still too uncertain.

    There is short-term information content in diurnal and seasonal air/sea temperature variation. I wonder if reports such as this might benefit the impact of corrected buckets/intakes exercises.

  16. Sylvain
    Posted Jun 1, 2008 at 1:16 AM | Permalink

    Excerpt from the paper:

    “as the data before ,1940 and after the mid-1960s are not expected to
    require further corrections for changes from uninsulated bucket to
    engine room intake measurements.”

    If the data from 45 to 65 are adjusted why shouldn’t they be adjusted after the mid-60 at the exact moment when the US% of observation decline to its minimum.
    If understood correctly the water intake was used by the USA and the bucket technic was used by the UK. If so then the bucket data from 65-70 should have more weight than ever before even if the adjustment to insulated bucket is less than adjustment to uninsulated or canvas bucket.

  17. Posted Jun 1, 2008 at 1:19 AM | Permalink

    Re# 14-Kim for poet laureate!

  18. KevinUK
    Posted Jun 1, 2008 at 3:18 AM | Permalink


    IMO what this sorry tale (pale?), and not just of bucket adjustments, tells us more than anything else is that the primary so-called ‘evidence’ for an increase in global temeprature (of which SST forms the biggest part) in the latter half of the 20th century as compared to the first half is at best dodgey, and at worst useless.
    [snip -policy]


  19. SOM
    Posted Jun 1, 2008 at 3:32 AM | Permalink

    Looking at the charts at the top if this thread why do the temps go down around 1900 just as the % UK Obs go down. Is this a coincidence?

  20. tty
    Posted Jun 1, 2008 at 5:00 AM | Permalink

    There is another possible bias with SST’s from 1939-45 (and 1917-18). Most merchant shipping moved in convoys during this period. In convoys ships were organized into columns, usually about 10 ships long. This means that unless a ship was first in column, any temperature measurement would be in waters that had been churned up by one or (more commonly) several other ships. It would thus not be a true surface temperature, but a sort of average of a several meters deep water layer. There might even be a slight warming effect by water that had passed through the condensors of other ship. This bias would of course affect to both bucket and intake measurements, though it might well different (or even have different signs).

    However it should be possible to check if this is a significant effect, since ships faster than 15 knots were allowed to sail independently.

  21. Paul Maynard
    Posted Jun 1, 2008 at 5:19 AM | Permalink

    SST Accuracy

    Referring to an exchange with Mark W on one of the threads below about the need for accuracy to 0.1C for climate as opposed to weather purposes. Ships measured water temperature for engine mamagement purposes not climate. Bearing in mind the circumstance in which the temps were taken whether by engine intake or buckets, it is hard to see accuracy better than 1C or even worse. See the comment on Watts Up… on this subject and the Independent article.

    So yet again we have statistical torture being applied to find a signal that is probably one tenth of the measurement accuracy. What I’m curious about therefore is whether the trends presented before and after “adjustments” are just various versions of the truth with the possibility that the actual trend is something quite different.



  22. Posted Jun 1, 2008 at 7:11 AM | Permalink

    Re# (22) Paul Maynard – It is the accuracy of the data that this issue has brought to light and I believe an error margin of +/- 1C may be too generous. One has to remember that as soon as a vessel gets out of continental waters all seas tend to roll and taking samples and measuring those samples in such conditions is not easy, even with by the most assiduous of technician. I have no idea how samples were taken at varying of depths of ocean and I’m afraid when asked to perform the near impossible there is a great temptation to produce fraudulent data, just to say the work was done.

  23. Bob S
    Posted Jun 1, 2008 at 8:48 AM | Permalink

    Considering the parallel discussions on the problems with the terrestrial stations, is it not possible that the entire surface temperature record of the last 150 years is bogus? It would appear that the only reliable indices of temperature change are the occasional historical reports on weather (freezing the Thames, grapes in England, year without summer, etc).

  24. Harry Eagar
    Posted Jun 1, 2008 at 11:07 AM | Permalink

    aegislaus, I am surprised that if temp is a process control, the measurements would be taken only every hour. A ship could enter a new water temperature regime a lot more often than that, wouldn’t it?

    Or do you mean that the temperature is monitored continuously but only entered in a log hourly?


    When I first became aware that there was a concern about global rising temperatures, something over 20 years ago, one of the first two or three statements I read concerned a Harvard researher who said he was applying to the Natural History Museum to borrow the bucket used by HMS Challenger to take sea surface samples. He wanted to determine the delta for convection losses/gains between the sea surface and the deck.

    I never heard how that came out. But my conclusion, not as a scientist but as a newspaper reporter, was that he was being way too fine, and that was the beginnng of my skepticism.

    Anyhow, although I have not bothered to do so, an inquiry to the museum should settle whether Royal Navy vessels took samples with wooden buckets in the 1870s.

  25. Roger Helvey
    Posted Jun 1, 2008 at 11:57 AM | Permalink

    I’ll try to comment and then go back to lurking. I’m a retired meteorologist. Back many years ago (before computers and satellites) when weather maps were plotted and analyzed by hand, we attempted to glean every bit of information to diagnose weather patterns out of the reported weather data. We were aware that certain individual observers were less conscientious than others. This was apparent in reported ship sea surface temperatures, as well as other parameters (including ship positions). For that reason the most careful analysts kept track of individual ships and observers (identified in the weather transmissions) as they progressed along their ocean routes. Some indication of the data quality could be obtained by comparison with neighboring ships and previous analysis of surrounding weather conditions. Suspicious sequences of unchanging values were sometimes obvious, implying that a lazy observer had just repeated or invented data. I’m not sure to what extent modern automated weather data quality control and analysis systems attempt to identify and correct for this type of ship/observer data idiosyncrasy.

  26. Kenneth Fritsch
    Posted Jun 1, 2008 at 12:35 PM | Permalink

    In doing an “audit”, if an auditor identified a situation where the assumptions did not warrant the conclusions (e.g. the IPCC hypothesis of a sudden and permanent changeover to engine inlets at Pearl Harbour), he would not substitute his own assumptions; he would reject the assumptions and throw the problem back at the authors. I’ve been pretty consistent about this in the proxy areas, avoiding the temptation to posit what really happened. I don’t want to go beyond this policy here either.

    Now I did a graphic showing the impact of abandoning the Pearl Harbour hypothesis under certain other assumptions. Please construe any such calculation as equivalent to a sensitivity calculation to show the effect of (say) excluding bristlecones, to illustrate the impact of certain assumptions, but not advocacy of a specific alternative. My language at the time may not have been explicit on this, but it’s the way that I do things.

    I do not expect any auditing purity from Steve M, nor do I require it for evaluating potential biases in his approach in these matters of climate science. I find half the enjoyment of these discussions and expositions owing to the nuanced auditing that Steve M refers to here as sensitivity calculations. I also look forward to seeing sensitivity calculations on the matter of SST adjustment/changing measurement techniques by Steve M or others who might be so inclined to attempt it here.

    In searching for an appropriate analogy here for Steve M maintaining his auditing purity in spite of sensitivity calculations, I could only think of those who might have maintained their (claim to) chastity despite some healthy and pleasurable experiences with sensitivity issues.

    Long live sensitivity calculations — here at CA.

  27. MPaul
    Posted Jun 1, 2008 at 12:36 PM | Permalink

    (1) What percentage of the samples were taken on sunny days vs cloudy days (effects radiative heat transfer)?
    (2) What percentage of the samples were hauled up on the sunny side of the ship vs the shaded side?
    (3) What is the distribution of the lag time between when the sample was taken vs when the temp was measured?
    (4) What’s the distribution of ‘fullness’ of the buckets (presumably half full buckets reach thermal equilibrium faster than full buckets)?
    (5) How does moisture content affect the insulating properties of the bucket?
    (6) How does wood species affect the insulating properties of the buckets?

    etc., etc., etc…

    Does anyone really believe that an adjustment method can be found for this?

  28. MarkW
    Posted Jun 1, 2008 at 1:53 PM | Permalink

    I believe this has been discussed before, but another issue that could affect any attempt to compare SST numbers taken during the war vs. those before and after. During the way, many ships travelled in convoys. As a result a smaller percentage of the ocean’s surface was being sampled.

  29. MarkW
    Posted Jun 1, 2008 at 1:58 PM | Permalink

    I’ve read that ships crossing the Atlantic like to catch a ride on the Gulf Stream. Not only is there a branch of the stream running across the N. Atlantic going west to east, there is also a southern branch running from somewhere around the Azores to near the Carribean that runs east to west.

    So ships running from N. America to Europe and from Europe to N. America can benefit from ridding the current.
    The problem is that the location of the current changes from day to day and month to month. Once satellites started taking pictures of the oceans, the job of locating where the Gulf Stream is today, got easier. As a result, ships should have been able to spend a greater percentage of their time in the stream.

    Now the Gulf Stream is warmer than the rest of the Atlantic, so …

  30. Paul Maynard
    Posted Jun 1, 2008 at 2:16 PM | Permalink

    And clearly the Gulf Stream is teleconnected to the Bristlecone pines so there we have the complete explanation for all warming. Apart from when the Gulf Stream flips as per Goreballs.


    Paul Maynard

  31. Willis Eschenbach
    Posted Jun 1, 2008 at 2:45 PM | Permalink

    Steve, there’s something badly wrong with the Thompson et al. figures for the percentages of observations by ships.

    The problem is most clearly visible in 1960. If the US contribution went from about half to about zero, the UK contribution must have represented a larger proportion of the remaining measurements. Ceteris paribus, it should have approximately doubled … but it didn’t change at all.

    Anyone have any ideas how that could have happened? It defies the laws of mathematics, unless someone else stepped in to fill up the US shoes for exactly that amount of time, and then stepped back out again … not bloody likely.

    To me, in addition to highlighting the lack of internal review by the authors, this just highlights the foolishness of our current peer review system. We need to make two changes, and urgently:

    1. The review should be double-blind, rather than single-blind as at present. That is, the reviewers should not know until the paper is published who the authors were.

    2. When the paper is published, the reviewers names and their comments should be made public. No more secretly giving your good friend’s paper a pass.

    Alas, I digress again …


  32. Phil.
    Posted Jun 1, 2008 at 2:58 PM | Permalink

    Both the British and US merchant fleets dropped because of the growth of flags of convenience, Liberia Cyprus etc. there’s no reason to suppose that there’s a cpmplementary relationship between the two.

  33. stan
    Posted Jun 1, 2008 at 3:11 PM | Permalink

    Glad to see someone mention the gulfstream. We now know that the oceans have bands of water at different temps and these bands, just like the gulfstream, change location. How do the researchers account for this?

    Temperatures recorded by sailors on ships in these circumstances have to be considered to have a pretty wide margin for measurement error. Slicing and dicing this garbage with sophisticated statistical methods may make the garbage look prettier, but given the extremely fine conclusions that the researchers come up with it still looks like garbage to me.

  34. Willis Eschenbach
    Posted Jun 1, 2008 at 3:18 PM | Permalink

    Phil, you say:

    Both the British and US merchant fleets dropped because of the growth of flags of convenience, Liberia Cyprus etc. there’s no reason to suppose that there’s a cpmplementary relationship between the two.

    Nope. We’re looking, not at absolute numbers, but at percentages of the total. The US went from 50% to zero % in what looks like one year. If that happens, either:

    a) Someone else picked up the slack (who and how), or

    b) Everyone else’s percentages must have doubled.

    Now, it’s possible that the whole US fleet went to foreign registry in a single year around 1964, and that half of them returned to US registry in a single year around 1968 … it just doesn’t seem very likely. However, stranger things have happened.

    Either way, the authors should have adjusted it (if it is from foreign flagging, it wouldn’t make any difference to the means of temperature measurement) or at a minimum commented on it.


  35. tty
    Posted Jun 1, 2008 at 3:35 PM | Permalink

    “During the way, many ships travelled in convoys. As a result a smaller percentage of the ocean’s surface was being sampled”.

    Yes and no. It is true that convoys restricted shipping to certain areas, but on the other hand in e. g. the North Atlantic where most shipping usually follows the great circle routes convoys were often diverted long distances to avoid U-boat wolfpacks, sometimes all the way up to Greenland or down to the Azores.
    Also during the war there was a lot of shipping moving in areas that are very thinly covered in peacetime (e. g. the Iceland-Murmansk routo, the area around the Solomons and New Guinea)

  36. steven mosher
    Posted Jun 1, 2008 at 3:57 PM | Permalink

    re 13. very interesting.

    I’m wondering that with the “adjustimications” Post 1945 and pre 1975
    if the change points wont vanish… and we get a more begign regime…

    A long while back I looked at STT-LAND.

    It was an interesting curve. The odd part of the curve was from 1945-1970 or so.

  37. steven mosher
    Posted Jun 1, 2008 at 4:09 PM | Permalink

    re 32. Willis. They need to Murder board this stuff.

  38. Pete
    Posted Jun 1, 2008 at 4:12 PM | Permalink

    I imagine the U.S. Senate debating the Lieberman-Warner carbon tax (I mean cap and trade) legislation this week will want to be sure the need is well grounded. If ocean data is the main contributor then bucket data quality should be on the table.

    I also imagine that they will fund an improved data collection program going forward to support whatever regulatory implementation comes out of it and perhaps even fund a detailed survey of all past SST data.

    My last bit of imagining is that little buckets become the symbol of the AGW skeptics.

  39. Dave Clarke
    Posted Jun 1, 2008 at 4:27 PM | Permalink

    CA readers who have not read the entire Nature article (available here on the Heat is Online) might be interested in these two excerpts:

    “It [i.e. the explanation of the 1945 cooling anomaly] is welcome news for climate modellers. The post-war temperature anomaly has been grossly outside the range of all computer-based climate reconstructions considered by the Intergovernmental Panel on Climate Change (IPCC).”

    “For example, the gradual shift since the 1970s from (warm-biased) ship-based measurements to (cold-biased) drifting buoys has probably led to a slight underestimate of SST warming, says Richard Reynolds of NOAA’s National Climatic Data Center in Asheville, North Carolina.”

    This suggests that, when the adjustments implied by historical changes in measurement techniques and data sources are properly applied, the past SST record will end up cooler on the whole relative to current temperatures, and will also provide an even better fit to climate models.

  40. Posted Jun 1, 2008 at 5:39 PM | Permalink

    A response to Real Climate’s dismissal of this issue:


  41. Kenneth Fritsch
    Posted Jun 1, 2008 at 6:35 PM | Permalink

    Re: David Clarke @#40

    Climate researchers can now start setting the twentieth-century temperature record straight. The abrupt drop in 1945 will then probably disappear, but what the corrected time series will look like is not yet clear

    David, the above comment that I excerpted from your link is probably more in tune with the thinking of many of the readers/posters at CA. We are not so quick to move-on. I suspect that there is much analyses to do on this issue before one can conclude how it will stand with the climate models. Steve M has sourced some suggested timings and magnitudes of transitions in the buckets (of several types) to inlets that could weigh heavily on any conclusions.

    Without a clear understanding of how the transition break will be allocated one wonders how some climate scientist are so sure that their models will redeemed — and would that be the original models or those adjusted for aerosols. Odd also that these problems, like evidently with climate models, was not well publicized or that the surface temperature data set keepers have suddenly become aware that their QC could allow large errors such as this one.

  42. Michael Jankowski
    Posted Jun 1, 2008 at 7:50 PM | Permalink

    Re#39, that the fix will “provide an even better fit to climate models” was also Gavin’s rally cry when the GISS Y2k error came up. When “consensus” climate science says it is right, it is right. To argue otherwise is “denialism.” When “consensus” climate science admits it is wrong, it is even MORE right.


    “Real Climate’s dismissal of this issue”

    – RC is the Black Knight of climate science. Lop off an arm, and ’tis but a flesh wound.

  43. jeez
    Posted Jun 1, 2008 at 7:55 PM | Permalink

    RE: 42, brilliant.

    Before: We know positive C02 forcing to be ginormous in order to overcome negative aerosol forcing. Our models demonstrate this precisely.

    Now: Aerosol forcing? ¿Que?

  44. David Smith
    Posted Jun 1, 2008 at 8:53 PM | Permalink

    The geographical imbalance in sampling may be a problem that’s as big as, or bigger than, changes in sampling methods like buckets vs intakes. Here, for illustration, are two maps of the ship data density used for Kaplan SST:



    Link to map-maker

  45. Steve McIntyre
    Posted Jun 1, 2008 at 9:47 PM | Permalink

    I’m trying to track down the provenance of the Wooden Bucket theory. Brooks 1926 mentions canvas buckets but not a whiff about wooden buckets. Aside from assertions by Jones and Folland, can anyone locate any evidence that wooden buckets were actually used in the 19th century for measuring SST?

  46. jeez
    Posted Jun 1, 2008 at 10:05 PM | Permalink

    I’m not sure if I’m being redundant here, but the file format the Gosta Atlas8 available on this data CD-Rom appears to include metadata for identifying measurements taken by Wooden Buckets. I’m not sure if any of these were in the 19th Century data.


  47. Sylvain
    Posted Jun 1, 2008 at 10:11 PM | Permalink

    Realclimate posted this photo the wooden bucket seems to have been used in pre-1900:

    Steve: I’m not asking what realclimate says – I already know that. I’m looking for evidence that these wooden buckets were actually used before 1900 for taking SST as opposed to swabbing decks.

  48. jeez
    Posted Jun 1, 2008 at 10:22 PM | Permalink

    The data appears to all be available here–not obvious as far as I could tell.


  49. jeez
    Posted Jun 1, 2008 at 10:25 PM | Permalink

    Sorry for the multiple posts, but it looks like they also provide a binary to ascii converter of their own.


    Steve: This only helps for Fortran users.

  50. jeez
    Posted Jun 1, 2008 at 10:34 PM | Permalink

    Nevermind, looks redundant–D’oh.

    Mr. Zamboni, clean up on aisle 5.

  51. Sylvain
    Posted Jun 1, 2008 at 10:44 PM | Permalink

    I found these papers that deal with bias between don’t know if it helps.


    Reconstruction of Monthly SST in the Tropical Pacific Ocean during 1868–1993Using Adaptive Climate Basis Functions

    S. D. Meyers*, J. J. O’Brien, and E. Thelin


    Bias Corrections for Historical Sea Surface Temperatures Based on Marine Air Temperatures

    Thomas M. Smith and Richard W. Reynolds

    “For the FP95 bias, it is estimated that the percentage of canvas versus wooden buckets increases linearly from 1856 to 1920”

  52. agesilaus
    Posted Jun 1, 2008 at 11:00 PM | Permalink

    Harry Eager:
    The watch went around hourly and recorded the readings on a paper logsheet. The themometer is not located where it can be easily seen either.

    This temperature device is a dial type bi-metallic thermometer stuck into a thermometer well in the circ water inlet pipe. Not a mercury type. At least when I was in. These read in degrees F not C and I don’t recall that we recorded the readings to closer than 1 degree. Probably not. So the readings would be taken to roughly a 0.5 deg C.

    I don’t recall calibrating those bimetallic thermometer either. Maybe during shipyard periods, about once every 5 years for USN vessels.

    In more modern ships it probably is a RTD that is wired to a computer and continiously monitored.


  53. Bill Norton
    Posted Jun 2, 2008 at 9:33 AM | Permalink

    FYI regarding Tsw inlet:

    In a steam-driven USN ship, it is the Main Condenser vacuum gauge that is continuously monitored, not Tsw inlet. Any lowering of the vacuum reading is immediately investigated, and Tsw in and out is checked as well as condenser hotwell level and valve positions.

    A salty watch officer or watch supervisor can feel the ship rocking as it passes the “wall” of the Gulf Stream and win a coke from the engine watch by betting that Tsw inlet has gone up (or down) at least 2 degrees (F) since the last hour.

    All deckhouse logs and main engineering logs are legal documents retained at the Navy Historical Society “http://www.history.navy.mil” but the engineering log that has Tsw inlet may not be. The main engineering log only records the overall plant data for midnight local. Many Navy logs since the time of Maury have been designed to gather dual-use data that someone, somewhere, collects for Navy operational/developmental use as well as research/scientific purposes. The FOIA coordinator at “http://www.cnsl.surfor.navy.mil” would probably be a good place to start for USN ships, although no one you would reach immediately would even understand the question.

    I also wonder what the USS Constitution does… I know it still takes logs, and may even have historical data, procedures, equipment, etc… hmmm… “http://www.ussconstitution.navy.mil/”


    /resume lurk

  54. Phil.
    Posted Jun 2, 2008 at 9:45 AM | Permalink

    Re #47

    I’m looking for evidence that these wooden buckets were actually used before 1900 for taking SST as opposed to swabbing decks.

    Even an expedition as meticulous in their descriptions of the apparatus and procedures used as the Challenger Expedition just says “It suffices to take a sample in a bucket, taking care that it is not contaminated with water, either from the scuppers, ……, to plunge a good thermometer into it, and observe it carefully.” There’s no indication in their report that they used anything other than the same buckets that they used for other purposes.

  55. Bill Norton
    Posted Jun 2, 2008 at 9:46 AM | Permalink

    re my #53,

    I sent a note to the USS Constitution PAO asking if they take measurements when doing the turnaround and if they have the historical procedures/data/equipment.

    Then it occurred to me that some US Coastie may know what Eagle does since she is underway all the time. Any Coasties?


  56. Tom Gray
    Posted Jun 2, 2008 at 10:53 AM | Permalink

    re 54

    This may be obvious to everybody but me but why would a canvas bucket be used instead of a wooden or steel bucket that would have other uses? I suppose that the sailors who read this blog would know the reason.

  57. dscott
    Posted Jun 2, 2008 at 11:12 AM | Permalink

    He found a sudden jump around 1941 in the difference between SST and all-hours air temperatures reported largely by the same ships.

    This statement just raised a question, remembering my history, the US was turning out ships via mass production methods during this time period in order to keep up with the Germans sinking them. What kind of thermometers were prevelent during this period? Given the mass production, disposable nature of those ships (very short life span) and buildem as cheap as possible, where alcohol based thermometers used? What is the error bar on alcohol thermometers versus mercury thermometers? http://www.is.mines.edu/ehs/policies/HgThemo.doc pay attention to page two on accuracy: In general, non-mercury thermometers have a lower usable temperature range than mercury-containing units. Non-mercury thermometers have scale divisions equal to mercury thermometers with most scale divisions ranging from 0.1° to 1° C. Be familiar with the accuracy of the non-mercury replacement thermometer you would like EHS to order. Accuracy decreases approximately ± 1º C for every 100º C increase in temperature. For non-mercury thermometers that measure up to 200º C, thermometer accuracy decreases ± 2º C.

    After WWII when commercial ships were built to last and carry large quantities of cargo, I would expect that mercury thermometers would have become the predominate sensing device.

    While hunting around for some historical info. I came across this paper detailing the weather of 1939/40 being a cold snap being described as the coldest in 150 years. http://www.seaclimate.com/2/2_11.html

  58. dscott
    Posted Jun 2, 2008 at 12:09 PM | Permalink

    “Accuracy decreases approximately ± 1º C for every 100º C increase in temperature.”

    That means a .3 C for every 33 1/2 C. Ambient air temp. being usually higher than water temp. (except for winter) would immediately introduce an error with alcohol based thermometers. Then comes the point of where the thermometer reading the air temp. is located on the ship. On the bridge some feet away from the nearest door???? What did a typical weather station of that era look like, because it was certainly all manual requiring someone to write down the readings.

    Here is a reference from Wattsupwiththat on this very subject by a retired Marine Engineer, a person in the know: http://wattsupwiththat.wordpress.com/2008/05/30/buckets-inlets-sst%E2%80%99s-and-all-that-part-2/ They were alcohol thermometers, so here is your .3C jump!

  59. Harry Eagar
    Posted Jun 2, 2008 at 8:32 PM | Permalink

    The ramp up of US merchant ship construction was barely under way in 1941. If there was a changeover, it happened a good many years — at least 3 — later.

    Liberty ships weren’t junked after the war, either. I used to pass a working LST while driving to work in Norfolk, Va., up to 1976.

  60. Posted Jun 3, 2008 at 1:06 AM | Permalink

    Re dscott says:
    June 2nd, 2008 at 11:12 am

    He found a sudden jump around 1941 in the difference between SST and all-hours air temperatures reported largely by the same ships.

    I don’t suppose there is any chance that the data is right? I see no real reason for this sudden conversion to a non-aerosol world other than a manipulation of the data to fit the model — the Folland and Parker correction was ‘found’ originally for just that purpose. ‘What does the data say’ must be a better and more scientific approach than ‘what can we squeeze the data into saying’.

    http://www.climateaudit.org/index.php?p=226 shows the F&P correction in 2 graphs, NH and SH. The uncorrected sharp upwards excursion in ’39/’40 is neatly smoothed away by F&P. The latest correction addresses the end of the WWII warm excursion but not the beginning. Why this asymmetry?


  61. Luis Dias
    Posted Jun 3, 2008 at 7:56 AM | Permalink

    I did say that it was blood for Real Climate and not otherwise. It is a revised trend upwards in the past, clearly contradicting someone’s theory that whenever there was a revision in the past, it was downwards, and in the present, upwards…

    What I also think is that I’m really even more confused than ever to the actual trends.

    The first place to suggest that there were problems in the WW2 adjustments and then that the post-WW2 downturn was an artefact of adjustments was Climate Audit, not Real Climate. My only interest is in reliable information and adequate due diligence, let the chips fall where they may. In this case, we proved that the due diligence had been inadequate by identifying a bias that subsequently warranted publication in Nature. It will be interesting to see the extent to which models are tuned to features like this in the record – what’s going to happen to aerosols?

  62. dscott
    Posted Jun 3, 2008 at 8:04 AM | Permalink

    #59, then if in 1940/41 a surge of Liberty ships came back into service from the mothball fleet left over from WWI ( http://www.usmm.org/libertyships.html ), I ask again the same question, what type of thermometers were used in the mass production of those vessels? They like the later Victory class ships of WWII were mass produced to overcome the sinkings by U-boats so I would be inclined to believe they would also have alcohol based thermometers as a cost savings measure given the mortality rate of the ships. Remember, we suppled tanks and ammo to the British under “lend-lease” ( http://www.loc.gov/exhibits/archives/worw.html ) in the run up to Pearl Harbor on Dec 7, 1941.

  63. Gerry Morrow
    Posted Jun 3, 2008 at 12:59 PM | Permalink

    Wrong method + Wrong data + Right answer = Wierd Scientists.

4 Trackbacks

  1. […] on May 31 in a head post here, I stated: A CA reader emailed me, observing that there may be relevant differences in insulated and […]

  2. By HadSST3 « Climate Audit on Jul 12, 2011 at 11:39 PM

    […] cold bias. I referred to these issues in the final post of my 2008 series on Thompson et al 2008 here. I did one more post on SST at the time: a short consideration of the ICOADS data set […]

  3. […] Lost at Sea: the Search Party […]

  4. […] Lost at Sea: the Search Party […]

%d bloggers like this: