Lyman et al have had to correct their paper on ocean cooling, as the effect that they observed has proved to derive in part from a bias from “Xpendable Bathy Thermograph (XBT) instruments”. They report
The rapid decrease in globally integrated upper (0—750 m) ocean heat content anomalies (OHCA) between 2003 and 2005 reported by Lyman et al. [2006] appears to be an artifact resulting from the combination of two different instrument biases recently discovered in the in situ profile data…
The second instrument bias arises in data from eXpendable BathyThermograph (XBT) instruments. These inexpensive probes were not designed to provide climate quality scientific data….
The other bias, however, appears to be caused by eXpendable BathyThermograph (XBT) data that are systematically warm [Gouretski and Koltermann, 2007] compared to other instruments. Characterization and removal of this bias will be required before the historical record of ocean heat content can be reevaluated
Perhaps they can now turn their attention to characterizing biases and uses of Recyclable AbathyThermograph (RAT) instruments (formerly known as “buckets”), also a form of inexpensive probe not designed to provide climate quality scientific data, used in an earlier generation of shallow ocean measurements, illustrated below. (As observed by commenters, the earlier generation of inexpensive probes measured shallow (abathos) water and not deep water (bathos)).
Figure: Recyclable AbathyThermograph (RAT) instruments, formerly known as “buckets”.
Climate Audit 
47 Comments
LOL. What do the Xpendable Blathys look like?
“…climate quality scientific data…”
That’s funny.
Here’s another oddity about the adjustments made in the RBT record, when theoretically uninsulated buckets were phased out:
The adjustment was almost three times as large in the Southern Hemisphere as in the Northern … what’s up with that?
w.
Once again it seems that anything that might even create doubt about the A in AGW is closely audited/checked – another hypocracy for bender’s list, perhaps?
Bender, do you have a count so far? Will Steve let you post a list? (assuming you are ready to “publish” this list)
Neil, he published it some time back, but I think he’s added a few items by now.
#3, Willis
Isn’t that the ratio of ocean area between the Southern and Northern Hemispheres? Not that that’s the right way to make the correction…
re: #4 Neil,
Why is specialization hypocracy? Steve’s one man. The IPCC is how many, a couple thousand? Considering that none of them try to look at the weaknesses in A in AGW, why is it a problem for you that someone else does?
And, BTW, Steve generally doesn’t allow discussion of aspects of AGW which he can’t monitor. So unlike the IPCCers who rarely do more than shake their heads when the extremists distort their findings, Steve keeps a somewhat taut rein on things here. “I.e. object to my science all you want, but keep on topic or keep the volume down.”
re: #7 Dave,
I think you mis-understand – I fully support Steve’s position on full disclosure and replicability/audability. IMO, anything less is not science, it’s speculation/propoganda/religion (pick whichever you prefer). My comment was directed at those who make these “adjustments”, who seemingly quibble over 100ths of a degree in the satellite data, yet ignore several degrees of unaccountable/inauditable/unknown adjustments in, for eg, HADCRU GTA data.
Politicians do this sort of thing, and it’s expected from them. Scientists should not, nor is it expected (at least in public announcements of imminent doom).
What I particularly like about this blog is that it’s based on data and methods, and anyone can post – even those who disagree with Steve’s position. This is sadly lacking in most climate blogs (RC being the extreme opposite, for example)
The oceans are different sizes, but they should heat/cool at approximately the same rates … which, as the graph shows, they do, except for the adjustment.
w
re: #8 Sorry Neil,
But I think you can see that your statement was ambiguous.
#1 jae :
XBTs are kind of tubes with electronics filling which are shot from a running ship, and while falling down through the water column, send up a record of the pressure/temperature profile through a wire. They are lost in the act; this is regarded cheaper than stoppping the ship for an hour or more in order to send down and haul a reusable device.
Steve : Neither specimen from your neat exhibit of buckets will serve for that purpose, since whereas you can lower a bucket to any desired depth (‘bath..’, Greek, stands for ‘deep’), when hauling it the contents will mix with water from the above layers. This is why buckets were never considered appropriate tools for measurements at depth, not to speak of other deficiencies. Good so, because such a practice would invariably fall victim to the brilliant scrutiny of CAuditors, and all was in vain.
Secondly, why didn’t you quote from the article the problems with the other instrument type, which poduced the
and which the authors list first? In that case, a software error was detected, something which as well may happen with satellite soundings too, I guess.
For the XBTs, as the cited article tells us, modelling of the falling rate is crucial, probably because the response times of the pressure and temperature sensors are different. So there is a mismatch between simultaneously recorded values. This needs a correction, which seems to have been inadequate in the past.
I only wonder, if there exist some temperature measurements that can be taken at face value, just raw data as they are.
Neil #8
In this case it’s business.
Steve, the Lyman correction derives from TWO biases. You edited nearly all reference to the other out of your quote, and wrote your article as if the other doesnt exist.
The major issue derives from a known software error effecting a subset of the modern ARGOS sensors. It is known exactly which records are affected by this error, and the major part of the Lyman correction is due to dropping those records from the analysis. Lyman reports that 30% of the incorrect records can be corrected exactly, and the remainder ‘approximately” and that data will be useful after correction.
The XBT ‘warm’ bias is less well understood, but seems, as Ulises alludes to above, to come from a combination of errors in the sink-rate profiles for the instruments, and different lag times on the pressure and temperature sensors. They have made a correction based on a bias analysis. As I understand it, doing an analysis on uncorrected XBT data through the modern period also yields warming where the modern cooling was reported. In any case this is one of two errors being corrected, and probably the lesser of the two.
Buckets are not any kind off Bathythermograph. They collect SURFACE water, whereas the XBTs and ARGOS readings extend to hundreds of meters below the surface – thus the “Bathy” prefix. But I can see how your recent penchant for snark over substance would lead you to ignore this obvious fact.
AND you also ignore that this analysis is an effort to get accurate ocean heat content numbers that are NOT dependent on issues of variability in surface sampling technique. IOW, they are doing exactly what you snark at them to do – trying to get more accurate values for oceanic heat content.
Lee, do you agree that the 19th century RBTs shown above are an “inexpensive probe not designed to provide climate quality scientific data”?
Unfortunately, nude data not exist…
In deference to commenters who observed that bathos is deep water, while the recyclable thermographs illustrated above probed shallow water (abathos), I have edited this post to refer to Recyclable AbathyThermographs, also known as RATs, instead of Recyclable BathyThermographs (RBTs) , as in the original post.
Sure, Steve. I’ll also point out that it is the data we have, and analysis of repeated coarse measurements can yield a finer result than that obtainalbe for each individual measurements by the limitations of the individual instruments.
And I’ll further point out that it is general practice and common courtesy, when editing a post on which others have commented, to indicate the specific edits and not just make a general comment that doesn’t actually mention that you have made edits.
And I’ll even further point out that your one-line response concentrates on continuing the snark you began here, and ignores essentially all the substantive commentary, including the addition of the information you edited out of your quote and failed to include in your article.
Lee, go away. You have no sense of humor, and you’re nothing but a troll.
Further you don’t know what you pretend you know.
That’s possible if there’s no bias in the instrument. But that’s precisely the problem with RATs. Now it may be possible to determine at least roughly what the value of the bias is, but that’s also precisely Steve’s point. The historical temperature records (at least the RAT-derived ones) need a major Augean effort.
Dardinger.
No. Steve’s “point” here is buried under snark, and a false presentation of the paper he cites. I have a fine sense of humor – I just don’t happen to think that a snarky misrepresentation of the science is all that funny in a context of supposed serious examination of science. Of course, tis possible that Steve has decided that humor trumps science, in which case I’ll know precisely how seriously to take him.
Also, burying an acknowledgment of an edit in the comments section, instead of making it clear in the post itself, would be considered dishonest, not humorous, at most blogs I take seriously.
re: #10 Dave,
No problem. In truth, I didn’t think it was ambiguous, but it clearly was given your response. Like all forms of written communications, there is always plenty of room for mis-understandings and providing all respondants remain calm and logical until positions are made clear to all concerned, I have no problem. On that point, I’d like to thank the few dendro people who *do* hang around for seemingly taking the same approach (eg twq).
Like many “lurkers” here that rarely post, I am trying to get my head around, not just the issues, but the science behind them. I’ll freely admit a lot of the math is over my head, but at least here I can ask “dumb questions” and get some pointers on where and how to improve my (limited) knowledge. And something to keep in mind, I think, is that my engineering background has lead me to a number of general conclusions that I’d like to share with you:
1) It’s easier to answer a dumb question than it is to fix a stupid mistake.
and
2) Taking the time to explain what you’re doing and why to someone with limited knowledge of the subject at hand, is often times a very useful way to gain an insight into the subject that you wouldn’t otherwise reach.
I feel that point 2 is often neglected in science – to many researchers detriment. My short time in research several years ago reinforced this strongly with me, and its apparent from Steve’s “outsider” impact on, for eg MBH9X, it’s something that is needed.
If you don’t like it here then go away and stay away, Lee! You’re a waste of time replying to anyway. I’d be interested in seeing what you’d consider funny. Not!
I used to hang out on the Mensa forum on Compuserve. I never joined Mensa though I was elgible but I did enjoy the reparte on-line even when I disagreed with the others which was usually. The point is that they were clever and others played off of anything clever which was brought up. I can recognize that sort of clevereness in Steve’s “snarky” posts here. I wouldn’t be as pliant as he is in changing things but I’m absolutely sure it’s not because he’s trying to fool anyone. Just as I’m sure he’s not making a “false pesentation” here. People have to do a little work and go read articles you know.
And I note that you managed, by turning the blame back on Steve, to avoid admitting that you were wrong in claiming that everything could be made well by using millions of measurements. Oh, well, enough wasted time.
Steve frequently asks readers to point out any errors, freely admits errors and then fixes them. I have yet to see Lee admit to or fix an error. But then, maybe he doesn’t make any.
guffaw…
Dardinger, I am one of two posters who have made substantive comments on the misrepresentations of the Lyman correction, and NO ONE has commented on the substantive content of either post.
I made the obvious point that many coarse measurements can yield a finer result, and that one has to work with historical the data one has. All measurements can be subject to bias – I never disputed that, and in fact that was one point of my post. The Argos system IS designed to yield climate quality data – and it has a bias that is being corrected.
The measurement errors that are described here seem both signficant in magnitude and, in retrospect, kind of obvious. Is there some accepted procedure in this field for establishing a “stopping condition” for measurement error correction?
Ok, Lee, one last remark and then I’m out of this “conversation”.
Steve M was NOT trying to “present” the Lyman corrections. He was using the article (which he linked to as I’m sure you noticed), to point out the deficiencies in the earlier data in a light-hearted way. For some strange reason you (and Ulises) chose to ignore what the post was about and try to use it as a way to attack Steve, as is your wont. The post was meant to be humorous as can be easily seen from the visual of the RATs and the naming of them. He didn’t consider the other correction as the post wasn’t about the corrections per se but simply the funny play on words. And then when he was given some information by Ulises as to the nature of the difference between the expendable probes and the old buckets he just HAD to change the post to match the new acronym which came into being. He didn’t mention the change in the post itself because the post wasn’t, strictly speaking, a serious one, where changes might mean something. It did have a bit of a serious point; the need to examine bias in bucket temps, but that wasn’t the reason for the post, IMO. The purpose was to lighten up a bit from a rather long run of serious posts the past few days. So what thread does Jokester Lee jump in on? QED
Lee, you say:
Are you truly surprised? You come in, insult everyone in sight, toss out a couple of trivial (albeit substantive) ideas and complain because no one comments on your ideas? Your main contributions so far have been:
LEE’S SUBSTANTIVE POST #14
1) You say Steve is trying to be sneaky by not mentioning both types of error discussed in the Lyman paper … what, you think those of us who care haven’t already read the Lyman paper? Steve is under no obligation to mention every detail of every paper he comments on.
2) Then you go on to tell us what is in the Lyman paper, as if you were summarizing it for a class paper. Why should we care about your summary of Lyman? You have not introduced a single new idea, just rehashed Lymans work … you’ll excuse me if I don’t comment on that “substantive content”.
3) You inform us that buckets are not bathythermographs, because they collect surface water, not deep water … my only comment on that substantive content is, have you ever heard of humor?
4) You say Steve is trying to “ignore that this analysis is an effort to get accurate ocean heat content numbers that are NOT dependent on issues of variability in surface sampling technique.” … I haven’t a clue what you’re on about here. In fact Steve is not ignoring their analysis, he is encouraging them to apply that same level of analysis to the sea surface temperature record.
Then we have LEE’S SUBSTANTIVE POST #18, which contains several complaints and one brilliant substantive insight:
1) Repeated measurements can increase the accuracy of a given instrumental result … whoa, be still my beating heart, that is so novel. However, it is also totally irrelevant to the question of the bucket measurements. You sure you’re following this story? The problem is not the accuracy of the bucket measurements, it’s the huge adjustment to the record in 1945.
Dave Dardinger, to his credit, did reply to the substantive part of your post #18. You, to your discredit, ignored his reply and then falsely said that no one commented on your substantive content.
Then you have LEE”S SUBSTANTIVE POST #20, which contains …
1) Nothing substantive at all, just complaints about how the blog is run.
That’s it. That’s your substantive comments. Buckets are not bathythermographs, and repeated measurements improve accuracy. If you truly are expecting a serious discussion of these substantive issues, I suggest you call up your grade school teacher. If you want to get some traction and some respect here, come up with some new, substantive ideas. If you don’t like Steve’s sense of humor, tough, go play somewhere else. Me, I find his comments funny. I also think that given the constant abuse that is heaped on Steve, both in print and by fools who come here and do nothing but complain, he has maintained a praiseworthy low level of snarkiness.
The thing you don’t seem to get is that the people on this blog are producing new work, and investigating new ideas, and asking new questions, and coming to new conclusions. UC, and bender, and Steve M, and Jean S., and a host of others here are actually doing the hard yards, reading the documents, re-running the numbers, examining the data, and seeing if the work was done right. When you start doing that, people will start commenting on your posts, because at that point you will actually have made some substantive points.
Until then, don’t expect much response to your “substantive points” like repeated measurements improve accuracy …
w.
Guys,
Lee is, as usual, just looking for a reason to be upset.
Lee’s complaints might be valid if the edit were to a core issue. But whether the probes in question are properly labeled bathos or abathos is hardly germain to the topic at hand.
everyone,
Yes, I got the ‘point” of SteveM’s post here. It was snarky, insulting, indirect, misleading on the cited content, and designed seemingly to ensure no substantive response on either the topic he snarks at, or the paper he cites. I responded in kind – although I’ll freely admit that I don’t do snark anywhere near as well as Steve does – but with an attempt to include more substance than the original post. Steve didn’t only just concentrate on just one part of the paper – he made a misleading claim about the contents of the paper, which I addressed by -yes- summarizing the parts that I felt he was misleading about and explaining why I felt it was misleading. I am hardly the first to do this, an effort I see often praised here. If you, Dardinger, are arguing that apparently misleading claims don’t need a response because people can just go read the paper, and that corrections of apparently misleading claims are not substantive, then what the hell is the point of any of this discussion at all? Y’all could just go read the papers you dispute, in silence, for yourselves. In fact, why don’t you all just go away and do exactly that (mirroring Dardinger, elsewhere)?
Yes, Dardinger, I am following this issue. Steve’s snarky comment here in the thread was about the measurements themselves, not adjustments to them. He didn’t ask if the adjustment was ‘climate quality’,’ He asked if buckets are “are an “inexpensive probe not designed to provide climate quality scientific data”?” I answered, and added the point about multiple measurements. I have seen people right here on climate audit directly ask how one gets values in the .1C range from raw measurements in the 1C range, so this seemed relevant, especially since Steve’s snark could be read to imply that this was an issue.
As Dave Dardinger observed, this post was not an effort to summarize the entire paper, as is obvious. I fail to see that I made any “misleading” claims about the paper. I wrote one sentence, which seems accurate to me and quoted from the paper.
In prior posts on buckets, I’ve spend considerable amount of time on the Folland adjustments, notably the assumption that all measurements were switched to engine inlets in 1941 – notwithstanding the fact that buckets continued to be the dominant form of known measurements as late as 1970. The adjustment here is as large as the effect. Under such circumstances, people should spend a lot of time analyzing and justifying the adjustments. That’s what I posted about – not the estimation of 0.1 deg C from a bunch of 1 deg C precision measurements. That’s a different issue, which I haven’t discussed here.
The Lyman experience draws attention to the need to carefully examine biases in measurement methods. Problems arose here with relatively well-controlled instrumentation. All the more reason to redouble efforts to examine potential bias with even more poorly controlled instrumentation and to properly account for such issues in confidence interval estimates. Your indignation would be better directed at the failure of peer-reviewed articles used by IPCC to adequate address such issues.
Lee, believe me when I observe that your most serious rendition of the subject matter at hand on this thread makes me no less concerned about large adjustments to data (relative to differences considered critical) then Steve M’s more humorous one. In fact I got a bigger chuckle out of your account since it appeared that you felt that perhaps a serious tone in the description of the adjustment would lend more credence to it.
I sometimes think you are much too concerned about the readers at this blog being misled by the tone of the discussion. In my view most of them have more than sufficient sophistication to see through all the peripheral stuff in evaluating the point being made and determining how well the evidence provided supports it.
Of course, having slick instruments, and full depth profiling, is no guaranty of statistically meaningful results:
Click to access pacific_abyss_v3.pdf
Lee, given your substantive claim that:
I thought you might be interested in this description of the first bathythermograph …
w.
I think not. The rest of 29 is pretty close to lunacy.
Re #33 : How ever this bathybucket was equipped, the ‘graph’ term does not apply, because it does no recording. This holds for the whole collection exposed at the top of the thread. And only one of those (the middle one in panel a.) deserves the ‘thermo’ term because it has a thermometer built in. So the RAT naming is not overly appropriate. I would also prefer ‘reusable’ over ‘recyclable’, since the latter in common usage involves some fundamental processing. The buckets are emptied (hopefully) before the next use, but not molten and newly forged.
The original bathythermograph, by the way, is torpedo-shaped (NO bucket resemblance and fully reusable) and is towed behind the ship. It records pressure/temperature by a mechanically moved pen which scratches a curve on a metal-covered glass slide.
In the end, I have to apologize for spreading misinformation in my #11 : The XBT does not have a pressure sensor, but rather depth is estimated via the falling-rate. Which makes it even more crucial…
Recommended reading : Gouretski, V., and K. P. Koltermann (2007), How much is the ocean really warming?, Geophys. Res. Lett., 34
#33,
I remember reading something about those early buckets having a cover that could be snapped shut by pulling on a second cord.
I don’t know how tight the seal was.
re: #35 Ulises,
Neither does a phonograph. And a telegraph, relies on a human recording the “graph” at the end, as does a RAT.
RE: #32 – Any comments, especially from stats experts, on this NOAA paper?
#36. I think that this problem was discussed in the song: There’s a hole in my Recylable ABathythermograph, dear Liza, dear Liza”…
From a link in Unthreaded #8 and post #354:
Click to access lindzen-nature-of-arguments.pdf
Lindzen explains the dangers of making corrections/adjustments to previously collected climate data. Very serious stuff.
#39,
Then fix it dear Willy, dear Willy …
Sure. There are scientists falsacionist (trial and error) and scientists falsificationist (I do not remember,in the absence of another explanation explanation, in accord with canonical and “independent” studies etc..). Only Mannist scientists a free error or correction.
hypocrisy. not hypocracy. although if we accept this malcoinage, it would mean “a government characterized by a its lack of rules” or, simply, “underruled”. Kind of like the Internet.
Bump:
Click to access pacific_abyss_v3.pdf
RE: #44 – I’m surprised that some of the statistically oriented folks here have not jumped all over this paper. So much low hanging fruit, so little time ….
Steve, I took a look at their paper a few weeks ago, and I couldn’t understand enough of what they were doing to comment. They say:
They go on to describe how they have adjusted for autocorrelation:
So we have data that has been:
1. Measured.
2. Linearly transformed.
3) Transformed by the equation of state, which is (to a good approximation) cubic in potential temperature, quadratic in pressure, and linear in salinity.
4) Low-pass filtered vertically with a 40-dbar half width Hanning filter.
5) Sub-sampled at 10 dbar intervals.
6) Interpolated every 0.033° latitude or longitude using a cubic interpolation at each pressure level.
Now, they say that the number of degrees of freedom for this is the sampling range (in degrees) divided by the “integral spatial scale”. This is referenced to “Von Storch and Zwiers 2001” … but they neglect to put the exact reference into the reference list at the end of the paper.
Because of this, it’s not clear what definition of “integral spatial scale” they are using. There are two main divisions of integral scales, Eulerian and Lagrangian. Within these divisions, there are several methods:
Area under the full autocorrelation function.
Area under the part of the autocorrelation function up to the first zero crossing.
Area under the part of the autocorrelation function to the point where it drops to 1/e.
Sum of squares of values up to one or the other of the above stopping points.
Depending on which one is chosen, the values will be quite different. In addition, their method for estimating the number of degrees of freedom does not seem to include the number of samples taken … but I may be just misunderstanding the paragraph immediately above.
Now, at the end of all of that, they are reporting a warming on the order of 0.05°C … wish I knew more about their methods, but that seems to be at or beyond the limits of detection. I’m not saying it’s wrong … just that it’s not substantiated. Remember that, because of the “equation of state” transform, that any errors in both salinity and pressure will appear in their temperature record. Remember also that they are reporting a change of two parts in a hundred thousand … like I said, the limits of detection or beyond.
So, I wasn’t ignoring the paper … just lacking data.
w.
RE: #46 – This seems to be even more contorted than BCP proxies!