ClimateAudit has posted about stations in Peru. Steve McIntyre complains about the large adjustments made by NASA GISS to some of the data series. In fact, he displays graphs showing some of those adjustments, the largest of which are for Puerto Maldon, which has more than a 3-degree adjustment between 1958 and 1988. As usual, a question is asked which includes a none-too-subtle (in fact rather obvious) implication of deliberate misdeeds by NASA GISS:
The annual variation in monthly temperatures is only a few degrees. Given this very placid temperature variation, how can Hansen simply assume that something happened that threw the measurements off by 3 deg C?
Perhaps I can ask a question: How can Steve McIntyre simply ASSUME that James Hansen would “simply assume”?
Here’s the data for Puerto Maldon (raw monthly data, anomalies):
Clearly there’s a sizeable cooling trend in the raw data. Why would adjustments be made? We already know the answer, and so does Steve McIntyre! Non-rural stations are adjusted so that the trend matches the trend in nearby rural stations. Let’s take a look at some of that data, from the two nearest rural stations that cover most of the time interval covered by Puerto Maldon, namely Cobija and Rurrenabaque (raw monthly data, anomalies), superimposed on the data from Puerto Maldon (Puerto Maldon in black, Cobija in red, Rurrenabaque in blue):
Hmmm… One of these things is not like the others. That would be Puerto Maldon.
We can see the same thing if we compute annual averages for the period of overlap:
And it becomes even more obvious if we add trend lines to this data:
In the same time period (1958-1988) during which Puerto Maldon raw data show more than 2 deg.C cooling, its neighbors show about 1 deg.C warming. To make the Puerto Maldon trend match its rural neighbors (which will include a lot more than I’ve shown here), it needs an adjustment amounting to about 3 deg.C during that time period.
And that’s what it has.
Hansen (or anybody else at NASA) didn’t “simply assume” that a 3-degree error occured at Puerto Maldon. THAT’S WHAT THE DATA TELL US. It seems that McIntyre is willing to criticize the data and to nurture conspiracy theories about how it’s adjusted, but the one thing he’s not willing to do with the data is pay attention to it.
The most amazing thing about McIntyre’s entire charade is that he and his ilk already know what the procedure is (to adjust non-rural stations so that the trend matches that of nearby rural stations) but despite his repeated protestations that he’s just trying to improve the data, it seems that he didn’t even bother to LOOK at it. That’s one of the reasons I believe that Steve McIntyre lacks both the honesty, and the competence, to audit surface temperature data.
560 responses so far ↓
chriscolose // February 26, 2008 at 12:35 am
What is incredibly worrysome, if that people take McIntrye serious. It is most unfortuante the audit and “watts” mission to destroy the temperature record has led people into accepting the blogosphere over institutions like NASA, and the peer-reviewed literature.
McIntrye has absolutely no intention to “improve” anything. He complains about a decade old paper, and runs a fraudulent mission to discredit those who dedicate their lives to the analysis of climate science, and where we may be headed. I am not sure that much constructive has come out of him, and if so, it has been strongly outweighed by him misleading people.
Patrick Hadley // February 26, 2008 at 12:45 am
Another interesting post, thank you. As far as I can tell Climate Audit mainly seems to find stations that have been adjusted so that:
- either the past is made cooler than the raw data would suggest;
- or the present is made warmer than the raw data shows before adjustment.
I suppose if there are also lots of stations where NASA GISS has made the opposite adjustment to the data (making the past warmer and present cooler) that would be a pretty effective rebuttal to the CA auditing.
[Response: Take a look at Tokyo.]
Hank Roberts // February 26, 2008 at 1:28 am
Is that the same as this?
http://climexp.knmi.nl/gettempall.cgi?someone@somewhere+84658+PUERTO_MALDON+
P. Lewis // February 26, 2008 at 2:09 am
As Jimi Hendrix once said “… Where are you on this ah hot cold summer?”
It’s a real pity “Hot Gets a Little Cold” was written and sung by Cyndi Lauper … rather than by The Three Degrees.
Simon D // February 26, 2008 at 3:23 am
It might be instructive to show some examples where the adjustment has gone in the other direction, simply to convince any skeptical readers of the impartiality of the adjustment process.
Dano // February 26, 2008 at 4:31 am
The most amazing thing about McIntyre’s entire charade is that he and his ilk already know what the procedure is (to adjust non-rural stations so that the trend matches that of nearby rural stations) but despite his repeated protestations that he’s just trying to improve the data, it seems that he didn’t even bother to LOOK at it. That’s one of the reasons I believe that Steve McIntyre lacks both the honesty, and the competence, to audit surface temperature data.
Yes.
And?
Oh, yes. There’s nothing else. Man, behind curtain.
Thank you for your work, making complex things simple. Here’s to a large audience one day.
Best,
D
Heretic // February 26, 2008 at 4:41 am
Same old same old on CA and surfacestations. Their vocations have always been clear. I grow less interested in the all bizwax everytime they try a new trick.
EliRabett // February 26, 2008 at 4:56 am
Worse, falsifying the corrections has no effect on the GISS global trend or the maps, because only the rural stations contribute. Of course, Peterson and Hansen have shown that using uncorrected non rural stations has only a minor effect on the global trends. Toot sweet.
Hank Roberts // February 26, 2008 at 6:05 am
So who filled in all the missing data on those graphs? Aren’t these the originals, with blank spaces where the data file shows missing data?
http://climexp.knmi.nl/gettempall.cgi?someone@somewhere+84658+PUERTO_MALDON+
By the way the online tools at that site for making comparisons are astonishingly varied. Look into this.
Hank Roberts // February 26, 2008 at 6:07 am
Viz:
Climate Explorer
Time series
PUERTO MALDON mean temperature (all)
Retrieving data from GHCN v2 (all) database …
PUERTO MALDON (PERU), coordinates: -12.63N, -69.20E, 266m (prob: 210m), WMO station code: 84658 PUERTO MALDON , temp from v2.mean nodup [Celsius], (postscript version, raw data, netcdf)
Raw data as text, note missing info:
http://climexp.knmi.nl/data/ta84658.dat
Martin Vermeer // February 26, 2008 at 7:15 am
This is probably a question that could be answered by looking at the original documents, but as there are many people on this list that have already done so, I’ll just ask :-) Is it true that in the GISS reduction
1) urban stations have their trend replaced by a weighted average constructed from nearby rural stations, and
2) independently, _all_ stations are cross-validated against their neighbours, and any outlier trends are replaced by a weighted average constructed from these neighbours’ data?
That’s the way I would do it.
It is of course sobering to realize that some stations contain error patterns on the 3 degree level. But then, they were never designed with climatology in mind.
Undoubtedly many more stations contain smaller error patterns that the cross-validation will never detect, so these get propagated into the final trends. The hypothesis is then, probably valid, that these error patterns are quasi-random and as often up as down.
fred // February 26, 2008 at 7:37 am
I wonder whether any of you have actually read the CA thread?
What the CA thread shows happening in Peru (and elsewhere) is not what is supposed to happen. What is supposed to happen is that rural stations = population less than 10k are to be used to adjust urban ones. Leave aside whether this is legitimate or useful.
What CA shows, with names and numbers, is stations which do not meet the criteria for ‘rural’ being used to adjust others which do not either. Go over and look. Or do I have to repost it before you will read it?
Eli and others making his point have to answer a fundamental question. If ‘using uncorrected non rural stations has only a minor effect on the global trends’ then what are we gaining from using any others? Why exactly are we doing it? We are making unjustifiable changes in some data series to gain nothing at all.
You cannot have it all ways at once. These are intellectual contortions in aid of defending the indefensible to no particular purpose anyway.
I have come across one case of what seems like a legitimate series of adjustments in the helpful references cited on the other thread. That was the adjustments to Quebec. They had detailed station histories, and using statistical analysis were able to tie down the element contributed by the detailed account of the changes of siting or surroundings.
The readings on the other thread made the point that to justify any changes to the readings, you have to have very good metadata. How good is the Peruvian metadata, where is it archived, and how was it used. What about Dawson? What about Vermont? Was the available metadata used to make the adjustments there?
Quebec is the standard here. I do not believe it is generally met elsewhere in the US, let alone the ROW. If you can’t do it right, stop doing it, because you are not adjusting, you are simply changing the instrumental record.
And tell me again, I didn’t hear the first time. It was about statistics. If the spread among adjusters is 3 degrees for some years, how, when you do your further processing, do you reflect this uncertainty when you pick on particular estimate, and has Hanson done that?
None of this proves it is not warming. We know it is. Those of us who have lived long enough know from experience. We also know it from plant data and other indicators. That is not an issue. The issue is just whether legitimate scientific method has been followed here in the treatment of surface temperature readings, and it has not.
[Response: I can tell that you actually paid attention to what's written here -- because it's making you squirm. The whole point of this post is that NASA GISS did in fact follow legitimate scientific method. THEY DID EXACTLY WHAT THEY CLAIM, IN EXACTLY THE WAY THEY SAID THEY WOULD, USING A PROCEDURE DESIGNED TO BE COMPLETELY OBJECTIVE.
I'll agree with you on this: the issue is NOT about whether some stations in Peru have incorrect population data. It's all about CA doing everything in its power to get people like you to believe that NASA is violating legitimate scientific method, and to suggest (by innuendo, the coward's way) that such violations are deliberate and malicious. It appears your belief is somewhat shaken; I get the distinct impression that reading here is making you uncomfortable, and I'll guess this is because you're so deeply invested in believing the crap coming from CA that when I show how really wrong they are it makes you feel as though maybe you don't really understand what's going on.
You should trust that feeling.
The degree to which you've been brainwashed is indicated by your plaintive "If you can’t do it right, stop doing it..." Here's an opportunity for you to benefit from an epiphany: the REAL GOAL of CA is to get us to stop doing it, because they really really hate what the truth is.
You've bought into it so much they've got you convinced that if the standard set in Quebec isn't upheld everywhere in the world, then we should just stop doing it. Forget about using all available information in the best way possible, don't even try. Don't bother continually improving the data handling to make it better. And if there's a single station anywhere in the world where we can take a picture with asphalt within a country mile of the temperature sensor, then the entire global temperature effort is a sham, so just stop trying. That's what CA really wants you to believe because, again, if you start to believe the *truth* that would make them miserable.
Are there stations with incorrect population assignments? Certainly. Is the temperature record flawless? Of course not. Is there any part of the GISS procedure which introduces a cooling bias into the analysis? No. Is there any valid reason to believe that the global temperature trend is any less than indicated by the GISS analysis? No. It might be less -- but there's just as much chance that it's MORE.
If you, or Steve McIntyre, really want to improve our understanding of global temperature based on what's happening in Peru, here's a suggested title for the upcoming paper you'll be submitting to a peer-reviewed journal: "Effect of Improved South American Population Data on Surface Temperature Analysis." I'll predict the exact time at which you, or McIntyre, will submit this work: never. Frankly, when LEGITIMATE SCIENTIFIC METHOD is applied to that issue, it's vastly more likely to come from James Hansen and/or his colleagues than from Steve McIntyre and/or any of his.
As a final note, I do wish that you and your friends could learn how to spell. There's no "o" in "Hansen."]
Peter Hearnden // February 26, 2008 at 9:58 am
Wrt Puerto Maldon*, is one question to ask (since the explanation presented wont convince any ‘auditers’): ‘Why does it show the trend it does while others around don’t?’. If that can explained then the ‘Oh, but how do we know the data needs adjusting?’ cries can be addressed.
* Is it actually called Puerto Maldonado? It has has an interesting history according to it’s wiki entry.
MarkR // February 26, 2008 at 10:17 am
Have you seen a picture of the weather station site? Find one and then tell me the weather station is “Urban”. By the way, as you only allow posts which support you point of view I’ll post a copy on Climate Audit.
PS Why not try doing a global Raw v Adjusted data comparison, and the see which regions have the largest adjustment?
EW // February 26, 2008 at 11:10 am
And why is Port Maldon cooling? There must be some physical cause, no?
Ade // February 26, 2008 at 12:31 pm
“We already know the answer, and so does Steve McIntyre!”
Errr, yes. And Steve points it out in his comments section.
It doesn’t answer the question *why* does the urban station need a 3 degree adjustment, though, does it?
Shouldn’t the UHI effect have biased the station in completely the opposite direction?
Don’t you want to know why?
Ed Davies // February 26, 2008 at 1:06 pm
Playing Devil’s Advocate:
1) What is the assumed physical cause of the cooling in this urban station (as EW asked)? At first glance one would expect a warming unless there has been an unusual change to the urban area.
2) If long-term trends of urban sites are unreliable, why not just discard them?
3) Doesn’t adjusting urban sites to follow local rural sites have the same effect as giving rural sites which happen to have a nearby urban site extra weight?
These questions are probably answered in the GISS papers but since what you’ve already posted here is presumably also described there it would make sense for you to complete your comments.
P. Lewis // February 26, 2008 at 1:07 pm
Re MarkR
A few observations:
(1) Anyone who actually frequents this blog would know “as you only allow posts which support you point of view” is clearly an error.
(2) Puerto Maldon is an airport serving Puerto Maldonado, a city near the Bolivian border, in the Amazon having a population between 10,000 and 50,000 (apparently)
Tony Edwards // February 26, 2008 at 1:26 pm
As far as I can see, this whole business of adjusting the data is fraught with potential error. As an engineer and machinist, I know that what you measure is what you measure, subject to the tolerance of your equipment . You don’t measure the diameter of something and then “adjust” the number for any reason before you machine the part. Similarly, the temperature that was taken would originally only have been in error by the built in error of the thermometer. Trying to hindcast for microclimate, site faults, UHI, etc. can never make the answer anything other than more dubious. Throw into the mixture a failure to show the exact operations that have been applied and the general public should have extreme doubts as to the accuracy of the final anomolys that get published.
null{} // February 26, 2008 at 1:29 pm
” … it needs an adjustment amounting to about 3 deg.C during that time period.”
Or maybe the instrument at Puerto Maldonado is going bad. Or maybe instruments at all three locations are incorrect. Or maybe all instruments are not in fact measuring the same quantity. Or …
But of course the *only* correct alternations of the data are those that produce a *trend* of increasing ‘temperature’ with time.
Given all the unknowns, way not rely on the LLN and random variations of noise?
Thanks for any insights.
Bob North // February 26, 2008 at 1:56 pm
Looking at the first Puerto Maldon graph, it appears that there is a discontiniuity around 1975. Similar discontinuities were present and much more distinct for the Cuzco and Pucallpa stations. Perhaps there was an equipment change around this time or a change in time of measurement. If that was the case, a direct shift of the data either before or after the discontinuity would seem to be more appropriate than a trend adjustment.
Bob North
R Clark // February 26, 2008 at 2:26 pm
Just a quick comment/question from a long time reader first time poster.
If the quality of data from a given station is so questionable that we have to make adjustments then isn’t the work being done by surfacestations.org even more important?
One would think their findings would only add to the justification for adjustments and support your position.
Just wondering…
fred // February 26, 2008 at 3:56 pm
Tamino, no I am not getting uncomfortable about anything to do with CA. I read it with interest, and sometimes with agreement. Like I read this. One of the most unattractive foibles of the AGW movement is to keep imputing all kinds of imaginary motives and feelings to people who raise questions. I thought you were above that stuff.
[Response: Are you really that naive? If you actually believe that McIntyre doesn't have a prejudicial purpose then you need more than an epiphany. You need to take your blinders off.]
It was also not CA who sent me to Quebec, it was a poster here on another thread who kindly gave us three interesting public references, one of which was an article about adjusting Quebec among other places. It was not CA who suggested that Quebec is a gold standard. I thought of that myself. It is. Now, whether everything has to be as good as Quebec, that can be left to the professionals to debate. But there is no doubt, its a good one, and I don’t have any worries about those particular adjustments, and it proves it can be done.
[Response: It was certainly CA that got you saying "Peru" just about every chance you got for the last several days, with strong statements here about how that effectively damns the surface temperature analysis. You made those statements based on McIntyre's posts, right? But I'll bet that *neither* of you bothered to look at the data enough to find out what the trend in neighboring stations might be. Did you?
As for Quebec, in your earlier comment you said:
The implication of that is quite clear: if it's not as good as Quebec, stop doing it. What else could you be suggesting? But now you say:
Which is it?]
McIntyre’s view is not the view you attributed to me. Don’t think its mine actually. His view is (and it seems to be also his view on proxies) that there is data in at least some of these records. That what we need to do is get at it. Perhaps by starting with a subset of records which are ‘gold standard’ and seeing how far we can move out from there.
[Response: I guess you really are that naive. That's what McIntyre repeatedly insists is his purpose -- but all he ever DOES is try to embarrass the surface record. Far more often than he insists his purpose is noble, he makes some suggestion, some innuendo, some implication, of either gross incompetence or deliberate deception or both by modern climate scientists. It's his bread and butter.
Actions speak louder than words.]
He raises the question, to which so far there has been no answer: If the adjustment is supposed to adjust by comparing to stations located in places with less than 10k population, why exactly are adjustments being made by comparing to towns with over 20k?
[Response: IF that's being done (I don't take McIntyre's word for it) then it's because the population classification is in error. If that's true then the classification should be changed.
That does NOT mean that GISS is violating the standards of scientific method, it just means there are errors in the metadata. But McIntyre has managed to persuade you that "The issue is just whether legitimate scientific method has been followed here in the treatment of surface temperature readings, and it has not." That's a direct quote from you.
So put up or shut up: in what way did GISS violate legitimate scientific method? Unless you answer THAT question, or admit that you were mistaken to make the claim, your credibility is nil.]
Let me tell you: I was never particularly sceptical about the surface temperature record until I started reading this site and encountering some of the tortuous defenses of truly crazed procedures with the data. Now, I am.
[Response: What truly crazed procedures? Just because CA calls 'em such and you repeat it, doesn't make it true. Answer the question or admit you're mistaken.]
I don’t know why GISS adopted and now retains the algorithms and methods it did. I don’t have much of a view of whether and in what direction they are distorting the trend. My view is, it is pretty clear they have to change, which is more interesting. Lets hope they are not in the same state of denial as most people here seem to be.
[Response: Let's sum up: 1. YOU DON'T KNOW why they do what they do. 2. YOU DIDN'T LOOK at the data from the rural neighbors (whether misclassified or not -- you didn't look at ANY of it). 3. But you still feel qualified to say that NASA has violated legitimate scientific method and that the procedures are "truly crazed." Did I get that right?]
And let us repeat the prime idiotic argument here. It makes no difference, is the cry. So tell me, if it makes no difference, why are we doing it?
[Response: You and Steve McIntyre can't think of a reason, so the idea must be "prime idiotic"? Well I guess I'll just have to post on that subject soon.
But not until I get answers from you. In what way did GISS violate legitimate scientific method? What procedures are "truly crazed"?]
Boris // February 26, 2008 at 3:58 pm
Nice post, Tamino.
Steve McIntyre has posted about 1,000 blog posts since his last publication. Apparently he hasn’t found anything worthy of sharing with the scientific community yet. (Keep in mind, even his hero Wegman said that blogs are not the proper place to discuss these kinds of issues.)
Also note that McIntyre referred to the folks at Real Climate as “vicious, little men” and you’ll perhaps see what I believe his true goal to be.
MarkR // February 26, 2008 at 4:14 pm
Clue for those who won’t check for themselves.
Google Earth for
Padre Aldamiz
Heretic // February 26, 2008 at 4:58 pm
“failure to show the exact operations that have been applied”
This statement shows that you have taken the whining at CA at face value without checking for yourself; incidentally, Tamino just showed the exact operations.
nanny_govt_sucks // February 26, 2008 at 5:35 pm
Why wasn’t the “nightlights” data used to adjust Peruvian urban sites, as was done in the USA? Why have two different adjustment methods depending on what flag the host country waves?
Adam // February 26, 2008 at 5:41 pm
“Wrt Puerto Maldon*, ….* Is it actually called Puerto Maldonado? It has has an interesting history according to it’s wiki entry.”
It is the same place, and it is an interesting one. The station is at the airport, also called “Padre Aldamiz”. The airport is about 1km-1m SE of the town. There is an object halfwat down the runway on the E side that *might* be the station (hard to tell from a sat photo).
The only history of the airport that I can find is that the main carrier stopped flying there in 1990 when it got closed down. Others now fly there, inc. the Peruvian military.
There is no info. that I can find on instruments, observers (even organisation), or possible moves.
As the station is/was in cleared forest land, it’s quite possible that it was affected by regrowth. But that is getting into guesswork. I would guess we’ll need a Spanish reader to find out more.
P. Lewis // February 26, 2008 at 5:48 pm
Fred, one has to be incredibly naive, incredibly stupid or incredibly mendacious if you believe
(1) The original data are almost invariably retained (though there are occasions when someone “boobs”). Why? Because people still want the original raw data for other purposes, at least one of which is to carry out more stringent/better homogenisation checks when such tests become available (or just plain different homogenisation checks, because one worker favours one suite of methods over another’s). That process has been going on for decades.
(2) The procedures outlined in Aguilar’s ppt presentation (or something similar) are carried out everywhere one wants to use the data to make comparisons between temperature series. There are literally hundreds, if not thousands, of papers in journals and conference proceedings to this effect.
(3) Do you think it’s only (near-)surface temperature data that are homogenized? Radiosonde data are, precipitation data are, atmospheric pressure data are, wind data are, …
dean_1230 // February 26, 2008 at 6:15 pm
In what way did GISS violate legitimate scientific method?
Well, if you assume that the sorting algorithm is doing what CA has determined that it’s doing (and there was some significant dialog on the FORTRAN code snippets as to exactly what it was doing), then the “legitimate scientific method” that was violated had to do with sloppy data handling. If Hansen wanted to use all sites with less than 10,000 people as rural, then there should have been some way to ensure that that was actually the case. When violations to that desire are found, they should be fixed.
Is Peru “cherry-picked”? Maybe (but CA says it wasn’t). Is Peru in conflict with what Hansen’s stated sorting algorithms? Yes, according to the data presented at CA.
So what do you do? You fix the errant datafiles and re-run the code. Steadfast refusal to do just that is NOT part of the scientific method. Likewise refusing to admit that there’s a problem is also not part of the scientific process.
[Response: It was NASA, not Steve McIntyre, who first indentified that there are problems using the population data supplied by GHCN. In fact that's the reason for using nightlights to classify stations in the U.S.:
(Hansen et al. 2001, J. Geophys. Res. vol. 106, noD20, pp. 23947-23963)
NASA isn’t refusing to admit that the population data has problems; they’re the ones who identified that the problem warrants a new approach to classification, and devised a method to do so. Here’s my guess: the improved method will be extended to the entire globe as soon as time, resources, and data permit.
So answer me this: Who has “steadfastly refused” to fix errant data files? Who has refused to admit that there’s a problem?
Steve McIntyre and his henchpeople have taken a problem which was identified by NASA, for which a new method was devised (by NASA), which is being applied in the U.S. (by NASA), done methodically (by NASA) rather than cherry-picked (by CA), and used it to slander NASA. That’s the BIGGEST problem in this whole mess. When will Steve McIntyre — and you, and Fred — stop “steadfastly refusing” to admit there’s a problem with CA?]
Phil. // February 26, 2008 at 6:35 pm
“It is the same place, and it is an interesting one. The station is at the airport, also called “Padre Aldamiz”. The airport is about 1km-1m SE of the town. There is an object halfwat down the runway on the E side that *might* be the station (hard to tell from a sat photo).”
Actually that’s the VOR station, they’re quite distinctive.
caerbannog // February 26, 2008 at 6:48 pm
It’s interesting to note that the Orland and Marysville USHCN stations are *still* featured front-and-center over at surfacestations.org, with the obvious purpose of giving the casual viewer the impression that new construction around the Marysville station (cell tower, air-conditioning units, etc) is responsible for the discrepancies in the temperature records.
A careful viewer would note that the temperature records for the two sites are strongly consistent with each other for the past 50 years or so, and would also note that the major temperature discrepancies occur prior to 1940 and pre-date all the man-made clutter around the Marysville station.
But careful viewers are not surfacestation.org’s intended audience.
Hank Roberts // February 26, 2008 at 6:50 pm
Nan, that’s been answered repeatedly. Are you just posting the question again for rhetoric?
Assuming you missed it before, you can tell where people live in the USA by how bright the area looks from space.
Is that true elsewhere? Look it up, e.g.
http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6VFX-4K1G554-1&_user=10&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=a7d1d93a878e2c2101b7e97ec4594ba1
ABSTRACT begins:
“Liberal reform programmes in developing countries were not designed with the need to extend electricity supplies to rural areas. This paper focuses on the Peruvian experience, examining the impact of electricity reform policies on the characteristics of rural electrification. In rural areas, electrification levels have traditionally been the lowest in the country – making them less or non-profitable for private firms. Only in 2002 did the government introduce a specific Rural Electrification Law, which was intended to promote electrification….”
NOTE that “rural” for “rural electrification” is not “rural” for temperature measurements.
Short answer is, there are no small towns in the USA that have no electric lights. There are many small towns in the rest of the world that don’t waste electricity sufficient to be lit up all night every night from space, and few countries tell the whole world where their electric grids are like the US does.
But you knew that, right?
MarkR // February 26, 2008 at 6:58 pm
The airstrip is 2.6 Nautical Miles from the town, and is mainly in “jungle”. The exact whereabouts of the weather station equipment is unknown. The station would appear to be rural, and therefor its down trending data should have been used to adjust the urban , and not the other way round as NASA GISS have done.
See how it helps to audit the data?
Now if some Spanish speaker were to email the airport, and ask for some photo’s of the weather station and environs, we could make some progress. A bit like Watts Gallery.
Tony Edwards // February 26, 2008 at 7:08 pm
[Response: I guess you really are that naive. That’s what McIntyre repeatedly insists is his purpose — but all he ever DOES is try to embarrass the surface record. Far more often than he insists his purpose is noble, he makes some suggestion, some innuendo, some implication, of either gross incompetence or deliberate deception or both by modern climate scientists. It’s his bread and butter.
Actions speak louder than words.]
Really, why does someone posting under a pseudonym have to be so exceptionally rude about someone who, so I believe, makes no money from his weblog, but is looking for answers which might get us closer to an accurate notion of what is happening to the world climate and why. Surely a serious scientist should be pleased that people are concerned enough about accuracy. As to Heretic’s suggestion that the operations “have been explained” why is it that the actual codes, operating procedures and algorithms haven’t, so far as I can ascertain, been put into the public domain?
Incidentally, the described procedure of adjusting urban areas to match up to adjacent (1000km is adjacent?) doesn’t seem to be a sensible way to get rid of any UHI. All it does is put spurious information, if you can call it that, into the mix. A better way, surely, would be to take the stations out of towns and site them in reliable locations.
Sometimes, when you are in a hole, the best thing to do is stop digging.
[Response: If Steve McIntyre were truly trying to "get us closer to an accurate notion of what is happening to the world climate," then I'd be one of his strongest supporters. I believe that claim is a LIE. McIntyre's real purpose is to discredit climate science.
You say that "the actual codes, operating procedures and algorithms haven't, so far as I can ascertain, been put into the public domain." You must not have tried very hard to "ascertain."]
Hank Roberts // February 26, 2008 at 7:15 pm
> (1000km is adjacent?)
Asked and answered repeatedly.
The scale takes into account the size of large weather systems. As those move through, temperature changes across a large area. You have to capture those changes at that scale to be able to tell them apart from climate trends.
Patrick Hadley // February 26, 2008 at 7:26 pm
As a global warming agnostic I am always happy to see comments such as those above pouring out scorn and bile on those who are investigating the theory.
When people who are genuinely unsure about AGW visit sites like this they will wonder why pseudonymous proponents of the theory find it necessary to be so rude about named people who are asking awkward questions. I have no doubt at all that the reaction of any fair-minded neutral would be to sympathise with those under attack.
[Response: If McIntyre were only asking awkward questions, he'd be helping. Alas, that's simply NOT TRUE. I question his motives because they are questionable. So here's a newsflash for you, Mr. Hadley: when, in my opinion, people behave in a despicable way, I'm gonna call it like I see it.
But I guess the case for CA is so weak that the only thing left is to resort to claiming that I'm rude, and pointing out that I post under a pseudonym.]
Tony Edwards // February 26, 2008 at 7:36 pm
You say that “the actual codes, operating procedures and algorithms haven’t, so far as I can ascertain, been put into the public domain.” You must not have tried very hard to “ascertain.”]
So, where are they, exactly?
[Response: These will get you off to a good start, and you may wish to examine references therein:
http://pubs.giss.nasa.gov/abstracts/1987/Hansen_Lebedeff.html
http://pubs.giss.nasa.gov/abstracts/1999/Hansen_etal.html
http://pubs.giss.nasa.gov/abstracts/2001/Hansen_etal.html
Here's the documentation:
http://data.giss.nasa.gov/gistemp/sources/gistemp.html
The programs are here:
http://data.giss.nasa.gov/gistemp/sources/
]
MarkR // February 26, 2008 at 8:00 pm
So, has anyone other than NASA GISS been able to run the NASA GISS programs?
You for example?
nanny_govt_sucks // February 26, 2008 at 8:06 pm
Well this is hardly justification for having two different methods when it comes to the USA vs. Western Europe or Canada. It seems Hansen just slapped this code together without too much thought or investigation into what might be an appropriate way to adjust for … whatever he’s adjusting for. That’s really not a criticism, as no one else seems to have put much work into this before Hansen, but there should be an openness and willingness to come to terms with these issues, and modifications to the adjustment algorithm as new information comes to light. Certainly there’s a better and more consistent way to adjust for any UHI.
David B. Benson // February 26, 2008 at 8:18 pm
Off-topic, Tamino, but I have recently been seeing quite a few comments about “the world has cooled for the last ten years”. I am hoping there is a link (somewhere) that I can suggest that the (crdious, naive) posters might care to read before repeating this claim.
Thanks.
[Response: Try this, this, this, and this.]
Hank Roberts // February 26, 2008 at 8:34 pm
Yeah, he could have used the flat-earth maps where the continents get wider the farther north you go.
Dean P // February 26, 2008 at 8:39 pm
“NASA isn’t refusing to admit that the population data has problems; they’re the ones who identified that the problem warrants a new approach to classification, and devised a method to do so. Here’s my guess: the improved method will be extended to the entire globe as soon as time, resources, and data permit.”
But wait, I thought the predominant belief here is that there’s no problem with the data? Your comment doesn’t square with there being no problem with the data.
I understand funding constraints. Since there’s a problem with the data, IDENTIFY & QUANTIFY the problem, don’t sit there and say that there’s no problem (as most of the people here are saying). If the error analysis shows that the problem is serious, then elevate the importance of fixing it.
[Response: You are simply mistaken.
The only reason you think anybody ever claimed that there's no problem with the data, is that you've been spoon-fed that propaganda (from CA, I wonder?). If there were no data problems, they wouldn't be extensively documented nor would procedures be devised to compensate for them. Why do you think adjustments exist? If there are no adjustments, denialists complain about errors in the raw data; if there are adjustments, denialists complain about the adjustments.
And they'll try very hard to make you (and others) believe that climate scientists claim there are no problems -- when it's the climate scientists who are working HARDEST to find what those problems are so they can be compensated in the most effective way possible.]
Tony Edwards // February 26, 2008 at 8:40 pm
Source code and documentation for GISTEMP software is available here. The programs are intended for installation and use on a computer with a Unix-like operating system.
* Download GISTEMP_sources.tar.gz.
This archive is approximately 2.2 MB. It was updated Sep. 10, 2007, to clarify the procedures of some steps; finally it was updated Oct 10, 2007, to simplify and speed up STEP2 (homogeneization)
Once you have gunzipped and de-tarred the archive, you will have a directory containing a text version of the documentation and several sub-directories, the latter each named STEP*. Please read the documentation and follow the instructions to go through the steps of the analysis, proceeding through each subdirectory.
The subdirectories contain software which you will need to compile, and in some cases install at particular locations on your computer. These include FORTRAN programs and C extensions to Python programs. Some Python programs make use of Berkeley DB files.
This is helpful? Just reading this would seem to suggest that you have to write half of the code yourself. And it says “updated” Sep 10, 2007, but was it available to the public before that?
[Response: You claimed that the procedures and programs are not in the public domain. I pointed out that you were mistaken. You asked where to find 'em. I answered.
But apparently it's beyond your ken actually to *use* them. Pity.]
jim // February 26, 2008 at 8:50 pm
David, this might be what you are looking for.
http://tamino.wordpress.com/2007/12/16/wiggles/
Hank Roberts // February 26, 2008 at 8:52 pm
Homework help for Tony:
http://web.archive.org/web/*/data.giss.nasa.gov/*
dhogaza // February 26, 2008 at 9:14 pm
Yes, it is. Sounds like they’ve included everything I’d need to make it work, if I were so inclined.
No, it says you’re going to have to compile the code yourself.
It’s not put together like a distribution of a commercial software product built for everyday software consumers.
What did you expect? Would you rather have them spend the money to do that sort of expensive (time is money) packaging or on improving the models? I know where I’d put my money if it were my choice.
Mitchell // February 26, 2008 at 9:37 pm
[Response: If McIntyre were only asking awkward questions, [...]. I question his motives because they are questionable.]
Let’s talk about science. Last time I checked, particular motives had nothing to do with science. If you judged, then accepted or refused, all the scientific work on the basis of the motives of its authors, where would that get you?
Patrick Hadley // February 26, 2008 at 9:47 pm
It is describing McIntyre’s behaviour as “despicable” that will raise flags in the minds of those reading this site who have not yet made their minds up about AGW.
How are you going to convince someone who comes to this site with an open mind, particularly someone who lives in a democracy where the right to challenge the authorities is revered, that there could possibly be anything “despicable” about McIntyre criticising the use of data and statistics by proponents of AGW?
[Response: Steve McIntyre has taken a problem which was identified by NASA, for which a new method was devised (by NASA), which is being applied in the U.S. (by NASA), done methodically (by NASA) rather than cherry-picked (by CA), and used it to criticize NASA.
What would you call that?]
Adam // February 26, 2008 at 10:08 pm
“Actually that’s the VOR station, they’re quite distinctive.”
Thanks. Did think it might be something like that. I’ve never actually spotted a square from a sat pic, so would have been surprised if I’d started now.
As for the GISS code, sounds like slightly less effort than half the OS packages I’ve installed on my system.
sod // February 26, 2008 at 10:19 pm
How are you going to convince someone who comes to this site with an open mind, particularly someone who lives in a democracy where the right to challenge the authorities is revered, that there could possibly be anything “despicable” about McIntyre criticising the use of data and statistics by proponents of AGW?
you did not understand what tamino said. someone who comes here with an open mind and really READS what is written, will come to a different conclusion.
no one wants to deny CA the right to “challenge” anyone. the problem is, that CA is leaving people with a wrong impression.
dhogaza // February 26, 2008 at 10:28 pm
What word would you use to describe someone who continuously insinuates that leading climate science researchers are engaged in scientific misconduct and fraud?
Are you aware of the seriousness of such claims?
McIntyre was given the opportunity to declare that he does not accuse climate scientists of scientific misconduct right here on Tamino’s blog, but refused to answer direct questions on the matter.
I’m sorry, but there’s nothing wrong with accurately labeling McIntyre’s behavior.
For Tony’s benefit, “OS” means “open source”, and in this case specifically those distributed in source form, therefore requiring compilation.
null{} // February 26, 2008 at 10:31 pm
Adam // February 26, 2008 at 10:08 pm
ok, report back when you’ve go it up and running.
Mitchell // February 26, 2008 at 10:42 pm
“the problem is, that CA is leaving people with a wrong impression.”
Why would you worry about the impression CA leaves?
Tony Edwards // February 26, 2008 at 10:50 pm
[Response: You claimed that the procedures and programs are not in the public domain. I pointed out that you were mistaken. You asked where to find ‘em. I answered.
But apparently it’s beyond your ken actually to *use* them. Pity.]
Never said I wanted to use them, my expertise lies in different fields, but there are others who have the requisite expertise who, so far, have failed to get the programmes to run.
sod,
“you did not understand what tamino said. someone who comes here with an open mind and really READS what is written, will come to a different conclusion.
no one wants to deny CA the right to “challenge” anyone. the problem is, that CA is leaving people with a wrong impression.”
What the hey does it matter what impression ClimatAudit leaves people with?
Up until recently, I was, indeed in most matters still am, an open-minded, reasonably intelligent and experienced person who is always up for debate. But harangue me, shout at me “The science is settled”, tell me that I am too stupid to understand the experts who are so far above me, and I dig my heels in. I can understand most things if they are explained. Saying it’s ex-officio and not open for debate brings total opposition.
As for “you did not understand what tamino said. someone who comes here with an open mind and really READS what is written, will come to a different conclusion.”, oh yes, I understand him/her but am not convinced.
Real data and logical justified exposition is what leads me to a conclusion, not climate models and postulates.
[Response: And how did you conclude that climate models (and postulates) can't lead you to a conclusion? Are they no evidence at all? If we were "debating" evolution, would you protest that genetic models and postulates are meaningless?
I very much doubt that you reached that opinion by examining real data and logical exposition. I have no reason to doubt that you are indeed a "reasonably intelligent and experienced person who is always up for debate." But you might be seriously in need of an open-mind check.]
Hank Roberts // February 26, 2008 at 11:26 pm
Tony writes:
> Up until recently, I was, indeed in most
> matters still am, an open-minded,
> reasonably intelligent and experienced
> person who is always up for debate
Science isn’t debate; you’re experiencing what’s called ‘hard argument’ — a quite different form of discourse. You’re expected to hold up your end with citable published studies.
All involved should remember or re-read this very thoughful post and remember you’re writing for the ages, not for the moment:
http://initforthegold.blogspot.com/2007/03/why-truth-is-losing-ground.html
…
the malefactors and their innocent sympathizers…. are trying, all too successfully, to convince their audience of several things including
* that there is a substantive scientific controversy (which delays and dilutes considerations of policy) about even the most well-established basic facts
* that the presentation of a consensus, even on totally unambiguously established results, is in itself an indication of dishonesty, bullying and arrogance
* that the motivation for all this fuss is a deeply corrupt scientific establishment that is motivated to lie because of huge and rapidly increasing grant money. (Don’t we wish…)
It all ties together into a pretty clever strategy:
1. say things that are exasperatingly ignorant
2. get real experts with little understanding of polemics to express their exasperation
3. cast that exasperation as arrogance and bullying
That’s not the whole strategy … but it’s emerging as a central feature.
The best counter is to avoid the appearance of arrogance in public forums. ….
——–end excerpt——-
And on the latter point, the best advice ever posted online, the last words from:
How To Ask Questions The Smart Way
Eric Steven Raymond
http://catb.org/~esr/faqs/smart-questions.html
How To Answer Questions in a Helpful Way
Be gentle. Problem-related stress can make people seem rude or stupid even when they’re not.
…There is no need of public humiliation for someone who may have made an honest mistake. A real newbie may not know how to search archives or where the FAQ is stored or posted.
…
Be humble and honest; set a good example for both the querent and your peers.
If you can’t help, don’t hinder. …
Ask probing questions to elicit more details. If you’re good at this, the querent will learn something — and so might you. Try to turn the bad question into a good one; remember we were all newbies once.
While just muttering RTFM is sometimes justified when replying to someone who is just a lazy slob, a pointer to documentation (even if it’s just a suggestion to google for a key phrase) is better.
If you’re going to answer the question at all, give good value. Don’t suggest kludgy workarounds when somebody is using the wrong tool or approach. Suggest good tools. Reframe the question.
Help your community learn from the question. When you field a good question, ask yourself “How would the relevant documentation or FAQ have to change so that nobody has to answer this again?” …
If you did research to answer the question, demonstrate your skills rather than writing as though you pulled the answer out of your butt. …teaching them research skills by example is teaching them to grow food for a lifetime.
———————–
Look back at this history from a century in the future. How did we do, dealing with what we barely were beginning ot understand? Did we leave the world better than we found it?
luminous beauty // February 26, 2008 at 11:32 pm
“…but there are others who have the requisite expertise who, so far, have failed to get the programmes to run.”
I scarce dare point out the oxymoron in this statement for fear one might construe the criticism as haranguing one as too stupid to understand.
The problem with creating a false impression is it is a lie.
That matters.
Competence, my friend, is measured by the limits of one’s capability.
dhogaza // February 26, 2008 at 11:50 pm
Perhaps their claims to having the requisite expertise is as overblown as their claim that a few photographs will overturn years of work by NASA scientists, etc etc etc?
Since you have no faith in models, I trust you don’t fly in modern airplanes?
David B. Benson // February 27, 2008 at 12:24 am
Tamino — Thanks for the links.
I’ll put all four to good use.
Patrick Hadley // February 27, 2008 at 12:39 am
The complaints that McIntyre has made against members of the AGW consensus are that they have made mistakes in selecting and analysing data. Irrespective of whether he is right or wrong in any of these complaints simply asking awkward questions and offering an opinion about possible errors cannot be “despicable”.
McIntyre publishes under his own name detailed descriptions of where he thinks there may be problems in the methods of some climate scientists. These scientists typically have serious resources behind them which they can and do use to respond to these criticisms. I really do not see how someone who is publishing research which he hopes will lead to a dramatic change in public policy and in the lifestyles of billions of people, should consider it “despicable” if someone tries to poke holes in that research. In a democracy that is simply bound to happen, and it is a very good thing.
dhogaza // February 27, 2008 at 1:25 am
He goes far beyond that, and as I said earlier, when given the chance to positively state that he does NOT believe that leading climate scientists are guilty of scientific misconduct and fraud, he chose not to make such a statement. He continues to imply misconduct by NASA scientists.
Beyond the big accusations regarding folks like Michael Mann, he’s directly accused Gavin Schmidt of violating government regulations on the publication of information due to his activities at RealClimate, and has directly accused him of misconduct by blogging there during work hours. He’s threatened to “expose” him to his superiors at NASA.
What does that have to do with “mistakes in data an analysis”?
Surely your ability to read english is not so impaired that you are unable to distinguish tabloid-quality dirt-flinging when you’re confronted by it in the writings of McIntyre and his band of buddies?
me no understand // February 27, 2008 at 1:26 am
Going back to your original post, could you explain why the correction to Puerto Malden stops at 1988 (I am looking at the graph on CA)? Do your 3 trend lines run parralel from 1988 onwards and what will happen if they diverge again in the future? What happened in 1988 to abruptly halt the need for a correction factor?
I am sorry if you have already answered these questions but the debate moved into snark mode so quickly and it is hard to concentrate on content when the insults start flying. Many thanks
[Response: To answer confidently, I'd have to track down all the rural stations within 1200 km of Puerto Maldon and compare the trend of the rural neighbors. But here's a guess:
GISS uses a 2-legged linear adjustment, with the "hinge point" allowed to vary, so that the trend at the non-rural station can be adjusted to give the best agreement of its trend with that of its rural neighbore. If you look again at the first graph, you'll see that around 1988 the behavior of the raw anomalies changes: the steady nosedive from 1958 onward finally comes to a halt. So it's not implausible that 1988 turns out to be the hinge point.]
Michael Tobis // February 27, 2008 at 2:07 am
It’s kind of awkward for me to post something intemperate here when Hank has yet again linked to my blog entry counseling patience and understanding of the nature of the game you are playing. SO I’ll duck by quoting someone else
There is a strong case that the game McIntyre et al is playing is not honest. P Z Myers at Pharyngula put it this way:
===
the need to defend the climate change denialist, McIntyre — so many of you, after carping that I’m not meeting your demands, are protesting that he’s not a denialist, and you aren’t denialists, and you’re all here in the cause of good science.
Bullshit.
My expertise is not in climate, but in biology, and I’m familiar with his type — it’s a common strategy among creationists, who do dearly love to collect complaints. There are people who put together a coherent picture of a scientific issue, who review lots of evidence and assemble a rational synthesis. They’re called scientists. Then there are the myopic little nitpickers, people who scurry about seeking little bits of garbage in the fabric of science (and of course, there are such flaws everywhere), and when they find some scrap of rot, they squeak triumphantly and hold it high and declare that the science everywhere is similarly corrupt. They lack perspective. They ignore everything that doesn’t fit their search criterion, and of course, they’re focused only on putrescence. They aren’t scientists, they’re more like rats.
And the worst of the rats are the sanctimonious ones that declare that they’re just ‘policing’ science. They aren’t. They’re just providing fodder for their fellow denialists, and like them all, have nothing of value to contribute to advance the conversation. You can quit whining that you and McIntyre are finding valid errors; it doesn’t matter, since you’re simultaneously spreading a plague of lies and ignorance as you go.
So bugger off, denialists. I am not impressed.
===
end quote.
It’s me again, polite and rational, reminding you that different folks play different games on the same playing field sometimes.
Boris // February 27, 2008 at 2:27 am
“McIntyre publishes under his own name”
McSteve doesn’t publish. He merely types. Do you see the difference?
[Response: McIntyre does have peer-reviewed publications, although all I've found (I really haven't looked that hard) are criticisms of the "hockey stick," many of which appeared in the highly dubious journal "Energy & Environment" (although there's at least one in Geophysical Research Letters).
Readers who have been following the series on PCA, in hopes that it will address hockey-stick critics: stay tuned.]
Hank Roberts // February 27, 2008 at 2:27 am
Chuckle. And I agree wholeheartedly, per Eric Raymond:
“… muttering RTFM is sometimes justified when replying to someone who is just a lazy slob, a pointer to documentation (even if it’s just a suggestion to google for a key phrase) is better….”
Give’em hell.
P. Lewis // February 27, 2008 at 2:46 am
So, Tony Edwards, are you saying that every casting or fabrication that has ever been machined in which an error has been made has gone to the scrap pile instead of somewhere else in the machine shop for rectification?
Are you telling me that no one has ever, say, reversed electro-discharge machine polarity to deposit, say, stainless steel in something like a gas turbine component that has been overmachined?
Such items were rarely scrapped in my day when I was in and around a tooling workshop, where errors were common, if not an everyday occurrence. Who in their right mind would scrap a casting costing on the order of tens of thousands of pounds and hundreds of hours of machine shop time when, with a little ingenuity and application, the problem can be ameliorated?
There are dangers in carrying out homogenisation incorrectly, but as a (suite of) statistical technique(s) it is a perfectly acceptable procedure that — at least as far as the SNHT procedure is concerned (and I’ve no doubt with other techniques, too) — has been tested on artificial data sets into which various inhomogeneities have been added and been found to be acceptable.
If you can’t see far enough, then try standing on the shoulders of someone who can, to paraphrase a rather famous scientist.
KH // February 27, 2008 at 3:40 am
There might be a good reason for the temperature trend at Puerto Maldonado. I visited Puerto Maldonado and area on a family vacation in 2005. I’m a bit surprised it has over 10,000 people. The roads are mostly dirt roads with lots of small motorcycles. The airport handles passenger planes (737’s). It is in a jungle area just across the Andes. When we arrived, the temperature was about 30 degrees Celcius. It stayed near that for a couple of days, then late one afternoon, there was a large thunderstorm and the temperature dropped to about 7 degrees Celcius and stayed near that all the next day. A mass of cold air had blown off the Bolivian Andes down into the jungle. I remember shivering in the open air airport waiting for the plane that we would be departing in, to arrive.
Puerto Maldonado appears to be in or close to a transition zone between a jungle climate and a mountain climate. Any trends seen in the climate there might not be very representative of either type of climate.
mmghosh // February 27, 2008 at 4:58 am
Excellent post Hank
Chris Colose // February 27, 2008 at 6:05 am
This discussion about climate audit vs. NASA does not appear to be very productive. The issue of surface stations, hockey sticks, and other such topics coming from his blog appears to get plenty of attention in blog protocol, and other secondary sources. McIntyre is obviously an interested and dedicated individual, but does not sit in the mainstream of research on these topics, has not done much refereed original research on it (primarily but not exclusively blogging), and is often (rightly or not) associated with one end of the political spectrum.
My advice to certain people here, is that if you are getting too much information from CA before checking with the relevant GISS publications on the topics, IPCC or NAS, etc then you are well off course for a dispassionate consideration of the issues.
There appears to be two audiences on subjects such as this–the ones that want to be entertained about the “controversy” that is AGW, and those that want to honestly improve every aspect of our understanding that we can.
The second crowd does not claim that every detail is solved, and no problems exist in the instrumental record. Life is not so easy. Understanding why adjustments or corrections take place is probably a good start rather than declaring that adjustements or corrections are a fraud.
To the first crowd, the details of individual stations are probably not relevant to you. I strongly suggest that people in this category look at the data from satellites, the surface record, air balloons, glacier balance records, ocean heat content, sea level rise, etc for a broad-brush picture of what is going on, and what the best scholarship has to say.
Patrick Hadley // February 27, 2008 at 9:53 am
KH, I don’t see how your suggestion that the cooling trend at Puerto Maldonado might be caused by its micro-climate could possibly be correct. Hansen has shown that the true trend in each weather station can be found by looking at the trend in stations up to 1200km away. The whole GISS methodology is based on this principle.
me no understand // February 27, 2008 at 11:37 am
Thank you for your reply but I am still confused. I want to know what happens to your 3 trend lines post 1988.
In your original post you justified the 3 deg C correction, applied until 1988 and then stopped, by reference to only 2 rural sites. With the benefit of up-to-date data for these 3 sites does the 3 deg C correction factor still seem reasonable?
What procedure is in place if the trends continue to diverge post 88? Will a new correction factor be applied post 88 or is the post 88 data considered accurate? Is this what you mean by a 2-legged linear adjustment?
What physical phenomenon, in your view, has caused the need for this correction factor?
Is the post 88 data available for these 3 sites?
I hope you don’t consider me presumptuous by seeking further clarification but the maths seems pretty simple and showing that these corrections make sense of the post 88 data is the first step to convincing a layman like me that these urban corrections are worthwhile.
[Response: A 2-legged adjustment refers to a line with a "turn" or "bend" at one point. So it's a "bent line" with the bend at a time which is determined by computing the best fit, but the line is straight both before and after the bend.
There's no post-1988 data for Cobija or Rurrenabaque. I selected those stations because they're classified rural in the GHCN, they cover the time period discussed in the CA post (the adjustment is nonzero from 1958 to 1988 but zero beyond), and they're not too far away. Then nearest rural station which covers the entire time span of the Puerto Maldon data is Cruzeiro Do S, more than twice as distant as either Cobija or Rurrenabaque.
If you look again at the very first graph, it's evident that the trend in Puerto Maldon changes significantly about 1988, from steeply downward to upward. Hence it's really no surprise that if you fit a "bent line" to the difference between that station and its rural neighbors, the bend will be about that time.
You shouldn't be too surprised that the correction after 1988 is exactly zero. Keep in mind that monthly average temperature is rounded to the nearest 0.1 deg.C. Hence if the Puerto Maldon adjustment post-1988 ranges from -0.049 to +0.049 deg.C, when rounded off the adjustment will be zero.
I don't know why the Puerto Maldon data 1958-1988 show such a strong cooling trend while its neighbors don't. My first guess would be that the trend is false, caused by undocumented changes in observing practice or instrumentation, or by land use changes. But it's not impossible that it's correct; contrary to what some may claim, "spatial coherence" does *not* mean that nearby locations share the same trend with certainty, just that it's far more likely than not. Therefore adjustments to bring the trends into agreement will give the overall most accurate picture of regional (and therefore global) trends, but *not* necessarily the most accurate picture of trends at an individual station. NASA emphasizes that the purpose of adjusted data is to estimate regional/global climate, but for studying individual locations it may be best to use the data without homogeneity adjustment.]
Tony Edwards // February 27, 2008 at 12:09 pm
P. Lewis // February 27, 2008 at 2:46 am
You have completely misunderstood my analogy. certainly, if an error is made, if possible, corrective action should be taken. My point was that if a measurement has been taken, say of a hole into which the part you are about to make should fit, you don’t then adjust the measurement to give something different.
So data that was taken a while back should be the start line and left alone. Recent data that has discoverable errors might need adjustment, but any correction or adjustment ought to be on a station by station basis not as a generic variation.
On the subject of station errors, I am amazed that the efforts being made by volunteers to investigate the state of the weather stations around the USA and collated on Anthony Watts’ site http://www.surfacestations.org
have drawn so much criticism. Surely any determination of defects in the data collecting system should be lauded and assisted. If decisions are to be made for the future they must be based on firm foundations, so why not stand on these efforts to get a better view of these foundations.
[Response: As I've said many many times, sincere and objective efforts to improve data quality are welcome. Prejudicial efforts intended to discredit rather than investigate, are not. See this.]
Adam // February 27, 2008 at 12:25 pm
“Now if some Spanish speaker were to email the airport, and ask for some photo’s of the weather station and environs, we could make some progress.”
Well you’d need photos going back to the 1960s to be useful. Plus it would help to have a station log including instruments, procedures, observers, etc. etc. Otherwise it won’t tell you much.
“ok, report back when you’ve go it up and running.”
What a silly thing to say. Why would you take my word on it?
Raven // February 27, 2008 at 1:09 pm
Chris Colose
“I strongly suggest that people in this category look at the data from satellites, the surface record, air balloons, glacier balance records, ocean heat content, sea level rise, etc for a broad-brush picture of what is going on, and what the best scholarship has to say.”
Virtually all of the metrics you indicate suggest that the world has stopped warming or that the rate of warming is nothing to be concerned about. For a long time GISS has stood out as the most alarmist of metrics. Perhaps it is time to do to GISS was Tamino has suggested be done to other data sets that do not match the others.
Before anyone responds by attempting to deny the obvious I suggest to look at this OLS trend for HasCRUT over the last 10 years:
http://rankexploits.com/musings/wp-content/uploads/2008/02/hadcrutandgiss7.jpg
The OHC, MSU and Sea Level metrics show similar trends.
[Response: What makes you think I haven't looked at the trend in HadCRUT3 over the last 10 years? You should pay closer attention.
You cherry-pick a HUGE el Nino event as your starting point, and the calculated trends are still positive. That's like asserting that people are not getting taller, and you'll prove it with the data -- but only if your trend *starts* with Shaquille O'Neill. Unfortunately for your case, Yao Ming walks in the door, so you then insist the trend is actually "negligible."
You need to read this and this. More important, you need to understand it. The "global warming stopped in 1998" mantra is perhaps the single most stupid claim of the denialist camp.]
Barton Paul Levenson // February 27, 2008 at 1:16 pm
Patrick Hadley writes:
[[The complaints that McIntyre has made against members of the AGW consensus are that they have made mistakes in selecting and analysing data. Irrespective of whether he is right or wrong in any of these complaints simply asking awkward questions and offering an opinion about possible errors cannot be “despicable”.]]
I agree. But lying about the motives of the scientists involved is despicable, and refusing to acknowledge refutations of flawed work is despicable. When I get a paper turned down by peer review, I don’t get all sulky and tell people there’s a conspiracy against me and the reviewers are arrogant. I try to fix the flaws in the paper.
Andy // February 27, 2008 at 1:44 pm
You didn’t comment wether eg Iquito is concidered by GISS as a rural site (population 400 000 according to Wikipedia).
Or why NASA keeps the temperature series for a station in different files, with 10 – 50 years data in each, instead of just having one file there?
Patrick Hadley // February 27, 2008 at 1:58 pm
I think I now understand why some people are so upset with McIntyre and Watts. It seems that they are deliberately going against the consensus by doing research or analysis that they hope to use to prove the establishment wrong.
If only we were in the time of Gallileo we would know what to do, or in Stalin’s Soviet Union when scientists who did not follow the party line were sent to Siberia.
After all science can only make progress if the established consensus is never questioned. How despicable Watt and McIntyre are to do such a thing.
[Response: People are upset with Watts and McIntyre because they are prejudiced, they take every opportunity to snipe at climate science based on innuendo, and all the while they are such hypocrites and cowards that they hide behind a veneer of "plausible deniability." One of the best examples of this is an infantile post by McIntyre, which he titles "Is Gavin Schmidt Honest?" He insults the reputation of a fine scientist (and a fine person), based on the flimsiest non-evidence, all over a childish peeve that there was a *delay* in one of his comments appearing at RealClimate. But you can bet he'll deny having insulted Gavin, insisting he was "just asking a question."
That's despicable.
Now you want to put them in the same class with Galileo? Because they disagree with the concensus? What a crock! For every Galileo who bucks the concensus because of his greater insight, there are ten thousand or more *crackpots* who claim that Einstein was wrong, Darwin was evil, cigarettes don't cause cancer, global warming is a fraud, or all of the above. You've put McIntyre and Watts in the wrong category.]
Boris // February 27, 2008 at 2:11 pm
Galileo and Stalin in the same post? You win!
kim // February 27, 2008 at 2:39 pm
And yet, it cools.
==========
Patrick Hadley // February 27, 2008 at 2:43 pm
It is a bit rich for an anonymous writer to describe Watts and McIntyre as “hypocrites and cowards who hide behind a veneer of plausible deniability.”
I have no doubt at all that anyone reading this thread with an open mind will be able to decide for themselves who deserves that description. I am not “trolling” i.e. deliberately provoking you into a negative reaction, but genuinely seeking to improve the way the debate is conducted. Unfortunately it seems you cannot stop yourself insulting your opponents. Surely you must know that does not win you any friends from the uncommitted.
[Response: It seems that YOU simply can't stop yourself from insulting ME.
But like McIntyre and Watts, you don't have the guts to say it without giving yourself an "out" by claiming that you're just trying to improve the way debate is conducted. This, in spite of the fact that your very first sentence is meant to imply that I'm a hypocrite and coward.
I have vastly more respect for someone who has the guts to say that I'm an idiot and a scoundrel, than for the coward who implies it while *claiming* to be working in the best interest of productive debate. Rather than take the coward's way, using innuendo by asking "Is Patrick Hadley Untruthful?" (while leaving myself the option to claim later than I was "just asking a question"), I'll say it outright: I don't believe you.]
Bob North // February 27, 2008 at 4:05 pm
Have McIntyre, Watt, and many posters criticized/insulted Schmitt, Tamino et al.? it would appear so. Have Schmitt, Tamino and many posters criticized/insulted McIntyre, Watt et al.? It would appear so. Nobody is a winner and nobody comes out smelling like a rose. The quote posted above by Michael Tobis ranks as one of the worst ad hominem attacks I have seen from either groups. Maybe it time to step away from the keyboard and take a deep breath.
Regards,
Bob North
kim // February 27, 2008 at 5:07 pm
Hey, Bob, most of that rant was PZ Myers, and he will live to regret it.
========================
Barton Paul Levenson // February 27, 2008 at 5:17 pm
It’s one of the sure signs of the crackpot pseudoscientist when they start comparing themselves — or the pseudoscientist they’re following — to Galileo. I like Carl Sagan’s comment — they laughed at Wilbur and Orville Wright, but they also laughed at Bozo the Clown.
Hank Roberts // February 27, 2008 at 5:20 pm
Like they say, it’s hard for passers-by to tell which one is the crazy one, once you engage on the sidewalk.
If you look in the science journals, you have no trouble telling them apart.
JCH // February 27, 2008 at 5:33 pm
“Hey, Bob, most of that rant was PZ Myers, and he will live to regret it. …” – Kim
What exactly does that mean? What is going to happen to him that will make him regret his “rant”?
kim // February 27, 2008 at 6:20 pm
He has compared climate skeptics to creationists. We know a lot more about genomes than climate.
=============================
Hank Roberts // February 27, 2008 at 6:24 pm
Chuckle. JCH,that’s because he’s attracting the “witnessers” — the people who make faith-based postings without citation. They, provably, will fill up any thread beyond the ability or patience of the reality-based to cite evidence or help people think for themselves.
The better the bridge you build, the more traffic your bridge gets, the more prolific the trolls that are attracted to live under it.
Gresham’s Law of Blogging. It’s what happened to Prometheus; it’s happening to Dot.Earth. The sites get so “popular” with the Witnesser crowd that nobody looking for fact cites goes there anymore.
Cautionary.
kim // February 27, 2008 at 6:24 pm
The belief that CO2 had much to do with recent warming is a ‘madness of crowds’. We are herd animals, and we’ve been stampeded into that belief. The globe is cooling, and CO2 continues to rise. Observe the reliable thermometers, the RSS and the UAH, and observe the quiescent sun. You’ll understand soon enough.
=====================
elspi // February 27, 2008 at 6:30 pm
Bob
There are SIX BILLION people in the world. I don’t have time to listen to each of them for even a second.
The first rule of APPLIED logic is that you don’t waste your time listening to someone who is not intelligent, honest, and well-informed. (notice that politeness is not on this list).
Shorter rule: Never give a known liar the benefit of the doubt.
You might think that you miss something by doing this, but if there is a legitimate argument to be made, then there will be some intelligent honest well-informed person making that argument.
McIntyre has argued dishonestly in the past (as the links on this very thread document) and therefore should be ignored.
What you ask, would honest argument look like? Did you see how the Climate Science community reacted when McIntyre (somebody they despise) found an error in the temperature record in the US? They acknowledged his contribution and immediate fixed the problem.
kim // February 27, 2008 at 6:31 pm
He is precisely backwards with the comparison. The hook that creationists have against the evolutionists is that the exact mechanism for the development of the so-called ‘irreducible complexities, such as hemostasis and flagellae, is still unknown. The analogue in climate science is that the exact mechanism by which small variations in solar output cause greater variations in climate is also unknown.
In fact, the ‘warmists’ of today will once be seen as creationists are now.
I realize that is a radical thought. Hold it your hand and your head, and turn it over a little.
=============================
Phil. // February 27, 2008 at 6:44 pm
““Actually that’s the VOR station, they’re quite distinctive.”
Thanks. Did think it might be something like that. I’ve never actually spotted a square from a sat pic, so would have been surprised if I’d started now.”
Adam, as a pilot you get quite used to seeing one of those little ‘upside-down’ ice cream cones from the air, I’d be surprised if anyone who hadn’t trained as a pilot would know what they were looking at.
Looking at the aviation plates for SPTU it’s clear that the VOR as shown on DAFIF /Google Earth is not in the same place as shown on the approach plate dated 2001 so it might be reasonable to suppose that there has been some reorganisation.
Hank Roberts // February 27, 2008 at 6:48 pm
Here’s how it’s done:
http://ams.allenpress.com/archive/1520-0477/74/2/pdf/i1520-0477-74-2-215.pdf
Bob North // February 27, 2008 at 7:01 pm
Serious question for Tamino and others -
As I noted above, there appears to be a step discontinuity in the Puerto Maldon data around 74-75 which is not present in either Cobija or Rurrenabaque data sets. A similar step discontinuity at about the same time was clearly noticeable in the Cuzco and Pucallba data sets referenced by McIntyre. To me at least, such step discontinuities are suggestive of some specifc change at a station rather than UHI type effects. If this is the case, would a single step adjustment make more sense than a trend adjustment (either adjusting everything before the discontinuity down by a set amount or adjusting everything after the discontinuity up by a set amount [eyeballing it in this case, I'd say about 1-1.2C]). If I recall, GISS does something like this when combining data from different sources at the same location.
Regards,
Bob North
[Response: Part of the reason there appears to be a step discontinuity is that there's a fair amount of missing data (months for which there are no values). But there is a case to be made for a step discontinuity.
In my opinion, this station needs a different treatment, in fact there's a case to be made for omitting it from the global calculation altogether. But before we go deleting it willy-nilly, bear in mind that one of the things that's strictly forbidden in data analysis, because there's just too much temptation to cherry-pick (even for the honest-hearted), is to adjust or delete things on a kill-as-you-go basis. Things MUST be done methodically. You don't get to delete Puerto Maldon because you don't like it or have suspicions. What you ARE allowed to do is establish a procedure which is objective. That procedure might be to apply some objective test for "unreliable," then omit those stations which fall into the "unreliable" category. You're even allowed to define "unreliable" using your experience with stations like Puerto Maldon, although it's important not to taylor your definition to match what you "don't like," or cherry-pick where the cutoff falls based on which individual stations it includes and omits.
Instituting a new *procedure* is a comprehensive effort requiring a massive amount of work. If someone who suspects that bad station data and/or metadata are biasing the trends, would omit questionable station data based on a strict procedure which enforces objective criteria, and apply those standards to the entire world rather than cherry-picked stations chosen to embarrass climate science -- now THAT would be a valuable effort. But as I say, it would be a MASSIVE effort. If you decide to do this, please don't start a blog to post all the unflattering examples you find even though you're far less than halfway done with only one region that's less that 2% of the globe.
And while I'm musing about possible worthwhile projects, here's another suggestion: compute the trend rate for the entire data set, and for the 1975-present time range, for both raw and adjusted data, for all stations worldwide. Then plot histograms of the results, including a breakdown by classification (rural, peri-urban, urban) and by continent. I'd be *very* interested in the results.]
J // February 27, 2008 at 7:02 pm
Ugh, it really depresses me when threads here degenerate into endless wrangling over the surface stations record.
It brings out the most unpleasant attributes of people on both sides, and it seems like a waste of time for Tamino to be trying to ride herd on it all, when he could be writing more of his amazingly helpful explanations of actual science.
That said, I’d note there’s a lot of disagreement over whether CA & McIntyre should be viewed as disinterested-seekers-of-truth or as propagandists-intent-on-misleading-the-gullible. I think, to some extent, McI tries to have it both ways.
I just wandered over to CA, and briefly skimmed through the most recent thread (“Antarctica”). Somebody (Ross McKitrick?) pointed out that some IPCC graphic of the atmosphere had blacked out model results for low altitudes (under 4 km) over the Antarctic. That led to a lot of snarky comments about “scissors”, with the clear implication that the IPCC was fraudulently covering up problems with models by crudely excising the data from graphs. Eventually somebody pointed out that Antarctica is mostly high elevation (over 2 km, and the interior plateau is over 3 km) … so the IPCC’s graph was omitting the lower few km of the atmosphere over Antarctica because it doesn’t exist …
Anyway, my point is that a lot of the daily banter over there revolves around casually implying that the IPCC and the climate science community in general is more or less constantly involved in various forms of fraud and conspiracies. Until somebody manages to radically change the tone and attitude of CA, most actual scientists who wander in there are going to react badly. McI needs to decide whether he wants to be a serious, constructive part of the earth system science community or the leader of an unruly mob of blog commenters. At some point the two roles become incompatible.
kim // February 27, 2008 at 7:03 pm
Hmmmm. Somebody’s being firm, patient, and silent. Thank you.
==================
Bob North // February 27, 2008 at 7:15 pm
Elspi – Not sure what in my post elicited your response, I did not pass judgments on the merits of either groups’ arguments or the weight of evidence in support of AGW (which appears substantial). I merely noted that flinging poo helps no one.
Regards,
Bob
EW // February 27, 2008 at 7:20 pm
Ok, honestly: when there are three temperature seriesfrom three stations, two running more or less in parallel and the third has the trend almost exactly opposite, do you think that “adjusting” makes from it a valid dataset?
I don’t know about you climatologists, but when I have three parallel experiments and one behaves differently, I don’t “adjust” the values to match – I throw them out.
[Response: If you want to throw out Puerto Maldon based on its disagreement with its neighbors, you have to throw out EVERY station that behaves similarly. And you have to establish very strict objective criteria for classifying stations as "bad." You're NOT allowed to kill Puerto Maldon because it looks bad -- you have to apply equal treatment to all stations. Anything less is cherry-picking (even if unintentional).]
cce // February 27, 2008 at 7:22 pm
For a long time, the consensus was that small changes in the atmosphere could not change the climate, therefore humans could not affect the climate in any meaningful way. But now, thanks to people like Tyndall, Arrhenius, Callendar, Plass, and a host of others, we know that the consensus was wrong.
kim // February 27, 2008 at 7:22 pm
Hank, is that the growling of the current, or an unhappy spirit I hear?
=========================
jorge a. // February 27, 2008 at 7:30 pm
dear Tamino: what do you think of Dra.Curry (georgia Tech)??? she is thinks like you, however, she has a good opinnion of S.Mc. is she mad??
[Response: If disagreement with me (on McIntyre or anything else) were criteria for madness, there'd be a lot of madmen running around the planet. Come to think of it, there *are* a lot of madmen running around the planet.
I've found Dr. Curry's efforts to understand the "war against science" waged by the anti-global-warming forces, very enlightening.]
jorge a. // February 27, 2008 at 7:30 pm
sorry is “think” no “thing”
jorge a. // February 27, 2008 at 7:31 pm
sorry again!!!! is “good” no go… i’m 63 years old!!!
[Response: I've changed the original post according to your corrections.]
elspi // February 27, 2008 at 7:53 pm
Wow
I just gave an example of someone who failed the “honest” part of the first law, and was wondering where to find the examples for “intelligent” and “well-informed” when kim joined the discussion.
I know what you are thinking; “Is he just a sock puppet that you are using for illustrative purposes?”
No, he is his own sock puppet.
Just for the fun of it let’s find how many blunders he has packed into a very small space.
1. The globe is cooling
2. exact mechanism by which small variations in solar output cause greater variations in climate is also unknown.
3. We know a lot more about genomes than climate.
Each totally false. The first two are debunked on this very blog.
I will eviscerate the third one myself.
The science on which global warming is grounded is basic 100-year-old physics. DNA on the other hand is a very new discovery. We have computer models based on the physics, which model the climate both historically and into the future. There is nothing of this sort in biology. There are no computer models that allow you to enter the genome of a species and its environment (including the genome of the other species) and predict how this species will evolve. It is unlikely we will ever get to that point. The Climate science of 2000 has already surpassed the biology of 20,000.
BoulderSolar // February 27, 2008 at 7:57 pm
Re Ravin and Tamino’s reply. I looked at the link which Ravin supplied for the graphof HasCRUT and Tamino’s links he supplied in rebuttal. The Has CRUT graph begins in 2001. Tamino said this was a “huge el nino” year. I thought that was 1998 not 2001. Then I actually hand calculated the five year averages from the temp graphs Tamino has in his rebuttal link. For the last 3 data points the temperature was flat and very different from Tamino’s own 5 year average shown in the same link. Seems when you actually view the links each person presented Raven’s point is validated.
[Response: You have made an error, plain and simple. Are you computing 5-year averages, or 5-year *moving* averages? Where did you get the data? Did you digitize the graphs? If so, how? Did you go to GISS or NCDC or HadCRU and get the actual data?
As for Raven's comment, he specifically stated "over the last 10 years." That's starting at 1998, not 2001.]
chriscolose // February 27, 2008 at 8:00 pm
EW,
why do you throw it out? Science would not advance is everyone just threw out stuff that did not look right, you need to make it right. I’m not sure about the hard statistics around data adjustment, but from a purely intuitive perspective it makes sense. Isn’t it the skeptics who talk about the urban heat island effect, and how that is not properly adjusted for?? Do we just throw out every stations with buildings nearby, or with lots of people?
George // February 27, 2008 at 8:04 pm
Bob North said
If that’s the worst Bob has seen, I’d simply have to say that Bob has not seen any of the insinuations or outright charges of fraud (or other scientific dishonesty) that have appeared on Climate Audit , like this comment (by Steven Mosher, who also sometimes comments here) or this comment
… which means, of course, that Bob has probably never read Climate Audit.
Well, maybe once, probably when the server went down and a “Climate Audit temporarily unavailable” message was posted in place of the the normal Climate Audit web page.
Note: I have linked to the reproduced comments here because the original ones were removed from Climate Audit after I pointed them out on this thread. I would guess that the comments in question would still be there (on Climate Audit) had I not essentially embarrassed McIntyre into removing them.
henry // February 27, 2008 at 8:17 pm
Boris commented:
“McIntyre publishes under his own name”
McSteve doesn’t publish. He merely types. Do you see the difference?
[Response: McIntyre does have peer-reviewed publications, although all I’ve found (I really haven’t looked that hard) are criticisms of the “hockey stick,” many of which appeared in the highly dubious journal “Energy & Environment” (although there’s at least one in Geophysical Research Letters).]
And where would we find peer reviewed papers of yours?
Adam // February 27, 2008 at 8:30 pm
“Looking at the aviation plates for SPTU it’s clear that the VOR as shown on DAFIF /Google Earth is not in the same place as shown on the approach plate dated 2001 so it might be reasonable to suppose that there has been some reorganisation.”
That’s interesting, Phil. I think from the little bits of history seen, and the discrete data jumps, there’s been a few over the years.
Barton Paul Levenson // February 27, 2008 at 8:30 pm
kim writes:
[[The belief that CO2 had much to do with recent warming is a ‘madness of crowds’. ]]
Radiation physics is a “madness of crowds?” Look, CO2 is a greenhouse gas. Put more of it into a substantial planetary atmosphere and that atmosphere must get warmer, unless some countervailing process exists.
[[We are herd animals, and we’ve been stampeded into that belief. The globe is cooling, ]]
No it isn’t. Check the past several threads, and I think you’ll find a couple on the claim that the globe is cooling. It isn’t.
[[and CO2 continues to rise. Observe the reliable thermometers, the RSS and the UAH, and observe the quiescent sun. You’ll understand soon enough.]]
We understand as is.
Hank Roberts // February 27, 2008 at 8:32 pm
http://www.ncdc.noaa.gov/oa/climate/research/Smith-comparison.pdf
—-excerpt follows—-
As a replacement to the existing method, this new analysis uses improved methods
that provide error estimates as well as the ability to perform analyses on finer spatial scales. Comparisons show only
minor global-average differences, and the two estimates indicate essentially the same trend over the historical record,
beginning in 1880. The two are most similar after about 1970, a period with a large change in the global-average
temperature. The uncertainty estimates computed here account for changes in sampling and for systematic bias
uncertainties. The means of the different analyses generally fall within the uncertainty estimates. The uncertainty computed here indicates that anomalies in the 19th century may not be significant, but the 20th century trends are significant….
…
A great deal of work has recently been done to improve the accuracy of NCDC’s surface temperature analyses. These comparisons show that the new analysis has essentially the same interannual to interdecadal variations in global averages. Where differences do occur they are primarily confined to the early years, when sampling is more sparse. An advantage to the new analysis is the availability of uncertainty estimates. The addition of more observational data could reduce these uncertainty estimates and make regional analyses more accurate, but these results suggest that more data would be unlikely to greatly change the average global temperatures over most of the analysis period. Besides uncertainty estimates, the new analysis also makes it easier to develop a wide variety of regional analyses because it produces an intermediate gridded analysis that can be averaged for any region.
Zeke // February 27, 2008 at 8:39 pm
Kim,
The one trick “watch the thermometers, the earth is cooling” pony might fly over at Revkin’s blog. Here, however, you at least have an obligation to attempt to address Tamino’s many well thought out posts on the subject before turning on the broken record.
As far as McIntyre goes, I think he has done some interesting and, in many cases, useful work in poking holes in some aspects of temperature records. The danger is that people looking to reinforce their preconceptions on the climate debate mistakenly conflate these holes with cracks in the fundamental science, something McIntyre does not do enough to discourage (even though he himself strays away from posting or agreeing with the nuttier things that crop up in his comment threads).
Hank Roberts // February 27, 2008 at 8:45 pm
When one of these posters is not like the others, this is a tool worth mentioning.
http://www.caveyourtrolls.com/about/about.htm
It’s a plugin, for any WordPress blog host.
——-excerpts———-
Give trolls what they need the most: attention. …. His posting is accepted.
…
Only you and I know that the troll is the only person who can see the improper posting. Other forum users can’t see the troll at all….
…
3. How it works
—————
When you click the “Cave Troll” button that you added by adding the template tag, the “karma” of that comment is set to -100. … only the original poster of the comment and admins that are logged in will see this comment ….
—————–
kim // February 27, 2008 at 8:45 pm
elspi and Barton, sure CO2 is a greenhouse gas. It’s effect on climate has been exaggerated and the effect of the sun has been ignored. You do know that climate sensitivity to doubling of CO2 is highly controversial, don’t you?
Also elspi, what do mean that my contention that the exact mechanism is unknown by which small changes in the sun’s output are magnified into greater changes in the climate has been disproven on this blog. Do you mean the mechanism is elucidated? If so, show me. Do you mean the sun-climate connection has been disproven. If so, I’m dubious.
I contend your models are circular, they don’t account well for clouds and water vapor, and exaggerate the effect of CO2. We don’t understand how climate is regulated; we do understand how genes are expressed.
===========================
R Clark // February 27, 2008 at 8:50 pm
OK now i’m confused.
Is there or is there not temp data for Cobija and Rurrenabaque after 1988? (as posted over at CA)
Not trying to take any side here just losing faith on what to believe.
[Response: I downloaded both the raw and adjusted datasets from GISS, and there's no data beyond April 1989. ]
kim // February 27, 2008 at 8:54 pm
Also, the globe is cooling. Not for the last ten years, as the strawman says, but for the last 2-4 years. See Tilo Reber’s graphs at Reallyrealclimate.blogspot.com. Sure, they only include the last six years. Why would you say six years is noise, but deny that thirty years is not noise, especially if you understand that the PDO was in a heating mode for the last thirty years, with El Ninos predominating? That oscillation has flipped, warming La Ninas will predominate for awhile, and the sun is hibernating.
My arguments are not compelling, but the temperature course over the near, perhaps long, term will be. Watch the reliable thermometers, the RSS and the UAH. You’d be a fool to take the word of a stranger; you’d be a fool not to trust your own eyes and senses.
======================
kim // February 27, 2008 at 8:58 pm
Oh, heck, I meant cooling La Ninas will predominate for awhile. Probably for thirty years, or so.
==================
FrancisT // February 27, 2008 at 9:06 pm
You wrote:
In the same time period (1958-1988) during which Puerto Maldon raw data show more than 2 deg.C cooling, its neighbors show about 1 deg.C warming. To make the Puerto Maldon trend match its rural neighbors (which will include a lot more than I’ve shown here), it needs an adjustment amounting to about 3 deg.C during that time period.
I’m not a climate scientist so maybe there is an esoteric step here that is being jumped over because it is “obvious” but what you are saying here comes across to me as “We know the answer is X, this data doesn’t give X so we fix it until it does”
cce // February 27, 2008 at 9:35 pm
If RSS is reliable, then the UHI effect is a myth.
If UAH is reliable, then Hansen’s projections of the future are more reliable than Christy and Spencer’s version of the past.
steven mosher // February 27, 2008 at 9:38 pm
Tamino, you have to be careful in talking about how the adjustments are done. The rural stations are first sorted from longest record to shortest record, then they are debiased relative to each other, then they are WEIGHTED as a inverse function of distance from the urban site.
This doesnt make it wrong, but the actual math is different than you suggest. I think it would a GREAT excercise to select a Urban station randomly and then look at the rural stations that are used to adjust it and walk everybody through the math.
When I first looked at the “results” of adjustments I had many questions and misperceptions. As I walked through the code
Dr. Hansen kindly posted, my questions and misperceptions diminished. I’m still left with a few questions.
me no understand // February 27, 2008 at 9:38 pm
Thank you again for your detailed response. I really appreciate it, but I am afraid I am just getting more confused. How can one make a 2-legged adjustment with a “hinge” point in 1988 with no post 88 data from the rural sites? It just doesn’t seem intuitive to me.
I am afraid that the absence of post 88 data for Cobija or Rurrenabaque is a real worry for me. To make an aggressive adjustment to Puerto Maldon without testing it against the post 88 trends to see if they have stopped diverging is…well…it’s a bit of a shocker really. Don’t you think it would have helped your case if you could produce graphs, 20 years post-correction, showing the trend lines all moving in the same direction. I hope you see how this failure gives ammunition to the denialists.
I got into this debate because a friend of mine gleefully sent me the CA graph showing a 3 deg C adjustment to Puerto Maldon saying “see…see…they are just making it up as they go”. I accept that this correction is not made up but you wouldn’t call it a robust methodology either would you? My concern is that come 2015 when the temperature record agrees with the IPCC predictions my denialist friend wont be saying “ah well…I guess I was wrong…Sorry guys” he’ll say “of course they agree, the data is cooked. Look at Puerto Maldon”
I’ll stop bloviating and ask you a direct question: Do you think the case for the NASA GISS warming trend is made stronger or weaker by the existence of Puerto Maldon? I think the answer is obvious and Puerto Maldon and its ilk should be removed from the record.
[Response: There's no data from Cobija or Rurrenabaque, but there is data from rural stations within a 1200-km radius. I haven't tracked them all down, in my opinion that's more work than it's worth for this blog post.
When non-rural stations are adjusted so that their trend matches that of rural stations, the regional and global trends end up being *defined* by the rural stations. So I expect that omitting Puerto Maldon won't make any substantial difference. Some researchers have computed what the global trend is using only rural stations, compared to using all stations (I'm not sure whether they applied adjustments and what kind), only to find that the rate defined using only rural stations is *larger* than that using all stations.]
sod // February 27, 2008 at 9:47 pm
elspi and Barton, sure CO2 is a greenhouse gas. It’s effect on climate has been exaggerated and the effect of the sun has been ignored. You do know that climate sensitivity to doubling of CO2 is highly controversial, don’t you?
the claim that CLIMATE scientists are ignoring the sun, is plain stupid.
talking CO2, i LOVE this experiment description:
http://www.espere.net/Unitedkingdom/water/uk_watexpgreenhouse.htm
lucia // February 27, 2008 at 9:57 pm
Actually, Tamino, the chart Raven linked was not constructed by cherry picking right after the El Nino.
Roger Pielke and David Stockwell pointed out that the current IPCC ‘predictions’ start at 2001. So, they wanted to see what things look like if we plot data starting in 2001.
The 2001 date was picked for that reason only, and was not influenced by the El Nino high of 1998. Not only is the first year not the El Nino high, the following two low years are also not on that graph.
So, that graphs is a very short snippet of this century. It’s even shorter than you imagined.
Right now, starting in 2001, HadCrut is absolutely flat. It’s a statistical fluke to come out totally flat, but it happened. GISS has a positive slope.
What does this mean? Possibly nothing. We all know we would need to wait a longer time to make conclusions about anyone’s claim we’ve reached the top. How long we need to wait to start making claims was actually the point being discussed in the blog post associated with that plot.
Will Richardson // February 27, 2008 at 9:58 pm
Dear Tamino,
I know you insist that “[t]here’s no data from Cobija or Rurrenabaque”, But McIntyre has posted the post 1988 temperature data for Cobija and Rurrenabaque at Climate Audit today.
Why the discrepancy?
Regards,
WJR
[Response: He didn't get it from GHCN or from NASA. Does it include adjustments for station moves, time-of-observation, instrument changes? Does Anthony Watts have photographs?]
MarkR // February 27, 2008 at 10:15 pm
The point is Puerto Maldon is treated by GISS as URBAN and adjustments applied accordingly (leaving aside whether they are correct). It turns out that the site is in fact RURAL. So as far as we know, many stations may have been wrongly classified as URBAN, and had their data adjusted, but in fact have been RURAL. That is a systemic failure.
[Response: I guess that's why NASA has switched to using nightlights data rather than GHCN population classification for the U.S., for which that data is available. I expect they'll use nightlights for all stations when time and data permit. Let's not pretend that Steve McIntyre identified this as a problem which needs addressing; as far as I'm aware, it was James Hansen and colleagues at NASA who did so, they're the ones who devised a procedure to replace GHCN population classification wiht satellite-derived nightlights data, they're the ones who implemented it in the U.S., and when this problem is finally met face-on on a global scale, I'll bet it's NASA that does the work.]
me no understand // February 27, 2008 at 10:16 pm
Oh ok. So the post 88 corrections were made with data from other rural sites, not the ones shown above. Bugger! It sounds like I owe my smug denialist friend a pint. When I originally saw your post I told him that you had a convincing explanation for the Puerto Maldon adjustments but it seems that you have only been able to explain them pre 88. The decision to stop the adjustment in 88 remains a mystery.
“When non-rural stations are adjusted so that their trend matches that of rural stations, the regional and global trends end up being *defined* by the rural stations”
Precisely, so can I ask my question again. Do you think the case for the NASA GISS warming trend is made stronger or weaker by the existence of Puerto Maldon?
Puerto Maldon hasn’t just cost me a pint. Do you realise how sceptical the average man in the pub is? Maybe I should stop trying to win the debate with rationality and rely on scare stories instead.
Thanks for your atention. I’ll sign off now, a bit more depressed than I was 2 days ago when I first raised this topic.
[Response: Console yourself with the fact that at least you found some actual data which is relevant to the issue, and although this post is hardly rigorous, is certainly establishes that the adjustments are not "made up" nor are they in violation of stated procedure.
So before you pay for the pint, ask your skeptical friends how many stations THEY got data for and computed the trends of.]
Vic Sage // February 27, 2008 at 10:29 pm
SOD: “talking CO2, i LOVE this experiment description:”
Why, it isn’t really an experiment. There should have been more permutations i.e. they should have swapped the lights instead of assuming they had identical outputs. They should have swapped the tanks instead of assuming they were identical as well. It proved nothing except that the ones doing the experiment don’t understand the scientific method.
geochemist // February 27, 2008 at 10:30 pm
I just checked CA and they came up with data for Cojiba from 1988 to the present. I think their question today is why this was not available in the data base used for calculating the global temperature. Who dropped the ball and why did no one notice?
[Response: It's a fallacy to assume that because there's data available which is not included in the GISS analysis, somebody dropped the ball. Perhaps finite resources (time, money, and people) simply haven't been sufficient to obtain completion yet. Or maybe somebody dropped the ball. But I haven't seen real evidence of that yet -- just innuendo.]
null{} // February 27, 2008 at 10:35 pm
At MarkR // February 27, 2008 at 10:15 pm
Tamino said:
“[Response: I guess that’s why NASA has switched to using nightlights data rather than GHCN population classification for the U.S., for which that data is available.”
Would that be the brightness-index in the GISS/NASA GISTemp files?
Thanks for the clarification.
Bob North // February 27, 2008 at 10:38 pm
Hank – Thanks for the link to the Smith et al. paper. It does seem that using different methods and approaches calculates similar temp anomalies and that a lot of this stuff doesn’t make that much difference. The smith paper indicated that it’s approach used a gridded GHCN dataset and it is not clear to me if that is the same as the dataset used by HadCrut, NASA or something slightly different. Also, although they discuss the calcuation of error estimates, I did not see any actual statement of what those error estimates were other than they were greater than those calculated for the Quayle method ( of course, I might have missed something.)
George – If you look at my original post, I indicated that both groups appear guilty of flinging insults and I certainly haven’t seen all of the insults that have been flung so I don’t know what is the worst. While we can certainly disagree as to what constitutes a worse insult, I would rank referring to someone (and this is paraphrased though all the phrases were used), ‘more like a sanctimonious rat focused only on putrescence and spreading a plague of lies and ignorance as you go.’ as faily high on the list. The bottom line is both groups are guilty and such ad hominem attacks by either side are non-productive, even if they are entertaining in a “Jane, you ignorant slut” sort-of-way (reference to old Chevy Chase-Jane Curtin SNL routine in case anybody did get the allusion).
Finally, if anybody can expound upon my question regarding a simple step-wise correction versus a two-hinged trend correction, I would appreciate it.
Regards,
Bob
MarkR // February 27, 2008 at 10:41 pm
It follows that if the data from the rest of the world cannot be classified properly into RURAL or URBAN, then the adjustments cannot be made correctly.
As it is the data from the rest of the world, excluding the USA that gives the warming trend, where does that leave the reliability of analysis based on that rest of the world data?
[Response: You've fallen prey to one of the most common fallacies in science. NO DATA IS EVER PERFECT. If we do our job well, it gets better and better, but if you're waiting until everything is "just right," you'll never get anywhere.
The suggestion (implicit in your comment) that the classification for the rest of the world is as bad as McIntyre claims Peru is, isn't accompanied by any evidence. And frankly, McIntyre's implications about the state of affairs in Peru isn't convincing. Where's the comprehensive survey of ALL Peruvian stations? All I see at CA is McIntyre highlighting a handful of stations that look bad.
When McIntyre stops sniping, and starts doing comprehensive WORK, ...]
sod // February 27, 2008 at 10:53 pm
this is going off-topic, so i will not reply anymore to this. but i can t let stuff like that stand: (and i still have the slight hope that the reply was some sort of sarcasm..)
Why, it isn’t really an experiment. There should have been more permutations i.e. they should have swapped the lights instead of assuming they had identical outputs. They should have swapped the tanks instead of assuming they were identical as well. It proved nothing except that the ones doing the experiment don’t understand the scientific method.
it is an experiment DESCRIPTION. not a ground breaking article in a peer reviewed paper. they might not have mentioned each and every run and variation they ever tried, you know?!?
are you trying to imply, that they only did this simple experiment once, EVER?
do you assume that results CONSISTENT with consensus knowledge are typically caused by differences in standard lab equipment?
the thing i love most about this experiment is how simple it is. so i look forward to reading YOUR results. remember to swap lamps!
PS: (sort of on topic) Tamino was obviously talking about GISS data. case closed.
None // February 27, 2008 at 10:54 pm
“The science on which global warming is grounded is basic 100-year-old physics.”
That’s just precious. The whole global warming “equation” is based on the idea that the climate is precariously balanced and that a little bit of extra carbon dioxide in the atmosphere (the total percentage being approx 0.038%, most of which is natural) will cause an increase in temperature, causing an increase in water vapour which will cause an even greater increase in temperature etc etc). To claim that is century old physics is nonsense.
[Response: The warming influence of atmospheric CO2 is indeed more-than-a-century-old physics. The current CO2 concentration is almost 30% antrhopogenic -- that's a pretty substantial fraction.]
There has been incredible hatred directed at Macinyre by the climate science community since he showed (and a later congressional hearing where a statistics expert Wegman agreed) that Manns “decade old paper”, as chriscolose likes to put it, was premised upon incorrect statistical science.
[Response: He showed no such thing. I'll be posting on that topic soon enough, so save your rebuttal until the post appears.]
It is extremely embarassing for the climate science field, as their own peer review totally failed to notice this, and that Manns work continues to be the most cited climate paper around.
Lets not even go into the “error bars” in his hockey stick graph, or the overwhelming weighting given to BCP proxies when their validity as a proxy is, lets just say not exactly unblemished.
[Response: This is not the place for you (or anybody) to sling mud at the hockey stick. As I said before, I'll be posting on that soon, so save it until then -- you'll have plenty of opportunity. Anybody else who can't hold his tongue until that time, will be disappointed at the disappearance of their comment.]
George // February 27, 2008 at 11:13 pm
Bob said:
Leveling an unsubstantiated charge of scientific fraud against someone (which I referred to above) is not simply an “insult”
It can damage (even ruin) a person’s reputation an career, whether it is true or not.
And it is not simply a “matter of opinion” that such charges are reprehensible.
It is part of the legal code in most countries.
ianric // February 28, 2008 at 12:02 am
It must be very difficult adjusting raw data. I have two out-door thermometers on my house. When I bought them, I checked a large number in the store, selected the ones which showed mean temp, and chose two which showed identical measure. Even if the thermometers (both located on the wall facing north) show temp 2-3 degrees C above freezing, there may still be ice on the car windscreen and on the ground. Nothing strange with that. The car is about 3 meters (10 ft) away from the house while the thermometers are about 5 cm (2″) from the window. Of course the close proximity to the house will have an effect. What puzzles me is the influence of the wind. On a quiet day, the thermometer on the second floor will read one degree C higher than the one on the first floor. When it’s windy, they will read the same. I’ve noticed that this blog does not look kindly on comments questioning the quality of the surface network. So I would like to just ask how the corrections are made. Judging from my own house there would have to be made hourly corrections relating to wind speed as well as difference between indoor and outdoor temperature. And I gather that a large proportion of the weather stations in the US do not meet the standards. (If I remember correctly, only 13% of US weather stations meet criteria.) So how can this possibly be corrected for? And why the hostile tone against those who critsise the current standard? Thanks in advance.
[Response: It's no surprise that your two thermometers (on two different floors) often register different temperatures. But what if you kept daily readings for both thermometers, then computed monthly averages? And rather than worrying about the raw data, you studied only the month-to-month *changes* indicated by the two thermometers? I'll bet that the two would be in excellent agreement on the long-term trends, just as they are in visible disagreement about the momentary reading.
The hostility is not against those who criticize the current standard. It's against those who do so by going to great lengths to cherry-pick the most embarrassing examples they can find, for the specific purpose of discrediting the science rather than advancing it. They only end up painting a deliberately biased, and VASTLY more flawed, picture of the state of things than the legitimate analysis from NASA, HadCRU, NCDC, etc.]
dhogaza // February 28, 2008 at 12:18 am
That’s simply not true. Climate scientists working on surface temperature trends are well aware of the difficulties with the data, including problems such as you cite. In fact, they’re the ones who’ve pointed out the existence of most of the problems in the first place.
What some of us don’t look kindly on are insinuations that:
1. They aren’t aware of such problems
2. They do not take such problems into account
3. They just make up data when the data doesn’t exist
4. They’re committing scientific misconduct
5. Claims that photographs are data
6. Conflating of temperature measurements and temperature trends
etc etc etc.
As has been pointed out above, when McIntyre found a small problem with the data from the continental US from 2000 on, his discovery was welcomed and the corrections made.
Do good work, and people will welcome it.
Post bullshit, and your post will be treated with all the kindness it deserves.
Henk Lankamp // February 28, 2008 at 12:18 am
I think a lot of readers don’t know where the raw data for stations originate.
The WMO has made regulations on how to publish data on weather and climate (WMO306-vol1A). For each country a national weather service is responsible.
For weather there are SYNOP and METAR bulletins. METAR bulletins are intended for airfields, and are often updated every (half) hour. SYNOP bulletins are intended for global and local weather analysis, issued at least every 6 hrs and sometimes every hour; once a day they contain minimum temperature and maximum temperature.
For climatological purposes there are the CLIMAT bulletins (FM71), issued once every month: mean maximum, mean minimum, mean daily temperature, total precipitation etc. If
there is no CLIMAT bulletin for a specific station in a specific month, perhaps it is marked as ‘MISSING’, ‘NOT AVAILABLE’.
Hank Roberts // February 28, 2008 at 1:52 am
Where else does one look?
This site — huge set of analysis tools
http://climexp.knmi.nl/gettempall.cgi?someone@somewhere+84658+PUERTO_MALDON+
gets their data here:
http://www.ncdc.noaa.gov/oa/climate/ghcn-monthly/
and NOAA says data’s from CLIMAT.
There’s missing data in this graph
http://climexp.knmi.nl/data/ta84658.png
That’s shown in the raw data table as -999.9
http://climexp.knmi.nl/data/ta84658.dat
Note the oddity that 1980 and 1981 repeat (missing data once, numbers once) and much data is missing thereafter.
I wonder how those charts from the original CA post got filled in as solid lines. Someone else’s data source?
Paul S // February 28, 2008 at 1:59 am
==dhogaza said:==
==What some of us don’t look kindly on are insinuations that:
==1. They aren’t aware of such problems==
There is an awareness in a general sense but there does not appear to be a serious, concerted effort to document specific surface site anomalies.
==2. They do not take such problems into account.==
How does one take into account the specific issues affecting a particular site when it appears there is little or no data chronicling the anomalies?
==3. They just make up data when the data doesn’t exist==
That’s foolish for anyone to state.
==4. They’re committing scientific misconduct.==
Foolish to state also.
==5. Claims that photographs are data==
Photos are data. Why is that concept so difficult to grasp?
kevin // February 28, 2008 at 2:14 am
Bob, I think you’ll find it was Dan Akroyd, rather than Chevy Chase, in that news skit with Jane Curtin.
TCO // February 28, 2008 at 2:19 am
We’ve seen this movie before. McI picks the most extreme adjustment in the direction he dislikes and then complains about it. He doesn’t say what the average amount of adjustments is. He doesn’t account for the issue of correction of noise superimposed on correction of UHI and the probability that this will sometimes give a non-intuitive adjustment. But that one has to do this to account for the converse adjustments. I’m not sure if this is stupidity or sophistry.
BTW, there could still be an issue if the “rural sites” are contaminated and the urban ones not. But he doesn’t even get to the point of framing it. [edit]
Stan Palmer // February 28, 2008 at 2:27 am
=========As has been pointed out above, when McIntyre found a small problem with the data from the continental US from 2000 on, his discovery was welcomed and the corrections made.
===========
The “small problem was that they were taking the data from teh worng source for 7 years and no one at NASA or the broader climate science community notices it
chriscolose // February 28, 2008 at 2:51 am
None,
Svante Arrhenius did not have modern CO2-temperature correlations to look at, nor did he have 850,000 years of ice core data to look at. His prediction that doubling of CO2 would cause 4-6 C of warming (which is just a bit higher than the best estimates today) was grounded in radiative principles.
As for your meaningless percentages, see http://chriscolose.wordpress.com/2008/02/13/just-a-few-more-molecules/ .
IF you’re looking specifically for history, then Spencer Weart’s page is probably the best start (the first half of the lecture by Naomi Oreskes that Tamino posted was also good on global warming history.)
I am looking forward to Tamino’s post, but the Gulledge Testimony might be a great read for people interested in the meaning of the hockey stick. The NAS report on surface reconstructions for the last 2,000 years support the conclusions by Mann et al. and other studies (See the graph at http://www.globalwarmingart.com/wiki/Image:2000_Year_Temperature_Comparison_png ) which unequiovocally demonstrate warming over the last century, warmer in at least 400 years, and most likely past the MWP.
dhogaza // February 28, 2008 at 2:54 am
It had a very minor impact on the trend data, which is why no one noticed. It’s not like data from the moon was being substituted. It wasn’t “the wrong source” in that sense. It was “the wrong source” in the sense of how it was being massaged before they cranked on it, and, yes, in the overall scale of things it was a small error.
Still, McIntyre gets credit for spotting it. 1934 is now statistically tied with 1998 as the warmest year on record in the lower 48. Just as it was before. That surely disproves all of climate science, right?
The temperature data itself includes information on anomalies, quantatitive information that can be subjected to statistical analysis, which is exactly what NASA does.
The claim is not difficult to grasp. The problem is that it’s so easy to refute …
Ellis // February 28, 2008 at 2:59 am
Tamino says,
Funny, that almost sounds like something Steve M. might say. So there is common ground.
Just a couple of questions in regards to that response. From page 7, Hansen 2001,
Does this seem methodical to you, perhaps you could hazzard a guess as to the algorithim for this adjustment?
[Response: If the discrepancy was identified by an evaluation which is applied to all stations, so that those in a similar situation are treated similarly, then it's methodical. If, on the other hand, someone had identified this situation, then made a blog post about it without having looked for similar situations in all the data ... that would be different.]
Of course, this isn’t really a big deal to me because at least he had the good sense to put the information out in the light of day, however, if you go to page 4, speaking on tobs for row, you get,
This is more insidious, with words like generally and most, the reader is not able to comprehend whether the adjustment is made or not. This does not seem objective to me.
Perhaps, Dr. Hansen has it right when he states,(pg 3)
Bold mine.
On another point made by elspi,
Seems has though you forgot one step, as in, ad hom attacks. You read thispiece of bile and tell me again how well climate scientists deal with having a valid mistake pointed out. This leads me to my final point, if you ask a policeman what he thinks of internal affairs, what do you think the majority will say? What about the head of finance and his internal auditors? Lastly, how do you feel about the IRS? Auditors are universally loathed, so my question to you Tamino, is your hatred of Steve due to the man or do you feel that no one should audit the Climate Sciences because the science has already been settled?
[Response: You're making the same mistake so many others have made. Hansen doesn't object AT ALL about having a valid mistake pointed out. What he complains about is using the presence of an error to give the utterly false impression that the entire surface temperature record is untrustworthy. It's really no different than showing that the U.S. census figures from 1910 have numerous identifiable flaws, so we can't trust the conclusion that U.S. population has increased over the last century.
So EVERYBODY please pay attention, because I've said this already many times and it's tiresome to have to repeat myself so often: we don't object to people finding mistakes, that's a welcome effort. We DO object to those who take every flaw they can lay their hands on and use it as a propaganda tool to discredit, rather than improve, climate science.
Future comments which repeat your mistake go straight to the trash can.]
Ellis // February 28, 2008 at 3:07 am
all
[Response: Am I correct that this is the word you intended to bold in your previous comment?]
elspi // February 28, 2008 at 3:55 am
Ellis
Did you even read my post?
Which part of
“Never give a known liar the benefit of the doubt.”
did you not understand?
I followed your link and found an honest man decrying dishonest argument. How could you imagine that I would find that objectionable? My only quibble was that none of the liars was boiled alive. (If I were god, there would be a shortage of liars, but a surplus of liar soup.)
How can you condone the sort of dishonesty that he was describing?
Are you a con man, and so have a stake in defending all dishonesty?
“Nothing is more contemptible than man who clings to an argument he knows to be false.” Gauss
George // February 28, 2008 at 4:01 am
paul S said:
“Photos are data. Why is that concept so difficult to grasp?”
I think what dhogaza meant is that they are not temperature data.
First, a photo of a barbecue grill sitting in proximity to a temperature station is meaningless. It is not evidence of “contamination” of the temperature record, as some have insinuated.
The only thing that it is evidence of is that “there was a grill sitting idle near a temperature station at the particular point in time when the photo was taken.”
Not a particularly enlightening piece of “data.” (except perhaps to someone doing an inventory of barbecue grills)
It does not mean the grill was ever used near the station, nor that it had any significant effect on the measured temperature even if it was used.
Second, a photo or series of photos at a single point in time is not at all useful in gaging possible effects on the temperature trend over past decades.
In order to gage the reality and magnitude of such effects one has to have a series of photos to gage changes to the site over time.
In addition, one has to have a very good idea of how much each particular potential “problem” actually affects the ambient temperature in the immediate proximity of the temperature sensor.
One has to take actual temperature measurements to determine that!
Mere guessing based on the application of some general “standard” is simply not enough. Far from it. Standards are written to minimize (if not eliminate) potential problems — ie, to make it very unlikely that they will occur.
A “violation” of a standard does not mean that a problem actually exists, only that it might/could. The only way one can answer the question of whether a problem actually exists is with actual temperature data meant to gage the magnitude of the effect (eg, of nearby blacktop) on the temperature sensor at the site.
Mike Fijne // February 28, 2008 at 4:15 am
“So EVERYBODY please pay attention, because I’ve said this already many times and it’s tiresome to have to repeat myself so often: we don’t object to people finding mistakes, that’s a welcome effort. We DO object to those who take every flaw they can lay their hands on and use it as a propaganda tool to discredit, rather than improve, climate science.”
Each and every flaws are one too many any scientist would be eager to correct, especially when public policies are at stakes and that decimals have been given so much importance. And finding them doesn’t discredit climate science, but improves it. What however discredit people is just themselves and their attitude. And at what point cumulative mistakes become critical to the decimal precision it is supposed to achieve?
carlG // February 28, 2008 at 4:24 am
Tamino, don’t you think it’s a bit ingenuous to talk about the “actual codes, algorithms, and operating precedures” being out in the open as if it was NASAs idea to do this? McIntyre asked for that information for at least a year before it was released, and there is little doubt that it was released for any reason other than his consistent asking for it.
[Response: I'm guessing you meant to say "disingenuos."
All the stuff is out there for anybody to download from the internet; nobody seems to dispute this. But keep insinuating that NASA is guilty of trying to hide the truth.]
henry // February 28, 2008 at 4:37 am
“What some of us don’t look kindly on are insinuations that…:
5. Claims that photographs are data”
Pictures certainly are data. People are proving that there’s been a change in the arctic ice by comparing two pictures.
If there had been pictures taken of surface stations 20 years ago…
Ellis // February 28, 2008 at 4:52 am
Pardon me sir, a mistake was made, small as it was, and I presume that we can all agree in wanting the best data possible, but that does not excuse Dr. Hansen from Usufruct. He plainly calls Mr. McIntyre a jester and alludes to captains of industry pulling his strings as though he were a puppet. All that for .015c. It seems to me that both sides need some work with Emily Post.
Tamino says,
There are no similar situations. Dr. Hansen adjusted rural sites with urban, periurban, and ocean? temperatures. This is a deviation and he mentions it, but is it methodical?
yes to all. Could you please add a preview to comments?
[Response: I read the pdf you linked to. Nowhere does Hansen mention anybody by name, let along McIntyre. He refers to "dogged contrarians who present results in ways intended to deceive the public" without suggesting who those individuals might be. The only reference to McIntyre I find is this: "We corrected the flaw in the program, thanked the fellow who pointed it out, and thought that was the end of it."
It's not in my power to add a preview function.]
dhogaza // February 28, 2008 at 5:31 am
I feel helpless. The troll to knowledge ratio has, unfortunately, made it difficult to respond intelligently.
I do remember a time, not far past, when Tamino chastised some of us (me, for instance), for being rude and impolite, and stated that he hoped this place wouldn’t be polluted in this way.
Yet, Tamino now is … direct and unforgiving.
Part of me thinks it’s sad, part of me welcomes Tamino to the real world.
Now I’m going to say something that would justify Tamino not approving my post, or editing it heavily:
McIntyre is an inherently dishonest person, who doesn’t even have the stones to stand behind the sidewise insinuations of scientific fraud that he coaxes out of his readers.
Because that’s his primary MO. Posts shit without saying directly “this is misconduct! this is fraud!”
And then let’s the choir post “fraud! fraud!” while maintaining deniability (“I didn’t say it! my readers did”) without admitting he’s flogged them to do so.
[Response: I had high hopes to maintain a polite dialogue. But unfortunately it's the *denialists* who can't take criticism; whenever they're criticized it seems that a hoard comes over here from ClimateAudit to assault this blog with both criticism and insults directed against climate scientists, especially James Hansen and Mike Mann. The real world is a scary place, but THEY don't scare me one bit. And unlike the cowards who use innuendo as their MO and their readers to do the dirty work, if I've got something to say I'll say it myself and there'll be no ambiguity about my meaning.
So I'm thinking of changing my psuedonym to "Hansen's Bulldog."]
Heretic // February 28, 2008 at 5:42 am
Boy, that thread has really got out there in a short time, influence of the terrain of choice at the moment for CA/Watts, perhaps. I noted these 2 pearls:
“How are you going to convince someone who comes to this site with an open mind, particularly someone who lives in a democracy where the right to challenge the authorities is revered” Patrick Hadley 02/26 @09.47
That’s truly funny when considering how much J. Hansen had to challenge the authorities who were so eager to shut him up. What’s more, those authorities had no scientific basis to justify censure, nor were they suggesting that he was actually wrong and should shut up for that reason. They were just trying to shut him up. And they were exactly the kind of authority that so many say we should be wary of (i.e. the executive branch, the Government). Yet he won’t be a hero to the admirers of authority-challenging behavior. Mc Intyre, on the other hand, never had anybody in an official position with actual power to even try to shut him up (as in preventing him from stating his opinion and fostering his views in the public place). Ironic, really.
The other one was from Kim:
“Sure, they only include the last six years.”
What the heck, who needs more? Especially if these years serve your purpose just right. Did you try to cut all the data in 6 years segments to see if you could find others like that? I mean, if you had a couple more 6 years period showing cooling it could really strengthen the argument, right?
“Why would you say six years is noise, but deny that thirty years is not noise,” (cut here, the rest being, well, possibly noise). Yes, why would you do that? It would truly make no sense. I suspect that many, whom you address this to, would not deny that 30 years is, indeed, not noise (as you stated yourself). You should really proof read, at least a little.
“My arguments are not compelling”
You reckon?
Hank Roberts // February 28, 2008 at 5:49 am
http://www.google.com/search?q=cache:nLSLBqD5nuoJ:www.columbia.edu/~jeh1/distro_realdeal.16aug20074.pdf+hansen+usufruct
Peaseblossom // February 28, 2008 at 6:06 am
So I’m thinking of changing my psuedonym to “Hansen’s Bulldog.”
Woof.
Heretic // February 28, 2008 at 6:26 am
Gosh, this is even worse than it seemed at first glance:
“Pictures certainly are data.” NO. At best, a picture is datum (that’s the singular form). One fleeting moment in time. The ultimate out-of-context quote, in a way. You kinda say it yourself: “If there had been pictures taken of surface stations 20 years ago”. Develop that to: and taken day after day, month after month, year after year. Then, it would be data. What the data would say is yet another question.
Carl G, why don’t you go to NIH to ask for code and data in any subject, in a readily usable form for the average user, less it be all a possible fraud. There is plenty of material there impacting people’s LIVES (and deaths) here and now, and affecting policies with money (public, private, whatever) involved aplenty. Are you interested in that? If not, why not?
And then, why not ask for the same kind of thing from FDA? No reason to stop there, you might as well ask also from Fermilab. And from CERN. And wherever.
But the contrarian/denialist/whatever-you-call-it effort is concentrated on climate science. Who would ever think of that kind of data bickering about clinical trials or physiological processes? Nobody. Why is that? Where does this intense focus come from?
Spare me the hypothetical trillions a la Baliunas/CFC, please.
Anyone truly skeptic should ponder this question.
Ron Cram // February 28, 2008 at 6:47 am
Tamino,
I would like to point out that GISS is being looked at by McIntyre and Watts for two reasons: 1. GISS data and methods are more readily available than CRU data and 2. GISS shows a greater warming trend than CRU data (and CRU is warmer than the satellite data).
Watts has a comparison of GISS, CRU, RSS and MSU data and the results are pretty interesting.
http://wattsupwiththat.wordpress.com/2008/02/27/a-look-at-temperature-anomalies-for-all-4-global-metrics/
fred // February 28, 2008 at 7:18 am
Been away for a while. Deadline. Tamino asks what evidence I had of unscientific behaviour.
Well how about:
1) stopping using data series in 1988 when a full series exists till today (documented on CA for Cobija, Rurrenabaque).
2) Classifying stations as rural that are in fact urban (documented on CA for Yurimaguas, Moyobamba, Chachapoyas, Lambayeque, Tarapoto, Cajamarca, Tingo Maria) and adjusting them accordingly.
3) Having a standard that says you should adjust stations by neighboring ones that are under 10k population, but in fact doing it with ones that are over 20k.
4) Running stations that are way out of spec as documented with photos on surface stations.
4) Changing the readings on the instruments in places like Dawson for years, such as the 1900s, sometimes by as much as 3 degrees, for which there is no evidence of any malfunctioning.
Wild abuse convinces no-one. I was not totally surprised by it, but I was disappointed.
I am not some kind of horned creature refusing to believe things that are obviously true.
I am not using this to avoid believing in AGW, which I have no emotional attachment to one way or the other, unlike many folks here. Surface station data could be being mishandled and misreported and AGW still be true.
But, the evidence is it is being mishandled, and this needs to be cleaned up. The funniest thing by the way is the abuse I received for following up an article supplied by an AGW believer, and being convinced of the merits of adjusting Quebec and of the method used.
That was excellent. And significantly different from making changes to station readings by using a sort of standard curve to take them up or down.
I continue to find it extraordinary that you can qualify as a denier simply by wanting the surface station record done right, though not particularly sceptical about what it shows. Or by wanting to see PCA done right on proxies, while being prepared to concede that the proxy record, when done right, may show unprecedented modern warming. Or by thinking Jones should have revealed the names of his Chinese stations without a FOI proceeding. Or that Thompson and others should archive their data, for future generations if not for us.
But I increasingly realize that you are a heretic if you fully accept the divinity of Our Lord, the Gospel account of all his doings, the manner of His birth, His ethical teachings, and that we should take Communion in memory of him, but are not totally certain about whether His flesh and blood are really present in the sacrament as one substance….
[Response: Let's begin by correcting your first sentence. I asked you "in what way did GISS violate legitimate scientific method?"
It appears that it's not GISS but GHCN which left the post-1989 data out of the combined data supplied to GISS. Maybe there's even a good reason.
Clearly it was not GISS but GHCN which classified urban stations as rural. GISS was sufficiently dissatisfied with the classifications provided by GHCN to devise a whole new method and apply it to the U.S.
Adjusting stations by comparing to other stations which have faulty population metadata is most certainly NOT a violation of legitimate scientific METHOD -- it's faulty metadata.
"Running stations that are way out of spec..." GISS doesn't run any stations at all.
It's not criticism of climate data or procedures that gets you labelled a denialist. Accusing GISS of violating legitimate scientific method based on errors (not violations of method at all), the vast majority of which don't come from GISS but from elsewhere -- THAT earns you the moniker "denialist."
Since you're clearly among friends in the denialist camp, I have some advice for you. Learn to love your enemies, 'cause you can't trust your friends.]
kim // February 28, 2008 at 10:25 am
Heretic, who uses double negatives, dies by double negatives. I had an extra negative in the text you criticize. Couldn’t figure that out, could you?
If six years is noise, are you sure thirty years is not noise. Particularly if the thirty years in question had a PDO in a warming trend?
Observe the reliable thermometers, the RSS and the UAH. Observe the quiet sun. Trust your senses, not me.
===============
EW // February 28, 2008 at 10:46 am
Re throwing out results:
1. If you want to throw out Puerto Maldon based on its disagreement with its neighbors, you have to throw out EVERY station that behaves similarly.
Yes, that’s what I would do. And have a look why it doesn’t match, possibly.
2. EW,
why do you throw it out? Science would not advance is everyone just threw out stuff that did not look right, you need to make it right.
WHAT??? If I set an experiment (or observations) to search for an influence of A to B, in three parallels and in the two the data match, whereas in the third they don’t, then I can’t torture these data to match, because obviously, there was something else in play.
So the only honest thing I can do is to attempt to find out what is the real cause of the difference, or, failing that, discard the series.
Nylo // February 28, 2008 at 11:05 am
Why would the fact of a place being urban or rural affect the temperature trend it shows? Urban thermometers will show higher temperatures, but I don’t understand why they are supposed to also increase their yearly temperatures in a different way than surrounding rural thermometers. I don’t understand why the warming is supposed to increase faster in cities.
It is different if we talk of stations that were rural and became urban, or were already urban but the population increased. But that doesn’t seem to be the criteria Hansen used to decide which stations to adjust. He just classified them according to the population in 1980. Is there a logical reason for that?
Furthermore, I think I have read somewhere that the adjustments so made in cities had changed the trend to more warming or more cooling in a 58%/42% ratio. If that is the case, why should any adjustments be made at all? This 58/42 seems to me that there is no clear effect in warming or cooling trends in cities compared to their rural surroundings, and definitely not one that can be explained for all cases.
Thanks,
Nylo.
Lazar // February 28, 2008 at 12:42 pm
Steve McIntyre’s criticisms are silly beyond belief. He ought to have mentioned the adjustments were following a statistical method applied to all stations, instead he frames the issue with his purported confusion;
and in his Waldo in Russia post;
… and by talking of a declining UHI in the above, and the original;
He has led some rather innocent readers to the false conclusion (Raven);
Note to Raven; if your method for compensating UHI is to set the trends of urban stations equal to rural trends, then you must do so for all urban stations including those which run relatively cooler (or you will introduce spurious cooling).
The ’silliness’ is the fact that James Hansen very probably does not think “the UHI in St Petersburg has declined in the past 15 years” or assume that “something happened that threw the measurements off by 3 deg C” in the case of Puerto Maldon, as such thoughts are irrelevant to the procedure.
What McIntyre is demanding in comments such as;
and;
is that NASA abandon the statistical method of UHI and homogeneity adjustments and do a fundamentally new analysis.
The logical response is; do it yourself.
That, or quantify an error in the current procedures.
Instead, as TCO notes, it’s just showing an extreme example as if that means anything. It doesn’t.
carlG // February 28, 2008 at 1:05 pm
Lazar: I don’t think McIntyre would disagree that you need to apply the same algorithm everywhere, he was merely curious why an urban station would have a cooling trend when rural stations around it were warming.
Secondly, your sentiments are at the heart of the matter of the Peru post: GISS is using two different measures of “urban-ness” worldwide without showing that they measure the same thing.
Barton Paul Levenson // February 28, 2008 at 2:44 pm
kim writes:
[[elspi and Barton, sure CO2 is a greenhouse gas. It’s effect on climate has been exaggerated and the effect of the sun has been ignored. ]]
Both statements are false. Any climate reconstruction or model of the past takes variations in sunlight into account. Variations in the distribution of sunlight are what account for the Milankovic cycles which govern ice ages. The effect of sunlight on climate has been known and accounted for for a long, long time, probably before most of the people on this blog were born.
Barton Paul Levenson // February 28, 2008 at 2:52 pm
kim says:
[[Also, the globe is cooling. Not for the last ten years, as the strawman says, but for the last 2-4 years. ]]
That’s too small a sample size to mean anything, and in the last year temperature went up (NASA GISS land-ocean mean temperature anomaly 0.54 => 0.57). Climate is defined as mean regional or global weather conditions over a period of thirty years or more. A sample size of 2-4 years tells you nothing.
Barton Paul Levenson // February 28, 2008 at 3:11 pm
kim writes:
[[If six years is noise, are you sure thirty years is not noise. ]]
Using a normal distribution, what is the level of confidence for a value with a sample size of six? For a sample size of thirty?
MrPete // February 28, 2008 at 3:26 pm
Tamino, you said
If only your actions matched your words. On this little tussle, Steve began by choosing a nation and stations essentially at random. He immediately found trouble. He wasn’t looking for the worst thing he could find.
You did not respond with welcoming praise.
Steve took a look at your analysis. He found that NOAA actually does have the data post 1988, and thus your statement about there being no data (at all) was false.
You did not respond with welcoming praise.
Now you claim that it is scientifically reasonable for GISS to pull data sets into their analysis with no investigation or validation of the source data, and as long as they use a consistent methodology, all is well.
[edit: comments on the "hockey stick" deleted, as I said they would be until my upcoming post on the subject]
Watts has shown, at minimum, that there’s a significant need to reconsider our use of the US surface station record. And so forth.
You do not respond to these significant efforts with welcoming praise.
All I see from you is defensiveness and derision. And never a humble admission that maybe we don’t know as much as we think we do. You’re a lot more gracious than one or two other bloggers out there whose words are pure poisonous trash, but your statements aren’t winning you any points with this guy who appreciates truth.
Can’t we all get on the same side of shining a bright light where there’s darkness?
I actually do have hope for that!
[Response: MrPete, you said, "He wasn’t looking for the worst thing he could find." I believe you're mistaken; he's ALWAYS looking for the worst thing he can find.
You said, "... your statement about there being no data (at all) was false." As another commenter has pointed out, it's OBVIOUS I was talking about data from the GISS website. I did make a mistake not stating this explicitly, and I made an even bigger mistake thinking that visitors from CA (including McIntyre) would have the common sense to get the obvious.
You said, "Watts has shown, at minimum, that there's a significant need to reconsider our use of the US surface station record." He has done no such thing.
You said, "... you claim that it is scientifically reasonable for GISS to pull data sets into their analysis with no investigation or validation of the source data, and as long as they use a consistent methodology, all is well." I submit that that's one *hell* of a stretch. It's not a true representation of my statement, and quite ignores the fact that GISS does quite a bit of analysis and validation of source data; that's the whole reason for adjustments, which is what this brouhaha is all about.
I'd love to see us all shine light in the darkness. If you really want that to happen, you have to stop following the blind.]
Hank Roberts // February 28, 2008 at 3:29 pm
> he was merely curious
Like so many great men in history, eh? It’s never their fault. Somehow they express their mere curiousity in terms such that people’s feelings are accidentally inflamed, their opinions accidentally transformed to certainty they know a new truth, and that their great leader calls them to war.
It’s those innocent, naive, charismatic, curious men, gathering these angry crowds of followers entirely by accident.
They have no responsibility for what people do in their name.
Riiiight.
Never their fault, never their intent, and they don’t know why people are making them into great leaders.
Yep.
MrPete // February 28, 2008 at 3:30 pm
Lazar notes:
Have you ever been through a corporate accounting audit? Their job is not to do your bookkeeping. Nor do they only seek quantitative errors. If a corporation can’t substantiate the numbers with appropriate documents and internal control processes, they flunk.
When I was in grade school, I had to show my work. Right answer, bad method still was wrong.
As far as I know, that’s still supposed to be true in science today.
cce // February 28, 2008 at 4:12 pm
The only thing interesting about that post by Watts is that it proves he has no idea what he is talking about. Not a clue. After having spent so much time attacking the temperature record, for someone to have to explain the definition of “anomaly” to him at this point is beyond belief. But who cares, right?
If you want to see all four on the same graph, with approximately the same offset, look here:
http://cce.890m.com/temp-compare.jpg
I heard Brit Hume last night quoting Watts that the drop was the largest ever seen by all four of the major analyses, yet there were greater drops by UAH and RSS from 1998 to 1999 (13 months) under similar conditions (strong El Nino to strong La Nina). The biggest drop this time was from NASA, which apparently means that Hansen’s nafarious UHI amplifying machine was turned off last month. There were even greater jumps for UAH and RSS from 1997 to 1998 (about 1 degree), another fact that has supposedly never happened. There was an increase in GISTEMP from February 1994 to February 1995 that was equivalent to the recent decrease, but I’m not sure if February 1994 is far enough away from Pinatubo to make a fair comparison (looking at the aerosol concentration, it looks like it had cleared up by then but there may be “inertia” issues).
And I keep hearing about how GISS is trending away from RSS. If anyone would bother to actually check, RSS and GISS have been converging over time. The fact that recent years have shown GISS moving away from RSS does not somehow negate the larger portion of the record when RSS was warming faster than GISS. But again, who cares, right?
Lazar // February 28, 2008 at 4:24 pm
MrPete;
what I wrote was
… based on ‘errors’ which he has not quantified. One can throw up an approximate infinity of alleged but unquantified ‘errors’ and demand a completely new analysis every time. The demands are unreasonable. This is not the way science is done. If you think there’s a problem with someone’s work, then you quantify an error, or you do your own work using what you think is a better method, and let the scientific community decide.
That is a vastly different demand than updating books.
And that is part of the problem with the ‘auditing’ approach to science.
Yet until you do some quantifying you still don’t know the method is ‘wrong’.
‘auditing’…
LadyGray #107 says;
Yet we’re already down to comment #160 and still no one has bothered to correct Raven in #50… what is the effect if you only adjust ‘warmer’ urban sites to match the trend of rural neighbours?
dhogaza // February 28, 2008 at 4:29 pm
I thought we were talking about science, not wall street … has Tamino changed his pseudonym to “Hansen’s Stockbroker” overnight?
[Response: ROTFLMAO!]
dhogaza // February 28, 2008 at 4:32 pm
It is true today. And, hey, guess what, bad method giving wrong answers is also a no-no.
Replacing statistical analysis with amateur photographs is a bad method that tells us nothing about the answer.
Cherry-picking, a McIntyre specialty, is a bad method designed to give wrong answers. Ouch. That one’s especially frowned upon by real scientists…
luminous beauty // February 28, 2008 at 4:33 pm
MrPete,
What SteveMc is doing bears no relation to a professional audit. It is nothing more nor less than an advocacy centered fishing expedition.
It is research by lynch mob.
Harold Pierce Jr // February 28, 2008 at 5:08 pm
There is a simple way out of this mess: use only rual stations that use WMO protocol. There are fair number of rual or remote stations but it rather tedious and time consuming track them down. And stations with records with before 1900 are rather sparse. I have found a few in Canada and I taking a look at thir records.
For the US places like Death Valley, Telluride CO, Tombstone AZ meet this requirement.
Lighthouse or lightstations are fairly remote and some have records that go far back in time. Maybe somebody should a study of these weather stations.
Hank Roberts // February 28, 2008 at 5:17 pm
> bothered to correct Raven
See the definition of incorrigible.
tetris // February 28, 2008 at 5:24 pm
Tamino
Is the .75 degree C drop in global temperatures recorded by GISS for the Jan 2007-Jan 2008 period based on raw or adjusted data, and where did the data come from?
JCH // February 28, 2008 at 5:24 pm
Corporate auditing is a rigorous intellectual exercise that will put your grade-school arithmetic and preschool calendar studies through the gol darn brain ringer.
Heretic // February 28, 2008 at 5:36 pm
Kim, why did you think I suggested you proof read a little? Could it be because you could then catch double negatives?
Incidentally, I’m not too worried about dying by double negative, and after all you used it first, if inadvertently, so perhaps you will precede me in that most dreadful end.
Have you done a little digging to determine if the trend observed over 30 years (and its extra fast rate) has not been extensively researched in order to be established as not-noise? There is considerable peer-reviewed litterature on the subject, you know. Some of it even produced by people who are also quite knowledgeable in ENSO, I hear.
Trust my senses eh? That’s the new investigative method you recommend? Even at that, I live at over 45 degrees lattitude, right now basking in sunshine at 60 degrees F in late February; it’s been like this for days. What would my senses suggest?
Lazar // February 28, 2008 at 6:33 pm
MrPete;
What Steve McIntyre is doing would be the equivalent of an auditing firm demanding a company change manufacturing procedures and methods. That is not the job of an auditing firm. Moreover he has not quantified any error (reason) for doing so.
As for population data errors these have already been acknowledged by Hansen et al.
I think the main point is that one cannot usually apply the ‘auditing’ meme to science and get reasonably intelligible results.
And here’s how auditing actually works, according to Nick Barnes;
None // February 28, 2008 at 7:04 pm
Chriscolose:
“Svante Arrhenius did not have modern CO2-temperature correlations to look at, nor did he have 850,000 years of ice core data to look at. His prediction that doubling of CO2 would cause 4-6 C of warming (which is just a bit higher than the best estimates today) was grounded in radiative principles.”
As you may or may not be aware the numerical inputs into Arrhenius’ formula were bad estimates. Recent calculations using his formula with more modern estimates for the inputs yield a 1 degree increase for a doubling of CO2, not 4 degree. Regarding the IPCC figures for a temperature increase of 4 deg plus they are based on a water vapour feedback calculations, not merely as a result of doubling CO2. So, still its clear, the science of global warming is not based on 100 year old physics.
The models used to predict these figures very clearly massage their inputs to match the 20th century record: check the peer reviewed paper by Kiehl, Twentieth century climate model response and climate sensitivity. So really, the temperature increases forecast are not based on century old physics, but on models where the parameters are adjusted so the particular algorithms used by the particular models best fit historic data…
MrPete // February 28, 2008 at 7:05 pm
That’s a great recipe for accepting spurious correlation. All the numbers can look just fine, and still be wrong. Methods and meaning are rather important.
There seems to be a general sense here that Steve’s “cherry picking.” I have to laugh! His biggest complaint to date is about the cherry picking that is so easily accepted.
Folks, this is not about taking potshots at the weak edges of various arguments. As someone has said, data is not perfect… nor are people. You can always find a typo somewhere to complain about.
The issue is whether we can all agree on, and strongly urge, a move towards adoption of “good science” in these subfields of climate science, as we have all come to expect in other scientific arenas. I’m saddened, but no longer surprised, that people so easily get away from that, to defend their favorite ‘champion.’
kim // February 28, 2008 at 7:14 pm
It was a little vague; I meant I die by double negatives, not you. It appears, as I thought, that you understood me despite the error. The warming PDO for the last 30 years explains the global warming over that period better than CO2 does. The recent flip, with onset of La Nina, explains the recent cooling better than rising CO2 does.
In general, and specifically in response to your last paragraph, I would suggest the use of common sense.
This is what will blow up the anthropogenic global warming myth(hoax). Common people won’t understand why the globe is cooling while CO2 keeps rising. They will then rightfully realize the sun has more to do with both heating and cooling of the globe. What an uncommon thought!
By the way, a Nobel awaits whoever figures out how small changes in solar output translate into larger changes in climate. It is probably through the oceanic oscillations and the behaviour of clouds.
Check out a recent paper by Gill at al, showing that the collapse of the Mayan civilization co-incided with a 200 year drought, associated with decreased solar influence.
===============
MrPete // February 28, 2008 at 7:15 pm
“places like Death Valley…”
“Rural” ain’t enough. Death Valley is a good case in point.
here is the USHCN station. Suffers from the usual modern challenge: MMTS sensors tend to be placed close to a building instead of out in a field like the older stations.
Here is a map depicting the location of the Badwater sensor “hottest place in the world” — except they placed it in a very poor location.
The 30,000 foot view can’t tell you such things. Should scientists care? I would think so.
MrPete // February 28, 2008 at 7:20 pm
I’d be more cautious about ROTFL about auditing and accountability.
Rather large scientific endeavors can come quickly crashing down once the digging begins at a high enough level.
Better to get our act together, quickly, from the inside than to keep deriding the concerned critics until it’s too late.
Raven // February 28, 2008 at 7:37 pm
Lazar says:
“Yet we’re already down to comment #160 and still no one has bothered to correct Raven in #50… what is the effect if you only adjust ‘warmer’ urban sites to match the trend of rural neighbours?”
Probably because people on CA realize that one must never adjust data without a good justification. The adjustments for removing spurious *heating* caused by urbanization should result in reduced trends in the urban records. Spurious cooling caused by urbanization has not be identified as a problem which means one cannot justify ‘correcting’ it.
If an algorithm to remove UHI effects results in urban sites having *heating* added then that is good sign that there is a big problem with the algorithm. In this case, the problem is likely a result of ‘rural’ sites that are not really rural.
The affect the fact justifications by Tamino (i.e. the site is obviously bad and should be adjusted) make no sense since removing UHI effects is the stated purpose of the algorithm.
[Response: Am I correct, that you stand by your claim that the right thing to do is to adjust urban station data ONLY if the adjustment is a cooling one?]
Harry Eagar // February 28, 2008 at 8:22 pm
Martin Vermeer sez: ‘It is of course sobering to realize that some stations contain error patterns on the 3 degree level. ”
Obviously, it isn’t, or Tamino would not have written this post.
Frank // February 28, 2008 at 8:33 pm
I am a CA and RC lurker, rarely post, and am glad to find another site to learn from.
I understand Mr. Watts’ role in this climate discussion is a whistleblower. Never mind the motivations. Everybody has motivations. Is there an issue with the facts?
We, the people, have contracted with our government, over the past decades to measure temperature at thousands of sites across the US. Specifications were drawn up by scientists to ensure that the temperatures would be as reliable and consistant as possible within the technological constraints of the day. There are documents which describe in great technical detail all of the aspects of choosing a site, equipment, maintenance, etc. , such as the WMO docs:
http://www.wmo.ch/pages/prog/www/IMOP/publications/CIMO-Guide/Draft%207th%20edition/Part1-Ch01FINAL_Corr.pdf
Mr. Watts is simply documenting the failure of the contractor to provide the services requested.
(i.e. a temperature record obtained using the specified means)
The data is believed to be damaged because the specifications were not met, in some percentage of cases. Dr. Hansen, Mr. Watts, and Mr. McIntyre will agree on this point, (if no other :>) ) because Dr. Hansen has crafted some imaginative means of attempting to extract the true temperature signal record. I do not say imaginative in a childish way, but in a creative, outside the box, way.
Maybe there is data that can be salvaged from the station data. Technology gives great power to pulling data out of noise, when the noise is describable (as in being able to be modeled). NASA is an incredible storehouse of creative and talented scientists and engineers. For years I thumbed through issues of NASA Tech Briefs. Lots of creative and competent folks.
Dr. Hansen has created a means of extracting a true temperature record, to his satisfaction.
Where are the research papers and articles documenting the experiments, their results, and how these results allow correction of the temperature records of these stations? Where are the papers that not just “explain”, but document smaller scale experiments that show correction of station data with population data, or “nightlights” brightness data, or by creating entirely new data for a site by a “model” which incorporates temperature data from locations “nearby”?
We are told that the GCM outputs match the temperature record. We then find out that this “temperature record” is actually the output of another “model”. So one model output matches the output of the other model. Is that correct?
Are there papers out there which experimentally confirm and document the effectiveness of these models?
Thank you for your time, thoughts, and opinions.
fred // February 28, 2008 at 8:36 pm
Hank Roberts, February 28, 2008 at 3:29 pm.
Hank, you have a problem here, and its to do with how debate progresses in open societies. People say things we do not agree with, find threatening, and when others find what they say convincing, we like it even less.
But it is how debate progresses in a free society. Gradually, a consensus emerges, arguments are weighed and found wanting, eventually people are convinced, and usually they get to the right answer in the end.
Steve M is neither a demagogue nor any sort of danger to our society. He’s expressing a point of view which, like Tamino’s, convinces some and not others.
Get used to it. Learn to tolerate differences of opinion. Also, it would help to stop using the terms denialism and denialists. They are part of your problem.
JCH // February 28, 2008 at 8:41 pm
“The recent flip, with onset of La Nina, explains the recent cooling better than rising CO2 does. …”
If there was no CO2 in the atmosphere as the result of mankind’s activities on earth; as in, a natural atmosphere, what would the current La Nina be like?
Bishop Hill // February 28, 2008 at 8:45 pm
I don’t really want to divert the conversation, but Nick Barnes’ understanding of what he is required to show an auditor (as quoted above) is woefully mistaken.
A UK auditor is allowed to see anything they think is necessary for their audit.
Rattus Norvegicus // February 28, 2008 at 8:47 pm
Kim sez:
“Check out a recent paper by Gill at al, showing that the collapse of the Mayan civilization co-incided with a 200 year drought, associated with decreased solar influence.”
Jeez and I always thought that the Mayan civilization began it’s collapse around 950 AD. Wasn’t that the MWP you so love to carry about like a talisman? You can’t have both a MWP, presumably caused by increased TSI, because no other forcing can really explain it, and a 200 year drought in Central America and the Yucatan, caused by decreased TSI occuring at the same time.
I know this won’t convince you of the absurdity of your position, but I hope that someone picks up the inherent contradiction and stops listening to the fairy tales.
Martin // February 28, 2008 at 8:53 pm
One stated purpose. The other is eliminating erroneous trends in individual station data that have nothing to do with UHI (“cross-validation”), as is done throughout the data, not only for urban stations.
If you only correct positive trends in urban areas, you will eliminate the urban heat island effect. You will also eliminate only the upgoing half of the station trend errors present on top of possible UHI. The downgoing half will remain, systematically distorting the average. (Whether it’s a big bias? I doubt it. But methodically it’s wrong).
This is really, really obvious to anyone with a background in statistics having been involved in this kind of thing, like me (well, tide gauge records and permanent GPS) …
luminous beauty // February 28, 2008 at 9:25 pm
“The issue is whether we can all agree on, and strongly urge, a move towards adoption of “good science” in these subfields of climate science, as we have all come to expect in other scientific arenas.”
One problem is that a hand full of corporate bean counters, whose field specific expertise is avocational at best and completely lacking for the greatest part, looking at a hand full of anecdotal cases from a single line of research from a multitude of independent, cross-correlated and corroborative sources, of which they can only say superficially seem counter-intuitive, is not sufficient evidence for a reasonable person to conclude that climatologists are using bad science.
I hope we can agree that the best scientific practice should always be used.
I should hope we can agree poorly substantiated accusations of sub-standard scientific practice are just so much crap.
Raven // February 28, 2008 at 9:33 pm
Tamino says:
“Am I correct, that you stand by your claim that the right thing to do is to adjust urban station data ONLY if the adjustment is a cooling one?”
That is not what I said.
If you a design an algorithm to remove spurious heating but find later that algorithm is adding heating then that is a sign that your algorithm has a serious flaw.
If spurious urban cooling is a statistically significant phenomena caused by problems at urban stations and removing it is part of the stated purpose of the algorithm then removing cooling can be justified.
This, of course, is seperate from the problem created by bad metadata which can render any algorithm useless.
[Response: This is what you said at CA:
Urban heating should be removed, but urban cooling not. This sure seems to me (and apparently, a lot of other folks as well) to be a very clear outright statement that if the homogeneity adjustment is a cooling one it should be kept, but if it's a warming one it should be discarded. If that's not what you meant, please explain how you can reconcile that with your own words.
If it is what you meant, do you stand by that claim?]
Harold Pierce Jr // February 28, 2008 at 9:34 pm
RE: JCH
Probably about the same. The ENSO takes place in the Pacific along the equator from S. America to Australia and Idonesia. It is like the 800 pound gorilla of the climate. Not a good idea to annoy it.
Lazar // February 28, 2008 at 9:48 pm
MrPete;
Disagree. That is not quite what I implied when I wrote;
The challenge is simple. Steve McIntyre is claiming the adjustments by GISS are possibly physically unsound, and is insisting therefore GISS change their methods. He can either demonstrate the adustments are unsound, or he can’t. If he can’t quantify an error, then he has nothing, and no reasonable person would demand GISS change their methods.
Barton Paul Levenson // February 28, 2008 at 9:54 pm
None writes:
[[The models used to predict these figures very clearly massage their inputs to match the 20th century record:]]
Nope. Wrong.
All that changes when they upgrade a climate model is the physics. If they’ve found some newer and more accurate way to represent a process, that’s what they put in. There is nothing from the climate record that goes into those models except basic geographical data like the elevation and surface albedo of a given grid square.
Harold Pierce Jr // February 28, 2008 at 9:57 pm
MrPete:
Go: http://john-daly.com, scroll down and click on “Station Temperature Data” and find the chart for Death Valley.
What is very interesting is that annual mean spring and fall plots coincide with the annual mean which is half way between the summer and winter plots. Preumably this occurs because Death Valley is below sea level and is sheltered from the wind and weather. And of course there are no humans around to mess up the weather station measurements.
MrPete // February 28, 2008 at 9:59 pm
luminous beauty says “I should hope we can agree poorly substantiated accusations of sub-standard scientific practice are just so much crap.”
You betcha! So that’s why you should be welcoming Steve’s work with applause. After all, he’s well qualified in his specialty (statistics), (which certainly applies to climate science BTW)… Wegman is probably more qualified than you are to ascertain the quality of Steve’s work — and Wegman vouches for Steve’s work.
“Poorly substantiated” has never been a serious accusation against Steve’s work.
I was part of the little expedition to validate or falsify the Starbucks Hypothesis (and a bit more). What we did was reasonably solid work. Even without further extensive analysis, our basic results have substantiated some of Steve’s claims. So far, scientists have quietly provided quite a bit of appreciation and no serious opposition to the work we did. (And to think we had a blast doing it!)
What amazes me is that I, who definitely fit the “field specific expertise is avocational at best” moniker, am able to easily produce field data that is at least comparable to what the specialists generate…I won’t claim our data is of better quality, although some others have made such allusions.
To me, that’s a concern. Why are people “doing time” in grad school etc, if they aren’t being prepared to care about the simple basics of metadata and data handling?
kim // February 28, 2008 at 10:03 pm
Harold, I wish I could attribute this but I’ve heard it said that the climate is the continuation of the ocean by other means.
Norway, you can lead a rat to the hawser, but you cannot make him flee. I expected you to at least sniff at the rope before you hanged yourself with it.
=====================
Lazar // February 28, 2008 at 10:03 pm
Raven;
I think you’re conflating the (one of the aims of) statistical effect of the algorithm with its effect on individual station records. It adds warming to an individual station record (method) and negates the statistical UHI effect (aim). If it did not warm some individual records, then it would add a statistical spurious cooling effect.
Martin // February 28, 2008 at 10:12 pm
Frank: what are you missing in http://data.giss.nasa.gov/gistemp/ and refs therein ?
Not quite like that. These “models” are entirely different beasts:
* One type, the GCM, models the large scale physical behaviour of the Earth’s atmosphere (and oceans) under the known forcings, and does not use observed temps as input, and
* the other type describes the successive reduction steps needed to get from the raw station data from all over the world to a global areal average temperature product that is as free from the effects of gross, systematic and random errors — and areal averaging uncertainties! — as it is possible to do based on inevitably imperfect data.
So they are not directly comparable, and the “match” — a very rough and ready agreement of the global mean temperature curves, within the inevitable uncertainty of the weather — actually does mean something, as in: comparing observations to theory.
BTW Hadley Centre do their own processing of global temperature data. They arrive at similar (and sometimes slightly but interestingly different!) results.
DocMartyn // February 28, 2008 at 10:46 pm
Barton Paul Levenson, I here much of the physics based approach to the role role of CO2 in AGW. May I ask if you understand the difference, thermodynamically, between an equilibrium and a (quasi)-steady state.
If you do understand the difference, could you be so good as to answer why the equations used in heat transfer models rely on equilibrium thermodynamics.
The biggest difference in the two approaches would be in terms of modeling the temperature of an iron bar, like a poker, placed into a hot fire (say 800 degrees). The steady state thermodynamicist would predict that the end in the fire would be hottest and the temperature would fall away in a stretched exponential manner. What temperature would an equilibrium thermodymicist predict the bar to be; (room + 800)/2 ?
None // February 28, 2008 at 10:47 pm
Barton:
“Nope. Wrong.
All that changes when they upgrade a climate model is the physics. If they’ve found some newer and more accurate way to represent a process, that’s what they put in. There is nothing from the climate record that goes into those models except basic geographical data like the elevation and surface albedo of a given grid square.”
I am not saying there is anything from the climate record going into the models, that is a straw man. I am saying the models are run, their output compared to the historic figures, then the input parameters adjusted in an effort to improve the match. It’s classic data mining.
Otherwise how to you explain the models with high sensitivity all having low aerosol forcing values, and the models with low climate sensitivity all having high aerosol forcing values ? They’ve picked the values that give the best match to the 1950-1970 temperature decline. Why don’t you take it up with Kiehl ?
As Kiehl himself put it:
“Climate model simulations of the 20th century can be compared in terms of their ability to reproduce this temperature record. This is now an established necessary test for global climate models.”
So, its necessary that models are compared to the historical trend.
So when they “improve” their models, the condition is not “Are the parameters correct ?”, the condition is “Is the historical record fitted better ?”
That’s why there’s no real effort by these modellers to get out and empirically get as accurate as possible a figure for the aerosol forcing, or other such parameters because to a large extent the accuracy of their models is based on a forced choice of parameters to get the best graph fit.
Anway, my main point was not that the modellers pick their parameters to get a best fit then project this forward to get their “likely future global warming”, but that this “likely future global warming” is not based on proven 100 year old science, but that its based on models.
Frank // February 28, 2008 at 10:52 pm
Martin // February 28, 2008 at 10:12 pm
Thanks for the link. I’ll study.
MrPete // February 28, 2008 at 11:40 pm
Harold writes, “What is very interesting is that annual mean spring and fall plots coincide with the annual mean which is half way between the summer and winter plots. Preumably this occurs because Death Valley is below sea level and is sheltered from the wind and weather. And of course there are no humans around to mess up the weather station measurements.”
For lurkers, here is the plot Harold’s referring to.
It’s possible that there’s some special significance to the general alignment of the spring, fall and annual data. Of course, the definition of spring and fall could be chosen to fit halfway between winter and summer.
Ever been to Death Valley? The fact that it is below sea level does not lead to a conclusion of being protected from wind and weather. Our tent was constantly buffeted by wind when we camped there, and we had a nice rain storm too. (I was glad for that — several bus loads of school kids showed up and stirred up huge clouds of dust!)
Here is a good starting place to see what’s actually in Death Valley — a lot of empty space, but also a lot of people, depending on the season.
Zoom out, and you will see links to some online-accessible weather stations. Here is a report from the station one valley further east — even more “protected.” Still windy, still rather dry.
Fun stuff, guessing the climate somewhere! Often quite different from what we experience at “home” (wherever that may be.)
Hank Roberts // February 28, 2008 at 11:41 pm
Tam, er, Dawg, where did you get the raw numbers to make the chart at the beginning?
“the data for Puerto Maldon (raw monthly data, anomalies)” — I haven’t found where the raw numbers came from yet.
Phil. // February 29, 2008 at 12:04 am
Re None
“As you may or may not be aware the numerical inputs into Arrhenius’ formula were bad estimates. Recent calculations using his formula with more modern estimates for the inputs yield a 1 degree increase for a doubling of CO2, not 4 degree. Regarding the IPCC figures for a temperature increase of 4 deg plus they are based on a water vapour feedback calculations, not merely as a result of doubling CO2. So, still its clear, the science of global warming is not based on 100 year old physics.”
Even if that is the case it’s still 100 yr old Physics (Chemistry) since it is based on the Clausius-Clapeyron equation.
Hank Roberts // February 29, 2008 at 12:16 am
Hobby horse stampede, just like dot.earth.
I’m outta here til playtime ends and class resumes.
Robert Wood // February 29, 2008 at 12:55 am
One data set is not like the others, therefore correct it!
How about just throwing out data sets that need correction?
carlG // February 29, 2008 at 1:04 am
Heretic: nobody asked for specially designed code that was easy to understand, although that would be hoped for. McIntyre asked, for over a year, for the exact code used to process station data and did not receive it until after said year. This would take little more than a few uploads to a FTP site. I am pretty sure that if I filed a similar FOI request that most organizations (including all of your examples) would comply more quickly.
But, as Tamino is implying, this is beating a dead horse since the code is now out there.
Lazar // February 29, 2008 at 1:08 am
… Raven, perhaps a better explanation. GISS assumes the average trend of nearby rural stations is generally more representative, of the natural local climate trend at urban sites, than is urban data contaminated by the UHI effect. So they chuck out the urban trend entirely, discarding components due to natural climate variability, UHI, microsite effects, instrumental errors, everything. If they do that just to ‘warmer’ sites, they introduce artificial cooling.
Harold Pierce Jr // February 29, 2008 at 1:34 am
MrPete:
No never been in a real desert. Went to grad school at UCI, but didn’t do too much traveling around. Came to BC in ‘72. This prov truly is “The Best Place on Earth” Enviroment is squeeeeeeky clean, lots of pure fresh water and air , and oodles of resources, and cheap electricity: 7.7 cents/kwh.
I went on several occasions to Kamloops for soccer tournys, which is arid (ca 12 in of rain /year), and desert like. There is lots of sage brush and really laid-back marmots. These guys sun themselves on conrete barriers right next to busy highways. Thanks for the link to wunderground.
Temp records from Deathy Valley could be used as a check for AGW. Since there is low moisture, the nighttimes get quite cold since there is no greenhouse effect. If CO2 is an important factor for trapping IR, then we would expect the min temperature to increase and to track the increase of this gas overtime from the start of record keeping. But we really don’t have to do this. The time-temp plots are flat lines; thus, there is no effect of CO2 on temperature at this site.
As Alfred E said: “One fact can change the consensus of one hundred scientists” Death Valley is that “one fact”.
Harold Pierce Jr // February 29, 2008 at 1:53 am
MrPete:
BC has pocket desert near the washington state border and it is the very end of the Great Basin Desert. There are weather stations at Osoyoos and I going to check out my hypothesis about min temp and CO2 conc. Stay tuned everbody!
Hank Roberts // February 29, 2008 at 1:59 am
Mr. Pierce, track cites forward:
http://links.jstor.org/sici?sici=0435-3676%281970%2952%3A3%2F4%3C160%3AANEAMB%3E2.0.CO%3B2-0&size=LARGE&origin=JSTOR-enlargePage#abstract
dhogaza // February 29, 2008 at 2:11 am
The warming or cooling or no-change of a single location is meaningless. I can’t believe someone can believe something as false as your claim.
Well, no, actually, the NW boundary of the hydrological Great Basin lies a bit east and north of Bend, Oregon.
There’s plenty of sage-steppe outside the Great Basin, however.
Lazar // February 29, 2008 at 2:51 am
DocMartyn;
Probably because of “local thermodynamic equilibrium”.
See IPCC AR4 chapters 8 and 9 for model-observation comparison.
With that, I agree with Hank Roberts;
… farewell.
MrPete // February 29, 2008 at 3:00 am
Lazar, perhaps we’re talking past one another? I hope not. “until you do some quantifying you still don’t know the method is ‘wrong’…He can either demonstrate the adjustments are unsound, or he can’t. If he can’t quantify an error, then he has nothing.”
What I’m saying is simply this: a method can be falsified without quantification. I agree that falsification is important. But the adjustments are built on more than a numerical method.
This is why I brought up spurious correlation. To paraphrase a famous example… If I show there’s an inverse mathematical correlation between historical railroad ridership and global temperature change does not mean I have proven anything at all. One could falsify it without any use of quantitative analysis.
This is about the meaning of numbers, and the metadata thereof, as much as the quantitative analysis itself.
Another example: in y2K days, I proved that embedded clock chips were not going to be an issue. What did it require? Going to the source and asking a few questions, enough to know that the statements made by the doomsayers were simply wrong. Not quantifiably wrong. Just wrong.
Yes, I could have generated a mathematical model of the accumulated difference between what people were estimating would happen in those chips vs the reality. But why? I could explain in a couple of sentences how they got it wrong. Their facts were wrong, their analysis was wrong. No reason to make it more complicated than necessary.
Saved some friends in the business a chunk of change. :)
Lazar // February 29, 2008 at 3:14 am
MrPete;
Yes, that’s fine, obviously there’s no causal relationship there.
On the other hand, we know very well physical processes by which temperatures correlate over small and large distances, and observe correlation to a great degree. When we’re talking of homogeneity adjustments to some of those temperatures, and the relationship to regional and global averages, in my opinion you’re going to need numbers if you’re to claim those adjustments are worse than leaving the data alone or any other approach you may choose.
dhogaza // February 29, 2008 at 3:23 am
Great, so you proved the METHOD used by the doomsayers was wrong without quantifying it.
To satisfy my own curiosity, can you tell me, in simple words, what METHOD the doomsayers had used to make their predictions?
Since you proved their METHOD was wrong, surely you know what it was?
Since methodology was the problem (this being the basis of your analogy), surely it wasn’t just a matter of them being IGNORANT of how the clock chips you were studying represent time, right?
Or are you suggesting that the NASA folk are so pig-ignorant that they don’t even know the weather stations use thermometers?
Greg // February 29, 2008 at 3:24 am
Yes, Lazar, we all understand that. But I can construct a sentence like this:
a) The Urban Heat Effect is a documented, proven, and widespread effect that tends to introduce a spurious heating trend to urban stations as opposed to rural ones, … b) therefore we introduce an algorithm than detects urban stations with a stronger upward trend than nearby rural ones, and we homogenize them downwards.
I can’t however complete this sentence:
a) ??????? … b) therefore we introduce an algorithm than detects urban stations with a lower trend than nearby rural ones, and we homogenize them upwards.
If you have a valid physical effect to replace the question marks above, that is likely to be shared at many stations throughout the entire dataset, please tell me.
Or put another way, Tamino questioned: “Urban heating should be removed, but urban cooling not. ”
Does urban cooling even exist? If it doesn’t then adjusting for it is just butchering the data.
Heretic // February 29, 2008 at 3:32 am
“I am pretty sure that if I filed a similar FOI request that most organizations (including all of your examples) would comply more quickly.”
Really? Why don’t you give it a try? And make it one where a fat, juicy medication market is involved, we’ll see.
Hank Roberts // February 29, 2008 at 4:16 am
> not quantifiably
Local weather isn’t climate.
http://lists.ucc.gu.uwa.edu.au/pipermail/ucc/2000-January/001370.html
http://www.talkaboutprogramming.com/group/comp.lang.cobol/messages/128158.html
fred // February 29, 2008 at 8:08 am
Tamino is right, in my remarks about unscientific behavior I have confused the behavior of those running stations, those collecting or failing to collect the data, and those using it.
The accusation to be levelled at GISS is that it has used suspect or incomplete data, and made unjustified changes to it, not that it has run bad stations or itself omitted available data from the series. I don’t know that this is much better, but it is certainly different. Sorry about that.
Martin // February 29, 2008 at 9:10 am
Greg:
What about this:
a) The Urban Heat Effect is a documented, proven, and widespread effect that tends to introduce a spurious heating trend to urban station…
b) therefore we introduce an algorithm than detects urban stations with a stronger upward trend than nearby rural ones, and we homogenize them downwards as opposed to rural ones,
c) However, in doing so, we will inevitably also eliminate upward trends in urban stations that are not due to the urban heat island effect, but due to, e.g., there having been something wrong with the station itself, its instrumentation, the way it was operated, or its immediate environment.
There is no a priori reason why such a station-specific error would not as often introduce a spurious downward trend as it would an upward one.
d) therefore, in order not to introduce a net bias, we also introduce an algorithm that detects urban stations with a lower trend than nearby rural ones, and we homogenize them upwards.
Renumbered your a), b) to c), d), rephrased a bit.
chriscolose // February 29, 2008 at 10:29 am
None,
the following quotes are from Arrhenius (1896);
“The selective absorption of the atmosphere is, according to the researches of Tyndall, Lecher and Pernter, Röntgen, Heine, Langley, Ångström, Paschen, and others, of a wholly different kind. It is not exerted by the chief mass of the air, but in a high degree by aqueous vapour and carbonic acid, which are present in the air in small quantities. Further, this absorption is not continuous over the whole spectrum, but nearly insensible in the light part of it, and chiefly limited to the long-waved part, where it manifests itself in very well-defined absorption-bands, which fall off rapidly on both sides.”
and from part 4 (italics from me)
“We now possess all the necessary data for an estimation of the effect on the earth’s temperature which would be the result of a given variation of the aerial carbonic acid. We only need to determine the absorption-coefficient for a certain place with the help of Table III. if (he didn’t capitalize :-) ) we know the quantity of carbonic acid (K = 1 now) (K referred to “carbonic acid” in Arrhenius’ piece) and water-vapour (W) of this place. By the aid of Table IV, we at first determine the factor p that gives the mean path of the radiation from the earth through the air and multiply the given K- and W- values by this factor. Then we determine the value of β which corresponds to pK and pW. Suppose now that the carbonic acid had another concentration (e.g. K, 1.5). Then we at first suppose W unaltered and seek the new value of p and p_1 (proper subscript were given in his text), that is valid in this supposition…(skipping a bit of stuff)…In consequence of the variation in the temperature, W must also undergo a variation. As the relative humidity does not change much, unless the distribution of land and water changes (See Table 8 of my original memoir), I have supposed this quantity remains constant, and thereby determined the new value W_1 of W. ”
……
Arrhenius evidently was well aware of water vapo(u)r feedback, constant relative humidity, and at least the overall concept of “solar in, IR out, IR absorbed, etc). This was not at all ignored in his calculation. I do not claim his work was perfect, but the physics was known (and improved upon). The big thing back then was that people were not convinced CO2 could rise at today’s rates (until Keeling, Suess, Revelle, etc came along). Tyndall showed that water vapor and carbon dioxide were the important substances exerting a strong influence on the transmission of heat rays.
Now I suppose Arrhenius was being slipped some money under the table from the eco-nazis (like all the other scientists today), or if you’re a really crazy conspiracy theorist you could just say he was trying to solve the ice age mystery.
I probably did the history an injustice, since the idea we can influence climate, and basic understandings of the greenhouse effect extend back to Fourier. As time evolved, the work continued to pile up, and the literature on the physics so well documented…until of course the likes of Channel 4 and the Marshall Institute disproved it all.
Kim,
like my CSI detective, we’re all looking for natural causes as well. The PDO explanation by venues like wattsup fails on many grounds– why is heat going in the ocean and not out, the radiative imbalance, why is the globe warming and not heat just being shifted around, why is the stratosphere cooling; I can go on and on). There are three proposed mechanisms whereby solar activity changes climate
1) Change in TSI
2) Change in UV, which effects heat in the stratosphere, may effect circulation
3) Change in cosmic rays and possible effects on cloud micro-physics
Now given that we have no real trends in solar since about 1950 (See Benestad 2005, Lockwood and Frohlich 2007) it is interesting to see how solar variability is now changing so much after 10,000 years of stable Holocene time (albeit some changes in solar), but now is causing a rapid warming signal, cannot be detected by satellites and not in line with what it should do to summer v. winter temps, the DTR, strat. heating, etc.
I truly wish I could disprove AGW. It might make me feel a bit better about the future, but I’d also have a Nobel and be famous. But, I guarantee I will not do it.– chris
Harold Pierce Jr // February 29, 2008 at 10:38 am
Hank Roberts
Thanks for the reference. After reading the first page i get the impression they are not really experimentalists. They did only one experiment
apparently. I ‘ll have to fetch the article from the SFU library. How did you know of this article?
BTW, please address me as Harold but not Hank for short. These were common given names back in the ’40’s and ’50’s, but are no longer used these days.
Harold Pierce Jr // February 29, 2008 at 11:03 am
dhogaza:
Don’t be so picky! In BC we call it a pocket desert. Really draws in the tourists especially since several wineries started up a few years ago. This little desert has a number of plants and animals found nowhere else in the world.
There are a great many deserts in the world. Death Valley is n=1. Australia has a number of weather stations in the desert many with records that start before 1900, e.g., Alice Springs.
Go to GRISTEMP and check the records for Tombstone and Yuma AZ. These look pretty flat to me. Go to http://www.john-daly.com and check out the temp records for AU. Click on “Station Temperature Data” to acess the data base.
Nick Barnes // February 29, 2008 at 11:21 am
Bishop Hill points out a misconception of mine:
My rant about the nature of auditing in business was based on my experience of business, not on a knowledge of the law. My essential point remains: an auditor is selected by the business, works closely with the business to identify and eliminate problems in the accounts, generally forms a close and cooperative relationship with the business, operates under the cover of confidentiality, and behaves in any number of ways utterly unlike CA.
EW // February 29, 2008 at 12:00 pm
Robert Wood said:
“One data set is not like the others, therefore correct it! How about just throwing out data sets that need correction?”
Weelll, that’s exactly what I suggested many posts above. For some reason, this approach isn’t the favored one, whereas applying various
“correcti0ns” even in case of entirely different trends apparently is…
Boris // February 29, 2008 at 1:40 pm
MrPete says:
“You betcha! So that’s why you should be welcoming Steve’s work with applause. After all, he’s well qualified in his specialty (statistics), (which certainly applies to climate science BTW)… Wegman is probably more qualified than you are to ascertain the quality of Steve’s work — and Wegman vouches for Steve’s work.”
Then why doesn’t Steve M write up a complete analysis of anything? An auditor actually completes the work and does not stop in the middle. Steve M’s blog is a mess with about, what 1500 posts on dozens of topics? Actual science gets published somewhere where it will be permanent and where researchers can find it. CA isn’t indexed anywhere, so how is some researcher in Argentina going to know if CA has found anything important? Steve M must know this, so why not publish?
In fact , the best suggestion that was made by Edward “no net global warming since 1998″ Wegman was that blogs were inappropriate places for discussing these types of issues.
Until Steve M actually takes the time to complete an analysis rather than showing there are some unspecified “problems” that have unspecified
results, then I think it’s more likely that he is sniping from the sidelines. Given the insults he has flung at the folks at Real Climate, sniping is what he does best.
(Yes, I know he had a paper in GRL. That was 2 and a half years and 2,000 insinuating blog posts ago.)
Barton Paul Levenson // February 29, 2008 at 1:56 pm
None writes:
[[I am not saying there is anything from the climate record going into the models, that is a straw man. I am saying the models are run, their output compared to the historic figures, then the input parameters adjusted in an effort to improve the match. It’s classic data mining.]]
No. You’re still wrong. The input parameters are not “adjusted in an effort to improve the match.” Do you understand the difference between a parameter and an algorithm?
Barton Paul Levenson // February 29, 2008 at 1:59 pm
Robert Wood writes:
[[One data set is not like the others, therefore correct it!
How about just throwing out data sets that need correction?]]
Because scientists would never do such a thing. You don’t throw out data with a bias, you correct for the bias. The fossil record is biased toward creatures with hard parts, but palaeontologists don’t throw out the fossil record. The deep space sky surveys are biased toward past time, but cosmologists don’t throw out the deep space sky surveys. The idea that biased data should just be thrown out is simply wrong. You would lose valuable data that way.
dean_1230 // February 29, 2008 at 2:07 pm
Nick,
You failed to mention the most salient point about auditors. Auditors are bound to use a verifiable set of standard practices to gauge how well the company is performing. Financial auditors are to use GAAP principles when auditing the company’s finances. Manufacturing auditors use ISO principles. These principles are well defined and deviation from these principles must be fully explained.
In the auditing of data, the use of standard, accepted statistical methods are required. Deviation from these methods require documentation as to 1) what the new method is in sufficient detail for others to judge the validity of the method and 2) some sort of explanation why the standard methods are not acceptable or are inaccurate.
fred // February 29, 2008 at 2:42 pm
Hans Erren has an interesting account of Arrhenius. Historical interest only. He was wrong about sensitivity to CO2, and I think that is generally accepted now. Hans Erren explains the detail. Not that we have to defend Arrhenius too, do we?
http://home.casema.nl/errenwijlens/co2/index.html
look at the links half way down the page.
steven mosher // February 29, 2008 at 2:59 pm
H2001 used 1980 population data to determine rural for nonCONUS stations. We can second guess this, but the better thing to do is understand some of the effects this choice can have on the adjustment algorithm and seek
out improved methods.
The approach treats a station that was Rural in 1980 ( pop lt 10000) as rural in 2008.
This means it gets no adjustment.
So, one thing we would want to look at is places that have population growth since 1980. This is especially important if we believe Oke. Oke established a log relationship between population and UHI. UHI=.73Log10(pop)
so, a town growing from 5000 people in 1980
to 15,000 today would see a Trend in warmth
due to population growth ( plus a trend in global warming) A urban city growing from 100000
in 1980 to 110000 in 2008 would have the same global warming trend, but less UHI trend. It would look like the rural were warming faster than urban. I would not suggest there are a lot of cases like this. I would look for them, and see how many. Then try to improve the algorithm.
EW // February 29, 2008 at 4:07 pm
“You don’t throw out data with a bias, you correct for the bias. ”
yeah, but first you has to know, what exactly for bias is in the play.
I’m a biologist. If I have three cultures, two behave similarly and the third is different, I have to know why if it is different. Maybe there was too few microbes inoculated, maybe there is a contamination, maybe there was a pipetting error. But if I don’t find out, I have to throw it out. I can’t just multiply the result by 3.7658793 and declare it homogenized.
Barton Paul Levenson // February 29, 2008 at 6:02 pm
EW — No, you can’t. But that’s not what climatologists are doing, so why even bring up such a ridiculous straw man argument?
Hank Roberts // February 29, 2008 at 6:22 pm
EW, do you submit research papers to peer reviewed journals?
Do you record that you threw out cultures before doing your statistics?
Please show us an example of how you account for this procedure in print.
Otherwise people will say you’re an example of scientists who lie to get the results they expect. Disclosing this is basic to publishing. Show us how.
MrPete // February 29, 2008 at 7:21 pm
(Just an FYI: for some of the responses to what I’ve written, I’m beginning to wonder if it’s worth responding to your tone, since you’re attempting to attribute attitudes and words to me that I try hard to never harbor nor communicate. You know who you are. Sure, I can take the heat, but it’s a waste of time for me to have to defuse your venting. You know who you are ;)
Boris asks, “why doesn’t Steve M write up a complete analysis of anything? An auditor actually completes the work and does not stop in the middle.”
One doesn’t have to be a rocket scientist to understand that Steve M has a bit of a challenge: normally, the audited party pays for the audit, willingly supplies whatever information is needed, bends over backwards to cooperate, appreciates the results, and listens quite carefully at the board level to all recommendations made.
Contrast that with the situation in this corner of climate science. Nobody’s being paid to do “contrarian” studies like Steve does. He receives no compensation for this work. The audited parties are rarely cooperative, providing very little of the needed data without a fight. The writeups (e.g. at AGU Fall ‘07) tend to get sidelined to a “poster” presentation rather than being received gratefully. And even his most well-endorsed audit recommendations are resisted and ignored, rather than welcomed and adopted.
You tell me: how many papers in 2007 still reference strip bark BCP data? It’s been thoroughly discredited, yet still used and referenced widely.
You tell me: what proportion of underlying data from Thompson, Jones, etc etc is fully archived today, as required, for papers published in 2007? 2006? how about 1998 (a decade ago)?
Demonstrate to me that there’s a viable accountability in this corner of the science world.
Remember, Steve’s not the bookkeeper. He should not need to publish scientific papers about these things. He’s not trying to propose alternate hypotheses. He simply wants those who are doing the work to do it well, to follow proper scientific procedure, to support their assertions with proper data, etc etc.
MrPete // February 29, 2008 at 7:22 pm
(BTW, I continue to say “this corner of climate science” because I in no way want to insinuate something about climate science in general. I’m even related to a climate scientist or two!)
MrPete // February 29, 2008 at 7:40 pm
Lazar, Dhogaza,
“Since you proved their METHOD was wrong, surely you know what it was?”
Yes: they presumed rather than doing the necessary research to discover facts.
Presumption is a common problem for us all, including scientists. We think we know how “it” works… far more than we really do.
So, if dendroclimatologists publish work based on the idea that tree ring records contain temperature history (without checking to see if that’s true), or if NASA folk publish work based on the idea that the surface station record is garnered from stations maintained to NWS standards (without checking to see if that’s true)… I’ll call that presumption any day of the week.
Got to be able to support your assertions. AFAIK, that’s good science.
Ian // February 29, 2008 at 7:53 pm
EW, for your 3 cultures analogy, I hope you’re thinking about sampling each culture on some measure over time, rather than just once. Then you can develop anomaly data – for climate, the difference in raw temp from station to station is not that interesting compared to the trends for the stations over time.
For your 3 cultures analogy, I think everyone here agrees that it would be *nice* to know the exact sources of any discrepancies. But suppose that instead of 3 cultures right in your lab, you have thousands of cultures in different locations, and you have some measure of interest for them extending over time. Perhaps there is no way for you to reconstruct the source of any discrepancies in the trend of the measure from the more distant past. But if you know that, say, there is a high probability of cultures that are physically close together to share their trends (“teleconnection” for climate), and there are physical reasons for the connection, then you have a decent basis for adjustment. Not ideal, perhaps, but far from worthless, and for the purposes of statistics often better than discarding the data altogether.
dhogaza // February 29, 2008 at 7:58 pm
Which, of course, they don’t, so what’s your point, exactly?
None // February 29, 2008 at 8:10 pm
Barton:
“No. You’re still wrong. The input parameters are not “adjusted in an effort to improve the match.” Do you understand the difference between a parameter and an algorithm?”
Of course I do. Yet all the models using slightly different algorithms, use different values for the parameters, and STILL all closely match the historical data.
Either some of the models are wrong, or they are using wrong parameters, yet they still match the historical record relatively closely. As Kiehls paper points out the models with high climate sensitivity use low aerosol forcing values and those with low climate sensitivity use high aerosol forcing.
Are you telling me they just happened to pick which aerosol forcing values to use which happened to compensate for each models sensitivity ?
Of course they didn’t, as Keihl said, checking the model against the historical record is an important and necessary thing to do to evaluate the model. When you do this either you are making random changes to the algorithm to fit the curve, or random changes to the parameters to fit the curve. Either way, you end up closely matching the historical curve with no extra chance whatsover your final model is going to be better at matching the future temperature trends.
Its exactly the same in models predicting the stock market. If you come up with an algorithm and keep adjusting the algorithm and/or the parameters until it matches historical data, all you are doing is data mining.
None // February 29, 2008 at 8:19 pm
Phil:
“Even if that is the case it’s still 100 yr old Physics (Chemistry) since it is based on the Clausius-Clapeyron equation.”
The models no doubt contain loads of old bits and pieces – I don’t know about the nitty gritty. However, Elspi’s original statement:
“The science on which global warming is grounded is basic 100-year-old physics. ”
is clearly meant to imply its a 100 year old widely acknowledged fact, that was known even back then. Either he knows he’s lying and is just trying to kill debate or he doesn’t really know what the models consist of. Hugely important questions, for example to what extent increasing cloud cover would affect feedback cycles, make a huge difference to what the models predict, yet these questions are all relatively recently being asked and answered. The science of global warming is only in the most abstract sense possible based on 100 year old basic physics.
Deech56 // February 29, 2008 at 8:33 pm
RE: MrPete // February 29, 2008 at 7:21 pm
“[Steve McIntyre's] writeups (e.g. at AGU Fall ‘07) tend to get sidelined to a ‘poster’ presentation rather than being received gratefully. ”
Ouch! OK, now I feel bad about all the abstracts I submitted to various large national meetings that ended up as posters. There’s one way to show them, though – finish collecting data and publish in peer-reviewed journals like I did.
Look, a poster presentation is nothing to sneeze at. I rarely miss the opportunity to visit key posters and talk to the scientists presenting their work. Much excellent work is “sidelined” to poster sessions.
JCH // February 29, 2008 at 8:53 pm
All I can say, as a former financial auditor, is there are some people here operating on some extraordinarily fanciful notions about the relationships between a corporation and their auditors – which is somewhat akin to don’t ask; don’t tell policy between a john and his prostitute: negotiate fee; render service; no kissing.
Barton Paul Levenson // February 29, 2008 at 9:19 pm
Mr. Pete writes:
[[Remember, Steve’s not the bookkeeper. He should not need to publish scientific papers about these things. He’s not trying to propose alternate hypotheses. He simply wants those who are doing the work to do it well, to follow proper scientific procedure, to support their assertions with proper data, etc etc.]]
Mr. Pete, if you think scientists are doing some procedure wrong, the thing to do is write up a study about it that can be submitted to a peer-reviewed journal. If you don’t do that, what you’re doing is not science and is not going to be helpful to scientists.
In the 1960s and 1970s, a wave of exosolar planet discoveries happened, mostly coming out of Sproul Observatory and the work of Peter Van de Kamp and Sarah Lee Lippincott. When George Gatewood studied their studies and thought they were subject to systematic errors, he wrote up papers about it and they were published in the Astronomical Journal. He didn’t start putting out a hand-printed anti-Van de Kamp journal on a ditto machine (that being the ’70s equivalent of the internet).
Barton Paul Levenson // February 29, 2008 at 9:24 pm
None posts:
[[ Elspi’s original statement:
“The science on which global warming is grounded is basic 100-year-old physics. ”
is clearly meant to imply its a 100 year old widely acknowledged fact, that was known even back then. Either he knows he’s lying and is just trying to kill debate or he doesn’t really know what the models consist of. Hugely important questions, for example to what extent increasing cloud cover would affect feedback cycles, make a huge difference to what the models predict, yet these questions are all relatively recently being asked and answered. The science of global warming is only in the most abstract sense possible based on 100 year old basic physics.]]
Read my lips: Svante Arrhenius’s paper estimating global warming under doubled carbon dioxide was published in 1896. It took water vapor feedback into account. This is the year 2008. Therefore the physics of anthropogenic global warming are 112 years old.
That’s if you don’t count John Tyndal’s experiments establishing that carbon dioxide was a greenhouse gas back in 1859.
Bishop Hill // February 29, 2008 at 9:33 pm
Nick Barnes
I see what you’re trying to get at, but there are a number of differences between what CA is doing and the audit of a private company. Firstly the science is all public sector work (NASA, CRU etc). Public sector bodies don’t get to choose their auditors as a rule. You are right that they try to work closely together, but McIntyre was told he couldn’t see the data, or that the data had been lost, or that he could go hang. Any one of these responses would have lead to an accounts qualification in either private or public sectors. Your point about confidentiality is also not quite right, since an auditor issues a public report on the body they have audited, and have to make public anything which might have lead to material misstatement of the accounts.
To make clear the analogy between CA and a financial audit, if a company’s valuation methodology is wrong (say they have used LIFO when the law says they must use FIFO) then the company can either correct it, or have the auditor say in his report that he believes it to be wrong. And note that the auditor is issuing an opinion on the accounts. So the company can say they disagree, but the auditor’s report stands.
cce // February 29, 2008 at 9:51 pm
The existence or non-existence of an as-yet unobserved negative cloud feedback or the exact magnitude of various aerosol cooling and warming influences does not erase the known existence of the warming influence of increased CO2 and the associated water vapor feedback. CO2 sensitivity is not as ridiculously low as skeptics like to pretend, but there remains a range of uncertainty with both CO2 and (especially) aerosols. Obviously, that is going to produce models with every combination of CO2 and aerosol forcing that matches observations, but yet remain plausible based on known physics. That’s why it’s called “uncertainty.” If all the models agreed, there would be little “uncertainty.”
If you want examples of curve-fitting and data mining, look no further than our friends who are torturing TSI reconstructions, cosmic ray counts, ocean oscillation, cloud cover, and who knows what else.
Hank Roberts // February 29, 2008 at 11:27 pm
> Nobody’s being paid to do
> “contrarian” studies
Utter nonsense. The big money in climate modeling is the proprietary models the oil companies use to model the climate that _created_ the petrolium they’re looking for.
Some of them have commented briefly about their work in various climate forums. But they can’t say very much.
If their models didn’t work, and work well, they wouldn’t be using them.
If they were turning up information that would challenge the publicly available models — considering the financial consequences — they’d certainly be publishing.
They aren’t publishing. They’re using politics and PR to try to convince people not to believe the science, in public, while relying on it for their own business.
steven mosher // February 29, 2008 at 11:40 pm
Here’s a piece of climate science code.
the urban adjustication thingy
After folks read this, everything will be clear.
DO 200 NURB=1,NSTAu
C**** Loop over all rural stations in memory; try 1/2 the radius first
CSCRIT=CSCRIH
RBYRC=RBYRCH
Rngbr=RngbrH
125 IS0=0
DO 130 N=1,NSTAr
ISR=IORD(N)
IYU1=MFSU(NURB)+IYOFF-1 ! subtract 1 for a possible partial yr
IYU2=IYU1+ILSU(NURB)-I1SU(NURB)+2 ! add 1 for partial year
C**** Find distance between center and station
CSDBYR=SNLATR(ISR)*SNLATU(NURB)+CSLATR(ISR)*CSLATU(NURB)*
* (CSLONR(ISR)*CSLONU(NURB)+SNLONR(ISR)*SNLONU(NURB))
X0=1950.
IF(CSDBYR.LE.CSCRIT) GO TO 130
DBYRC=0.
C**** The arc is replaced by the smaller chord (to avoid using ACOS)
IF(CSDBYR.LT.1.) DBYRC=RBYRC*SQRT(2.*(1.-CSDBYR))
IS0=IS0+1
WTI(IS0)=1.-DBYRC
ISOFI(IS0)=ISR
lenis(IS0)=len(n)
130 CONTINUE
C****
C**** Combine the station data
C****
DO 150 M=1,IYRM
WT(M)=0.
IWT(M)=0
URB(M)=XBAD
150 AVG(M)=XBAD
IF(IS0.EQ.0) THEN
IF(CSCRIT.EQ.CSCRIH) THEN
write(*,*) ‘Trying full radius’,RngbrF
CSCRIT=CSCRIF
RBYRC=RBYRCF
Rngbr=RngbrF
GO TO 125
END IF
WRITE(79,’(” NO RURAL NEIGHBORS FOR ”,I9)’) IDU(NURB)
GO TO 200
END IF
ioff=MFSU(NURB)-I1SU(NURB)
DO 151 M=I1SU(NURB),ILSU(NURB)
c if(M+IOFF.gt.IYRM0) stop 231
151 URB(M+IOFF)=RDATA(M)
write(*,’(a,i9,a,i4,a,2i5,f9.0)’) ‘urb stnID:’,idu(nurb),’ # rur:’
* ,is0,’ ranges:’,MFSU(NURB)+IYOFF,ILSU(NURB)+ioff+IYOFF,Rngbr
C**** Start with the station with the longest time record
IS=ISOFI(1)
IOFF=MFSR(IS)-I1SR(IS)
nuseid(is)=nuseid(is)+1
C**** First update of full data and weight arrays
DO 160 M=I1SR(IS),ILSR(IS)
c if(M+IOFF.gt.IYRM0) stop 244
AVG(M+IOFF)=RDATA(M)
IF(RDATA(M).LT.XBAD) WT(M+IOFF)=WTI(1)
IF(RDATA(M).LT.XBAD) IWT(M+IOFF)=1
160 CONTINUE
write(*,’(a,i5,a,i4,i6,i10)’) ‘longest rur range:’,
* MFSR(IS)+IYOFF,’-',ILSR(IS)+ioff+IYOFF,LENis(1),idr(is)
C**** Add in the remaining stations
DO 190 I=2,IS0
IS=ISOFI(I)
IOFF=MFSR(IS)-I1SR(IS)
write(*,’(a,i5,a,i5,a,i4,i6,i10)’)'add stn’,i,’ range:’,
* MFSR(IS)+IYOFF,’-',ILSR(IS)+ioff+IYOFF,LENis(i),idr(is)
C**** Extend the new data into a full series
DO 170 M=1,IYRM
170 DNEW(M)=XBAD
DO 180 M=I1SR(IS),ILSR(IS)
c if(M+IOFF.gt.IYRM0) stop 259
180 DNEW(M+IOFF)=RDATA(M)
NF1=MFSR(IS)
NL1=ILSR(IS)+IOFF
C**** Shift new data, then combine them with current mean
CALL CMBINE (AVG(1),WT,IWT, DNEW,NF1,NL1,WTI(I),
* IDR(IS),NSM,NCOM)
write(*,*) ‘data added: ‘,nsm,’ overlap:’,ncom,’ years’
IF(NSM.EQ.0) GO TO 190
nuseid(is)=nuseid(is)+1
190 CONTINUE
C**** Subtract urban station and call a curve fitting program
IY1=1
191 NXY=0
N3=0
N3F=0
N3L=0
NXX=0
TMEAN=0
if(IY1.eq.1)
* write(66,’(a,i9)’)'year dTs-urban dTs-rural StnID=’,idu(nurb)
DO 195 IY=IY1,IYRM
IF(AVG(IY).NE.XBAD.OR.URB(IY).NE.XBAD) NXX=NXX+1
IF(AVG(IY).EQ.XBAD.OR.URB(IY).EQ.XBAD) GO TO 194
if(IWT(IY).ge.NRURM) then
N3L=IY
N3=N3+1
if(N3F.eq.0) N3F=IY
end if
if(N3.le.0) go to 195
NXY=NXY+1
TS(NXY)=AVG(IY)-URB(IY)
F(NXY)=TS(NXY)
TMEAN=TMEAN+F(NXY)
YR(NXY)=IY+IYOFF
X(NXY)=YR(NXY)-X0
W(NXY)=1.
if(IWT(IY).ge.NRURM) THEN
NXY3=NXY
TM3=TMEAN
END IF
194 if(nxx.gt.0.and.IY1.eq.1) write(66,’(i4,2f10.2)’)
* IY+IYOFF,URB(IY),AVG(IY)
195 CONTINUE
IF(N3.LT.NCRIT) THEN
IF(RBYRC.NE.RBYRCF) THEN
write(*,*) ‘trying full radius’,RngbrF
RBYRC=RBYRCF
CSCRIT=CSCRIF
Rngbr=RngbrF
GO TO 125
END IF
WRITE(79,’(a3,i9.9,a13,i5,a15,i5,a50,a5)’) CC(NURB),IDU(NURB),
* ‘ good years:’,N3,’ total years:’,N3L-N3F+1,
* ‘ too little rural-neighbors-overlap – drop station’,’ 9999′
GO TO 200
ELSE IF(FLOAT(N3).LT.XCRIT*(N3L-N3F+1.)) THEN
IY1=N3L-(N3-1)/XCRIT
WRITE(79,’(a3,i9.9,a17,i5,a1,i4)’)
* CC(NURB),IDU(NURB),’ drop early years’,1+IYOFF,’-',IY1-1+IYOFF
GO TO 191
ELSE
TMEAN=TM3/NXY3
NXY=NXY3
CALL GETFIT(15)
C**** Find extended range
IYXTND=NINT(N3/XCRIT)-(N3L-N3F+1)
write(*,*) ‘possible range increase’,IYXTND,N3,N3L-N3F+1
n1x=N3F+IYOFF
n2x=N3L+IYOFF
IF(IYXTND.lt.0) stop ‘impossible’
if(IYXTND.gt.0) then
LXEND=IYU2-(N3L+IYOFF)
IF(IYXTND.le.LXEND) then
n2x=n2x+LXEND
else
n1x=n1x-(IYXTND-LXEND)
if(n1x.lt.IYU1) n1x=IYU1
n2x=IYU2
end if
end if
write(78,’(a3,i9.9,2f9.3,i5,5f9.3,I5,a1,I4,i5,a1,i4)’) CC(nurb),
* idu(nurb),(fpar(i),i=1,2),nint(fpar(3)+X0),(fpar(i),i=4,6),
* (rmsp(i),i=1,2),N3F+IYOFF,’-',N3L+IYOFF,N1X,’-',N2X
END IF
200 CONTINUE
nuse=0
do n=1,NSTAr
if(nuseid(n).gt.0) then
write(77,*) ‘used station ‘,idr(n),nuseid(n),’ times’
nuse=nuse+1
end if
end do
write(77,*) nuse,’ rural stations were used’
STOP
END
Hansen's Bulldog // February 29, 2008 at 11:51 pm
Someone (I don’t recall who) posted a comment which contained a link to an article from RC, and a quote from same, which talks about the use of climate models by the petroleum industry to assist in the identification of possible oil fields.
In moderation, I deleted that comment by *accident*. Please re-submit.
caerbannog // March 1, 2008 at 12:06 am
(Cleaned up a bit — Tamino, feel free to delete/edit this and/or previous post of mine as you see fit.)
Utter nonsense. The big money in climate modeling is the proprietary models the oil companies use to model the climate that _created_ the petrolium they’re looking for.
A little googling turned up this: http://tinyurl.com/2pa2r5
Resource Relation AAPG Bulletin (American Association of Petroleum Geologists) ; Vol/Issue: 74:5; Annual convention and exposition of the American Association of Petroleum Geologists; 3-6 Jun 1990; San Francisco, CA (USA)
Description/Abstract:
Numerous investigators have examined the potential use of numeric climate models and paleogeographic reconstructions to predict the deposition and preservation of organic-rich sediments, which may ultimately develop into hydrocarbon source rocks. These studies have concentrated on the Mesozoic and Cenozoic eras.
Although geologic conditions during these periods were different than that of today, they do have many similarities. In contrast, the boundary conditions associated with the Paleozoic are dramatically different. For example, no significant land plant cover is assumed in pre-Devonian simulations. In addition, for many of the simulations the bulk of the land mass was situated in the southern hemisphere at high latitudes. This compares with the Mesozoic and Cenozoic distributions that exhibit nearly coequal land-sea distributions in the two hemispheres.
An examination of the results of paleoclimate simulations for time slices in the Paleozoic reveal significant changes in spatial distribution of marine conditions that would favor high levels of organic productivity and organic preservation through time. The authors study of the stratigraphic record, though incomplete, has revealed a favorable correlation between organic-rich black shales, capable of acting as hydrocarbon source rocks, and those regions that had both high preservation efficiencies and elevated levels of organic productivity. These results suggest that numeric climate models can be effectively used to predict source rock distribution throughout the Phanerozoic.
Hank Roberts // March 1, 2008 at 12:16 am
Woof!
http://www.realclimate.org/index.php/archives/2006/04/gray-on-agw/#comment-13064
excerpt:
…. A lot of my own recent work is on Neoproterozoic climate, and without careful geological studies by Paul Hoffman and others (ironically, the same kind of field studies that petroleum geologists use to spot promising formations) my whole subject would be toast, for lack of data to help constrain the models. Going more recent, the record of the Pleistocene (relying on both ice core and marine sediments) is a big part of our understanding about how the climate system responds to CO2 and other forcings. The LGM is in fact a major test of the model sensitivity to CO2, as has been discussed in a number of places on RealClimate. In between you have hothouse climates like the Cretaceous, which cannot be explained without the greenhouse effect of elevated CO2 (and so far have not even been completely explained with it, though the gap may be more due to data than the theory). … As an indication of the extent to which the petroleum industry accepts the very same climate physics used in understanding global warming, at least one major oil company is using climate simulations of the Cretaceous (based on CO2 induced warming on top of geography changes) to spot promising oil formations. In fact, one of Paul Valdes’ industry-funded Cretaceous simulations was more or less embargoed from publication for five years because of its potential value to exploration. I hasten to add that Paul is one of the most respected climate modellers in the business, and his case shows that it is entirely possible to get some funding from the fossil fuel industry without compromising one’s research. The paleogeographic atlas project at U. of Chicago has also gotten funding from the oil industry from time to time. …”
—– end excerpt—-
and the inline response in the next post, excerpt follows:
http://www.realclimate.org/index.php/archives/2006/04/gray-on-agw/#comment-13070
… I don’t know much about petroleum geology, but my recollection is that the organic precursors to oil are in high-productivity marine environments, and those are less affected by the continental interior problem. For the gradient problem, I believe the idea was to set the CO2 level at something that achieved a reasonable match to the available Cretaceous data, even if it somewhat overestimated the gradient, in the hopes that the oceans would still be close enough to the real thing to provide some guidance as to high productivity regions. I have sent Paul some email to see if there is any work in the petroleum exploration literature which explains how this is done — for all I know, the details may be proprietary. I am also checking up on the publication status of the specific Cretaceous simulations I referred to, and will post the references once I hear from Paul. –raypierrre]
—-
See Ray’s recent topic on the work on this era for more and more recent info.
Phil. // March 1, 2008 at 12:30 am
MrPete
“You tell me: what proportion of underlying data from Thompson, Jones, etc etc is fully archived today, as required, for papers published in 2007? 2006? how about 1998 (a decade ago)?”
You tell me what this ‘requirement’ to archive is.
Hank Roberts // March 1, 2008 at 2:17 am
Caerbannog, thanks, good reference.
> organic-rich black shales, capable of
> acting as hydrocarbon source rocks
Interesting. The black shale is the sediment from the low-oxygen great extinctions.
So fossil carbon from the great extinctions in the past is what we’re pumping up now and putting into the atmosphere.
Coincidence?
fred // March 1, 2008 at 7:43 am
Clausius-Clapeyron equation and also the immortal
“Read my lips: Svante Arrhenius’s paper estimating global warming under doubled carbon dioxide was published in 1896. It took water vapor feedback into account. This is the year 2008. Therefore the physics of anthropogenic global warming are 112 years old.”
No, this is not true, and unfortunately, as Hans Erren has shown in a detailed and scholarly analysis, Aarhenius was mistaken in his calculations.
There are, as I posted in the other thread, three independent questions here.
The first is whether CO2 doubling will raise temps and by how much solely due to the fact that the extra CO2 will absorb heat. The answer to this question is a little under 1 degree C.
The second is whether the increased warmth in the atmosphere will increase the atmosphere’s capacity to carry water vapour. The answer is yes, and the amount is given by application of Clausius-Clapeyron. But this is about capacity. It is not about what happens.
The third is whether when we find out fully how the atmosphere works, we will discover that the increased warmth from increased CO2 will in fact lead to increased water vapour, or whether something will happen to increase discharges of water vapour so that it stays the same or declines. Clausius-Clapeyron does not say it will rise or fall. It just says that its carrying capacity has increased. Whether this carrying capacity results in more vapor being carried is not a matter of physics but of the ‘engineering design’ of the atmosphere.
For the avoidance of misunderstanding, the part in quotes was a metaphor.
I do not know why it matters to you all whether the science is 100 years old, 200 years old or 10 years old. Its correctness would seem more to the point. But the answer seems to be that like most science, bits of it are old and bits of it, fortunately, are recent. In a sense most modern scientific theories whether right or wrong have been several hundred years old because they’ve used calculus. Does it really matter?
None // March 1, 2008 at 10:41 am
Barton:
“Read my lips: Svante Arrhenius’s paper estimating global warming under doubled carbon dioxide was published in 1896. It took water vapor feedback into account. This is the year 2008. Therefore the physics of anthropogenic global warming are 112 years old.”
Like I already said, you can get away with claiming some of the science behind AGW is 112 years old, but then you have to accept that this science (using correct values as inputs, rather than the original erroneous ones) results in 1 deg per doubling of CO2 – and this science is just a basic model, not a testable law of nature.
Modern IPCC estimates of up to 6 deg for doubling is NOT based on that 112 year old physics, in the sense that the majority of the 6 deg effect is not a result of old physics.
So again, playing the “its old known science” card is dishonest. The size of the effect is what is important, not that there is one, and the current estimate of the size of the effect is more a result of modern models estimates rather than anything else.
MrPete // March 1, 2008 at 12:21 pm
Phil, the intent and wording are pretty clear at every publication I’ve examined. Here’s the first example I’ve come across in going back to other links to fulfill your request.
From PNAS,
The same reference provides further requirements for various kinds of data. For several scientific specialties, a specific archive site is specified. Obviously, the intent is that readers should be able to replicate the analysis of the data. That’s good science — no secrets. I don’t see why people would argue about this!
Here are the appropriate similar links for Nature and Science
Again, one might try to weasel around specific wording, but they all make clear the intent: at publication time, readers must be able to obtain the data underlying the analysis. That’s a very reasonable expectation. And it is generally followed in most scientific arenas. Questioning and/or doubting such things in this subfield leads the reader to wonder about real intent. (By the way, of these three only Science specifically lists “climate data” but all of them speak of the underlying data in general. If you want to give a pass just because a particular type of data is not mentioned, well, that doesn’t impress me a whole lot.
Unfortunately, too many of us have presumed that “good science” is defined simply by “appearance of a paper in a peer-reviewed journal.”
Unfortunately, poor science practice too easily gets published, and reasonable skeptics are too easily stymied.
Perhaps climate science mirrors the weather-vs-climate “thing.” I.e., if we can’t assume good science in any individual paper, nor even a decade or two of papers (the data and methods ultimately prove to be unsupportable), at least over the long haul hopefully we get
good science from the peer-review model, because the garbage eventually gets filtered out.
If so, then I guess in 2020 we’ll see which 1990’s papers have stood the test of time.
20/20 hindsight isn’t much fun for those of us who want to base policy decisions on current science, but hey, mankind has never been good at important predictions :).
Jim Clarke // March 1, 2008 at 2:50 pm
Tamino,
I read your take on this matter with interest until I got to this:
“The degree to which you’ve been brainwashed is indicated by your plaintive “If you can’t do it right, stop doing it…” Here’s an opportunity for you to benefit from an epiphany: the REAL GOAL of CA is to get us to stop doing it, because they really really hate what the truth is.”
At his most emotional, Steve McIntrye never comes close to such blather. While some on his blog may assume that Hansen is intentionally up to something, McIntrye constantly reminds them that there is nothing wrong with making adjustments as long as the reasons fit sound, statistical methods and our scientific understanding. He is a quality control analyst and much of his nit-picking has resulted in an improved understanding of several climate related issues.
The fact that you rally against McIntyre demonstrates that the ‘truth’ is not as important to you as ‘defending the faith’.
In this particular example, you argue that the patterns of one station do not fit the patterns of two nearby rural stations, so that station is adjusted. This seems fair enough if you don’t really care about accuracy.
We have learned that many stations listed as rural, by population only, have their sensors in areas that strongly mimick urban settings, surrounded by buildings and pavement that has increased over the years. Also, I can not think of any physical reason why a growing urban area would record a ‘false’ cooling trend of 3 degrees C over a 30 year period. Can you?
There may be legitimate explanations and the adjustments may be valid. McIntyre says as much. But they do not appear valid and the reasoning given does not explain why we should accept them as valid. A good scientist would tend to question what does not appear to fit reality, while a bad scientist would simply accept the pronouncements of scientific authority and defend such authority (with rants and unsubstantiated ad-hominem attacks) against those seeking a better understanding.
In this regard and on many other issues raised by CA, Mr. McIntyre appears to be doing the job of scientists, while scientists appear to be doing the job of advocates!
[Response: Suppose, for the sake of argument, that McIntyre is just "doing the work of scientists."
Then one really has to wonder, why would he have blogged about Peurto Maldon at all? 'Cause let's face it, the only thing it really shows is that GISS adjusted this station data exactly how they said they would. Was he trying to demonstrate that the homogeneity adjustment doesn't always reduce warming but sometimes increases it? Newsflash: we already knew that, GISS even says so in their documentation. Is he trying to demonstrate that the homogeneity adjustment increases warming often compared to reducing warming, so much so that it calls the entire procedure into question? If so, then as a "scientist" he should know full well that a handful of stations isn't proof of that, or persuasive evidence, or even any evidence at all. We'd need something like a comprehensive, methodical, totally objective survey even to get a *clue* -- but that seems to be the one activity McIntyre isn't willing to do. Furthermore, if he *did* show this then he'd have demonstrated that urbanization leads to anomalous cooling about as often as anomalous warming. Do you believe that's the case? Is McIntyre suggesting that when a homogeneity adjustment reduces warming it should be kept, but when it increases warming it should be discarded? If so, he'd be advocating "cherry-picking," a flagrant violation of the most fundamental tenets of proper data analysis.
Has he collected together his findings on this topic into a thoroughly documented analysis and submitted it for peer review? Now THAT would be "doing the work of scientists." Frankly, I would *love* to see a thorough analysis of the number, and distribution, of cooling vs warming introduced by the homogeneity adjustment, published in the peer-reviewed literature. I'm not holding my breath waiting for McIntyre to supply it.
The *reason* for urban adjustments is to compensate for the impact of urbanization on temperature data, so a procedure has been devised to do so. I merely showed that for Puerto Maldon the procedure was followed. But McIntyre, very early in his post, states
I'll "simply assume" that McIntyre is not so ignorant as to be unaware that the adjustment was in fact made according to stated procedures. If that's true, then his question carries an implication of gross incompetence or deliberate malfeasance, which McIntyre, apparently, hasn't got the GUTS to say outright -- then he might be held accountable. It seems to me that he'd much rather incite his readers to do the nasty work; again, it enables him to deny responsibility. Is that "doing the work of scientists"?
This much cannot sanely be denied: I *do* speak for myself, and I don't use innuendo I say it outright. So here's my opinion: you have one helluva bizarre notion of what the proper work of scientists is.]
Hank Roberts // March 1, 2008 at 2:59 pm
http://scholar.google.com/scholar?q=%22hans+erren%22++%2Bclimate
Citation please?
Hank Roberts // March 1, 2008 at 3:10 pm
Fred, you’re basically channeling blogging, not climate science. The idea was popular with fans of Lubos, and is well refuted here among many other places:
http://rabett.blogspot.com/2007_06_01_archive.html
john // March 1, 2008 at 3:57 pm
I just don’t trust peer reviewed science.
The Y2K computer bug myth was supported by peer reviewed science, being a bit of a machine code junkie I knew it was overblown rubbish from the start, but the whole of scientific academia either went along with it or stayed silent.
You have no credibility in my eyes.
Hank Roberts // March 1, 2008 at 4:14 pm
John, that’s “weather not climate” and you’re grossly missstating the climate of opinion around y2k. See comp.risks for the best compilation as it happened.
dhogaza // March 1, 2008 at 4:41 pm
This is simply a false statement.
1. It wasn’t a myth. Businesses spent huge sums of money rewriting old code in preparation for Y2K, and as a result most made it through the new year without problems. But if they’d listened to you and your “it’s a myth” line, many businesses would’ve been facing chaos in their business systems.
2. the Y2K “myth” wasn’t based on “peer-reviewed science”. For starters, IT isn’t science. Secondly, it was being trumpeted largely through the popular media, through books written by folks like Ed Yourdon. There’s nothing “peer-reviewed” about the popular media.
3. Why would “scientific academia” weigh in on arguments over forty year old COBOL programs? What would a biologist, for instance, know about such things that would give her credibility?
Thomas Huxley // March 1, 2008 at 5:13 pm
“Not far from the invention of fire… we must rank the invention of doubt.” [Collected Essays vol 6, viii]
Lazar // March 1, 2008 at 5:15 pm
Jim Clarke and others…
‘We’ (at least, Hansen et al. and climate scientists who use the data) are concerned about how UHI effects the averaged (regional and globally averaged) temperature trends.
The effect of the GISTEMP adjustments on these trends is the ’same’ (as near as makes no difference) as using rural stations alone, making the average temperature trends better representative, and the trends of individual station records better representative on average. They may make the trend of individual station records less representative in a smaller number of cases. Be that the case, there is no effect on average trends compared to the case where, which presumably you have no problem with, just rural stations are used. It is the average trends which GISTEMP and climate scientists are interested in. If one is interested in just a few stations, they have the optiion of using uncorrected data. You are correct that we don’t know in any individual case why an urban station shows a cooler trend relative to rural stations. Neither do we know why any individual station shows a warmer trend, which may be due to UHI, but might just as well be a genuine climatic effect, instrumental error, or microsite bias. Steve McIntyre has shown adjustments to one station, and not quantified any error. He needs to quantify errors for many stations and their impact on regional and global averages. Until then, it’s just handwaving.
dean_1230 // March 1, 2008 at 5:32 pm
john said…
‘I just don’t trust peer reviewed science.”
Sigh… independent of what side anyone here is on, this comment is … what word am I looking for…
abhorrent…
Naive…
Silly…
Disgusting…
Nah. None of those do justice to that comment. So much of what we have in our lives is due to the application of peer-reviewed science that the statement is just unbelieveable. How someone can dismiss the peer-review process en-masse is beyond me.
If john ever had credibility in this discussion before, it’s gone now.
chriscolose // March 1, 2008 at 6:09 pm
None,
the IPCC AR4 never cites 6 C at 2x CO2. “By 2100″ is a possibility, but that is different than “2x CO2″ because of emission scenarios. The 2x CO2 range is from 2 to 4.5 C.
Jim Clarke // March 1, 2008 at 7:00 pm
Tamino,
Thanks for your response. If I get you correctly [Response: No, you don't, not even close.], you are arguing that Hansen followed the procedure he outlined, so we can not question the results of that procedure, or the procedure itself. I don’t think that is a very robust argument. Like McIntyre, I can not imagine a scenario where an urban location would experience a significant cooling, while two nearby rural locations had a warming trend. A step-cooling would be possible with changes in instrumentation or location in the urban area, but the data shows a fairly consistent trend. Perhaps there is a physical explanation for the readings at this location and the many other similar locations now being recognized at CA, but no one seems to be offering one. In fact, you seem to say that it doesn’t matter what is physically happening or whether the methodology used is appropriate, only that there is no intentional fraud. You are defending Hansen, not the science or methodology in question. McIntyre is questioning the methodology. Your counterattack misses the mark!
Your definition of a scientist seems to be confined to ‘one who gets published in peer reviewed journals’, which I personally find bizarre. Currently, the journal method is the most common way for scientists to exchange ideas while trying to maintain quality control, but there is nothing about science itself that requires this particular method. It has many shortcomings and not doubt will evolve to a better method in the years ahead.
You scolded McIntyre for not jumping to conclusions and accusing Hansen of fraud. You claim he is gutless, but part of being a good scientist is avoiding the temptation to jump to conclusions. So if he behaves like a good scientist (or data analyzer), he is gutless. Following your argument, you need to add ‘being gutless’ as part of your definition of ’scientist’, along with getting published in a peer reviewed journal!
[Response: Do you really not get it? I scold McIntyre for MAKING accusations of fraud. I further scold him for doing so by innuendo and implication, while letting his chorus use the dirty words; it's the "do you still beat your wife?" method of inquiry. And judging from your protestations, the strategy is working -- on you.
It's just like McIntyre's Is Gavin Schmidt Honest? post. Do you deny THAT is an accusation of fraud? I'm not asking a rhetorical question, I want to hear your answer -- and it's not an essay question, either "yes" or "no." If you've got the guts.]
The fundamental point of the whole debate is that the method used to ‘homogenize’ the data, may not be a good representation of what is actually happening in the real world. There are areas where the temperature data seems to run counter to our understanding of rural and urban locations. This may be the result of poor data or a flaw in the methodology, or both. Either way, it is worth exploring, for it will ultimately improve our understanding of temperatures and climate. Even if you do not think it is ‘important’, I do not see the value in ridiculing such an exploration.
[Response: I think it's a very important issue, and as I said in the response to your previous comment, I would love to see some *real* analysis of the overall impact of the homogeneity adjustment on temperature records. What fraction of urban stations end up being made warmer rather than cooler? Is there a geographic pattern to that? Are most of the adjustment-induced coolings larger (in absolute value) than the adjustment-induced warmings, or vice versa, or neither? How do the results differ between the U.S. (using nightlights data) and the rest of the world (using GHCN population data)? There might be a wealth of insight to be gained from such an analysis, but from what I see, McIntyre not only hasn't addressed these questions, he has no intention to.
In which case, it's tragicomic to refer to his efforts as "doing the job of scientists."]
I found your original post helpful in explaining some of what was done and why, but so much of what followed seemed like nonsense and vitriol.
Thanks for the discussion!
MrPete // March 1, 2008 at 7:32 pm
I’ll agree strongly with dhogaza on y2k, add a couple of possibly-interesting points, and then I think I’m done with this thread. Gotta get back to the Real World :-)
The y2k panic was based on social hyperbole — “myth” if you will. The real y2k challenge was quite real — in certain arenas. It’s just that, as dhogaza notes, a lot of work was done to address the issue.
Doomsayers presumed the problems couldn’t be fixed. The reality: almost everything was nicely completed well in advance.
Perhaps the biggest remaining issue in the end was things like certain telephone systems in some remote developing world nations.
One lesson I learned was how presumption spirals into panic. And that, perhaps, is a precautionary tale for the global warming issue.
Power expert A would expect zero trouble and prudently prepare for the possibility of a brief power glitch (a few seconds), just in case. Observer B would say “A’s expecting a glitch! We better prepare for a one-hour outage.” C would say “B’s expecting a one-hour outage; we’ll prepare for a week.”…until some were expecting years of trouble and the End Of Civilization As We Know It.
Big Sigh.
Petro // March 1, 2008 at 7:45 pm
john stated:
“I just don’t trust peer reviewed science.”
This statement equals with
“I don’t trust reality.”
How come you are called denialists?
Harold Pierce Jr // March 1, 2008 at 9:33 pm
RE: Peer-reviewed science papers.
It means diddly squat. In any given area of research, everybody knows everybody, and it is very much of “I’ll scratch your back if you scatch mine.” There are some many journals out there, that you can always get your work published somewhere. In academia all that counts is that the results of your research no matter how mundane or trivial, which is about 99% of this type of research, is published in a journal of any kind. Grant officers just look to see if grant research get published.
Joke: How do you get your research published?
Ans: Start your own journal.
There are many commercial journal publishers who are quite willing to assist you, because they make a bundle selling the journal to research libraries.
I have reviewed papers and have totally trashed some of them , and I always signed my name to review form. And I have never gotten a nasty phone calls or emails from the senior author. Usually I get a letter from the ed thanking me for my efforts.
dhogaza // March 1, 2008 at 10:08 pm
If true then McIntyre shouldn’t have any problem getting published as often as he makes an earth-shaking discovery that climate science is a fraud, right?
So why isn’t he doing so? And why do he and others need a non-peer reviewed journal like “Energy and Environment” handy to publish their earth-shaking discoveries?
In actually, the bar isn’t nearly as low as you portray it, but it is relatively low, sure. Low enough that a brilliant man like McIntyre should have no problem clearing it, if he’s serious.
Bernie // March 1, 2008 at 10:49 pm
Tamino, you asked:
It’s just like McIntyre’s Is Gavin Schmidt Honest? post. Do you deny THAT is an accusation of fraud? I’m not asking a rhetorical question, I want to hear your answer — and it’s not an essay question, either “yes” or “no.” If you’ve got the guts.
Many thanks for the link to the “Is Gavin Schmidt Honest?” thread.
I had not been reading these climate sites in 2005 and so had not previously seen this string. I just read. My answer is definitively that Steve McIntyre did not accuse Gavin Schmidt of “fraud”. What he did raise was the question of how serious Gavin Schmidt was in being open to informed but contrary opinions. Could someone take exception to the “business fraud” allusion, I suppose IF you were looking to take exception.
I respectively suggest you re-read the interchange and then reassess your view of Steve McIntyre if this is a primary piece of evidence.
chriscolose // March 1, 2008 at 11:37 pm
Bernie, look at the title of the thread!! Don’t just throw in the later caveat “at wanting to seriously discuss, etc.” Apparently no one is capable of reading through the lines– read the posts on Hansen too.
To be honest, the only thing I got out of that thread was McIntrye stomping his feet because his comments did not get moderated right away.
dhogaza // March 2, 2008 at 12:03 am
If McIntyre’s being misunderstood, then why didn’t he set the record straight when asked here, directly, in a thread in which he was participating, whether or not he believed some leading climate scientists are guilty of fraud?
He refused, you know.
Nor as he ceased posting his constant sniping innuendo regarding the honesty of Mann, Schmidt, Hansen, Thompson and others.
Hank Roberts // March 2, 2008 at 12:10 am
Yet if there were some challenge to be made to the published science, it would be published in a science journal.
Scientists get attacked. It’s done by PR, blogs, letters to the editor, attacks on academic standing, and riling up people willing to make threats of personal violence or take actions that are real violence.
Opposition in science journals doesn’t start down the slippery slope. That’s why it’s respected.
Hank Roberts // March 2, 2008 at 12:13 am
And, yes, there are people who take publications in science journals as violent attacks on their positions.
Evolution. Stem cells. Climate change. Each has set off people who feel that scientific publication is an attack on their diety, their diety’s possessions, or their place in the economy.
Those people need a warning label:
Contents under pressure.
Do not uncork.
Chris O'Neill // March 2, 2008 at 1:13 am
None:
This is blatantly wrong. As well as models, there are calculations of climate sensitivity from empirical observations listed here. (BTW, Annan’s 2006 paper contains one result that is derived purely from observations.)
Heretic // March 2, 2008 at 2:08 am
Harold Pierce Jr. on peer-review: bullshit.
Same type of argument that “skeptics” make all the time: it’s not perfect, so it’s useless. Peer-review is a minimum standard. Not everything peer-reviewed deserves attention and there is no absolute guarantee it’s been done honestly and ethically. But it’s more likely to be so. And it will receive more and more scrutiny. And eventually will prove to be useful and honest, or not. Your statement seems to imply that Einstein would very well have dispensed of publishing his photo-electric effect paper or the special relativity paper, because peer-review doesn’t matter, and he knew he was right anyway. Well, he did not do that. Why?
He did just like anyone who is confident enough in what they have to say. Even the skeptics.
As for your remark “start your own journal,”, I find it rather funny, when E&E isn’t even peer-reviewed and it was in fact started for the purpose of “giving a platform” to skeptics (Sonja Christens’s words). In the case of Soon-Baliunas, they should probably have started their own, because even going through the Legates backroads did not eventually prevent them from being exposed.
Now, enlighten us. What papers studying AGW have been published in started-for-the occasion journals? Does that weigh anything compared to the papers published through Science, Nature, AGU, etc, etc? Comparative quantitative analysis for that?
Paul // March 2, 2008 at 2:22 am
It is interesting that the official census data from 1981 show a population for Puerto Maldonado of 12,693. The census data in 2005 show a population of 51,349. Yet most of the adjustment made in the GISS data occur prior to 1981 and virtually no adjustment is made after 1981. It’s also interesting that the raw data show a cooling trend inspite of rapid urbanization. Read all about at at http://temperaturewatch.blogspot.com
kim // March 2, 2008 at 2:30 am
Peer review is better than nothing, even though unpaid, often unaccountable, and often superficial. What did Wegman say about peer review in this particular discipline, which is the pertinent one?
===============================
MrPete // March 2, 2008 at 3:41 am
dhogaza, Steve’s reasonably consistent in his complaints about scientific process and data handling.
I find it illuminating that as soon as someone substantiates an issue here (such as the lack of archiving), which Steve has been pushing on for quite a long time, the subject gets changed.
A few questions for the published scientists among us:
* Do you agree that proactively enabling reader access to data, as required by the journals, is a good thing?
* Do you agree that scientists should not be allowed to ignore such requirements?
* Since we all agree that “sniping innuendo” is to be avoided, here’s a serious question: When it is proven that scientists are ignoring the requirements, and the journal editors are letting them get away with it, what would you do? What kind of peer-reviewed article would you write to expose this in a “validly scientific” way? What journal would you submit it to?
I believe these are serious questions and serious issues that need to be addressed. Do you really believe that exposing such practices is “sniping innuendo?”
I sure don’t. Nor do quite a few scientists I’ve spoken with. Once they realize what is going on, they’re horrified–on behalf of all of science, because the few who “get away with it” tend to tarnish the entire “science industry.” Particularly when there’s so much publicity surrounding this particular area of inquiry.
Deech56 // March 2, 2008 at 3:42 am
So now we’re trashing peer review based on Y2K and Wegman’s thought that the researchers are a tight little group? Amazing. Just wondering – did Wegman look at other scientific disciplines, or is it just climate science that has this phenomenon? At least peer review and publication allow ideas to see the light of day. With all due respect to our host, blog publication (while it has its place) is not quite on the same level.
Bernie // March 2, 2008 at 5:12 am
dhogaza and chriscolose
C’mon guys, there is a lot of emotion around these issues and nothing in that thread approaches some of the vitriol and ad hominems on Real Climate or for that matter on some Climate Audit threads especially when JEG almost deliberately stirs the pot with what passes I hope for Gallic wit. By contrast, you cannot possibly read the “Is Gavin Schnidt honest” thread and maintain that Steve McIntyre did not raise a potentially significant issue, accurately noted events, acknowledged that his initial comments were out of line and then left open the issue of whether all legitimate (or for that matter illegitimate) comments at RealClimate were/are treated equally. He did not waiver from the main issue and certainly did not go much further than mild sarcasm and irony. Was the original title of the thread overwrought — perhaps. It probably reflects the frustration of a researcher who raises a substantive issue and is repeatedly ignored by those who are part of the establishment.
My personal experience is that some moderators at Real Climate appear reluctant to let those with contrary perspectives post, ergo, I can appreciate Steve McIntyre’s frustration.
I suggest an experiment. Adopt substantive but contrary positions to those central to Real Climate and Climate Audit – pursue the point with 2 or 3 follow up posts and find out which one you get censored for first. My money is on Real Climate. Alternatively, post denigrating and insulting comments about those with contrary viewpoints on both sites and see which site puts the lid on you first. For this one, my money is on Steve McIntyre, who regularly snips out of bounds comments by all comers.
[Response: I read RC all the time, I've seen plenty of CA, and of course I'm familiar with this blog. The only one of the three that does anything like a decent job blocking vitriol and ad hominem, is RC. They're far from perfect -- but neither CA nor this blog comes close to the degree of civility they enforce.]
Hank Roberts // March 2, 2008 at 5:26 am
> horrified–on behalf of all of science
I see a t-shirt coming.
Ron Cram // March 2, 2008 at 5:28 am
Deech56,
If you have complete faith in peer-reviewed research, then you are not paying attention. The World Conference on Research Integrity to Foster Responsible Research met in 2007 and the two disciplines most criticized were medical research and climate science. Note the quote by Howard Alper about climatologists in the second link below.
http://www.esf.org/ext-ceo-news-singleview/article/world-conference-on-research-integrity-to-foster-responsible-research-318.html
http://www.esf.org/ext-ceo-news-singleview/article/experts-deconstruct-research-misconduct-from-global-and-institutional-perspectives-320.html
If you do not know the standards of science regarding data archiving and sharing, you can read these Wikipedia articles and then review the original sources the article references. Climatologists actually seem proud of their refusal to hold to these well established standards of science. Skeptics will never be persuaded until the standards are upheld in practice.
http://en.wikipedia.org/wiki/Scientific_data_archiving
http://en.wikipedia.org/wiki/Data_sharing
fred // March 2, 2008 at 8:35 am
Hank R, I have read Eli’s remarks, read them in fact the first time they came out, and they don’t seem relevant to the point, which is mostly a logical one.
If all other things are held constant, and the CO2 level of the planetary atmosphere rises from 300 to 600ppm and nothing else changes, this will result in the atmosphere warming because more heat will be absorbed by the atmosphere.
You seem to be arguing that the warming that will result from this alone will exceed 1 degree. Is that what you are saying? If so, could you provide a reference? I am under the impression that the number of about 1C from this cause alone is ‘just physics’ and depends on the amount of radiation there is in certain spectra, and the characteristics of CO2 in terms of which spectra it is sensitive to. But maybe this is wrong?
Martin // March 2, 2008 at 8:35 am
Steve Mosher:
Actually the essential code is here (if the wordpress formatter doesn’t mess up things):
SUBROUTINE TREND2(xc,A,LEN,Xmid,BAD,MIN1,MIN2, SL1,SL2,Ymid,RMS,
* SL,Y0,RMS0)
C**** finds a fit using regression analysis by a line
C**** with a break in slope at Xmid. Returned are the 2 slopes
C**** SL1,SL2 provided we have at least MIN1,MIN2 data.
C**** Linear regression data are also computed (for emergencies)
REAL*8 xc(*),A(*)
REAL*8 sx(2),sxx(2),sxa(2),sa,saa,denom,xnum1,xnum2
INTEGER kount(2)
sl1=bad
sl2=bad
Ymid=bad
sa=0.
saa=0.
do k=1,2
kount(k)=0
sx(k)=0.
sxx(k)=0.
sxa(k)=0.
end do
do 100 n=1,len
if(a(n).eq.BAD) go to 100
do 100 n=1,len
if(a(n).eq.BAD) go to 100
x=xc(n)-Xmid
sa=sa+a(n)
saa=saa+a(n)**2
k=1
if(x.gt.0.) k=2
kount(k)=kount(k)+1
sx(k)=sx(k)+x
sxx(k)=sxx(k)+x**2
sxa(k)=sxa(k)+x*a(n)
100 continue
ntot=kount(1)+kount(2)
denom=ntot*sxx(1)*sxx(2)-sxx(1)*sx(2)**2-sxx(2)*sx(1)**2
xnum1=sx(1)*(sx(2)*sxa(2)-sxx(2)*sa)+sxa(1)*(ntot*sxx(2)-sx(2)**2)
xnum2=sx(2)*(sx(1)*sxa(1)-sxx(1)*sa)+sxa(2)*(ntot*sxx(1)-sx(1)**2)
if(kount(1).lt.MIN1.or.kount(2).lt.MIN2) return
sl1=xnum1/denom
sl2=xnum2/denom
Ymid=(sa-sl1*sx(1)-sl2*sx(2))/ntot
RMS=ntot*Ymid**2+saa-2*Ymid*(sa-sl1*sx(1)-sl2*sx(2))+
* sl1*sl1*sxx(1)+sl2*sl2*sxx(2)-2*sl1*sxa(1)-2*sl2*sxa(2)
RMS=ntot*Ymid**2+saa-2*Ymid*(sa-sl1*sx(1)-sl2*sx(2))+
* sl1*sl1*sxx(1)+sl2*sl2*sxx(2)-2*sl1*sxa(1)-2*sl2*sxa(2)
C**** linear regression
sx(1)=sx(1)+sx(2)
sxx(1)=sxx(1)+sxx(2)
sxa(1)=sxa(1)+sxa(2)
sl=(ntot*sxa(1)-sa*sx(1))/(ntot*sxx(1)-sx(1)**2)
Y0=(sa-sl*sx(1))/ntot
RMS0=ntot*Y0**2+saa+sl*sl*sxx(1)-2*Y0*(sa-sl*sx(1))-2*sl*sxa(1)
return
end
---- [Ymid becomes knee below]
subroutine adj(info,idata,sl1,sl2,knee,sl0,iy1,iy2,iy1a,iy2a,
* iflag,m1,m2)
dimension info(*),idata(*)
if(iflag.ne.0.and.iflag.ne.100) then
C**** Use linear approximation
sl1=sl0
sl2=sl0
end if
miss=info(7)
m1o=m1
m2o=m2
m1=-100
m0=12*(iy1-info(6)) ! Dec of year iy1
do iy=iy1,iy2
sl=sl1
if(iy.gt.knee) sl=sl2
iya=iy
if(iy.lt.iy1a) iya=iy1a
if(iy.gt.iy2a) iya=iy2a
iadj=nint( (iya-knee)*sl-(iy2a-knee)*sl2 )
do m=m0,m0+11
if(m.ge.m1o.and.m.le.m2o.and.idata(m).ne.miss) then
if(m1.lt.0) m1=m
idata(m)=idata(m)+iadj
m2=m
end if
end do
m0=m0+12
end do
return
Fortran makes my toes curl upward, but yes — read this and everything will be clear :-)
(This code — in STEP2 — was apparently updated Oct 10, 2007. Some folks do maintain their code. Now and then, at least.)
Martin // March 2, 2008 at 8:38 am
WordPress did mess up above. Imagine textbook-style structural indenting.
fred // March 2, 2008 at 1:19 pm
If McIntyre is reporting the facts correctly, 1419 or 74% of US surface stations have had their measurements changed by GISS.
[Response: Aren't there only 1221 US surface stations? How would 1419 be 74% of that?]
Perhaps I am ignorant of normal scientific procedure. Is there any other field in which one has put out a series of instruments to record data, and then changed the recordings of 75% of them to arrive at the real data? It seems a very odd way of going about things.
Rather as if in paleo one were to decide in 75% of cases that contrary to the dig records, some of the bones were not in fact found at the places and the levels the records stated.
dhogaza // March 2, 2008 at 1:48 pm
I’ve read McIntyre’s allegations regarding Thomspon and IMO, they’re just his typical bullshit. Right up there with “Gavin Schmidt’s violating federal regulations by not having his blog posts vetted, and he does it during work hours”.
MrPete // March 2, 2008 at 2:28 pm
dhogaza, if you really think Steve’s allegations about Thompson are BS then we’ll all be very appreciative when you demonstrate how to obtain access to his data :-)
steven mosher // March 2, 2008 at 2:33 pm
Tamino, RE Fred’s comment on 74%. It’s 74% of the all the stations in the dataset. The USA (1221) + ROW.
Bernie // March 2, 2008 at 2:38 pm
dhogaza:
This is really simple. OK, if you are right about Thompson where is his archived data? If you cannot find it, a polite retraction would be in order.
Bernie // March 2, 2008 at 2:49 pm
Tamino:
I have no idea as to what is blocked and what is not, nor do I think do you unless you sometimes act as a moderator at Real Climate. I do know that empirical comments I have made have been blocked and criticisms of the blocking of those comments have been likewise blocked at Real Climate. The question is about the proportion of blocking given a particular line of argument relative to the general position of the site or the target of the ad hominems – Skeptic or non-Skeptic. I can’t see how the argument can be proved without such an experiment.
steven mosher // March 2, 2008 at 3:03 pm
corection tamino, fred read the chart right. i was wrong. i’ve posted question to steve mc on it
null{} // March 2, 2008 at 3:39 pm
The following is a true fact easily verified.
The GISS/NASA GISTemp process and procedure does in fact eliminate/omit some temperature data from the analysis.
So much for you don’t throw out data.
I can post the files here if you can’t find them.
Hank Roberts // March 2, 2008 at 3:56 pm
Fred, you’re talking about — instantaneously — twinning each carbon dioxide molecule in the atmosphere, doubling the number from the pre-fossil-fuel level to twice that, with no other change whatsoever then or consequentially.
It can’t happen. You understand this?
fred // March 2, 2008 at 4:32 pm
Hank
Yes, I understand that in the real world a rising level of CO2 and consequent heating will have all kinds of effects.
I am just talking about distinguishing what is due to what. If I understand it correctly, and may not, the pure effect of heating which is attributable only to the increased amount of CO2 and its absorption of heat is about 1 degree. Other effects are then hypothesized to add more warming.
Hank Roberts // March 2, 2008 at 4:47 pm
Null, you’re conflating recordkeeping — keeping all the original raw bits — with analysis. You know better than that.
fred // March 2, 2008 at 4:49 pm
Tamino, I’m no expert on this. I think the various sources of data and networks are discussed here. It seems that the post I referred to is talking about more stations than the USHCN. I do not know whether this is right or wrong, or whether a higher or lower proportion of adjustments happen on USHCN
http://www.climateaudit.org/?p=2034
Hank Roberts // March 2, 2008 at 6:36 pm
> Other effects are then hypothesized
> to add more warming.
And one looks for ways to test such hypotheses, both in the deep past and the present. So far, climate sensitivity looks like about 3 degrees C.
What’s to argue with about this?
Martin // March 2, 2008 at 7:01 pm
Fred, I find 1.1 deg using Dave Archer’s 1-D online model (tropics, no clouds), which seems plausible. This would be the value to use on a waterless planet.
BTW on can even consider the possibility of absolute humidity decreasing as temperature goes up. Weird, improbable to the extreme, but not physically illegal :-)
null{} // March 2, 2008 at 8:45 pm
list.of.stations.completely.removed.txt
115624640010 HURGHADA lat,lon 27.3 33.8 omit: 0-9999
134652010000 LAGOS/IKEJA lat,lon 6.6 3.3 omit: 0-9999
134652360000 WARRI lat,lon 5.5 5.7 omit: 0-9999
134652430000 LOKOJA lat,lon 7.8 6.7 omit: 0-9999
205549450010 JUXIAN lat,lon 35.6 118.8 omit: 0-9999
207433330002 PORT BLAIR lat,lon 11.7 92.7 omit: 0-9999
210476960010 YOKOSUKA lat,lon 35.3 139.7 omit: 0-9999
219415600005 PARACHINAR lat,lon 33.9 70.1 omit: 0-9999
303824000000 FERNANDO DE N lat,lon -3.9 -32.4 omit: 0-9999
314804440000 CIUDAD BOLIVA lat,lon 8.2 -63.5 omit: 0-9999
403717300040 RUEL,ON lat,lon 47.3 -81.4 omit: 0-9999
414762200010 CIUDAD GUERRERO,CHIHUAHUA lat,lon 28.6 -107.5 omit: 0-9999
414762580020 QUIRIEGO, SONORA lat,lon 27.5 -109.2 omit: 0-9999
414763730000 TEPEHUANES,DG lat,lon 25.4 -105.7 omit: 0-9999
414766950010 CHAMPOTON, CAMPECHE lat,lon 19.4 -90.7 omit: 0-9999
414767750030 CANTON, OAXACA lat,lon 18.0 -96.3 omit: 0-9999
440785260010 ANNAS HOPE, ST.CROIX VIRG lat,lon 17.7 -66.7 omit: 0-9999
425724910030 HOLLISTER USA lat,lon 36.8 -121.4 omit: 0-9999
425725130010 FREELAND lat,lon 41.0 -75.9 omit: 0-9999
425725210010 PHILO 3SW lat,lon 39.8 -81.9 omit: 0-9999
425725970120 CRATER LAKE NPS HQ lat,lon 42.9 -122.1 omit: 0-9999
425726710020 ROCK SPRINGS FAA AP lat,lon 41.6 -109.1 omit: 0-9999
501947880000 KEMPSEY lat,lon -31.0 152.8 omit: 0-9999
Hank Roberts // March 2, 2008 at 9:06 pm
Exactly. And if you want to do an analysis using those, there they are, you can look them up. You understand this, don’t you?
john // March 2, 2008 at 9:31 pm
I feel dismayed that some still think the Y2K computer bug existed, what they should ask themselves is which piece of code was altered.
If anyone had found one, we would have heard all about it! I never found one and neither did anyone else! I found some really rubbish code but that is quite normal.
As to the role of science in this sorry tale, well I have kept the advice of the UK governments Chief Scientific Advisor, a professional scientist of the highest standing, he told us it was one of the Greatest Challenges we faced and a threat to our civilisation! this sounds depressingly familiar.
The Institution of Electrical Engineers produced a 1000 page manual on the subject, full of learned opinion, rules, standards and comment, but not one line of machine code, so it was entirely useless.
Harold Pierce Jr // March 2, 2008 at 11:04 pm
ATTN: dhogaza
The internet is a quite new and easily allows an individual to set up a personal “e-journal”. Scientific publishing is a very cumbersome , tedious, lenghty and expensive process.
First you write a ms and check it to make sure it is in the format that the journal requires, and no two jounals have the same format especially for citation of references. Formatting a reference list is such a tedious and time-consuming process that software was developed that will take references from your data base and format them to the style required by the journal. But this is also pain in the rear: you have to put your reference data base in a format that this software requires. Until recently journals required such high-quality drawings and art work that it could only be prepared a highly-skilled graphic artist and this was quite expensive.
After you submit your manuscript, it is sent by the editor to reviewers, usually to those that you have listed who have the expertise in the subject and who are your friends. Most journals request that you provide a list of potential reviewers, and you most certaintly do not list your enemies or do you send your ms to a journal whose editor is your mortal enemy.
After the editor receives the ms. it sent out to review, a process that can take several or more months. Some reviewers will return the ms with their review promptly while others take their sweet time and few don’t even bother to acknowledge receipt of ms or to review it.
If you have prepared a rock solid ms and it has passed the first round, the editor will send the ms to the publisher. A few months latter you will recieve the galley proofs which you have to edit to make sure there are no errors. You send back the galley proofs and then you wait again. It can take many months before the printed journal is issued.
If you ms does not pass the first round and rquires revision, the whole process starts over again.
Many journal now permit submission of a ms in various formats for use with computers. I would never do this because if a reviewer has your ms in any digital format, it can be sent to anyone and everyone with a click of the mouse. This happens occasionly even though reviewers are supposed to treat the ms as privileged. And sometimes you just might find much to your surprise that you have been “scopped” a few months later by your enemies.
I always prefer hard copy so I can detach the tables and figures and refer to them will reading the ms and so I can make editorial changes on the ms such as corrections to grammer, sentence structure, typos, etc.
Many journals published by scientific organizations require steep pages charges and charge for reprints. This can be avoided by starting your journal or blog.
[Response: In my opinion, the use of LaTeX has made the scientific author's job vastly easier. But then, I'm a mathematician, and the formatting of equations may be the strongest aspect of the LaTeX system.
There are indeed different requirements for different journals, but within a single discipline there are often similarities. If you do astrophysics and publish in an American journal, you're very likely to be on firm ground using the aastex system -- and after you do it once it becomes straightforward to repeat the process.
I've found that there are *extreme* differences between journals, and even within the same journal, regarding response time for reviews. GRL, bless 'em, makes good on their promise to provide prompt review. But I once submitted a manuscript to a highly reputable journal (which I will not name), and only after numerous complaints to the editor and 11 months did I finally get a decision. Then -- they wanted major revisions! I won't be submitting to them again.
There are two unfortunate aspects of self-publication. First, it's absent peer review, which really is a *minimal* quality control measure. Even if you're a crackpot, and a sloppy one at that, you can still self-publish. By no means does peer review guarantee quality work, but it sure does increase the likelihood. Second, if you want your work to come to the attention of the scientific community, you can't just put it on your blog and expect people to track it down. Put it in a peer-reviewed journal (especially a major one) and the scientific community will see it.]
cce // March 2, 2008 at 11:35 pm
If there was no Y2K bug, I applied a lot of imaginary Y2K patches to operating systems, databases, and applications.
Martin // March 2, 2008 at 11:36 pm
Fred, tamino has written about this whole thing earlier:
http://tamino.wordpress.com/2007/05/11/best-estimates/
which compiles quite a lot of background. The total number of stations is 7364 going into Step 2. Now I don’t know if the claim was that 1419 stations or 74% was modified, but it cannot be both…
From the file v2.inv I find that there are 1959 ‘U’rban stations, 3996 ‘R’ural ones, and 1409 marked ‘S’ (semi-urban?) totalling 7364. Now we know that all urban stations were modified, so adding ‘R’ (1959) and ‘S’ (1409) yields 5405, or 73% of the total. (This is after the removal from the computation of whole stations or parts of station time series, or individual months, found to be in gross error.)
Makes sense. But as tamino explains (and I did too in an earlier post somewhere), this is perfectly legit, and even the only proper way of getting the UHIE trends out of the urban records without introducing new systematics.
Hope this helps.
Harold Pierce Jr // March 2, 2008 at 11:40 pm
ATTN: HERETIC!
I am a scientist (organic chemist) and always a skeptic. I am absolutely outraged that you and your hopelessly and perpetuality clueless ( i.e., downright ignorant) crowd portray us skeptical scientist as evil and sinister people. We are mostly are not! We do what we always do: question anyone and everyone, and anything and everything.
[Response: Skepticism is a necessary quality for the advancement of science.
But skepticism doesn't include denying the obvious, refusing to believe the evidence, cherry-picking, or repeating sham arguments long *after* they've been undeniably refuted. That's why I much prefer the term "denialist." Skepticism is a badge of honor, denialism one of shame.
I myself am a skeptical scientist. "My" crowd has never portrayed me as evil and sinister.]
Martin // March 2, 2008 at 11:44 pm
Oops: 1959 + 1409 = 3368, only 46%, not 74%. Some mystery remains… probably part of the rural stations were modified too (time-of-observation bias?)
Time to go to bed…
MrPete // March 2, 2008 at 11:49 pm
john, please don’t dig yourself any deeper. To say “I feel dismayed that some still think the Y2K computer bug existed, what they should ask themselves is which piece of code was altered.” Is simply to demonstrate that you know little if anything of the subject. Banks, phone companies, etc spent years correcting the problem. A friend worked hard fixing every single Chevron/Texaco drilling and refining platform worldwide. Huge lists of compliant and non-compliant computer software, medical products, etc were created. It was quite real. No, not a trillion dollar issue. And by the time y2k showed up, not a disaster.
Engineering, science and more really do help us move forward as a civilization. We all need respect for what we do and do not understand. All of us.
Hank Roberts // March 3, 2008 at 12:04 am
You can proclaim how it was in the worldline you were in at the time, and we’ll believe you were there then. You can look up how it was in this one. Don’t get them confused.
http://xenia.media.mit.edu/~cwren/y2k/risks-20.72.txt
http://catless.ncl.ac.uk/Risks/19.18.html#subj7
Greg // March 3, 2008 at 12:35 am
There was still an enormous disconnect between the reality of the Y2K bug and the alarmist hype. Countries such as Italy and Russia wisely didn’t spend a cent. The US comically issued ‘travel warnings’ against them. Nothing happened.
From 1999:
“The year-2000 crisis is about to become an international incident. The U.S. State Department on Sept. 15 will release its biannual consular advisory — this time including a list of countries expected to experience severe year-2000 issues.
Industry analysts expect that Brazil, Indonesia, Italy, Russia, Saudi Arabia, and Thailand are some of the candidates for the list. That means companies operating in those countries and sending business travelers to them need to begin thinking about developing action plans. “
Deech56 // March 3, 2008 at 1:00 am
RE: Ron Cram // March 2, 2008 at 5:28 am
Deech56, “If you have complete faith in peer-reviewed research, then you are not paying attention. ”
The first links you provided were not some serious study of the pitfalls of peer review in climate science but were what appeared to me as one scientist’s opinion. I would rather see something more concrete.
RE: Harold Pierce Jr // March 2, 2008 at 11:04 pm
Yes, peer-review has its pitfalls, but Tamino provided the best arguments in favor of peer review in his in-line response to you. I would just add that loading up a database in a program such as EndNote is not that difficult. Most of the manuscripts I use are indexed by PubMed (I am a Medical Researcher, more in administration now, but still writing manuscripts) and it is easy to download the citations from within the program.
I am a scientist and a skeptic, but at some point we have to understand that sometimes we scientists do discover how things really are, and that as more evidence, including independent observations, arises that bolsters a position, we need to sit up and take notice. I am “open minded” but I do set the bar high for contrarian views – my work in vaccine and HIV research has given me a dim view of denialists.
joe // March 3, 2008 at 1:43 am
I’m normally just an observer in this ongoing food fight. But, I find it interesting that nobody will answer Mr. Pete’s simple yes or no questions. Here they are again so that you don’t have to go looking. I’m curious to hear what the answers are.
* Do you agree that proactively enabling reader access to data, as required by the journals, is a good thing?
* Do you agree that scientists should not be allowed to ignore such requirements?
* Since we all agree that “sniping innuendo” is to be avoided, here’s a serious question: When it is proven that scientists are ignoring the requirements, and the journal editors are letting them get away with it, what would you do? What kind of peer-reviewed article would you write to expose this in a “validly scientific” way? What journal would you submit it to?
[Response: OK, here's my opinion.
#1 Is proactively enabling access to data a good thing? Yes.
#2 Should scientists be allowed to ignore such requirements? No in general, although I can't rule out the possibility of extenuating circumstances, and the amount of work involved to do so should be considered.
#3 What journal would you submit it to? I'd give first opportunity to the journal which has "let them get away with it." I'd be sure to have solid evidence, and frame the criticism as constructively as possible. If such proof were ignored and the practice continued, perhaps PNAS (or the relevant national science academy of the nation in which the journal is published).]
dhogaza // March 3, 2008 at 2:09 am
What makes you think that the claim that this has been proven is correct? I’ve read the thread in question regarding Thompson, and I disagree that it’s been proven.
The requirements aren’t being met in the way that McIntyre imagines they must be met.
There’s a lot of things that McIntyre imagines that the rest of the world doesn’t agree with, for instance his belief that the admins of Real Climate have to approve his posts on Christmas Eve rather than spend time with their family.
TCO // March 3, 2008 at 2:20 am
Let me answer:
1. yes. The absolute best is to just publish the actual raw data itself in the archived journal of record. This is not always practical. but any other fashion (SI, database, private means) is helpful. In crystallography for instance (and by this point most chemical, biological and physics journals where crystallography is reported of new structures) it is a requirement to archive the data properly in a common use database and reviewers will actually look at it and attempt to solve for structures or note aspects of the solution that are questionable. However, note that many physical and chemical measurements (for instance conductivities of substances) are not archived and one has to rely on them being in the paper, the SI or contacting the author. Saving of such data is not perfect. However, chemical type experiments can be repeated.
2. Scientists should be more encouraged to archive their data.
3. In terms of practical methods to change the behavior (in climate science), I’m not sure the best method. I think requiring the data to be archived PRIOR to publication and be available to reviewers is the most practical method of enforcement. I would also say that this is not the only problem in the field. Fixing it won’t make everything dasies. And leaving it so, won’t make everything awful.
3.a. In terms of writing a paper, I think you would need to do a very careful study of practices in climate and in other fields with some extensiveness. Would need to avoid just citing a few popular case studies. In other words, the harder thing would be to settle down…be objective…emotionally step back from the “wars” and just examine the problem. You ought to at least be open to finding that the problem is not so extreme or important as one might think if one read a lot of CA. (Not to say that you won’t find it bad, but an attitude going in of neutrality and curiosity would be important.) And doing a good study and being calm and fair about it would be the harder thing in getting a paper than deciding which journal to put it in. However, appropriate journals would be either the broader climate journals themselves or a “methods of research” journal.
Bernie // March 3, 2008 at 2:54 am
dhogaza:
You really can’t say where Thompson archived his data can you? Why all the bluster and avoidance just say it. The fulminating is not flattering.
Thompson ice core data is crucial to a number of assertions. His research was and is NSF funded. There can be no justification for not providing access to the data.
[Response: I managed to find it rather easily when the subject came up a while back. Have you actually looked?
It's not archived in the manner, or the detail, that McIntyre and others would like -- but apparently in sufficient detail to satisfy the journal editors.]
Hank Roberts // March 3, 2008 at 3:02 am
You can look up the terms of those grants; they specify cost sharing for those researchers who want to make use of the results of the funded expeditions, in this case ice cores. Want to split the cost of the expedition? It’s on record. NSF will be glad to get some of their money back from your research grant. Got credentials in the field? You know where to apply, if so. If not, you’re just copying and pasting.
dhogaza // March 3, 2008 at 3:41 am
The dog makes the point:
Exactly.
The only thing I would add is that the claim of unacceptable lack of archiving comes AFTER THE FACT.
Not only is McIntyre insisting that his own notion of the rules be followed, but he is insisting that scientists publishing in the past should’ve followed his interpretation of the rules long before he decided what they are.
Even if we were to make McIntyre God and allow him to establish his own interpretation of the rules as being binding in the future, it’s a bit stupid to assume that anyone who failed to meet the future interpretation in past papers is somehow guilty of fraud, etc.
Phil. // March 3, 2008 at 3:43 am
Bernie // March 3, 2008 at 2:54 am
dhogaza:
You really can’t say where Thompson archived his data can you? Why all the bluster and avoidance just say it. The fulminating is not flattering.
Thompson ice core data is crucial to a number of assertions. His research was and is NSF funded. There can be no justification for not providing access to the data.
Read his PNAS inaugural paper, he links to the data there!
fred // March 3, 2008 at 8:27 am
Martin, what is being said is that there are 1921 stations in the GISS database. Of these, 1419 are said to have been adjusted in some way, whether up, down or both. 353 are said not to have been adjusted, and 149 not used. This would imply that they must be using 1921 – 149 for the US, which is obviously more than one had thought.
From the previous posting I linked to, it appears that that total given includes MCDW stations and maybe also sum of the day stations.
Steve M’s point about the adjustments was not that they necessarily themselves led to a bias of any sort. Indeed, his view is that they do not seem to. It was rather that they did not seem to him to serve their announced purpose of eliminating UHCI effects. Whether this has the effect of leaving UHCI in the record, rather than eliminating it as hoped, is a question on which he is to write later.
I am puzzled about the number of stations. Maybe it will be cleared up this week.
fred // March 3, 2008 at 8:36 am
OK, now think it is clearer what is being said. There are “1221 USHCN stations, but GISS has 1921 stations. The proportion of adjusted stations shown here is the proportion of 1921.” So GISS is using more than the USHCN stations in its plots.
So the question remains, if GISS is using 1921 stations, and adjusting 74% of them, does this level of changes have any other parallels in science? It seems very strange to take a series of sensors, and then change some of the readings on 74% of them, before going on to plot the results.
Surely this has to be treated very gingerly when presenting statistics and error bars?
null{} // March 3, 2008 at 11:19 am
So Hank, are you now saying that it’s ok to toss out data.
If so, why not junk the stations that are known to have problems? After all it seems that GISS/NASA does that very thing.
All along here we’ve been told that data are not to be tossed out??
null{} // March 3, 2008 at 12:48 pm
ok, here’s an exercise left for the students. Why is Crater Lake data junked and not used and Puerto Maldonado data are not junked and used?
Is documentation of the GISS/NASA process/procedure sufficiently detailed that this question can be answered by anyone independent of GISS/NASA? Or must we contact GISS/NASA whenever situations like this are encountered?
And for those here who are reading the coding in depth for understanding, kindly show us the criteria used to make these two decisions.
MrPete // March 3, 2008 at 2:20 pm
This is sad. Some of you really are arguing not with Steve but with those who set up the clearly-stated journal publishing data availability principles. And unfortunately, the journal editors themselves are being caught not following their own principles.
“The only thing I would add is that the claim of unacceptable lack of archiving comes AFTER THE FACT.”
That’s kind of inherent to the issue. We’re not going to discover unacceptable lack of archiving BEFORE the fact, are we.
“Not only is McIntyre insisting that his own notion of the rules be followed, but he is insisting that scientists publishing in the past should’ve followed his interpretation of the rules long before he decided what they are.”
No, he’s insisting that scientists, and editors, follow their own principles.
“It’s not archived in the manner, or the detail, that McIntyre and others would like — but apparently in sufficient detail to satisfy the journal editors.”
And that’s the problem here. Even the journal editors are kowtowing, and ignoring their own principles. The principles are quite clear, and very reasonable: other scientists need access not just to the post-analysis data, but to the pre-analysis data, so that the analysis itself can be studied. Otherwise, we end up with the equivalent of cold fusion, etc.
Does someone wish to argue with that? If this were any other arena, I’m quite certain there’d be a dogpile (speaking of woofers :-)) to encourage a healthier appreciation of good science.
Oh, and the problem with unarchived data is not that the resulting publications are fraudulent. The problem is that they’re useless. Can’t be replicated, can’t be verified, can’t be falsified. It’s not really science without the data, is it?
MrPete // March 3, 2008 at 2:29 pm
Just to make further contemplation easy:
“As a condition of publication, authors must agree to make available all data necessary to understand and assess the conclusions of the manuscript to any reader of Science.”
“An inherent principle of publication is that others should be able to replicate and build upon the authors’ published claims. Therefore, a condition of publication in a Nature journal is that authors are required to make materials, data and associated protocols available… Any restrictions on the availability of materials or information must be disclosed at the time of submission of the manuscript…”
I don’t see a lot of wiggle room. They say it in different words, but the meaning is rather clear:
* any reader must be given access
* intent is to enable replicating (assessing, etc) the analysis/claims/etc
* any data restrictions must be disclosed at publication time
These principles are not being followed by several scientists in this area.
Other areas of science have MUCH better practices.
I’m told some universities (including my own alma mater) have standardized on a practice that some of you might find familiar: data is supplied along with analysis scripts in “makefile” form– the reader can reconstruct the analysis from input to output through a single unix/linux command! Talk about ensuring your analysis is replicable :-)
steven mosher // March 3, 2008 at 2:43 pm
Martin in v2.inv the label R only means Rural
in the ROW, in the US Rural is determined by the nighlights data feild.
dhogaza // March 3, 2008 at 2:47 pm
In other words, despite your bluster, you agree that the data HAS been archived, just as I’ve said.
McIntyre’s complaint is that it’s the analyzed data, not the raw data, that’s been archived. He, like you, though, argues that his means that the data’s not been archived, that Thompson and the journal haven’t followed their rules, etc.
All bullshit.
It’s very dishonest for you to post over and over again that “the data’s not archived” when you know it HAS been archived, simply not in a form you like.
The ice cores themselves are archived.
MrPete // March 3, 2008 at 2:55 pm
(I should have said unix/linux/ windows/mac — “make” is available on just about any platform one can imagine.)
Hank Roberts // March 3, 2008 at 3:31 pm
Science usually starts with collecting one’s own data, with one’s own instruments. That’s original work. That’s also replication. This is how it’s done. Many of the most interesting discoveries come from failed replications (one of the famous ones discovered an insect growth regulator that was present in commercial paper towels after the entire papermaking process — the paper towels used in Canada were from fir trees, as I recall, and the paper towels used in a British lab replicating the study were from a different plant — and that was the difference that changed the outcome).
This is one reason there are dozens of climate models — people replicate work. Everyone who’s built a climate model so far comes up with climate change. Funny that.
So the people who want to say it can’t be happening aren’t building models. Even the big oil companies that do build proprietary models and rely on them to find oil don’t also build other models for public relations use to try to prove warming isn’t a problem.
Guess why?
Pick any data. Give it a try.
Hank Roberts // March 3, 2008 at 3:36 pm
Oh, and for access to the ice core work, as I read the grant terms a few years ago when this was being bruited about, all you have to do is acquire a track record in the field (be a researcher) and split the cost of collecting them (your granting agency will be delighted to grant you only half as much money as it would have cost if you’d gotten Thompson’s original grant and gone out on the ice yourself to do the drilling. And no risk of frostbite or death for you. Cozy, eh?
Why not make a name for yourself? Get the degree, do the publications.
Or you can just copy and paste.
MrPete // March 3, 2008 at 4:22 pm
Three thoughts.
One, I respectfully disagree that “original” and “replication” work are identical in always or even mostly starting with new field data and instruments. If that were true, then why are any climate data sets ever reused, and why do people like Mann et al do zero field work? By Hank’s definition, wouldn’t Mann, Hansen, etc be susceptible to the accusation that they are not scientists doing scientific work?
Two, suppose Hank’s memory is accurate, that Thompson’s grant terms prohibited sharing data. If so, he was required to reveal that as part of his journal submissions and publications. The rest of the field could adjust their efforts in concert with that. (Hopefully reasonable analogy: various standards bodies require unrestricted access to patentable ideas before they are incorporated into further work. When hidden or undeclared restrictions later emerge, a huge hullabaloo ensues. The result: either the restriction is eliminated, or the standard is revised to avoid use of the restricted material.)
Third, much of today’s work involves statistical analysis as much, if not more than, field work.
How do you replicate field work? By doing the same kind of field work using the same methods published, as Hank nicely reminds us. No surprise there.
How do scientists replicate statistical analysis of source data? Don’t they take the same source data and do their own statistical analysis? Thus, to validate or falsify Thompson’s statistical analysis, one needs access to his data.
Many hard sciences work this way. Much of climate science as well. Why so much reticence here?
If it *really* is simply a question of money (and I don’t see evidence of that, since we’re talking about the analysis not the data collection)…but if it is all about the money, then fine. Let’s get some policy/money people involved. After all, isn’t splitting the cost of Thompson’s work peanuts compared to the policy expenditures envisioned?
Do you know of other arenas where unvalidated data analysis is knowingly permitted to influence significant outcomes?
Is “trust me, it works” (without validation or verification) scientific?
MrPete // March 3, 2008 at 4:24 pm
(On field replication, I could make a Martha Stewart joke here, but I won’t :-) )
MrPete // March 3, 2008 at 4:43 pm
Dhogaza — ok, let’s assume the ice cores themselves are archived and can be accessed for a price. Does that enable replication of the analysis?
Thompson originally took thousands of samples from the fresh cores, obtaining detailed isotope and chemistry info. See here for Steve M’s discussion including specific quotes.
The only data archived: ten year averages, using unknown methods. No access to the original data.
Now it is 2008. You pay whatever entrance fee is required and use new instrumentation to analyze the cores.
Suppose your results come out different. How do you know what the problem is? It could be any of:
* Thompson’s instrumentation
* Thompson’s data collection system
* Thompson’s analysis
* Core degradation since archiving
* Your instrumentation
* Your data collection system
* Your analysis
You have no way to know. No way to verify, validate or falsify Thompson’s claims.
I remain saddened by such spirited defense of poor scientific practice.
Martin // March 3, 2008 at 4:51 pm
Steve — thanks!
But the nightlights field is also given for ROW in my version (Oct. 07) of the file. So — do they plan on using it? Wondering.
(And anyway there seems to be a quite strong correlation between the two, as one would expect if both were roughly valid.)
Martin // March 3, 2008 at 5:21 pm
Fred:
So, when stating the number of changed records, are we talking only about the trend adjustment? Or does the count include all changes to records made, like the time-of-day correction?
Curious.
Martin // March 3, 2008 at 5:48 pm
Fred:
Not really. I feel that you are misunderstanding what it means to “measure” something. Mostly we don’t “measure results”. We measure something that is in some way related to the results, and extract the results by processing.
In geodesy (my field), typically 100% of measurements are adjusted. E.g., if you have a triangle network, you will adjust the angles of every triangle in order to make them total 180 degs, as theory says they should. In a levelling network, you make the sum of height differences around every loop equal to zero, as they theoretically should [well, not precisely true, but that's for theoretical purists].
The processing happens in two steps:
1) statistical testing, and
2) least squares adjustment.
Consider a simple triangle where three angles were measured using the same instrument and method. First you test: add up the angles and compare with 180 degs. If you find, say, 185.375 degs, you know there’s something wrong. A gross error. You have to throw out this data, remeasure it if you can and you don’t find the mistake.
Now, if you find 179.997 degs, you know that your measurements are OK, but imprecise. So you adjust. You add +0.001 degs to each of your angles, and now the triangle closes precisely. And each of the angles has been modified but has become a little bit preciser. As you can verify by computing its post-adjustment standard deviation — this is premised on the small imprecisions in the data being random.
Additionally there may of course be systematic errors, like atmospheric refraction. You try to model or eliminate them somehow. These are the most dangerous.
I recognize in Hansen et al.’s procedure elements of the same approach. Any proper analysis technique takes all different types of errors into account. So, e.g., you cannot use a technique for eliminating UHI if it makes also existing random effects in the data systematic. You won’t be able to eliminate all error; the point is minimizing their effect while keeping the error propagation analysis tractable.
And adjusting “only” 74% of the data is in my field atypically low!
MrPete // March 3, 2008 at 6:58 pm
Great example, Martin!
BTW, in your field, would it be valid to assume:
* You keep your unadjusted data
* Measurement errors tend to be randomly dispersed above and below the target total value.
The “nice” thing about this example is there’s a known, very precise target value. You can know exactly how far off you are. If only it were that easy in other arenas :)
luminous beauty // March 3, 2008 at 8:06 pm
MrPete,
You and SM, are demanding, not requesting, Lonnie Thompson to retro-actively comply with a rather selective and skewed interpretation of 2006 archiving standards for a 2003 paper quoting a decadally averaged data series from a 1997 paper for the sole purpose of checking his arithmetic, and implying that if he doesn’t immediately comply with your demands he must be guilty of scientific misconduct.
I guess you cannot see how trivial, arrogant, insulting and absurd that is.
MrPete // March 3, 2008 at 11:02 pm
luminous beauty,
This did not begin as “demands.” It began quite politely. Even now, the heated statements I read, even in this thread, are from those who are defensive about poor scientific practice. Not from those who are asking for better practice.
With words like “selective and skewed”, you apparently disagree with my reading of the PNAS/Science/Nature policy as requiring making all data necessary to replicate analysis available to any reader. I don’t just apply it to Thompson, I’d want that for any scientific publication. That’s my reading.
Since you disagree, please explain how you read the PNAS, Science and Nature archiving requirements in such a way that Thompson’s archiving (and any other similar work) is compliant.
Personally, I’m not picking on Thompson. I simply believe what the requirements say, and would want to see them sustained for any scientist who publishes. I want to see good science done.
Last, the purpose is not simply “arithmetic checking”.
To compare Thompson’s data vs others, when his is decadal and others are annual… would it not make sense to use his underlying annual data?
Do I think it trivial, arrogant or absurd to ask that data for 1997 work be still available in 2008? No. That’s what archives are for. That’s why the archives contain a lot of data collected in the 1990’s. And much earlier. Why do we set archiving standards, and maintain archives, if not to enable access to prior data?
luminous beauty, your statement sounds like this, mirrored back to you. Read this and perhaps you will understand why I would not make the kinds of claims you are making.
I am not trying to say the following. Only expressing a ‘mirror’ form of lb’s complaint. You and others are demanding, not requesting, that realistic skeptics should rely on a rather selective and skewed interpretation of long-time scientific archiving standards, for certain papers published by a select set of climate scientists, for the sole purpose of enabling those papers to be accepted and referenced by others without normal verification or replication study, and implying that if we do not comply with your demands we must be guilty of scientific denialism. I guess you cannot see how trivial, arrogant, insulting and absurd that is.
If this type of circle-the-wagons attitude could be isolated to a trivial aspect of science, without implication to the rest of the world, I’d simply go wait for the natural course of events to take over. Unfortunately, this area of science has visibility, PR, influence and resources. So it’s rather important to clean up the mess before it gets worse.
In that sense, yes, this is like y2k. I could not stay quiet. I had to say yes when senior leaders asked for help discerning the truth. There was just too much garbage being publicized.
Am I demanding compliance? Nope. Impossible. I have no authority and little influence to make such demands. I’ve been around the block long enough to know that such demands would be useless. And more harmful to my integrity and reputation than anything else. All I can do is suggest and ask. And I can ask others to do the same.
LB, how about you? What kinds of requests are you making to see better science done? When’s the last time you did something to contribute positively to improved the data, methods or results of scientific inquiry?
I believe good science gets better only when people push in that direction. Left to atrophy, it will get worse. I believe we should all be nurturing better science rather than defending poor examples.
And as always, I hope what I’m writing is not trivial, arrogant, insulting nor absurd. How you interpret it is up to you of course.
Bernie // March 3, 2008 at 11:12 pm
LB
As you put it yes it does seem trivial, etc., but then that is not what happened. Apparently those decadal averages shift around without any obvious explanation – that is one of the things that caught someone’s attention as various multi-proxy studies were deconstructed. Also the initial requests were requests, then they became repeated requests, then emphatic requests and probably then they must have seemed like demands.
Hank Roberts // March 3, 2008 at 11:13 pm
Butter wouldn’t melt in your mouth, bless your heart. l
Ian Forrester // March 3, 2008 at 11:21 pm
Mr Pete said: “What kinds of requests are you making to see better science done? When’s the last time you did something to contribute positively to improved the data, methods or results of scientific inquiry?”
I tell all my friends that are interested in genuine climate science to say away from junk science at sites such as CA, WUWT, TCS, HI, CEI, AEI, SEPP et al.
How can anyone who is interested in improving science use these sites for inspiration and information?
john // March 3, 2008 at 11:40 pm
Dear Mr Pete,
I will dig my hole a little deeper about Y2K , I too drew up long lists of compliant and non compliant software that were set against a set of rules drawn up by the experts.
However the non compliant stuff didn’t go wrong at the millenium, so the rules were wrong.
This Myth was restricted to the English speaking world only, I dealt with colleagues in Italy and Greece they did absolutely nothing and absolutely nothing went wrong.
You seem unable to accept something that is self evidently true, but if it makes you feel comfortable to believe in the infallibility of science then please carry on. I am afraid I don’t believe.
MrPete // March 4, 2008 at 12:48 am
y2k john, I’m amazed to find myself in this position, but hey…
Here’s my answer to your three assertions:
“However the non compliant stuff didn’t go wrong at the millenium, so the rules were wrong.”
Fallacy #1: believing uninformed people who said everything would break at the same time. The experts always knew that was untrue. Banks probably had the issue first — due to mortgage calculations going off. You saw the credit card issue in real life: credit card expiration dates became shorter and shorter, because they couldn’t handle four digit expirations. As that got cleaned up, the credit card expiration periods again could be lengthened (my newest card now expires in 2011; for a time in 1997-98 I was receiving one-year cards!)
This Myth was restricted to the English speaking world only, I dealt with colleagues in Italy and Greece they did absolutely nothing and absolutely nothing went wrong.
Fallacy #2: believing that everything is the same for everyone, everywhere, and that everyone lives under the same circumstances.
Most people were not in danger of trouble, ever. My recommendation to most people was “check X, Y and Z. As long as those are ok, you’ll be ok. To be careful, do what you should do to be prepared for any kind of trouble, simply because other people may panic.” For example, only a small proportion of PC’s had any issue at all, and most of those simply needed a reboot after 1/1/2000.
On the other hand, friends using certain brands of Kidney Dialysis machines, anywhere on the planet, had to get them serviced or they would no longer be able to do their life-sustaining dialysis.
“You seem unable to accept something that is self evidently true, but if it makes you feel comfortable to believe in the infallibility of science then please carry on. I am afraid I don’t believe.”
Fallacy #3: the self-validating proposition. Apparently you believe that lack of y2k trouble proves y2K “scaremongers” (and more) were untrustworthy, and therefore science is (partly, mostly, completely) untrustworthy.
This is a great example of how many PR flacks attempt to sway public opinion today. Sadly, it is a technique explicitly recommended as a strategy for winning progressive/environmental campaigns. (I won’t promote the guy with a link, but google “How To Win Campaigns” and find the book. Some may like it; it makes me wanna…)
Here’s how the Self Validating Proposition pitch goes:
You want people to believe C. Find A and B that at least on the surface are plausibly related to C, and where people will likely believe “If A then B”. Doesn’t matter at all if C is related to A and B, nor whether C is true.
Then simply say (with a flourish) “You know, B because A! So, C!” You want people to think “huh, I never thought about it that way! I guess you’re right!” but NEVER to consider whether A or B actually implies C.
In this case
A = y2K was not a disaster
B = y2K promoters were untrustworthy
C = scientists are untrustworthy
“Y’know, y2K promoters were untrustworthy. After all, y2k was not really a disaster at all, was it. So obviously, you can’t trust scientists.” Wow, I never thought about it that way. You must be right!
It’s easy! It’s fun! You can convince many people of anything you want this way.
I’d apply it to the current conversation here, but I’m tired. See y’all later!
“Y’know, MrPete is a denialist. After all, he keeps begging out of Tamino’s blog whenever the heat turns on. So it’s obvious that anybody who is skeptical about climate science claims is full of it!” Wow, I never thought about it that way. You must be right!
:-) :-)
Lazar // March 4, 2008 at 2:33 am
Did Lonnie Thomson archive to these standards?
Did Gavin Schmidt post in working hours?
Did James Hansen document to requirements?
Is posting on RealClimate in breach of DQA?
Did James Hansen really think A or B?
And what does red ink mean in climatology?
For it’s Tommy this, an’ Tommy that, an’ “Tommy wait outside”;
But it’s “Special train for Atkins,” when the trooper’s on the tide-
The troopship’s on the tide, my boys, the troopship’s on the tide,
O it’s “Special train for Atkins,” when the trooper’s on the tide.
Yes, makin’ mock o’ uniforms that guard you while you sleep
Is cheaper than them uniforms, an’ they’re starvation cheap;
An’ hustlin’ drunken soldiers when they’re goin’ large a bit
Is five times better business than paradin’ in full kit.
– Tommy, Rudyard Kipling
Hank Roberts // March 4, 2008 at 2:47 am
comp.risks 20.13
Date: Wed, 9 Dec 1998 23:03:55 +0000
From: Phil Pennock
Subject: British Government admits Y2K missile problem
From The Times, 9 Dec 1998, p2:
Bug threat to missile
The Ministry of Defense admitted for the first time that the millennium bug
could have left Britain vulnerable to air attack. It discovered that the Rapier anti-aircraft missile would have failed to retaliate. The problem was identified inside the field equipment which activates the missiles and it would have made the system inoperable. The threat to Britain’s defenses posed by the computer bug was outlined by George Robertson, the Defense Secretary.
[Well, at least the failure mode was to not fire.]
————-end quote———-
I’m sure in a few generations people will be hearing that all the predictions about GW were exaggerated too, no matter what conditions are like. People take for granted the conditions they’re born into.
luminous beauty // March 4, 2008 at 2:54 am
MrPete,
I have no problem with NAS’s archiving requirements. But they are new. It doesn’t make sense to re-do all the research for the last century to bring it all into compliance with current standards and it is just stupid to accuse a researcher of non-compliance with a standard that didn’t exist when he first published.
Steve Mc was already accusing Thompson of bad science before the subject of Guliya came up, demanding data from a study that was never published and unrelated to temperature proxies in the first place. He was never polite about it.
Steve has burned his bridges. He makes uninformed insinuations of fraud, ignores well founded criticism of his own poorly reasoned notions and doesn’t understand why the recipients of his insults and tantrums want nothing to do with him.
Typical Steve Mc apology: “I’m sorry you are uncomfortable with my questions.”
Well, I’m sorry you don’t understand the problem.
Greg // March 4, 2008 at 3:07 am
Anyone with a pair of eyeballs and a functioning brainstem knows that Y2K was not one-fiftieth of the problem it was beaten up to be. That people here can’t admit that speaks volumes about the kind of alarmism we’re dealing with.
Hank Roberts // March 4, 2008 at 3:57 am
No, because you’re talking about journalists and politicians, and we’re talking about scientists.
Nobody here who cares about getting the science right is cutting any slack for the people — from any political direction who can’t cite their sources and describe them accurately. The believers of all sorts are equally vague and unable to say what they base their beliefs on, why they consider their sources reliable, and how they’ve checked the footnotes.
Trust the cites — after you verify them.
Ian // March 4, 2008 at 4:08 am
Greg said “Y2K was not one-fiftieth of the problem it was beaten up to be.”
Interesting wording – “was beaten up to be” by whom?
Greg, you aren’t just giving us the gist of what you now remember, are you, especially after the Y2K links above?
Greg // March 4, 2008 at 5:18 am
No, I’m not giving you the gist of what I remember. I’m a computer scientist and an engineer, I know what went on. It was a real (but minor) problem, but we had the general public being told planes would fall from the sky and their VCRs would stop working, in ominous tones by some clown of a narrator in BBC documentaries. I wish I’d taped some of those ’specials’ on Y2K, they would be very amusing to watch today.
And that isn’t a million miles from Al Gore telling us that Tuvalu is being evacuated due to sea rises, or subtly linking the drying up of Lake Chad or Hurricane Katrina to AGW. Same rubbish tactics, new scare.
If you were right Hank, if people like that really weren’t cut any slack, then the ‘deniers’ would have a lot less ammunition. They’d be forced to debate just the dry science, rather than puffed up hype. Both sides should clean up their own yards first.
At times there seems to be a tacit acceptance of lying to the public in order to ‘motivate them’ to meet the AGW menace.
Hank Roberts // March 4, 2008 at 5:38 am
Greg, would you point us to your postings about chlorofluorocarbons and ozone? Were you active on sci.environment, for example? It’d be interesting to know what you thought was going on, at that time. And what you think about stratospheric ozone chemistry now.
dhogaza // March 4, 2008 at 5:41 am
No, it’s not just stupid. It is dishonest. Totally dishonest, and intentionally dishonest, and SM and (apparently) Mr. Pete are devoted to this dishonesty.
I’m both of those, and its being “minor” involved an expenditure by a large number of businesses that totes up, in total, to the “billions” range.
Deniers don’t simply refute the hype surrounding the science, they refute the science, as well, AND ACCUSE CLIMATE SCIENCE, IN TOTAL, AS BEING A FRAUD.
Actually, Gore was very careful in his wording.
But it doesn’t [edit] matter in the least. Gore is a politician who wrote a book, made a movie. He has no impact on the science.
McIntyre’s an [edit] who has a blog, and has no significant impact on the science.
Name one such lie that our host, or me, or other posters here accept in order to motivate the publics to “meet the AGW menace”.
One, just one.
State the lie. Pure and simple unambiguous English being preferred.
[Response: Please no profanity.]
dhogaza // March 4, 2008 at 5:44 am
So you admit, that unlike climate science (a fraud), that Y2K was real, you only disagree with the magnitude of the problem?
So, now, why do you buy into the supposition that climate science is inherently fraudulent?
Because if you believe more than about 1% of what McIntyre posts, that’s exactly what you’re doing.
Not to mention Watts…
Martin // March 4, 2008 at 5:48 am
MrPete:
1) Yes, you keep (i.e., don’t physically throw away; typically archive) your
unadjusted data.
2) What do you mean by “target value”? The “true value”? Hmmm yes, we call that the “expected value”, and assume measurement error (random error) to be randomly spread around this (“measurement noise”).
(In addition however, there may be systematic errors and gross errors. This is where statistical testing comes in, to be done before any adjustment.)
3) No, you don’t actually. Look back at the triangle example: we don’t know what the target value of each angle measurement A,B,C is. We do know what the target value of their sum A+B+C is precisely 180 degs.
So the a priori knowledge we have is not on the unknowns or observations themselves, but on relationships that exist between them (“condition equations”). In the temperature record case, the analoguous situation is our knowledge that geographically close stations will display very similar temperature variations. This knowledge can be used to
(a) identify gross errors or major systematic distortions in one of these station time series (if it differs from several others around it; “outlier detection”), and
(b) to adjust away the differences in variation between stations.
The confusing thing for many people seems to be that what Hansen et al. call “adjustment”, actually falls under (a). The (b) thing is the areal averaging done afterwards.
Brian D // March 4, 2008 at 6:05 am
Setting aside that Greg’s position amounts to “Scare tactics were used by people who weren’t entirely correct then, therefore anyone using scare tactics must always be wrong” (a fallacious line of reasoning if there ever was one), and that he’s misrepresenting Gore (he didn’t link Katrina to GW; he used the damage that a large hurricane caused when it hit an urban center as an example of… what happens when a big hurricane hits an urban center), there’s something else he’s missing.
Namely, that we actually acted on Y2K. We understood the problem, and we took the action predicted by the real experts in the field (essentially amounting to a multitude of adjustments before a particular deadline). I know firsthand as well: Although I wasn’t qualified to do much at the time, I spent my Y2K new year’s in a government server control room, monitoring incoming reports from across the province. (There were a rare few mishaps, but most were minor.)
We also had the naysayers who denied there was a problem, most of whom relied on armchair analysis, even as people were scrambling to be certain that adjustments were made. (Note: This is different from Greg’s position, which is that the problem was overblown. I’m not strawmanning; there’s a reason I’m doing this.) These people were only fed by the apparent lack of flashbangs come January 1, 2000 — especially because the news doesn’t tend to report this sort of thing if there isn’t something dramatic. (Meaning if the action taken is to *prevent* something from happening, as it was here, news coverage will be minor. Who wants to publish the headline Another day just like yesterday? Unless the disaster was something spectacular like an incoming meteor or something, I doubt we’ll see anything like that.)
Aside: To this day I find people who claim the whole thing was a hoax to increase government control of our lives. I don’t feel any mind-control rays — *gasp* I guess that means it worked!
…On a more serious note, I have also met people who use *precisely* this same logic when citing the CFC/ozone ’scare tactics’, including proclaiming the whole thing a hoax on account of… not hearing about it in the news today. I wish I were joking.
Getting back to the main point, and trying to segue back to the original topic, I must ask: How is this different from what we have today on the climate change debate (apart from the state of activity and the fact that no one on the take-action side is saying we *won’t* see effects, even if we act)?
We have a problem (the globe is warming), we (collectively, through the relevant experts) understand the solution (lowering emissions), we know what action is needed. We have people denying that we have a problem even as people/governments are scrambling to implement it. We have people like Greg, who insist that it’s overblown (although unlike Y2K, each subsequent credible prediction paints a darker and darker picture…). We have a media not likely to report more business-as-usual if it works.
Is the difference the clear deadline on Y2K? I don’t know about you, but I, for one, find the *lack* of a definite deadline more provoking than a given deadline (even before you factor in the procrastination penalty).
Is it the lack of distinct personal risks to the individuals? Have people not read WG2? Have they heard the US military’s statements on this? (There’s a few. I can provide links if need be.) Can they see a world the rest of us can’t, where everything will turn out just fine if we ignore the problem? ( Are there unicorns? )
The key question I ask these folks (and Greg, for that matter) is: Why risk it?
Are you so convinced that the damage won’t be that bad (let’s use the IPCC report as a benchmark here, for convenience, even though it’s been accused of being too conservative) that you’d gamble on it all being a bunch of scare tactics to motivate you to act on something far less sinister, when the best evidence to support that comes from people like McIntyre?
Because, let’s face it, that’s basically what CA tells us when it posts things like this — that it’s all natural, we can continue to do nothing, go back home and sit still, it won’t be too bad. To me, given the probabilities and the risks that I see in the peer-reviewed science, that seems like a bet I’d not be willing to take, even if I were a gambling man.
Forgive the rambling. I tend to be verbose, but I hope I got the point across.
MrPete // March 4, 2008 at 6:06 am
luminous beauty,
How new do you think these requirements are? Don’t read any further before you answer with what you believe the answer is.
Do you honestly believe this is some kind of new expectation that emerged with the rise of the Internet? How old are you? :)
OK, now let’s go take a look at what we can discover.
I don’t have good library access, but I’m not too bad at searching the Internet. An interesting process… mostly showing how litigous we’ve become.
The earliest recorded PNAS website is from April 1997. The requirements existed then (with much less legalese — what is now a multipage authors’ doc was then a single simple page.) The wording expanded over the years, but even then authors were *required* to archive their data and provide accession keys at time of publication.
How about Science? Same thing, but they were a bit slow :) — their first site went online in May 1997, and… yes, the requirements were there. (OK, at that time, only “investigators” had to be given access to data. The wording was broadened in 2000 and further broadened later. That’s just legal nitpicking though. Even a 14 year old can be a scientific “investigator.”..
Sad to say, I can’t find ancient archives of Nature online. Perhaps someone else would like to do the research if they care to add to this little romp through recent history.
Luminous, I’m happy to say that data sharing as a requirement for good science practice is not a new idea. And it has been spelled out for quite some time.
Why do you presume it’s something new? Perhaps because now we can quickly zip the data across the Internet?
Maybe you’re not old enough to remember, or have forgotten, the effort involved in data sharing a decade or two ago. In 1997 we shipped data cartridges (20MB!) all over. In 1987 we shipped tapes. In 1977 I knew people who shipped punched cards. (Yechhhh! I was able to shove those out of the way as I went through industry.)
This isn’t anything new.
Greg // March 4, 2008 at 7:10 am
Explain, someone please explain, why countries like Russia and Italy experienced no Y2K problems to speak of, and spent precisely @#$@ all by comparison. A lot of people made a lot of money from the scare.
[I]BrianD: Greg’s position amounts to “Scare tactics were used by people who weren’t entirely correct then, therefore anyone using scare tactics must always be wrong”[/I]
Wrong. My position is that climatologists should be embarrassed by the scare-mongering done in their name, based on their science, and would have a lot more credibility if they helped to shoot it down and set it straight.
[I]dhogaza: So, now, why do you buy into the supposition that climate science is inherently fraudulent?[/I]
Big giant strawman. I’ve never said anything of the sort. My views are on the ‘Hansen’s Bulldog’ post, go have a look.
[I]dhogaza: Name one such lie that our host, or me, or other posters here accept in order to motivate the publics to “meet the AGW menace”.
One, just one.
State the lie. Pure and simple unambiguous English being preferred. [/I]
You are excitable. I said ‘tacit acceptance of lying’. Look up the word ‘tacit’. And you just did it, excusing Gore on Tuvalu. “That’s why the citizens of these pacific nations had all had to evacuate to New Zealand.” – wrong, wrong, wrong. Didn’t happen. A lie, or an error. And an error defended in the face of fact becomes a lie. Go to http://www.realclimate.org/index.php/archives/2007/10/convenient-untruths/ and you will see what I mean by ‘tacit acceptance’. If someone gets it wrong, the climatology community should call them on it. Just be honest, or you will lose the public in the end.
mmghosh // March 4, 2008 at 7:52 am
Mr Pete,
I published papers on my scientific research in the early and mid 1990s. There was no requirement made by any publication (in my field) to archive raw data for future audit. Most of the data was held in Excel spreadsheets and was basically a mass of numbers outputted from a machine. The raw data was averaged, again using Excel. What was published was the results of statistical tests performed on that averaged data. Looking back at the work, I see that the data files are retained, but that is is simply because they haven’t been deleted from the computer, and similarly I think they may still be there in my old department computers if no one has deleted them.
For example, for the Human Genome Project, I do not think any of the hundreds of millions of blotting test raw data is actually archived centrally in any computer. That does not invalidate the HGP science.
Reputable scientists archive the relevant data that is needed for their present work and for future analysis, and the people working in the field tend to know what to archive and what not to. Casting doubt on their archiving process is essentially accusing them of fraud. All scientists average raw data all the time.
Of course one can legitimately accuse anyone of fraud. But the correct way to do that is to take the concerned scientist to court. Alternatively, one can write a letter to a particular journal and both state and prove the fraud. If scientific fraud is proven then Nature or any other worthwhile publication will delete the concerned article.
Hank Roberts // March 4, 2008 at 8:26 am
http://nsidc.org/agdc/
If you’re a researcher, you know how to request access.
MrPete // March 4, 2008 at 1:16 pm
mmghosh,
I never said “archive for future audit.” The key is ability to replicate the analysis for whatever purpose.
I also have no reason to accuse them of fraud; something far simpler produces the same sad result of unusable (or vastly less-usable) analysis that can’t be relied on. To put it most simply, I’ll repeat something I often say to computer engineers: document your work carefully, because the person you’re documenting it for just might be you, a year or three from now.
Most of us are experts at administrative sloppiness; it only becomes willful when we attempt to hide it and are unwilling and later unable to share our data and methods. I happen to know that’s true for certain climate scientists, including some who are now deceased and can no longer even be asked to explain what they did.
The value of scientific work can be destroyed by this lack of care in the same way that computer software codes can become useless — fail to document and preserve your work, and nobody else can make use of it.
“Reputable scientists archive the relevant data that is needed for their present work and for future analysis…all scientists average raw data all the time.”
Exactly. In your field, simple low-level data averages are necessary and obvious. You’re not tossing out crucial information. You’re not losing the inputs to the transformations you perform that depend on your interpretation of the data, and which others (or even yourself in five years) might wish to modify based on their own interpretation or on improved methods.
“…the people working in the field tend to know what to archive and what not to. Casting doubt on their archiving process is essentially accusing them of fraud.”
Hardly. How many scientists have important data and method papers lying about their office, not yet filed, let alone not yet archived? :-) That’s not fraud.
You might like to refine the following to be more realistic, but here’s a straw man stab at some categories. This is stream-of-thought stuff; someone can organize it better if they like:
Clean heart, clean office: (Empty mind? :) ). How many working scientists maintain a workspace that has no clutter in their office, their computer, and everything’s always filed or reported perfectly? I’d call this someone with too much time (or money, or help) on their hands. (I know a person like this, but not a scientist. He even cleaned the chrome under his sinks!)
Best practice: Works with humility, knowing that future collaborators, even they themselves, may falsify some aspect of the work. Archives data needed to replicate or falsify each important step of the work, and does so in conjunction with publication if not before.
Average: doin’ ok. Got a few publications in process. Wants to get it right, does good work, publishes regularly, complies with guidelines more or less on time. Sure, could spend time on niggly details but not worth it in most cases. Besides, their work is only seen by a small circle of colleagues and has little influence beyond that circle, at least for the next few years.
Over confident: colleagues love their work, they get writeups in major magazines and newspapers, politicians quote them regularly. More publications means more grants, more grad students, more influence. Why be concerned about the future? “Obviously, a success.” — Easily leads to sloppiness in documenting and archiving their work. Too much important stuff going on to address such niceities. (Until they’re hit by a car, their hard drive blows up with no backup, a grad student finds a crucial error in their results, and nobody can repair the damage because the foundation blocks were never saved.)
Defensive: Used to be over confident, until some people began to ask significant questions that the scientist couldn’t answer. Doesn’t have the time to go check on the details. Can’t be bothered. Stop bugging me, I’ll get around to it. Things like that.
Offensive: Silent treatment. Who are you to ask me something like that. You have no right to ask that. I’ll tell you when I’m good and ready. Obviously you’re incapable of understanding my breakthrough analysis technique.
Fraud: If we just toss out these kinds of data, the results will look exactly the way our financial sponsors want. And then we’ll have the funds we need to hire more people, do that big cool experiment, and save the world. I’m not concerned about what others will pay for this; I’m right, they’re wrong, and they deserve to pay!
(For onlookers: fraud is not at all the same as error: the difference is intent. Legally, it involves intentionally false representation of a past or present face, action relying on that representation, and damage suffered as a result.)
I’d guess that most scientists fall in the “average” category, perhaps slightly “over confident.” I’m sensing that a few are solidly “over confident” verging on “defensive” or possibly “offensive.” Hard to know without really knowing them… which will be my bottom line point:
These kinds of evaluations are actually impossible to judge well in an online environment. Screen to screen, we are who we are willing to reveal. It’s also true face to face, but much more difficult to hide your reality from people who spend a lot of face to face time with you. All we really can judge here are words, actions, consequences, etc. I believe intent is too subtle to evaluate online.
My tome for the day, now time for the Real World :)
Barton Paul Levenson // March 4, 2008 at 2:03 pm
Greg writes:
[[Anyone with a pair of eyeballs and a functioning brainstem knows that Y2K was not one-fiftieth of the problem it was beaten up to be. That people here can’t admit that speaks volumes about the kind of alarmism we’re dealing with.]]
No, Greg. Y2K was not a problem because people spent billions of dollars and millions of man-hours fixing the code before the crisis happened. Your statement is like saying “Well, the house at 10050 Cielo Drive never burned down, so there was never a danger from forest fires.” There was, but a bunch of guys working 12- and 14-hour days dug a firebreak protecting the neighborhood.
dhogaza // March 4, 2008 at 2:08 pm
And, of course, you have no proof that the combined archiving of ice core samples and processed data is not sufficient to do this.
If MacIntyre had the raw data, he’d then be claiming that it’s wrong and that the ice cores need to be accessible to regenerate that.
And if that didn’t overturn the science, he’d demand new cores.
steven mosher // March 4, 2008 at 2:18 pm
Its funny a while back I made a comment on RC that it is SOP to tweak observation data and bunches of folks descended to say that I had accused all scientists of fraud. raw data gets filtered, rounded,adjusted,Qced for outliers. This is all routine. The point is it should be completely documented. so third parties can check the work, or build on the work, or correct the work.
Hank Roberts // March 4, 2008 at 2:43 pm
> replicate the analysis
Drill another core and measure that. That’s what replicating scientific work means.
To get your hands on the untouched archived half-cores in the freezer you have got to be _really_ well funded and professional. That material is literally irreplaceable.
If you doubt the researcher’s work on the half that was taken and measured, seeing the paper copy of the notes won’t satisfy you. But if you get your hands on the remaining ice — you better be really good because you’re using up a very expensive data source.
This is why you’re expected to get funding to cover part of the cost of acquiring it, if you can’t get funding to acquire a new core.
Amateur reader’s opinion here, that’s how I read it at the ice core archive pages. You can read it yourself.
mmghosh // March 4, 2008 at 3:16 pm
Mr Pete
Your comments are valid to some extent. However scientific work is not like developing software programs.
Scientists work collaboratively in a group. this is not because they wish to hide data or collectively defraud their peers, as was insinuated by Dr Wegman in his report. There is an extensive network of PhDs and postdocs who tend to discuss data, data collection and data archiving pretty radically and continuously among themselves.
Also the people we are talking about are some of the best scientists in their own fields. they do not get there by defrauding their main reading public, namely their peers.
In the academic world there is keen competition for grants, getting the best students etc. Anyone who kept fraudulent data pretty soon gets found out. I trust PhD students and post-docs, frankly to catch errors much more than any other group in my experience.
I do not know what the standards of data recording and maintenance for Dr Thompson are. I would tend to believe that he would archive his data according to the best practices of his time – in general if I have to hear credible criticism of him, I would give it more credence if it came from his peers in the field, rather than Mr McIntyre.
In my experience, good scientists, and by those I mean who have the respect of their peers, are not in the least sloppy – rather they tend to ne methodical and careful. And the best criteria for good science is repeatability. Dr Thompson’s work will be repeated by some other scientist. If his data and analysis are correct then his work will stand up. In my experience, scientists are much more concerned about this kind of verification of their work, not “audit”performed by amateurs.
fred // March 4, 2008 at 3:56 pm
Martin, 74% of the stations have had some changes to some of their readings. The amount by which the readings have been changed varies from year to year and station to station. I do not know what proportion of readings have changed. It might be higher, lower or the same. The thing that is certain is that it is not linear. That is, we are not changing the reading on one station by the same amount for its whole career. We are changing different readings from the same station by different amounts in different years, and we are changing different stations by different amounts.
It doesn’t seem like your case – where we know for sure that the triangle must have 180 degrees and so the readings must add up to that. In the present case its a bit more like we get several hundred observers to take readings every so often. Then we make changes to some but not all of those readings. And we also exclude some observers. But we do this without any real knowledge that their readings are wrong. They just do not agree with some other observers readings. We don’t have any other evidence to suggest that back in March 1917, it was really on average 3 degrees warmer or cooler than the guys at the time wrote down. Just that some other guys within 1,000km wrote down something different. Well, maybe so….
Maybe there are lots of other cases in science where we do this sort of thing. It would be interesting to know how, if we do that, we take account of what we’ve done in progressing into further statistical analysis.
I can understand correcting for an instrument which reads high. Or for an instrument even which we have shown by calibration reads both too high at highs and too low at lows. After all this is what you sort of do when you mark up the dials. I can’t understand what is actually being done however. And I particularly cannot understand how the process can deliver results accurate to tenths of a degree for the whole planet. It seems likely we need fewer better stations, and a lot less changes to their readings.
John Cross // March 4, 2008 at 3:56 pm
Greg: A couple of points. I agree that the climate scientists have more creditability if they help to set the science straight which is one of the reasons I like RealClimate. Read this post from a while ago. They could have accepted the 10C rise or just not commented on it, but they went out of their way to provide what they see as the issues and problems and state that they don’t think it reasonable.
In regards to your Tuvalu example, it is not as clear cut as you would like to believe. Let me quote the context from the very link you post.
“In the movie there is only one line that referred to this: “That’s why the citizens of these pacific nations have all had to evacuate to New Zealand”, which is out of context in the passage it’s in, but could be said to only be a little ahead of it’s time.
I really don’t see how you can equate tacit approval with the phrase “out of context”
Regards,
John
luminous beauty // March 4, 2008 at 3:59 pm
MrPete,
Blah. Blah. Blah.
Promoting quality practice is good. Using it as a blunt instrument to throw around presumptive and unfounded suspicions of deception directed at individuals, not so good.
Bottom line. Thompson has archived his data, online, in form suitable to do analysis. Not sufficient to check his arithmetic on the presumption that he is hiding something or is incapable of correctly computing a running average, though. Too bad. If you needed the raw data in order to do original and meaningful research, I’m willing to bet he would be quite helpful, even if you were only 14. Demanding him to prove he isn’t still beating his wife is not in that category.
Bye the bye, Gore has admitted and corrected mis-perceptions he generated vis~a~vis the gradual and on-going emigration from Tuvalu, not all melting ofKilimanjaro’s glaciers due to GW, etc., and natural scientists have been critical of his slight errors, minor mis-statements and trivial exaggerations, none of which lead to conclusions at odds with his thesis.
http://www.lomborg-errors.dk/Goreacknowledgederrors.htm
I haven’t seen the current version of Gore’s presentation, but if he hasn’t changed from using fig. 7d to fig. 7c from Thompson, et al., 2003, then he truly is an idiot, but just making such a goof only shows he is human.
Gore is willing to accept genuine criticism. Climate scientists are willing to change their views as the data dictates, but the CAuditors seem incapable of comprehending, much less admitting, that the nits they pick have minimal significance, and do not add up to anything resembling either malfeasance or incompetence, much less the falsification of AGW theory. It is just an ad hom rhetorical distraction.
An even tempered skepticism is a good thing, but it needs be an informed skepticism. When one begins with the presumption one is being deceived and looks at every new piece of information one learns about such a complex subject as the terrestrial climate though the lens of thinking it must be wrong, an unbalanced skepticism quickly devolves into cynicism on it’s way to becoming paranoia.
Lee // March 4, 2008 at 4:51 pm
Mr. Pete.
“The earliest recorded PNAS website is from April 1997. … even then authors were *required* to archive their data and provide accession keys at time of publication.
Mr Pete, I published in PNAS in that time frame. I archived my data. My archived sequence data consisted of the base ‘calls’ – not a jot of the raw or original sequencing data got archived. The graphic output from the sequencing machines probably still exists in my lab notebooks, but I cant guarantee that. It certainly isn’t in any archive, nor was it ever expected to be.
My PNAS paper also included analytic data for a set of very messy biological assays. My report of the results consisted of mean, variance, and statististical tests. The raw data was recorded, digitized into a computer, marked, measured, collated, and then analyzed – and not a jot of that raw or intermediate data got archived, nor was it required to be. The process was described in the paper , along with the summary results.
This was also true of papers I published in that same time frame in Genetics and J Neuroscience.
I think you have an awfully idealized picture of what kinds of data got archived.
Lee // March 4, 2008 at 4:59 pm
fred says:
“It seems likely we need fewer better stations, and a lot less changes to their readings.”
On more freaking time, THIS IS HISTORICAL DATA!!!!!!!!!!! It is what we’ve got. And the corrected analyzed results of that data are pretty damn good – comparison with satellite data in the period and regions of overlap, looking at annual variations shows that, yes, it gets it right within a tenth or two. For trend, also they agree within a tenth or two . You may not understand how they can possible be getting that accuracy from these stations – but they DO.
Deech56 // March 4, 2008 at 6:19 pm
RE: Lee // March 4, 2008 at 4:51 pm
Thank you for injecting some realism into the debate. One gets the feeling that the “auditors” have never collected primary data, and the examples that are in the PNAS instructions are the types of data (sequences and the like) for which central repositories have been established.
My data from the late 1990s is hand written (antibody titers and spirochete counts under a scope) and archived in lab notebooks at a company. The chances of all of that being public is about nil, but there is plenty of information in the manuscripts that were generated, and in the sequences that were publicly archived, to replicate the studies, which one of our colleagues did.
The basic idea for providing information about the studies is for replication, which we can all agree is part of the way science is done.
MrPete // March 4, 2008 at 6:46 pm
Stopping in briefly, I only have a couple of comments that might contribute to the conversation.
1. With respect to sloppiness, I’ve seen an enduring difference of perspective in many if not most professionals between being meticulous about the data I’m “focused” on vs the data that’s less important to me right now. I agree that much scientific work doesn’t require a painstaking level of data archival and access.
However, when doing work that’s novel, with novel analysis techniques, and doing work with large immediate societal implications, what’s “known to be appropriate” changes. Jones, Thompson, etc are involved in such novel work. The statistical analysis techniques used are often not run of the mill techniques nor are the assumptions based on well developed underpinnings.
As Lee (?) said, it does come down to questions of professional judgment. Perhaps what many fail to recognize is that climate science, like so many fields, is more multidisciplinary than we think. We tend to presume that our experience, methods, practices are sufficient and applicable to new situations. And that’s when outsiders with a fresh perspective can bring something helpful to the table. Tough getting that conversation going in a constructive way, isn’t it! Especially when the “field” is pre-charged with societal and financial implications.
2. There seems to be a perception that Steve M is an amateur… to wit, “audit”performed by amateurs.
Steve’s credentials and analytical work as a professional statistician have not been questioned, and have been confirmed by Wegman (who I would hope is a statistician with solid credentials, else why does he hold his position.) Steve’s “audit” work is well within his area of specialty… hardly “amateur.”
Not trying to drag up the academy vs business arguments. And I trust nobody is saying that anyone outside of the academy is an amateur. Let’s at least have that much respect for one another.
I’m saddened but not really surprised by the “circle the wagons” perspective that says Steve is some kind of amateur nutcase picking at nits to try to destroy the solid work of respected professionals. To get past that divide will require a lot more give and take and face to face interaction. Steve’s a real person, intelligent, with integrity. Yes, “real” includes the fact that he has “buttons” that can be pushed although he’s learning not to react as much to the “sticks and stones” hurled in his direction.
This conversation began with a question about appreciation for those who are following in the good tradition of attempting to improve the science and shed light on weak work. We went through a time of good, high level discussion.
Clearly we’re heading toward a level of detail that is hard to sustain online (because we’re not seeing eye to eye and I don’t have the time/energy/interest capacity to dig into your strengths at a detail level, and you don’t have the time/energy/interest capacity to dig into Steve’s strengths at a detail level. Me, I have no strengths. Oh — in case someone wonders, I’m definitely NOT here to “represent” Steve or anyone else. I’ve not told others what I’m doing here, nor have they said anything to me. I’m an independent person acting independently. :) )
One of these days, we’ll have to sit down face to face to take this further. For now, I hope the conversation we’ve had has been helpful to at least get some of us thinking more clearly about some of the issues on the table. Thanks for the opportunity!
MrPete
John Cross // March 4, 2008 at 7:14 pm
Lee and Deech56: You have caused me to think back and now that you mention it, a great deal of my data can be found on 5 1/4 disks (for the older stuff) and on a 3/4 inch magnetic tape (and I think the computer it was written on is long gone).
Time and technology moves on
John
dhogaza // March 4, 2008 at 7:54 pm
I hope you didn’t hurt yourself moving those goalposts such a long distance, Mr. Pete.
dhogaza // March 4, 2008 at 7:58 pm
I don’t think “nutcase” is the description I’d use for McIntyre. His goal is clear and his methods dishonorable, but there’s nothing nutty about his approach. He wants to convince the world that climate science’s conclusions regarding AGW are a result of sloppiness and outright fraud.
harold // March 4, 2008 at 8:55 pm
No dhogaza there is no fraud, just a disinterest in detail and a lot of arrogance.
Ian Forrester // March 4, 2008 at 9:07 pm
Mr Pete said:
“Steve’s “audit” work is well within his area of specialty… hardly “amateur.”
Good grief he is a business man. He is more closely related to Enron and what happened there than the study of scientific data.
Heck, I know what two plus two equals but I could never audit a company’s financial activities. It is a joke to even suggest that SM knows anything about science let alone climate science. Get real.
Armagh Geddon // March 4, 2008 at 9:11 pm
“He wants to convince the world that climate science’s conclusions regarding AGW are a result of sloppiness and outright fraud.”
Lets say that your statement is true (I happen to think that it is not). Climate Scientists have a choice as to how they respond. They can engage in a fruitful discussion, explaining why they have done what they have done, and demonstrate openness and transparency.
Or, they could stonewall, fling ad hominems, obfuscate, and generally fail to provide a credible account in the face of the questions being asked.
The risk that climate scientists face is that the more they engage in the latter approach rather than the former, the more will they lose credibility in the eyes of the public. It seems to me, perhaps not a dispassionate observer, that there already has been a significant loss of credibility, due primarily to the response strategies chosen.
MrPete // March 4, 2008 at 9:25 pm
dhogaza, I’m all about pragmatic and reasonable. My goalpost is good science, by which we actually learn something (that won’t be refuted in a few years) and can pass it on to others. Poor science is science that later must be retracted when it is found to be non-replicable, incorrect, etc etc.
No need to re-prove that 2+2=4 every time.
OTOH, some want to believe that the work behind IPCC’s reports is likewise “settled” and not in need of significant refinement and rework. That the data, methods and conclusions are more or less rote and not likely to produce different results if subjected to careful analysis and correction.
Steve, as an expert in such things, senses there’s a LOT of rework needed. I have similar expertise in info/tech-related process. I sense a LOT of rework needed.
It can be demonstrated mathematically as well. There are plenty of “don’t go there” topics for the GCM folks. Example: a realistic CI around the GCM’s etc would show how little we really know.
As noted above, I don’t sense you want to actually discuss such things. You want the current “consensus” to be accepted and acted on. I’m certainly not the one who will influence you.
Martin // March 4, 2008 at 9:36 pm
But fred, it does deliver just that — compare with Hadley and the satellite results. And I do understand how and why it does. As does tamino/HB and undoubtedly many others.
This is not a claim that Hansen et al. are infallible. Only that whatever flaws there may still hide in their processing method, none of them are allowed to distort the result by more than — just what you mention, 0.1 – 0.2 degrees per century or so. It is this control of error propagation that makes the results as good as they are, not any attempt at “fixing” the data.
It’s a bit of a challenge to transfer this insight to people not backgrounded in this type of geospatial statistics. It’s frustrating, I wish it were different, but it cannot be helped. Still, fred, I’m willing to give it a try if you’re worth it :-)
Phil. // March 4, 2008 at 10:00 pm
Re the Tuvalu situation as I understand it 16% have left for NZ already. Other islands are considering similar moves:
http://www.countercurrents.org/cc-marks271006.htm
MrPete // March 4, 2008 at 10:07 pm
Deech56, Lee, Hank: I missed this comment… One gets the feeling that the “auditors” have never collected primary data, and the examples that are in the PNAS instructions are the types of data (sequences and the like) for which central repositories have been established.
No reason to fret over this. CA folk have had some fun collecting primary data. And Steve’s quite familiar with the repositories. Quite.
Many who voted for CA as “Best science blog of 2007″ explicitly said they did so because it’s not just a bunch of armchair gunslingers, but people who dig in and do some real work, whether in the field, or computer/statistical analysis.
CA is serious about good science. If there were a way to avoid all political/policy/etc discussion, I’d love it. Let’s learn something real without all that hanging over our heads!
Hank Roberts // March 4, 2008 at 10:12 pm
Look, people change. If you take seriously the ocean pH physical chemistry numbers — even if you don’t believe in the atmospheric physics numbers — then we have the moral responsibility to find common ground with everyone involved and not waste time tripping one another up.
Give the man credit for trying to present a face that works in a scientific environment, as he does here:
http://www.climateaudit.org/?p=2708
He will always provoke. But the nice thing about personalities is, like standards, we have so many of them. As he says there — as he presents himself there — he’s acknowledging the need to change how the blog operates.
I don’t want to see this one attract the people no longer welcome to stir up shit at CA, I don’t want them to move here and keep doing the same thing.
Work with the aspects of everyone that could make the world better.
If we lose this one and the oceans go sour, there will be no need to hold trials. We’ll all have been on trial and failed together.
Hansen's Bulldog // March 4, 2008 at 10:17 pm
For the record, I’ll state my opinion. Steve McIntyre is not naive about numerical analysis, as are so many of the usual denialists. I can’t *imagine* him making such an embarrassing mistake as recently done by Anthony Watts.
I believe he’s made fundamental mistakes on several issues (the upcoming installment on PCA will address some) and that he’s motivated by ideology rather than science, but I disagree with claims that he’s nothing more than a statistical ignoramus.
null{} // March 4, 2008 at 10:32 pm
Martin // March 4, 2008 at 9:36 pm
This is very strange:
“Only that whatever flaws there may still hide in their processing method, none of them are allowed to distort the result by more than — just what you mention, 0.1 – 0.2 degrees per century or so. It is this control of error propagation that makes the results as good as they are, not any attempt at “fixing” the data.”
If the flaws are hidden, how can the effects possibly be known?
Do you have documentation or coding examples of ‘control of error propagation’?
Why are the documentation and coding such a black box that the processes cannot be known?
mmghosh // March 4, 2008 at 10:44 pm
Mr Pete
I apologise for giving the wrong impression about Mr McIntyre – I did not suggest he was the amateur in question, I was referring to other amateurs in the field who post on several blogs. I accept Mr McIntyre’s expertise in his own field of mathematical analysis, which has been published in the literature.
However, I feel strongly that statistical expertise alone, and in itself, does not confer professional validity in a particular field. Statistical analysis of data is the last stage of a scientific process. Until one has a complete understanding of, and experience in the science itself, statistical expertise itself is not enough to make one a professional in a field.
I have read Dr Wegman’s report where, in various places, it has been suggested that Dr Mann’s data analysis is unacceptable because it has not been validated by a professional statistician.
Surely this is an incorrect contention. Most leading scientists are pretty good (amateur) statisticians in their own fields, and few scientific papers, in my experience, are validated by professional statisticians. In fact, many scientists I know write their own software to do their own specialised statistical analysis. This does not invalidate their science.
JCH // March 4, 2008 at 11:03 pm
Who was Einstein’s professional statistician?
When I read that part of Wegman’s report my warning bell for “federal dollars for unemployed statisticians” started ringing.
TCO // March 5, 2008 at 12:58 am
JCH: Don’t get too exercised. Some of what Wegman is saying goes back to a philosophical article of Hotelling (I hope we can ALL agree that he was a mensch.) There is a valid point of view that says that people can do their own statistics, etc. There is also experience with problems related to that. My personal opinion is that people who are inherently mathematical (physicist, etc.) should feel free to work on new statistical methods. However, that it is useful if these new methods are validated and are looked at theoretically. Heck…they might have usefulness in other fields! Also, that some mistakes might be found.
But there is a problem when new methods are covered as part of a problem but are not covered in the primary methods hournals. If I use a new method of mathematical crystallography as part of a study in Ebola virus formations…I may create problems. The journal that I submit to will probably get a single crystallographer to be a reviewer, but that will be it. We won’t have length to discuss the theoretical aspects of the crystallography in the Virus journal. And we won’t get good scrutiny of the method (or even transfer of it’s use to ther fields.) Also, it would be really GOOD to examine a new method of crystallography on a KNOWN structure. (You can tell…I seen this one play out for real…)
Just two cents. Well…maybe 4.
MrPete // March 5, 2008 at 1:14 am
mmghosh, I appreciate what you said; thanks!
I would suggest a tiny but significant modification in the logic of one of your statements:
“suggested that Dr Mann’s data analysis is unacceptable because it has not been validated by a professional statistician.”
AFAIK, those are two separate statements:
* Dr Mann’s data analysis is unacceptable (as proven statistically)
and
* It has not been validated by a (qualified) statistician.
Those who have expertise in cryptanalysis perhaps can relate via this analogy: it’s not easy to develop new, valid, cryptographic methods. If you’re not prepared to subject a new method to rigorous analysis it’s much better to lean on expertly-approved methods– your “new” method most likely has disastrous hidden flaws.
AFAIK, the complaint about Mann’s analysis is essentially that it uses “novel” statistical techniques, which the experts have ultimately demonstrated to be invalid. And, as with cryptography, the experts strongly encouraged Mann to use known, approved statistical methods.
dhogaza // March 5, 2008 at 1:32 am
HB, I thought you were going to excise posts that raise the hockey stick until you finish that post of yours.
But, since it’s been dragged in here again…
No, not invalid, and the NAS report specifically said so. It’s not the method preferred by Wegman, but “not the best analysis” is not the same as “a wrong analysis”.
[Response: Oops! You're right. Those who MUST discuss the issue, please wait, it won't be long.]
Hank Roberts // March 5, 2008 at 1:41 am
Look up what was actually said.
You don’t even have to read, just watch and listen. And then tell the truth about what’s said:
Monday, September 04, 2006
North on the hockey stick
My colleague Jerry North was the chair of the National Academy panel that investigated the hockey stick. Last week he gave an interesting seminar to our department about the experience. You can view the seminar here.
http://www.met.tamu.edu/people/faculty/dessler/NorthH264.mp4
[Techincal details: it's a 40 MB file, so it'll take some bandwidth. It's in mp4 format --- if you have a recent version of quicktime on your computer, you should be able to view this. It runs just over an hour, so grab some popcorn and enjoy!]
http://sciencepoliticsclimatechange.blogspot.com/2006/09/north-on-hockey-stick.html
null{} // March 5, 2008 at 1:56 am
Hansen’s Bulldog // March 4, 2008 at 10:17 pm
said:
“I believe … and that he’s motivated by ideology rather than science, …”
What are your objective criteria and measurable metrics for arriving at this belief? Did you measure both this blog and CA?
Hank Roberts // March 5, 2008 at 2:22 am
Here’s how to tell them apart:
> ideology
Posts opinions, turns only one direction
> science
Publishes in refereed journals, cites other scientists’ work, gets cited by other scientists, contributes to new knowledge.
Here’s the state of things as I see it, based primarily on the straightforward ocean chemistry changes, which is clear, inarguable, and critical within the century. The ocean pH change from increased CO2 is vitally important to address, even without considering the more complicated atmospheric physics.
I’m quoting Tom Athanasiou:
————-excerpt————
“… if you accept the analysis here, and I do, then we need a major reduction in the atmospheric carbon concentration from its current level. The only real issue is how fast we can do it. We are, in the lingo, facing a “peak and decline” future in which the rich and the poor, somehow, someway, work together to bring the concentration far below where it is today.
There are a million variables in all this. You ask me to specify an necessary emissions reduction and a time period. But if you want a very simple story, it’s probably better to think in terms of specifying a peak year and a subsequent rate of decline.
This is very different from how things have been thought about in the past.
In terms of the “sharing the world” problem, also known as the “what will be left for the South?” problem, well it’s the key.
—————-end excerpt————
From: http://gristmill.grist.org/story/2008/2/5/172318/6361
Picture this:
http://www.globalwarmingart.com/images/thumb/a/a3/Carbon_Stabilization_Scenarios.png/300px-Carbon_Stabilization_Scenarios.png
If you’re reading this and you still don’t know at which level the ocean chemistry changes beyond anything tolerable to life as we know it, you haven’t been paying attention. Look it up.
Hank Roberts // March 5, 2008 at 2:26 am
Dang, I’m being arch again. Let me be more helpful:
http://pangea.stanford.edu/research/Oceans/GES205/Caldeira_Science_Anthropogenic%20Carbon%20and%20ocean%20pH.pdf
“Anthropogenic carbon and ocean pH
The coming centuries may see more ocean acidification than the past 300 million years.”
TCO // March 5, 2008 at 2:55 am
The failure of McI to finish analyses, to publish in journals or AT LEAST to publish white papers is SHAMEFUL. Stuff like that makes me worry about skepticism. It’s wierd but he has this tendancy to want to have his cake and eat it too. He says that his blog is “just a scratchpad”, but then he calls himself an important player in the field and implicitly and occasionaly explicitly says that people should be aware of his arguments on the blogs. But he does half assed stuff. doesn’t finish stuff. Doesn’t quantify things. Etc. etc.
However, he does raise some interesting questions, does do some good reading of papers and playing with data sets. Just a shame that he argues in a confused mishmash, with occasional sophistry and not finishing things.
TCO // March 5, 2008 at 3:22 am
The North discussion is very “in crowd” in terms of the assumption that all good people are Democrats and are AGW beleivers. Also, I’m not that blown away by the level of work that North did. He decided to change what he answered. Instead of really pushing home and inspecting the micro-problem (Mann paper), he decided to change the question and do a general paleo review. By doing that, he flubbed on going deep on Mann.
North concentrates on the debatible and extraneous points that Wegman brought up in terms of social networking and advisability of using real statisticians…but he is unfair in not discussing the substantial portion of the report where WEgman discusses Mannian PCA itself (an area where Wegman actually goes in more detail than North).
To be honest, while I think there were a lot of issues with Wegman’s report and I don’t be all McIianisms, North (listening to him) does not give me the impression either of a great intellect OR of a diligent investigator. He’s not Feynmanesque.
Also, in the discussion…he says that statistics is too abstract and does not even cover it. That’s a flop.
Hank Roberts // March 5, 2008 at 3:31 am
Hey, look — it could be it’s the best he can do. Judith Curry commented that there were a couple of people in the CA group who understood statistics, a while back.
Not all ore is high yield.
TCO // March 5, 2008 at 3:45 am
He can do better. He is a little slack, hanging out not working, etc. Also, he is a little dishonest in refusal to admit fault, answer questions, be sophistic. But he’s not all bad. I sometimes wonder if there is something a bit twisted in his head, in all seriousness.
JCH // March 5, 2008 at 3:56 am
TCO,
I suppose today Albert might need a professional statistician. I’m leery of a situation where statisticians/auditors could act as gatekeepers of science.
TCO // March 5, 2008 at 5:00 am
I don’t think that “gatekeepership” is being proposed. The point is really a philosophical one about disciplines…and it would be useful if you had read the very classic Hotelling essay. Have you?
MrPete // March 5, 2008 at 5:14 am
“I’m leery of a situation where statisticians/auditors could act as gatekeepers of science.”
What do you do when more and more science is computer analysis work rather than lab work?
As a computer “expert” I too am leery. There are a million ways to get it wrong, and not many ways to get them right.
More professionals would do well to heed Brian Kernhigan’s Law (he’s co-inventor of the ‘C’ language, among other things): “Debugging is twice as hard as writing the program, so if you write the program as cleverly as you can, by definition, you won’t be clever enough to debug it.”
Martin // March 5, 2008 at 5:15 am
null{}:
Actually they are not. That’s an urban legend. Go look at the GISTEMP page.
If they’re a black box to you, that would appear to reflect more on your abilities/background in this field (no insult intended; we all have our different skill sets) than on the matter itself.
MrPete // March 5, 2008 at 5:21 am
Hank — so, what’s the CI on this work? the graphs and short comm’s have no indication.
mmghosh // March 5, 2008 at 5:53 am
JCH
You are correct about Einstein. But seriously, if any scientist is about to start something new, they usually get a statistician’s opinion to vet my proposed analysis *before* they start. Certainly I do, and all the people I know do too.
Otherwise you look really stupid when a reviewer either writes “this author should have checked his basic statistical method” or letters start in the journal about how the stats are incorrect. Not to mention the time and funding wasted in redoing the experiments or increasing the number of observations.
As I said previously, not to damage one’s reputation is a powerful corrective in scientific research.
mmghosh // March 5, 2008 at 6:13 am
Mr Pete
Mr McIntyre may be perfectly justified in his critique of Dr Mann’s choice of statistical method and the conclusions derived as a result. Having followed the debates on CA for some time, I believe his techniques are being examined by climate scientists.
But that is not the point. The point is whether that analysis was appropriate for that particular study. I believe that the majority of climate scientists accept that it is appropriate. Not being an expert in the field, I would tend to trust the scientists in this situation.
The correct criticism, in my view, would come from a combination of climate scientists and statisticians.
If, at any point, the majority of climate scientists feel that Dr Mann’s analysis is either incorrect or inappropriate then I am pretty sure that it will be dropped from the literature cites.
MrPete // March 5, 2008 at 6:38 am
Martin et al, “It is this control of error propagation that makes the results as good as they are, not any attempt at “fixing” the data.”
When dealing with systems where extensive calculation looping is used to operate on data sets, it is amazingly tricky to ensure your calculations are actually correct. I’m not saying anything new here. I *am* saying that I’m cautious about presuming the calculations are correct.
Here’s a simple demo that anyone can do in Excel. Some (many? all?) will quickly recognize the computational error I’m touching to generate this effect. There are quite a lot of these, depending on the kind of analysis one does.
a) Fill sheet1 with 100 columns (A through CV) and 1000 rows with: =ROUND(10*RAND(),1) [i.e., a number between 0 and 10, rounded to tenths]
b) in sheet2 (second tab), fill a1 with =round(sheet1!a1,0) [i.e. the value in the first sheet, rounded to an integer]
c) copy that formula to fill 100 columns and 1000 rows (same as sheet1)
Now you have 100 columns of 1000 random values in tenths, and the same rounded to integers in the second sheet
A few more steps on sheet1 and we’re done:
d) fill a1001 with =ROUND(AVERAGE(A1:A1000),1)
e) fill a1002 with
=ROUND(AVERAGE(Sheet2!A1:Sheet2!A1000),1)
f) copy those two cells to fill rows 10001 and 1002 out to all 100 columns.
Now you have the average of each column in both sheets, rounded to tenths.
Let’s find what kind of errors have accumulated from all this rounding.
g) fill a1003 with =A1002-A1001
h) copy a1003 to fill row 1003 out to 100 columns
i) set a1004 to sum(a1003:cv1003)
Voila. You have the sum of the differences between a hundred 1000-number data sets rounded before averaging or maintained to tenths while averaging.
Many would think all this rounding is no big deal — should just “average out”. [Lest one presumes so much rounding never happens, consider the frequent use of text/csv files to store intermediate results...]
Anyone surprised? The more rounded ’sheet2′ has an upward bias.
No, I didn’t cheat. If you don’t know why this happens, read up on rounding algorithms. Don’t blame Microsoft, by the way. By default, most systems round with a bias. Even in 2008.
As noted above, this is only one of many ways to get analysis wrong. Non-computer folk tend not to know such subtleties.
fred // March 5, 2008 at 7:54 am
Lee, I think I understand very well what they are doing. Whether it results in reliable data is an open question. What exactly was the temperature in July 1917 at a given station in Texas? Was it what is written in the record? Or is it higher or lower by 3 degrees? And how do we know?
Of course one realizes that date is in the past. That is not a reason for changing the recorded data to something that was not recorded.
I continue to feel we need to use fewer stations, but with more consistently rural siting and environmental factors, and with good station histories, and no changes to the data. What objection can there be? It may not make any difference, but it may. If it doesn’t, we would have found something out.
fred // March 5, 2008 at 7:59 am
Ian Forrester writes:
“Good grief he is a business man. He is more closely related to Enron and what happened there than the study of scientific data.”
As far as I am aware, Steve M has never been accused of any sort of financial irregularity of the sort that took place at Enron. Even had he been, it would make no difference to the accuracy of his contentions on climate.
What on earth is this stuff about? This issue is about the science of climate. Its not about Enron, Bush, Oil, God, Creationism, Evolution, Tobacco….all these other crazed hobby horses you guys keep going on about. I just do not understand why you are all in a continual state of excitement about irrelevancies. Isn’t climate difficult enough?
Martin // March 5, 2008 at 8:20 am
null{0}:
What is known is not the effects, but the largest possible effects of remaining errors. For many types of errors (gross, random, per-station systematic) this can be analyzed (and I suppose I should write up something on this sometime).
For unknown for-all-stations systematic errors it is obviously tricky, and your scepticism appropriate. Then you do external comparison. Like the one with Hadley, which shows that, whatever may be wrong with either of those processing methods, the total effect is less than their stated precision, some tenths of a degree per century.
GISS and Hadley of course use the same surface data, so any systematics in there will go into both. That’s why it’s good also to compare with satellite results, which are based on different data sets. Scientists never stop intercomparing, they never really believe they have “caught” all of the error sources… but in this case they appear to have.
john // March 5, 2008 at 8:14 pm
Dear Mr Pete,
I like you, you don’t give up!
How much code did you change, the kidney dialysis machine had two digit date sensitive software code, well that is a surprise!
And I never said Climate Science was wrong,
I was just saying in a perhaps not too elegant way, that Science, in my opinion, is no better than any other area of human endeavour, we are human beings.
Science usually challenges our inherent primitive human beliefs, in the case of climate change it supports our primitive belief, we have always believed our actions affect the climate of the planet, and sacrifice has to be made.
MrPete // March 5, 2008 at 11:59 pm
john, I went back to check my own footnotes.
Unfortunately, the primary reference material was web-based and even the web archives don’t have the pages. But according to my notes, 2% of kidney dialysis products required an update/upgrade, or they would stop performing automated schedules and would only function if manually programmed.
Me? I didn’t do any y2k programming. For that project, I was a (reluctant) technical analyst, strategist and advisor to leaders in various places. (It was a lose-lose situation: no pay, lots of spitballs, and in the best case situation people would invest some amount of resources so that nothing would happen.)
Sorry, I think we need to be done with y2k. I was tired of it in 1998 :)
Hank Roberts // March 6, 2008 at 12:15 am
“People ask me to predict the future, when all I want to do is prevent it.” — Ray Bradbury
Ian Forrester // March 6, 2008 at 12:47 am
Fred, you didn’t see my obvious criticism of SM.
In science one needs to knowe what the numbers mean and how they are related scientifically before agreeing on a statistical
method to use.
How can a business man know anything as complicated as that without taking the required courses and spending a lot of time (years) working with that sort of data?
That is why what he is doing is not accepted by scientists.
MrPete // March 6, 2008 at 2:30 am
Ian, professional number crunchers get a “nose” for how data works. All kinds of data, used for all kinds of purposes.
Likewise, I’ve got a bit of experience with computer technology. I have a “nose” for systems that are done well or poorly, and a sense of what is likely to be wrong with any given system–even though I may know nothing about the specific application of the system.
An equally valid question a lot of experienced gardeners in Colorado are now asking: how can dendroclimatologists do valid statistical analysis of Almagre (or Sheep mtn, etc) stripbark BCP growth without understanding the limiting growth factors in the region? Without doing any of the field work themselves? Some highly visible climate scientists have no background in plant biology let alone horicultural expertise in the mountain regions. And their results are literally laughable even to local gardeners and horticulturalists.
No degrees necessary. Just green thumbs.
Ian Forrester // March 6, 2008 at 3:37 am
Seems to me there is an old saying that the less some one knows about a subject the more they think they know.
That applies to quite a few here including Fred and MrPete.
Hank Roberts // March 6, 2008 at 3:49 am
> their results are literally laughable
Literally as in you’re making this up?
Or literally as in you can cite a source?
dhogaza // March 6, 2008 at 5:37 am
Yet in that same state, ornithologists are amazed to see that american robins now are year-round residents in areas where they were migrants just a couple decades ago.
No degrees necessary, just binos and a notebook, and yes, you can look this up in Bioscience, though the reference is a couple of years old.
All around the globe we’re seeing similar phenomena in the biological world. Early spring migration of golden eagles in NE Canada was first reported on in the literature about a decade ago. I could spend weeks doing nothing but point out biological evidence of warming.
Bioscience also reported on surveys of various sets of people in the UK, where gardeners and birders have a long tradition of avid record keeping.
Things like … first bloomings, first mowings, first observations of migratory birds, etc.
All moving drastically forward, a month or more, since WW II.
We don’t need climatologists to tell us the world is warming, we only need them to tell us WHY.
Tell us if your gardeners in CO are telling you that, on average, planting times for peas are moving forward or backwards in the year?
And tell us their names so we can audit your lie.
Martin // March 6, 2008 at 2:34 pm
fred:
About the rural-only thingy, that has been done. It doesn’t make any difference. See Peterson et al. (1999): “Global rural temperature trends”, Geophysical Research Letters 26:3, pp. 329-332, or here:
http://www.agu.org/pubs/crossref/1999/1998GL900322.shtml
mmghosh // March 6, 2008 at 3:53 pm
Mr Fred
I think you are conflating absolute temperatures and temperature anomalies.
For example, I may have a faulty thermometer that is reading 5 degrees C too high because it has been badly calibrated, or of defective manufacture, or because it has been set up inside a garden shed.
In that case – my absolute values will always be wrong, because they will be 5 degrees C too high. But, and this is important to understand, that I can still use those measurements over 30 years to find the temperature anomaly correctly, because the difference from the baseline will be the same every time – it will include the error. So my faulty daily temperature readings will still give a correct anomaly reading.
Suppose I realise after 30 years that my thermometers calibration was faulty – say, over reading by 5 degrees. I can then apply a “cooling calibration” by subtracting 5 degrees from each reading. But this will have no effect on the anomaly reading – which will remain the same, because the same error will be deducted.
Do you see, now, why it does not matter from the point of view of affecting the GW case what the absolute values are?
Now, I am not suggesting that weather stations should not have accurate instrumentation, regular calibration and location in areas where fluctuations should be minimal. I believe, though, that climate scientists are fully aware of these problems, and are trying as hard as possible to improve them.
dhogaza // March 6, 2008 at 5:01 pm
Mr. Fred has been told these things many times at this site, and he keeps repeating his basic misunderstandings repetitively.
There comes a point where it’s no longer worth bothering trying to correct people’s basic misunderstandings. If they don’t want to learn, they’re not going to.
George // March 6, 2008 at 5:17 pm
Mr pete asks: “What do you do when more and more science is computer analysis work rather than lab work?”
I don’t think that is a fair characterization of the situation at all.
It is simply not true that computers are replacing or displacing experiment.
They are allowing scientists to look for patterns in the data that they never would have recognized in the past.
A computer is a scientific tool and to suggest that it has “become” or “replaced” the science is simply not accurate.
P. Lewis // March 6, 2008 at 6:07 pm
Ah, George, but you forget … science split into two shortly after the invention of the computer and the Computer Science branch got to play with their tools a lot more than the other lot did.
[With apologies to most computer scientists, who I'm sure don't play with their tools more than is good for them.]
mmghosh // March 6, 2008 at 6:20 pm
dhogaza
I think it is important to explain the difference between absolute values and anomalies to people. I don’t know if you if you talk to many people who are non-scientists; the people I meet have quite a lot of difficulty grasping the concept.
I think this blog does a great job of clearing up many of the mathematical problems that we, with a science background have. But if you asked 100 people what the combined IPCC temperature graph shows, I’m pretty sure 95% think it shows absolute temperatures.
And do you know what? I think that that is actually the scientists’ main fault – not understanding the layman’s problem.
I think the reason why laymen go on about “microsite issues” because it appears so intuitively correct – surely a site that has old-style thermometers, or Russian thermometers (which must be infiltrated by communists), or Chinese recorders (who we know are uneducated or stupid cultural revolutionaries) or is sited in an parking lot surely must have thermometers that are reading too high or too low or be otherwise full of errors and are therefore untrustworthy? I have heard many reasonable people make these and other similar arguments.
I reiterate, the fact that the anomaly trend can be worthwhile even though the absolute readings may have problems is a counterintuitive idea.
dhogaza // March 6, 2008 at 7:30 pm
I’m not saying it’s unimportant.
I’m just pointing out that this particular person has had that fact pointed out to him by a dozen or so posters, and he continues to blather away as though it has never been said.
I do actually talk to a lot of people about stuff like this, and as long as someone’s truly interested in learning I’ll explain things to them with the patience of Job.
But some people simply aren’t interested in learning, and Mr. Fred has amply made it clear that he’s one such person.
john // March 6, 2008 at 8:27 pm
Mr Pete,
Agreed, Y2K all done! it was the most I have been paid for watching fireworks!
I have learnt things on SM’s website about GCM’s, temperature records, proxies etc. that I couldn’t easily find answers to elsewhere, because when you ask a question on say RC you are immeadiately lept on as a skeptic, and you don’t get a lot of sense from that point on.
I watched Al Gore’s propaganda film, and I watched the “GGW Swindle” propoganda film, but am amazed that leading scientists align themselves with Gore, so I don’t trust them any more.
Hank Roberts // March 6, 2008 at 10:44 pm
John, think about who’s lying to you.
Hint: it’s the people telling you what you want to hear.
http://scienceblogs.com/stoat/2007/03/the_use_of_damon_and_laut.php
luminous beauty // March 7, 2008 at 2:22 am
“I watched Al Gore’s propaganda film, and I watched the “GGW Swindle” propoganda film, but am amazed that leading scientists align themselves with Gore, so I don’t trust them any more.
It may be that Gore’s popularized presentation is an honest, if less than complete and perfect, reflection of the science, and TGGWS is full of distortion, misrepresentation and sheer fabrication of fact.
But you may be right. They might be doing it just to challenge your credulity.
[Response: The swindle "documentary" is truly reprehensible. Read about it here.]
Hank Roberts // March 7, 2008 at 5:27 pm
Remember that was a movie about Gore, a snapshot of what he presented a couple of years ago. Part of what it showed is that he’s updating the information in between each presentation.
I’d love to see a diff of what was in the movie compared to today’s version, and I think his people are missing an important opportunity by not making the current changes available online.
But I know from those who’ve seen it more recently than the movie was made that he has kept updating it.
joe // March 7, 2008 at 7:51 pm
dhogaza // March 6, 2008 at 5:37 am
“Yet in that same state, ornithologists are amazed to see that american robins now are year-round residents in areas where they were migrants just a couple decades ago.”
Hogwash! I had robins living in my crabapple trees throughout the winter more than thirty years ago in Idaho. Rest assured that the winters are far harsher there than they are in Colorado. So, it is no great surprise that robins winter in Colorado. Nor is it necessarily indicative of climate change.
As for the other examples you cite I really don’t know. Can you be certain that they are attributable to global warming? Surely as a scientist you know that correlation is not causation. See http://en.wikipedia.org/wiki/Correlation_implies_causation if you need an explanation. Believing that because A is happening and B is happening that A must be caused by B is faulty logic.
Besides, I thought that everything that was going to result from global warming was bad. I, for one, enjoy seeing robins in the river bottom during the winter.
dhogaza // March 7, 2008 at 9:35 pm
Where in Colorado? High up in the rockies where robins neither winter nor breed? Say, above tree line? Say, up where the glaciers are, where rosy finches live?
You’re going to say “hogwash” when I tell you there are parts of Colorado where robins are never seen?
We have decent winter presence/absence data that stretches a long ways back in most of the country, as far back as 1900 for NYC, in the form of CBCs. Not to mention records kept by individuals.
That’s DATA. Not OPINION. Nor HOGWASH.
Hank Roberts // March 7, 2008 at 10:03 pm
> I thought that everything that was
> going to result from global warming
> was bad
Bzzt. Troll.
TCO // March 9, 2008 at 10:11 pm
Colorado doesn’t have any true glaciers. It has “ice fields” which last for 100 years or so. But no persistent glaciers.
TCO // March 9, 2008 at 10:15 pm
Maybe it has some teensy ones. See this link: http://glaciers.research.pdx.edu/colorado.php
I weas just repeating something i heard before.
TCO // March 9, 2008 at 10:38 pm
Glacier picture. Poor Utah, NM and AZ are “ass out” on glaciers.
TCO // March 9, 2008 at 10:44 pm
http://glaciers.research.pdx.edu/states.php
MrPete // March 10, 2008 at 3:13 am
george said “A computer is a scientific tool and to suggest that it has “become” or “replaced” the science is simply not accurate.”
I agree 1000 percent! I wasn’t suggesting that. I was saying that more and more of science is computer-based rather than labwork-based (or fieldwork-based.)
To that end, detailed understanding of computation can be important. Computer have various imprecisions that are meaningless under normal circumstances but quite significant under the kinds of stresses introduced by certain calculations. My example of rounding error was just one.
How many typical scientists know what they can and cannot trust about their computers’ ability to perform accurate calculations?
Today,that’s an increasingly important issue (among several others.)
TCO // March 10, 2008 at 3:33 am
I guess it depends on the kind of work you do. I do science and business and have never had an issue with rounding error. The Excel seems to work just fine for me.
dhogaza // March 10, 2008 at 4:08 am
Rounding error can be easily bound (in the mathematical sense), and this forms one basis of the IEEE floating point standard adopted about 20 years ago.
So, please give us some detail as to how rounding error really effects the data at hand.
I can’t answer that question, but since considerable effort has gone into developing hardware and software to set standards to set bounds, I’d hope it is low.
Because, after all, it was scientists and engineers who pushed the computer design and software hackers in the direction of creating a floating-point representation that would allow for error bounds estimations (based on underlying hardware/software limitations) in the first place.
If it weren’t for such things 20 years ago, I doubt you’d even be aware of the problem …
MrPete // March 10, 2008 at 1:14 pm
TCO, as I said, under most circumstances, digital imprecision and rounding errors will not bite you. Some examples that can hurt:
* Repeated calculations done without careful floating point setup (my rounding demo showed that even Excel doesn’t properly set up FPU rounding; last I checked few compilers do either; might be different now.) The code I’ve seen so far in GCM’s and proxy modelers doesn’t exactly induce confidence in their care with code.
* A variety of algorithms work perfectly in theory, using infinite-precision real numbers. However, in the bit-limited digital world, single-bit issues add up over time unless one takes great care to handle the edge cases. One that I analyzed intensely involved algorithms that depend on the intersection of two lines. In a computer, the actual intersection can rarely be represented with perfect accuracy, and sometimes that’s an important issue.
dhogaza, I’ve been doing computing quite a bit longer than 20 years :-) Some of us wrote low level floating point code that ultimately helped influence the IEEE standards. Now we have those standards and you mostly don’t have to worry about the bits behind “real” or “double.” Yet as I demonstrated with Excel, one can’t assume the standards are properly used even in business software supported by a huge workforce.
It’s easy for people to presume that their equipment is perfectly capable of doing “good enough” calculations. Yet while your garden variety spreadsheet — or footran compiler — will generate acceptable results for basic analytic algorithms, we’re not dealing with such basics here.
We’re dealing with:
* Long term simulations involving zillions of iterations over the data
* Statistical analysis of small variations where the signals and noise are close to indistinct
* Numerical analysis of cause and effect where the fundamental physical basis of our analysis is not known.
Please don’t misunderstand: I’m not even close to suggesting that imprecision in calculation is a huge issue. I see ignorance and/or lack of attention to this as just one “canary in the coal mine” topic that tells me there’s a lot of work to be done. Does it make sense that just as digital calculation can be less “cookbook” than most people assume, so too statistical analysis is often less “cookbook” than most people assume?
Bottom line practical summary: what’s the probability of error in the calculations? As you said you “hope it is low.” Me too. But I have enough background to observe what’s going on and be a cautious Skeptic rather than a Believer. I think science and engineering have much to learn from one another.
L Miller // March 10, 2008 at 4:38 pm
“We’re dealing with:
* Long term simulations involving zillions of iterations over the data”
While that’s common in weather predictions I know of no cases where climate science depends on long term iterative simulations. Climate models certainly do no t fit that description because they do not attempt top predict the state of the atmosphere at a given time.
“* Statistical analysis of small variations where the signals and noise are close to indistinct”
Without iterative calculations it’s unlikely that you will hit rounding error in any reasonable situation. The required precision simply doesn’t allow for it. Clearly there are some exceptions, but I can’t think of any that would apply to climate science.
“* Numerical analysis of cause and effect where the fundamental physical basis of our analysis is not known.”
How may times have people told you climate science does not depend on pure statistics, it’s working with underlying physical models.
“Please don’t misunderstand: I’m not even close to suggesting that imprecision in calculation is a huge issue. ”
Then why mention it at all? It seems to me you are simply throwing **** against the wall hoping someone won’t take the time to scrap it off. As the loony toon brigade demonstrates regularly you throw out reasons why 9/11 was a hoax, we never landed on the moon the earth is growing far faster then real science can debunk them. This is because they do just what you are doing here, throw out “problems” without ever documenting whether these are really an issue.
This is why it’s customary in science for the person leveling a criticism to put up or shut up. If you think something may be wrong, go out and prove it by fixing the error and showing what the right answer is. Science is about better answers, and it’s up to the person doing the complaining to come up with a better answer. This is how the conspiracy theories are separated from real science.
dhogaza // March 10, 2008 at 5:51 pm
Cool. I was a professional compiler writer for 25 years, before getting bored with it and going off to do other stuff 10 years ago.
Why, yes, some of *us* did.
One of the arguments that has been used to keep FORTRAN alive has been that traditionally, mainstream implementations of that language have paid better attention to issues regarding the handling of floating point expressions than languages originally developed for systems programming (such as C). By “better attention” I mean in regard to optimizations.
Ian Forrester // March 10, 2008 at 10:22 pm
MrPete said:
“I was saying that more and more of science is computer-based rather than labwork-based (or fieldwork-based.)”
Do you know any scientists? Have you ever been in a real lab? Computers are just expensive paperweights without someone getting information (data) either by conducting lab experiments or doing field work to plug into them.
Keep it up , you are just showing how much you lack in actual scientific knowledge.
MrPete // March 10, 2008 at 10:23 pm
dhogaza, maybe we’ve met somewhere :-) …
L Miller sez “Science is about better answers, and it’s up to the person doing the complaining to come up with a better answer. “
Falsifying an hypothesis is generally considered a perfectly valid contribution. Knocking off rabbit trails is always good fun! :-)
L Miller // March 11, 2008 at 4:37 am
“Falsifying an hypothesis is generally considered a perfectly valid contribution”
To truly falsify a theory still requires you to advance positive evidence that it is false. This means your own evidence could in turn be tested and falsified.
Even so, falsification of theory still only matters if there is a competing theory available. We don’t throw out theories just because they are falsified after all, we only discard them when a better theory comes along and sometimes not even then.
MrPete // March 11, 2008 at 6:34 pm
Ian, I certainly hope you are not suggesting that Michael Mann is not a scientist? He does neither lab nor field work in his studies.
You’re being nit-picky. Of *course* the data comes from somewhere. What I’m saying is that scientists are able to do much if not all of their work without ever going into a lab or the field. To a certain extent that’s fine. It also leads to the possibility that a scientist loses, or never obtained, that valuable “sense” about the physical world.
Analogy: my daughter was only allowed to engage in college-level computer art courses after she first learned in depth to create physical art. They’re both art.
Unsure why this is contentious.
mmghosh // March 12, 2008 at 2:45 pm
http://www.meteo.psu.edu/~mann/MRG/index.html
Why should you think Dr Mann has no contact with actual data collection?
Are you familiar with his research group?
MrPete // August 9, 2008 at 5:30 pm
mmgosh (sorry, I don’t visit here often)… you ask for evidence of my suggestion that Mann might be one of those who is “able to do much if not all of their work without ever going into a lab or the field.”
Look at his publishing record, look at his data archiving record.
Mann’s impressive publication record derives not from his own field or lab work, but from his analysis of data provided by others.
His entire data publishing contribution consists of being part of one team that published 24 cores from a forest in the rolling hills of Virginia. Interestingly, the data was used to confirm precipitation records.
That tells me Mann may have a sense of how tree coring is done, and some of the challenges of interpreting core data into a correlated record. (Compare with a Graybill, who has archived many data sets… and even Graybill was notorious for keeping poor records.)
Now, please explain how such an experience gives Mann insight into interpreting the ecology, climate or other aspects of BCP growth at 10,000+ feet in the mountains… areas that are far more precipitation-limited than the eastern seaboard of the US.
Anyone who has spent much time in both areas knows they are incredibly different. Enough that even typical east-coast gardening advice (where people are in control of many factors) is of little value for people in the west.
Back to the point: without understanding the data itself in context, it is easy for analysis-based science to get out of whack.
This should not be a controversial statement.
cce // August 9, 2008 at 5:52 pm
I’m sure Malcolm Hughes is fully aware of all of these issues.
dhogaza // August 9, 2008 at 6:00 pm
But one doesn’t need personal experience of a particular area in order to gain knowledge about the ecology, climate, etc. There are other ways of gaining insight, and it’s apparent from the published record that Mann et al have gathered the required insight, so what’s your point, exactly?
MrPete // August 10, 2008 at 1:45 am
dhogaza, I suppose we’ll have to agree to disagree. What I see in Mann’s published record is an inability to see that stripbark BCP’s do not make a good climate proxy. So while I agree some are able to learn lessons indirectly, I don’t see it happening in this case.
It’s quite clear that Mann and others are doing a lot of statistical number crunching on dendro data with no physical basis for the climate connection. They’ve had lots of time to clean up their act.
Spend a few hours in a stripbark BCP forest reflecting on relationship between what you physically see and the wild data from the cores… any sane person would run away as fast as they could with respect to using such data for climate proxies.
Comparison with Hughes is perhaps helpful. In private conversation Hughes was quite humble about the same subject. I have no crystal ball but my sense is he’s working hard to come up with something much better.
I’m not saying it is impossible to do good science without going into the field. Just that it is easy to get off-base.
Pat Frank often argues that you eventually have to get back to a solid physical basis for your hypotheses… and he’s got a good point.
Ray Ladbury // August 10, 2008 at 3:04 am
I always find it interesting that anti-science types will sieze on a single result or investigation and completely ignore the mountain of evidence favoring the consensus science (which after all is why it is the consensus). Mrpete, why do you suppose that is?
MrPete // August 10, 2008 at 10:47 am
Interesting question, Ray!
To me, the distinguishing factor is not pro vs anti science. The distinguishing factor is whether a person is inclined to focus on what we do or do not know… on our growing areas of understanding or on the gaps in our understanding.
Of course, most people would tend to fall into one group or the other. A bit like city dwellers who prefer to strengthen the civilization we already have vs pioneers who always need to break new ground.
Thus, anti-science types can fall in both camps. For example (not comprehensive but to illustrate), they can decry action based on what we’ve learned (a bit of a stretch but consider Postman and Technopoly), or (I think more common) they can note the arrogance of those who are unduly confident that the gaps are not much different than what we know.
At the same time, I’ve learned to be cautious about “the mountain of evidence favoriing” anything that’s relatively new. People tend to run in herds, and that includes scientists.
When it comes to climate, AFAIK most scientists would say 20-30 years of measurement is barely enough to move from “weather” to “climate.” Clearly, we are just now hitting our stride with respect to current measurement (thus the need for so much prior-data adjustment), let alone certainty in proxy analysis. Over a 20-30 year period, I see quite a lot of change in the consensus perspective, which tells me we should not be too quick to proclaim the winners.
What I also find interesting are the admittedly biased PR campaigns that ultimately drag science into the gutter. PR campaigns have been funded and are being carefully followed, to lock the consensus in place independent of the truth (cf IPPR). Many are following insidious advice on how to win by lying (find “Self Validating Proposition” in “How To Win Campaigns”). Ray, why do you suppose that so many people, scientists included, cooperate with such schemes?
Seems to me we’re at a real crossroads for science.
MrPete // August 10, 2008 at 10:54 am
(I should have said “decry unwise action” when referring to Postman. He’s no Luddite.)
So much for my philosophical side. :-D
Lazar // August 10, 2008 at 11:43 am
MrPete,
Yes they do.
Red line is a linear model based on local monthly temperature and precipitation. My comments in open threads #4 and #3 for more.
Ray Ladbury // August 10, 2008 at 12:48 pm
MrPete, Dude, do you even know any scientists? Where the hell are you getting your information–the role of CO2 has been known for ~150 years. I suppose you reject evolution, relativity, plate tectonics, modern genetics and quantum mechanics, too?
You owe it to yourself to read Spencer Weart’s history:
http://www.aip.org/history/climate/
Also, you cannot manufacture or force scientific consensus–it forms because the ideas, concepts, etc. of the consensus form because it is the most productive way forward. That is why the denialists have such a dismal publication record.
MrPete // August 10, 2008 at 1:07 pm
Lazar, I didn’t find anything in #4. I did find you in #3 here.
First, your linear model is based on temp and precipitation. I could stop right there: I find little if any argument against the idea that dendro growth is precip-limited. You need to remove precip from your climate model to be analogous to what MBH et al are doing.
Second, now that you’ve found some nice short term correlation, I suggest you do a bit of verification testing. Try Ababneh’s data (net1) and the 2007 Almagre update (net2). You might find the data for tree #31 of particular interest. Steve M would love some help finding a statistical model that can account for temporary growth excursions exceeding 300 percent.
Working through the statistics to find a correlation is obviously a lot of hard work. Validating the correlation to form a useful proxy is a bit harder. This is where Steve M’s expertise is coming into play.
You might find his current work on multivariate calibration of interest.
Me? I’m less than a lightweight. Peanut gallery at best. :-D
MrPete // August 10, 2008 at 1:40 pm
Ray, I was in high school when plate tectonics was transitioning to acceptability. My best friend’s dad was a geologist with a significant role in bringing in the evidence. It was a big battle. AFAIK our current models still don’t fit all the data. (Way OT but the craziest related idea that intrigues me: what if the planet is growing more than shifting? Haven’t had time to dig into that…)
Sad that you don’t read what I write. I never suggested consensus is forced. In more detail: consensus is formed, broken, re-formed, etc. The problem is that people are great at concluding that we Finally Have The Answer… but we’re pretty bad at having the humility to recognize when there’s another layer of data that just might invalidate much of what we “know.”
I won’t stoop to asking if you know any scientists. Insults never help.
dhogaza // August 10, 2008 at 1:51 pm
Where’s your evidence that precip and temp aren’t correlated in the Great Basin?
No sane person staring at the physical world would imagine quantum mechanics to be true, either, but all this exercise does is to make clear the limitations of direct observation and “common sense”.
Which is why scientists analyze rather than sit and stare at forest, for instance. Sitting a few hours in bristlecone pine forest isn’t going to tell you squat about whether or not growth patterns are a good proxy for temperature.
Lazar // August 10, 2008 at 4:47 pm
MrPete,
No… MBH98 uses a combination of temperature and precipitation proxies under the assumption they provide linear approximations of large-scale temperature variability.
Different trees, different stands. Climate response is highly sensitive to local site factors such as aspect, gradient, soil type.
I have been thinking of verification, but looking for longer instrumental records in the USHCN that are hopefully homogenous and extend back to 1880.
luminous beauty // August 10, 2008 at 7:18 pm
MrPete,
Wronger than wrong.
Chris O'Neill // August 10, 2008 at 8:14 pm
Like saying any sane person would throw out likely useful data at the drop of a hat. Doesn’t sound very sane to me. BTW, what was that about insults never help?
MrPete // August 10, 2008 at 11:02 pm
1) About removing precip data
“MBH98 uses a combination of temperature and precipitation proxies under the assumption they provide linear approximations of large-scale temperature variability.”
I didn’t say remove proxies, I said remove precipitation from your model. I.e., you should not be incorporating precip data in your comparison. MBH did not include precip data in theirs. The hockey stick is a temperature diagram after all. If you want to analyze precip, go for it. I’m confident you’ll obtain better results that way. You just aren’t comparing apples to apples.
2) About learning something from field examination of the trees
As for whether studying trees in the field will tell you anything… look at the Tree #31 link I gave you above. Examine a few stripbark trees in the field and a few things become obvious:
* They grow disproportionately more in the middle of the whole bark and less near the stripped portion.
* Much of the stripping is a mechanical process
Our data set includes all the angles necessary to fully quantify this (and quite a few multi-core trees). Someone could pick that up as an interesting grad student project. In the meantime, the reality just stares you in the face when walking among stripbark BCP’s, because their growth is so incredibly lopsided… You just don’t get that kind of detail from the historical ITRDB data sets.
There’s plenty of reason to avoid strip bark. I’m certainly not the first to suggest this.
3) re: verification, “different trees, different stands”
Actually Lazar, in Colorado they are identical trees. EXACTLY identical. Photographed matching tags available, with 25 years of newer data. It’s all cataloged and made public through the CA site. (AFAIK the lab’s date-matching on some samples has gone slowly which is why it is not yet archived elsewhere. An interesting process, discovering sources of error, uncertainty and delay in production of dendro data samples!)
And if you’re only going to go back to 1880… well, I hope that gives you something useful. I think you’re going to need to go further back to recognize whether your methods are particularly useful or not.
Yes, late 1800’s is about all you can do to compare with thermometry. But you’re building on some very big assumptions. I know the assumptions are widely held, but that doesn’t mean they’ve been validated. What I mean is this: when the samples from a stand of trees, or even multiple samples from a single tree, do not match well… the CI for the data gets blown out. You could easily miss that looking at only a short time period of a hundred years or so. You could also miss that if your samples have been cherry picked (as happens too often.)
What we’ve seen of stripbarks is that they can grow in a boringly “smooth” manner for quite a while, but then there’s a sudden excursion. An excursion not seen in nearby wholebark trees. Seems to me that someone needs to validate the assumption that stripbarks are seeing “climate” rather than simply seeing “damage”. Because the damage is what is visible.
I’ve yet to hear of anyone proposing a physical theory for stripbarks being better temp proxies than wholebark BCP’s. The only reason they’re used is because of the shape of the data.
4) Throwing out “likely useful data”.
Chris, you don’t know how much I agree with you. Please explain that to the dendros who literally throw out any data that doesn’t match their preconceived notion of what they want to find. Misfits need to be explained or incorporated into the noise model, not ignored.
Either use, and/or explain, the entire data set, or chuck it. But don’t say “I like samples A, F and Q, and we’ll ignore the other fourteen samples because they don’t have the ’signal’ we want to see.”
Literally, that’s the kind of thing that’s been going on in some arenas of climate science.
Chris, you caught me on the “insane” insult. Although it was not aimed at anyone in particular, I do apologize. Let’s say it dispassionately: data selection to fit the “desired signal” falls into the Texas Sharpshooting fallacy. That’s a simple and accurate description. Personally, I think such practices should stop.
[Oh, LB you're barking up the wrong tree so to speak :). Yes, sphere's better than flat. Ellipsoid is better than sphere and geoid is better than ellipsoid. I'll take geoid over sphere any day thank you very much. Keeps you 'n me from getting lost while wandering the planet. Keeps airplanes from crashing too. What are you trying to defend? Let's "move on" to better data and models, and quit trying to defend the indefensible. ]
OK, time to run. Hope I can peek back in sometime soon, but it’s a busy week ahead.
BTW, good luck on your analysis, Lazar. It looks like a fun project! And do check in on the multivariate correlation work that Steve M is developing. You might get to be one of the first to apply it in a useful way. The code is all available.
[Response: It's the denialists who want to "throw out any data that doesn't match their preconceived notion of what they want to find." That's why they hate BCP data so much. They talk about it so much because they haven't got anything substantive to say.]
Chris O'Neill // August 11, 2008 at 1:42 am
<blockquoteThere’s plenty of reason to avoid strip bark.
“Plenty” perhaps, whatever that means, but not sufficient. OTOH, there’s sufficient reason to leave them in, i.e. without them, you don’t have any reconstruction of comparable skill before 1428.
MrPete // August 11, 2008 at 2:34 am
Stopped in briefly.
Tamino, if you really mean this, then you should be the first to insist that the rest of the BCP data needs to be incorporated. Don’t throw out the sequences that don’t fit. Insist that the dendro folks stop chucking the “no signal” data, the “random” data, the data that doesn’t look right…
Chris…”Comparable skill” — if the data has no “skill” it is not much use? Just because we know the tree is that old (which is VERY cool of course ;)) does not mean you have learned anything about the climate from it. This is where we have to be very careful about assuming we’ve learned something.
Here are two samples from one tree. Grind that through your analysis:
(Click on the image to see the related discussion, warts and all. Yes, it’s messy, with various folk including me coming up to speed…)
[Response: Why is it that almost every post ends up with denialists talking about bristlecone pines? Maybe because they really don't have anything substantive to talk about.
As for links to ClimateAudit, take it somewhere else.]
nanny_govt_sucks // August 11, 2008 at 3:34 am
This argument always baffled me. By this logic any series that provides “skill” should be included, yes? Baseball batting averages, relative proportion of pirates in the population, Dow Jones averages, whatever, right?
dhogaza // August 11, 2008 at 3:36 am
Right. Got it. Bad data should not be thrown away if it interferes with your “climate science is one big fraud” hypothesis.
No analysis, no QA, nothing should be done, because you don’t like the results.
We’re reading you loud and clear.
dhogaza // August 11, 2008 at 3:37 am
Oh, and why the scare quotes around “random” etc?
You seem to be infatuated with McIntyre’s attempts at analysis, yet any analysis that rejects “random” (your scare quotes) is wrong.
Your bias is transparent.
Petro // August 11, 2008 at 2:24 pm
nanny confessed:
“This argument always baffled me. By this logic any series that provides “skill” should be included, yes? Baseball batting averages, relative proportion of pirates in the population, Dow Jones averages, whatever, right?”
No wonder you are baffled, since you can not distinguish climate data from sports statistics.
Chris O'Neill // August 11, 2008 at 2:36 pm
I pointed out how facile your “There’s plenty of reason to avoid strip bark” argument is. If you want to disingenuously head off on a false tangent then I’m not interested in following.
Chris O'Neill // August 11, 2008 at 2:44 pm
No, that’s not what that logic is saying. The logic is saying that without that series the skill is nowhere near as good. It doesn’t mean you can’t test the series. Some people think we should just leave it out and be happy to live in an information vacuum.
nanny_govt_sucks // August 11, 2008 at 3:14 pm
OK, I get that. But isn’t that part of science? You gather data, do the analysis and let the chips fall where they may. If there’s no skill without a questionable series, then heck, there’s no skill and that’s that. Am I off base here?
Maybe it should be left out because it has a very odd way of growing such that a sample taken from one angle gives a completely different story than a sample taken at a different angle – on the same tree! I mean, you wouldn’t sample Baseball averages to get a climate signal, would you? Why sample the strip bark BCPs?
It is better to live in an information vacuum than to live with bad information. At least we’ll be able to honestly say “we don’t know”. Admitting this is a part of science as well.
Lazar // August 11, 2008 at 3:58 pm
MrPete,
They did. Available precipitation proxies were used in each temperature reconstruction step. E.g. the 1400AD step includes Stahle precipitation series. Check the inventory and Wahl & Amman’s code if you don’t believe me.
MrPete // August 11, 2008 at 4:05 pm
My bias CANNOT be that climate science is one big fraud. How could I have any respect for my climate-scientist friends and relatives? I do know people who do good work.
You guys are sure suspicious! I’ll keep trying hard to eliminate terminology relating to motivation behind actions. It’s the poor science I’m concerned about…the process, not results. How can ANY truly scientific result be unacceptable (as a description of reality)? It is what it is! My worldview says the truth is what we start with. Sometimes we have the ability to change reality (e.g eliminate smallpox), but blinding ourselves to the truth doesn’t help.
Honestly, I am happy to see any well-formed outcome that actually teaches us something. Too bad you still can’t see that. Why are you defending results rather than trying to improve process?
My data bias is that co-collected data needs to be explained together, rather cherry picked for the items we think will match what we are looking for. That’s the problem with stripbark. There’s no physical basis for using it, and yet we cherry pick the specific samples that match what we want to see.
Tamino, what’s so scary about primary data, even if at the CA site?
I’m still amazed we cannot agree on simple concepts like: data has a meaning even if we do not understand it; data should not be tossed just because it disagrees with our assumptions; data should not be accepted just because it agrees with our assumptions. We cannot draw conclusions from data if the data is not explained nor understood.
I think I’ve mentioned this example before, as a good personal illustration of the limits of science. Consider a medication that has been proven effective, even though nobody understands how it works. We can prove that it has the desired medical impact, even though we do not know how it works or why it works. That’s perfectly fine — I’ve used the medication myself for several years. However, very rigorous testing was required to proved the medication is effective. No analogies, indirect assumptions, proof-by-authority, testimonials, etc. Clear and specific double-blind measurements demonstrating that the medication is doing what is claimed. Until such testing is accomplished, insurance companies will not pay for the medication.
Likewise, climate data needs to be rigorously tested, without cherry picking, etc. And clearly we’re not there yet.
OK, time to start my week :)
(PS: Scare quotes? Don’t remember hearing that term before. Sorry, being brief. Further explanation: just as “no signal” data from a physical source is not necessarily devoid of information, so too “random” data from a physical source is not necessarily random.)
[Response: Climate data is rigorously tested, it's part of the ongoing process. You're saying it isn't, doesn't make you correct.
As for ClimateAudit, in my opinion it's simply untrustworthy. Steve McIntyre has an axe to grind a dn no amount of logic or evidence is ever going to change his mind.]
dhogaza // August 11, 2008 at 4:47 pm
Well, given that many believe the ball travels further in warmer air, perhaps you would? Though there are too many variables to control for (they’ve been building smaller ballparks, teams have many more pitching specialists, etc).
But, in principle, as long as there’s a physical linkage between something you can measure and the thing you’re trying to learn about, the thing you can measure is potentially a useful proxy.
MrPete:
Why is it that you guys can’t talk about MBH98 without accusing the authors of scientific misconduct? Intentional cherry picking of data in order to support a conclusion opposite to that which you’d get from using the full dataset meets that criteria, after all.
So, you say …
Then repeatedly accuse them of cherry picking data to fit the result they’re looking for.
You still haven’t answered an earlier question of mine, which is what evidence do you have that there’s no correlation between temperature and precipitation in the Great Basin (which is the basis for your claim that BCP data might correlate to precip but not temperature).
Again, you claim that you’re not accusing climate scientists of fraud, but you repeat your claim that climate data is cherry-picked. You’re not fooling anyone with your disingenuous language, I’m afraid. McIntyre fools some people but he’s a bit more clever with his insinuations and disingenuous asides than you are …
Hank Roberts // August 11, 2008 at 6:44 pm
Wossname went and drilled his own tree cores and is expected to show up as a coauthor whenever that data is published. Wait for it.
dhogaza // August 11, 2008 at 8:27 pm
Were they permitted or did they cowboy themselves some samples? The old pines are protected …
Hank Roberts // August 11, 2008 at 10:05 pm
Howzabout talk about it somewhere else, off topic here ennit? I’m sure somewhere else will appreciate the attention.
dhogaza // August 11, 2008 at 10:29 pm
Well, gee, it’s not like you didn’t bring up the fact that they drilled their own cores. Sorry if my simple question offended. The whole damned BCP distraction is off-topic for this thread in the first place.
Hank Roberts // August 11, 2008 at 10:39 pm
Yup, that’s what I meant to say — that until something comes out in the science journals, there’s no new science to talk about. Let’s just pretend I shut up several comments back, like I was smarter (grin).
Chris O'Neill // August 11, 2008 at 11:54 pm
You’re rather tediously repeating what’s already been said.
Assuming it is bad information for the purpose you want to use it for. Some people assume that because there are some confounding variables in some circumstances that data cannot be used for any purpose that they don’t want it used for.
Chris O'Neill // August 12, 2008 at 12:03 am
You’re not going to establish any credibility while repeating a lie over and over.
nanny_govt_sucks // August 12, 2008 at 12:16 am
Sure. So where’s that physical linkage with respect to stripbark BCPs?
MrPete // August 12, 2008 at 12:25 am
(End of day brief comment: “cherry picking” is not an accusation. It is what they themselves call it. I do try to use the terminology of the experts. Cites, references are all published, but you aren’t interested. Obviously, we’re not communicating here, folks. I’ll go do something more productive for a while.)
oh, and dhogaza, just think for a minute and you’ll understand the lack of correlation between temp and precip, particularly for a place like the Great Basin. If you can’t figure it out, go to WeatherUnderground.com, pick any place (I suggest KELY – Ely, close to G. B. Nat’l Park), and look at the historical temp and precip info.
That’s all the help I have time for. See ya.
Dano // August 12, 2008 at 12:41 am
If you go to the dendro fora run by Grissino-Mayer, one may read repeated posts of exaperation at the simplistic statements about temp-precip signals in tree rings.
It ain’t that easy, folks.
Way back in the day in Fire Ecology, we’d get all historian on our rings, talking about good years and bad years and why, when we were only looking for fire scars for dating, and to center us to look around at the site with better eyes.
Anyway, after a couple instances of getting all expert historian & aren’t we smart, invariably the PhD student would come by and tell us what needed to be done in the lab to verify what we just Galileoed in the field. We’d shut up and go look for some obscure plant after that.
———-
So the work that needs to be done is take the rings from the trees up off of Windy Ridge or whatever that place was on the other side of the pass from Breck, analyze them, apply the results to your hypothesis, and lets see the paper get submitted.
May I also suggest there’s an easy hike up the west side of Pike’s Peak to ~ 11k ft that will take you to a couple strip-barked bristlecones, altho they are likely only 650-800 yo, complicating things a bit, necessitating consults and acknowledgments and the like (I made a comment earlier today – not accepted – about discussing things with colleagues that is germane here, alas:) that side of the Peak is much wetter and you can test your null hypothesis with these, I’d assert.
So to reiterate the point: if it’s so wonderful, discuss it with the world. That is done via the journals, not blogs. That’s how the world works.
Best,
D
Hank Roberts // August 12, 2008 at 12:48 am
> where’s the physical linkage
Look it up. Good starting point if you want to do some serious reading:
http://www.pages.unibe.ch/download/PAGES%20Paleoclimate%20Book/H.%20chapter6.pdf
“Fig. 6.15. Nevada Division 3 precipitation (July-June) reconstructed from a network of lower forest border stripbark bristlecone pine (after Hughes and Funkhouser 1998). The series has been smoothed with a 50-yr gaussian filter. 1 standard deviation unit equals 4.4cm, mean = 18.3cm. Map shows the location of tree ring sites (red + signs) and of Nevada Division 3 (green line).”
The climate of the last millennium
RS Bradley, KR Briffa, J Cole, MK Hughes, …… – Paleoclimate, Global Change and the Future, 2003
Remember, it’s a science journal. It’s not the last word, it’s a place to start serious reading. Read the footnotes; read the citing papers:
Cited by 38
http://scholar.google.com/scholar?hl=en&lr=&safe=off&cites=9482211581003928539
Lost and Confused // August 12, 2008 at 1:25 am
In regards to cherry picking, I think everyone should be able to agree it happens in climate science. Climate scientists themselves admit it. Allow me to offer a few examples showing what MrPete said. D’Arrigo told an NAS panel if you want to make cherry pie, you have to pick a few cherries. An interesting comment comes from Esper et al (2003),
“However as we mentioned earlier on the subject of biological growth populations, this does not mean that one could not improve a chronology by reducing the number of series used if the purpose of removing samples is to enhance a desired signal. The ability to pick and choose which samples to use is an advantage unique to dendroclimatology.”
Perhaps the most clear statement comes from Jacoby, who not only admits cherry picking, but suggests it is a good thing, “If we get a good climatic story from a chronology, we write a paper using it. That is our funded mission. It does not make sense to expend efforts on marginal or poor data and it is a waste of funding agency and taxpayer dollars. The rejected data are set aside and not archived…. I maintain that one should not add data without signal. It only increases error bars and obscures signal…. Fifteen years is not a delay. It is a time for poorer quality data to be neglected and not archived.”
[Response: You really are confused. Cherry-picking is leaving out data because it implies an undesired result. Leaving out data because it increases error bars and obscures signal is not. If you want to argue that leaving out data that contributes only noise is a bad idea, then do so.
If you continue to accuse honest researchers of cherry-picking, expect your comments to end up in the trash heap. That's only noise, no signal.]
dhogaza // August 12, 2008 at 3:22 am
Stupid. Warm air blowing over water can pick up more moisture, leading to more precip later.
Always? I’m not claiming that, but your absolute negative statement that there’s never such a correlation is just stupid, stupid, stupid and wrong.
But one of the expectations here in Portland, OR is that global warming will lead to wetter (and warmer) winters. Just for instance. Take that into your “there’s not correlation between temp and precip” and smoke it. Oh, wait, we have historical data which supports predictions, warmer/wetter winters vs. cooler, drier ones. Hmmm. Gosh, not all models, who woulda thunk it?
Well, MrPete, since you have earlier argued that personal experience would lead one to reject conclusions made by inexperienced academics like Mann…
You’ve argued that a couple of hours sitting on your ass in a great basin BCP stand would make it *obvious* (whatever data might show) that they’re a useless proxy.
Apparently those couple of hours didn’t teach you that a place like Ely will have very different rainfall patterns than the Snake Range (Great Basin NP).
I have personal experience of living an accumulated total of about 2-3 years at 9,000 ft within sight of the Snake Range (Wheeler Peak being the most obvious feature), and nothing in my experience would indicate that what you say is true.
And regards Ely vs. the Snake Range, on our mountain there’s about 20″ per annum, while on the nearby Bonneville Salt Flats there’s … less. You look it up. I know it’s less. You can argue as to how much less, but it’s less. It’s so much less that BCPs don’t grow within a couple dozen miles of the Bonneville Salt Flats – i.e. on our mountain.
And I’ll tell you that in my experience, that wet weather in the portion of the Great Basin in which the Snake Range lies, in fall, at least, is very dependent on the strength and moisture content of storm systems coming from the Pacific.
Except when it’s being strongly affected by monsoon leftovers from AZ/MN (late summer, early fall, most typically).
So, again, where is your evidence that precip and temp aren’t related? You need to show us that the strength and moisture conflict of pacific storms aren’t affected by temp, etc etc, to make your point believable.
The fact that you present Ely as being representative of *mountain* precip patterns in the Great Basin makes it pretty clear that your real-world experience in the area is perhaps a bit limited.
J // August 12, 2008 at 12:02 pm
“Lost and Confused” does certainly seem to be confused.
Let’s say that you’re interested in dendrochronology of tree species X. And let’s suppose that only a very small fraction of the individual trees of this species have ring widths that are well-correlated with climate (perhaps just those individuals growing on marginal sites or whatever).
Now, here are two different research questions you could ask:
(A) What is the mean correlation coefficient between ring widths and climate variable Y?
In that case, you’re trying to describe the overall population of tree species X. Obviously, keeping only data with a high correlation coefficient and discarding the rest would lead to a highly biased result. This would be “cherry-picking”.
(B) What is the climate history of Site Z, as evidenced by tree ring widths?
In this case, you’re not trying to characterize the entire population of tree species X; you just want to use it as a tool for studying the climate of some site. In this case, you should use only the data from individual trees that are most highly correlated with whatever climate variable you’re studying. That’s not “cherry-picking” any more than buying the best piece of laboratory instrumentation is “cherry-picking”.
Of course, in case (B) there are other ways you could engage in cherry-picking. But simply being selective about which trees are used, labeling some as “good data” and others as “bad data”, is not in and of itself wrong, any more than choosing to buy a higher-quality gas chromatograph instead of a lower-quality one is “wrong”.
I’m not surprised that L&C was confused on this point; if you read CA uncritically, it’s easy to become confused.
Lost and Confused // August 12, 2008 at 1:55 pm
“And let’s suppose that only a very small fraction of the individual trees of this species have ring widths that are well-correlated with climate (perhaps just those individuals growing on marginal sites or whatever).”
The problem is how do you tell which trees are well-correlated?
J // August 12, 2008 at 2:59 pm
The problem is how do you tell which trees are well-correlated?
Is that a serious question?
http://en.wikipedia.org/wiki/Correlation
Lost and Confused // August 12, 2008 at 3:25 pm
Yes it is J. If only a few out of many trees are well-correlated with a signal you do not know, how do you tell which trees are well-correlated?
Dano // August 12, 2008 at 3:57 pm
Mr Pete misunderstands:
End of day brief comment: “cherry picking” is not an accusation. It is what they themselves call it. I do try to use the terminology of the experts. …Obviously, we’re not communicating here, folks. I’ll go do something more productive for a while.
Why this is done is found in the forum I mention in my comment previous to this one. And in basic uni-level forestry classes.
By this italicized statement, I can tell someone won’t be getting their manuscripts accepted any time soon.
Best,
D
Dano // August 12, 2008 at 4:01 pm
Ah. I missed J’s comments at 8/12 12.02 pm. Yes, this is what is found in basic forestry texts and any dendro discussion board (where you’ll find discussions asking questions that get responses like J’s).
I would add to J’s comment that “Mr Pete” uses ‘cherry-picking’ as “L& sowing C” does. That is: wrongly.
Best,
D
MrPete // August 12, 2008 at 4:18 pm
In deference to our gracious host’s desire for brevity, here’s a shorter response to dhogaza:
I wrote about lack of correlation.
dhogaza writes about relationship.
We agree more than is suspected.
Precip and temp are not (statistically, mathematically, etc) correlated. They do have a complex relationship, e.g. very high temp results in lower precip.
Hard part: growth = f(temp, precip, etc). Can’t factor precip out of the equation because precip is not a function of temp.
To use growth data as a temp proxy, you need such an equation. Doesn’t exist. Big problem.
harold // August 12, 2008 at 4:42 pm
I believe L and C was framing the question as an experiment, first you give a rationale then you do the testing. J first does the test and then picks samples which correlate. I believe the experimental method is more interesting, although the epidemiological method is a valuable tool in proto-sciences.
Hank Roberts // August 12, 2008 at 5:34 pm
> which trees
More like which species, in which locations. It’s not one tree vs. the next tree.
dhogaza // August 12, 2008 at 7:11 pm
You can repeat this as often as you want, Doing so doesn’t make you right.
Here’s one abstract that reaches the opposite conclusion for the Great Basin, for instance:
Pluvial lake advances consistently correlate with warmer temperatures and retreats with colder temperatures, implying that precipitation trends followed temperature trends in the western Great Basin during the Late Pleistocene.
There’s a bunch of other work done on the subject, specifically in the Great Basin.
Your bald-faced assertions, not backed by evidence, are getting tiresome.
J // August 12, 2008 at 8:01 pm
Lost and Confused writes:
Yes it is J. If only a few out of many trees are well-correlated with a signal you do not know, how do you tell which trees are well-correlated?
There’s nothing unique to trees about this. Consider any potential proxy. Before using it to estimate temperature, precipitation, or whatever, you compare it to observed climate data. If the correlation is high across a wide range of observed data, it’s probably a good proxy. If the correlation is low, it’s probably not.
Let’s say I want to know what the climate was like at site X at some point in the distant past. There are several potential proxies to choose from. Upon examination, one shows a good correlation with the observed climate record of the past century. Another one has very low correlation with the observed data. Using the proxy with the better correlation isn’t “cherry picking”, it’s common sense.
MrPete // August 12, 2008 at 11:05 pm
dhogaza, changing the subject is also tiresome.
Yes, it is wetter during an interglacial, drier during a glacial. Also drier during the Hadean ;). Does Dickerson have something that will help find growth proxy function g(temp) from f(temp,precip,etc) in his research? Nope.
There’s nothing for me to disprove. Show me the function. Be famous.
dhogaza // August 13, 2008 at 2:25 am
MrPete, you’ve claimed there’s no correlation between temp and precip at all (not just the Great Basin, but everywhere).
You provide this as evidence:
Which of course suggests a correlation between temp and precip.
You’re not even reaching the level of denialism, here.
Of course there is, if you’re going to disprove such correlations you’ll have to … disprove them, duh.
Again, your handwaving assertions, not in the least supported by science, are boring.
luminous beauty // August 13, 2008 at 8:59 am
Tree rings -precipitation – ENSO – temperature
Cook, E.R. 1992. Using tree rings to study past El Niño/Southern Oscillation influences on climate. In Diaz, H.F., and Markgraf, V., eds., El Niño: Historical and Paleoclimatic Aspects of the Southern Oscillation. Cambridge University Press, New York: 203-214.
D’Arrigo, R.D., and Jacoby, G.C. 1991. A 1000-year record of winter precipitation from northwestern New Mexico, USA: a reconstruction from tree-rings and its relation to El Niño and the Southern Oscillation. The Holocene 1(2): 95-101.
Lough, J.M., and Fritts, H.C. 1985. The Southern Oscillation and tree rings: 1600-1961. Journal of Climate and Applied Meteorology 24(9): 952-966.
Stahle, D.W., and Cleaveland, M.K. 1993. Southern Oscillation extremes reconstructed from tree rings of the Sierra Madre Occidental and southern Great Plains. Journal of Climate 6(1): 129-140.
Swetnam, T.W., and Betancourt, J.L. 1990. Fire-Southern Oscillation relations in the southwestern United States. Science 249: 1017-1020.
Woodhouse, C.A. 1993. Tree-growth response to ENSO events in the central Colorado Front Range. Physical Geography 14(5): 417-435.
Gavin's Pussycat // August 13, 2008 at 7:17 pm
MrPete // March 5, 2008 at 6:38 am:
Yeah… not a good idea to round in two steps ;-)
Actually the rule to follow is: never round prematurely. You’d think Hansen et al. would know that, wouldn’t you?
Actually there are a number of places in the surface station processing chain where similarly subtle errors could break you up, as described in gory detail in the docs on GISTemp. This is actually well known, and the saving grace is: massive redundancy.
When working with values that have been yearly averaged, they display a strong long range correlation over distances as large as 1000 km. Only a fraction of the stations would suffice if not for error checking.
Massive redundancy allows you to do lots of tricks: test each station individually against its neighbours; split the data in two and compare the halves; process urban and rural separately; and what not. And they did all that and then some, as any scientist would (and not in Excel, shame on you!)
There was this incident with the 0.07 degs error in the US data (0.003 degs global influence IIRC). Hansen was rather defensive about this, but he needn’t have been: this is as bad as it gets. Any larger systematic error of this kind would stick out like a sore thumb.
The story somewhat repeated itself with the recent ocean sampling bottle revision: anything bigger would have been found long ago. I wouldn’t be surprised if more similar (but smaller) issues are hiding out still (Siberia anyone?).
As for non-systematic (individual-station) errors, I am sure there are lots of them, all small-ish (as the large ones got found by testing as described above). But as they are not systematic, they drown in the large numbers.
BTW the stated precision of the GISTemp trend is a posteriori, based on comparisons using the above mentioned redundancy, rather than theoretical error propagation.
I am pretty confident (and would have been even without the Hadley replication) about GISTemp. It’s not rocket science. Now, about the satellite
results… no massive redundancy showing any individual-instrument bias in your face — with the embarrasssing result we all know about. Anyway, they now too agree with the surface data :-)
MrPete // August 13, 2008 at 7:23 pm
Trying to stay close to our ever-patient host’s topic: what does the data tell us. Thus, I’m not interested in loosening my concrete assertion about correlations.
Nth time: strip bark BCP growth = g(precip,temp,etc). Where is precip=f(temp) [let alone etc=f(temp)] that allows us to eliminate non-temp factors?
dhogasa: Epochal changes in (precip,temp) do not answer. Useless to remove precip as a growth factor.
LB: yup, nice refs! They show: tree rings help see ENSO precip record, fire, etc. (Frost kill too, BTW.) Just not growth as f(temp). I wish it were so! ‘Twould be much simpler.
Find a way to remove precip as a growth factor in the equation. I’ll not hold my breath. In the meantime, Lazar is not doing a temp analysis as long as precip (and etc) is left in there.
Remember, this is the easy part. Intra-tree data with 300%+ growth excursions is harder. (cf Almagre tree 31)
Hank Roberts // August 13, 2008 at 8:14 pm
> where is …
Google those terms you use; they’re from skeptic/septic bloggers, not from the dendrochronology journals.
Try asking your question using the terms used in the research papers.
dhogaza // August 13, 2008 at 8:22 pm
Huff ‘n puff, huff ‘n puff.
Sorry, MrPete, when it comes to science, I pay attention to scientists. When it comes to handwaving obfuscation, I pay attention to you.
Lazar // August 13, 2008 at 8:56 pm
MrPete,
There is no need relative to assumptions of MBH98.
Dano // August 13, 2008 at 9:41 pm
Thus, I’m not interested in loosening my concrete assertion about correlations.
I seem to recall inviting commenters making such bald assertions on CA to make them on a dendro forum several years ago.
I offered to sign them up and get them going, and let them loose on the crackpotosphere that comprises dendroclimatology people.
Not a single taker. Lots of excuses tho.
Not sure the offer of Henri still stands, but I could ask him again.
Mr Pete? You game, big fella, armed with your copious and spatially diverse data?
Best,
D
P. Lewis // August 13, 2008 at 11:44 pm
I find use of the words “remove precip as a growth factor” telling, if taken literally (and it seems to me that that is the intention here). Superficially, this (above-quoted) notion is all very scientific. Then you realise it isn’t very when applied to vegetative matter.
Everyone surely knows that if you remove precipitation you won’t have a tree to analyse at all! But perhaps that is the idea. Next, maybe, you will want the light turned out for 50 or 60 years to remove that tropic effect. Result? No trees! Quelle surprise!
Limiting stands. Now there’s a notion that I find a little more scientific than removing precip as a growth factor.
MrPete // August 14, 2008 at 2:11 am
P Lewis, don’t misunderstand. This subtopic began when Lazar included precip in his proxy analysis. I fully agree precip IS a factor. The problem is he’s trying to do some partial replication of MBH… which does not include precip.
This discussion would not exist if the result of MBH were “here’s how trees have grown historically?” The data does say that. What the data doesn’t say is what MBH concluded: “here’s what tree growth sez about temp.”
This whole conversation is made unproductive by assuming I am saying things that I’ve not said nor intended.
Dano, the time may come when I’m ready to join a dendro forum. In the meantime, there’s plenty of serious debate going on at a higher level, eg Pielke’s latest. I’ll gladly let others slug it out. See ya.
dhogaza // August 14, 2008 at 3:20 am
His bullshit depends upon his (unsubstantiated, counter-physics, counter-observations) belief that temp and precip aren’t correlated.
We should just all ignore him. Apparently, he’s on the path to being a CA hero who will have absolutely no affect on the science, which will make a few dozen libertarians happy without affecting anything else at all.
luminous beauty // August 14, 2008 at 7:58 am
Upper bristlecone tree line temperature limited
Lower bristlecone tree line moisture limited.
Compare the difference and voila! Temp and moisture indices disaggregated.
Of course, MrPete knows this, he just doesn’t want to understand it.
Dano // August 14, 2008 at 3:05 pm
Dano, the time may come when I’m ready to join a dendro forum. In the meantime, there’s plenty of serious debate going on at a higher level, eg Pielke’s latest. I’ll gladly let others slug it out. See ya.
Shorter Mr Pete: I can’t defend my baseless assertions.
———-
For the lurkers out there, the dendro forum alluded to is the highest level of debate.
These are dendrochronologists discussing their work, not amateurs and non-climate scientists “debating” the work of others.
This is all we need to know, folks. The italicized is the language of posers and poseurs. Yet another BSer making stuff up. They have nothing.
Best,
D
Hank Roberts // August 14, 2008 at 3:51 pm
Ya know, D, it’s performance art for these folks. They don’t want to participate in a science forum, let alone submit an article to a refereed journal.
I’ve been looking at the blogs linked from people posting some of the dumbest questions for Tamino lately, or googling them by name and favorite phrase. Most all of them have big popular climate blogs full of links to one another and CA and the like.
There are a lot fewer people participating in Tamino’s blog here than meet the eye.
And a lot more chaff and smoke and fog being thrown up.
I guess it’s a measure of his competence that so many people are trying to mess up his teaching.
Harrison Bergeron effect, I think.
Dano // August 14, 2008 at 6:37 pm
Flip side, Hank, is places like RC or here that delete spurious/specious/spammish comments can easily be malcharacterized as ‘biased’ or ’stifling debate’ or whatever phrasie-phrase the AEI-Heritage-Exxon axis releases this week.
It’s all they have, really – parroting the script that the AEI-Heritage-Exxon axis releases for consumption by the gullible-duped-denial axis.
Best,
D
Dano // August 14, 2008 at 6:41 pm
BTW, Hank, speaking of Harrison Bergeron, would the Handicapper General AEI-Heritage-Exxon axis?
Favorite Vonnegut quote: ‘The good Earth – we could have saved it, but we were too damn cheap and lazy.’
D
Hank Roberts // August 14, 2008 at 7:25 pm
> delete … accused of stifling
Yup, that’s why Pharyngula’s “Dungeon” idea is useful — instead of wasting time on perennial popup, they’re set aside.
Not censored, no matter how great
http://pbfcomics.com/archive_b/PBF216-Thwack_Ye_Mole.jpg
the temptation becomes
Just set aside so they don’t keep glopping up the discussion threads with the same old stuff.
Hank Roberts // August 14, 2008 at 7:56 pm
Just to give a complete pointer:
http://scienceblogs.com/pharyngula/plonk.php
“… Pharyngula filter file …… My killfile is more powerful. Being entered here means your comments won’t make it to the pages of Pharyngula, and no one will see them. Just to be fair, though, if such people have a web page or blog of their own, I’ll note it here. You can go elsewhere to read their foolishness….
… if you want your own personal killfile for your browser, try Daniel Martin’s killfile script for Firefox.
http://userscripts.org/scripts/show/4107
harold // August 14, 2008 at 8:57 pm
@ Dano and Hank
I agree that these ideas are fun to think about, really using them would make you a Monster IMO.
‘The good Earth – we could have saved it, but we were too damn cheap and lazy.’ This idea is used by Monbiot, his idea is to govern someone to make him act in a moral way:
http://nl.youtube.com/watch?v=mas-8Vf301k
;-)
Hank Roberts // August 14, 2008 at 10:11 pm
Harold, nobody’s going to _force_ you to talk to scientists or publish in science journals.
It’s up to the host of any individual blog lwhether they entertain, sideline, or decline your submitted comments.
Where do you see a monster, er, “Monster” in anything here?
I’m aware of the arguments along the lines that “vaccination / contraception / labor unions /gun laws / bank regulations / catalytic converters / unleaded gasoline / home insulation / water restrictions / etc. etc. are unwarranted constraints on personal individual libertarity” — if that’s what you meant.
Lazar // August 15, 2008 at 12:14 am
MrPete…
We seem to be talking past each other. Or you don’t believe me.
Here is a plot of the MBH98 1400-step temperature reconstruction (in black). Overlaid is the temperature reconstruction created with the precipitation proxies in the original removed from the input data file. See there is a difference?
There are 22 series in the original, 5 of which are of precipitation.
They are;
(RECORD, TYPE)
Quelccaya Ice Core summit, Ice accumulation
(x2)
and
Southeast U.S-N. Carolina, Dendro ring widths
Southeast U.S-S. Carolina, Dendro ring widths
Southeast U.S-Georgia, Dendro ring widths
… in the inventory.
Which are columns 3 (quelc1-accum.dat), 5 (quelc2-accum.dat), 11 (seprecip-nc.dat), 12 (seprecip-sc.dat), 13 (seprecip-ga.dat) in the input file as per the labels.
Do you still have doubts that MBH used precipitation series to reconstruct temperature?
I am not
These two questions need to be disaggregated…
1) Are the assumptions of MBH98 correct, that is, can local precipitation predict larger scale temperature?
2) Is variation in bcp series related to temperature and/or precipitation?
I’m trying to answer question 2.
If the answer is yes, there are no grounds for excluding bcp series from MBH98.
Lazar // August 15, 2008 at 8:23 am
MrPete,
Phrased the above poorly.
Should read
“Overlaid is the temperature reconstruction after deleting the precipitation proxies from the original input data file. See there is a difference?”
Gavin's Pussycat // August 15, 2008 at 10:23 am
Lazar, this is great. Thanks.
You are obviously on top of the science, and prepared to do the footwork. Which is more than I can claim for myself. Sure I understand PCA climate field reconstruction, but life’s too short to keep putting self-appointed ‘experts’ in their place…
BTW how does the recon look from only precip proxies? Or are those five too weak on their own? Just curious.
TCO // August 15, 2008 at 11:31 am
What is the cause of the delay with data from the McI field expidition? Yes, I know “the lab is taking long”, but specifically why?
TCO // August 15, 2008 at 2:24 pm
I feel like picking at this SM/CA ball of yarn. Think it will come apart like 99% honest John Edwards.
1. What exactly is holding “the lab” up?
2. Why hasn’t he just gotten his samples back and sent them to a different lab?
3. Does SM have some data from the sampling program that hasn’t been put up yet on his site? We know he made an initial disclosure BEFORE completion of the lab work. I want to know if he’s got some more stuff that he’s sitting on.
Dano // August 15, 2008 at 2:33 pm
Yes, I know “the lab is taking long”, but specifically why?
Maybe they can’t falsify the hypothesis.
Best,
D
TCO // August 15, 2008 at 2:43 pm
Maybe they lost it. maybe they broke it. Maybe they are a bad lab. Whatever. Why are we guessing? There should be more update than “lab taking a long time”.
Hank Roberts // August 15, 2008 at 2:44 pm
And you are asking this _here_, why?
You think we care? Touching, but alas, not so likely.
Antics, that’s all, ’til something’s published.
Same as any scientific paper, the author should wait and not talk about unpublished work, it interferes with publication because editors suspect that kind of behavior.
Lazar // August 15, 2008 at 3:38 pm
Hey, thanks GP!
This is a fulltime hobby. I do a dolt dayjob so that I have the time and energy remaining. Still splashing around learning to swim.
Umm… this is the 1400 step with precipitation-only proxies. Most of the variation seems to be from the quelccaya ice cores. It fails verification with an RE score of -0.37, and fails calibration (-0.33).
Dano // August 15, 2008 at 6:46 pm
There should be more update than “lab taking a long time”.
Your ’should’ presumes they were serious about doing actual work.
There is exactly zero historical precedent, so why now the free pass for the posers?
Best,
D
Gavin's Pussycat // August 15, 2008 at 8:33 pm
Lazar,
thanks. What I semi-expected.
TCO // August 16, 2008 at 2:08 am
Hank R.: I agree. The rapid publication of initial graphs, with an interminable wait for the rest of them is a perfect example of why it is a waste of time to even TRY TO ENGAGE with McIntyre or Watts work.
P.s. To be fair, I also don’t like Tamino’s spreading stuff over multiple days (PCA analysis, etc). In this way, I find him and CA and Watts to be very similar in an annoying mystery story style of posting.
P.s.s. I also think that SIs should be up at the time of publication. Think that data should be archived prior to review (I think this is the practice in crystallography and it has had a stellar impact in getting good data archiving.)
dhogaza // August 16, 2008 at 2:56 am
Oh, come on, he has a JOB, and posts here as he has time. And he doesn’t post in a “mystery style” of posing, rather it’s more like chapters in a textbook.
Equating him and Watts is idiocy. Try staying sober for longer periods of time (since you seem to take public pride in your lack of sobriety).
TCO // August 16, 2008 at 3:30 am
I’m equating them on a stylistic point, not in toto.
The point on the style thing is that the “hook” is set and then the answer is not delivered for several days. Meanwhile discussion rages (and suffers from not having the whole article). That’s fine that he has a job. But he could wait with publishing all the parts until he’s done. Otherwise we suffer from the mystery style junk.
Gavin's Pussycat // August 16, 2008 at 9:59 am
A more important difference is that Tamino doesn’t even claim to present new science (even when his analyses are often original). What he presents is classical, even textbook science. It’s all out there, you can do the footwork if you’re sober enough :-) — or wait for Tamino to spell it out and rub it in.
Ray Ladbury // August 16, 2008 at 12:12 pm
TCO, I’d suggest that Tamino’s style–even if driven by his need to put bread on the table and kibble in front of the cat–has an advantage. It allows us to digest what he says in one post and perhaps anticipate the future development before moving on to the next.
BTW, all of us have day jobs. I’ve even put to work some of the stuff I’ve learned (or been reminded of) here in mine. Tamino is providing the US taxpayers a service by providing on-line training to civil servants!
dhogaza // August 16, 2008 at 1:17 pm
Now if we could only get the uncivil nonservants like Watts to learn …
Hank Roberts // August 16, 2008 at 2:43 pm
> Meanwhile discussion rages (and suffers ….
There’s your problem:
“Doctor, it hurts when I do this.”
I prescribe a movie — “Day at the Races” (Marx Brothers). You’ll be cured by this line:
“Then don’t _do_ that!”
TCO // August 16, 2008 at 3:44 pm
I can blow off reading the 2-3-4 parter post until they’re all out there. That does not change the social dynamic of people commenting on them, touting them, seeing them as some sort of proof….when the real argument is saved for the last post.
And it’s the exact same phenomenon at Climate Audit, Open Mind and Watts Up. It’s as if they all ape each other a bit. Thinking that’s hwo the blogosphere should work.
Well it’s not some hairy huge deal. It’s really a minro complaint. But I still have it. And dhog or Hank just making social style rebuttals has done nothing to give me new insight to change my view.
Hank Roberts // August 16, 2008 at 4:53 pm
I agree with you about the clutter, realizing that our host is very reluctant to block the stuff that fills up the threads (that’s why I pointed to the Pharyngula ‘dungeon’ idea — and why I keep suggesting at dot.earth that they need a science thread, and a kibitzer parallel thread, and to keep the stuff separate from the nonsense).
MrPete // August 16, 2008 at 9:30 pm
Dano, silence is not that I have nothing, but (surprise) I have a life. Can only afford small bits of time outside Real Life. Blogging is more visible but less productive in short time. For example, had a good and productive face to face conversation with scientists at LTRR in AZ; 100x more ROI (at least) than similar time on this blog.
Lazar, AFAIK, your graph link is meaningless to your quest. From what I understand, the vast majority of the data feeding into the meat grinder does not contribute to the “HS” picture. Even though it is all supposedly proxies for climate. On the other hand, a small number of the proxies (guess which) produce the HS picture. If they’re all temp or precip proxies, they should correlate as such. You continue to ignore a reasonable suggestion: additional updated data is available for both the Sheep Mountain area and the Almagre area, the key sbBCP data sources. Why would it not make sense to recognize that the new, more updated data ought to have bearing on your question?
Various on lab delays: last I heard, there was no problem with the raw data. That’s been available all along. It’s dating issues, which are a common problem for the field. Some people set aside any data that the computer can’t auto-date. Others doggedly go back to the cores as much as needed to work out the puzzle. (Personally, that would be my bent.)
Still waiting for those missing functions. temp = f(precip); temp = f(etc). Still waiting for proposal of how to incorporate 300% variant growth a few inches apart on a single strip bark BCP.
Vigorous discussion but not productive here. My silence is simply the real world calling. I’ll continue to slow down here (and at CA too!)
dhogaza // August 16, 2008 at 9:48 pm
This is good, really good. MrPete, I humbly suggest you have a lot to learn before you succeed in overturning the entire field of climate science …
Dano // August 17, 2008 at 5:41 am
Mr Pete:
Heard that, been there.
Why f’n bother to borrow an increment borer if you’re not going to publish?
Why bother to pose as know-it-all if you can’t defend it to the folk who know?
Poseur.
Go away now, lad. You got nothing.
Best,
D
TCO // August 17, 2008 at 2:01 pm
Well then Mr. Pete: Since you all ALREADY published partial data, go ahead and publish whatever you’ve gotten dated so far. Heck, publish the raw undated data as well. BTW, the issues in this study and the (probable) need to resample sort of validate something Dano advised you guys of several years ago. Litt;le less self-congratulation on the coffee to tree transit. Little more delivery. You all have already shown that this is not quite so easy as you thought. More King Log, less King Stork from you critters…
MrPete // August 17, 2008 at 9:33 pm
This is an interesting process… the responses I receive here are similar to responses I got during y2k. Commitment to existing knowledge vs exploring new avenues. I realize you think I’m committed to a particular perspective; in reality I’m trying to reopen the “new things to learn” can by pointing to elements that don’t fit the “consensus.”
TCO, not sure what you are pushing for:
a) Publication: previous discussion on publishability of “data only” paper resulted in strong opinion that it is not. You disagree? Great. I’d love solid refs for this.
b) Why a need to resample? We obtained a large collection of good data. Automated crossdating failure on some samples is 100% normal.
c) ‘Twas as easy as expected. Never imagined we would get this far. Finding exact-match Graybill trees was considered “nice”; amazing it happened. Going back for 900+ photos and further provenance details was only because of the intense interest.
Brief notes on goals and objectives:
1) Primary Almagre goal: how hard is it to collect data. Starbucks Hypothesis was proven. Goal accomplished.
2) Found exact matches for important Graybill trees (not just same stands, same trees), in field and data sleuthing. An unlikely goal, more than accomplished; provided valuable info to dendros.
3) Surprised that our data is usable, and our process sound. Open source citizen science became a second objective. Goal accomplished when we began publishing all data as it emerged from the process. Nothing has been nor will be hidden.
4) Discovered our provenance data had value. Made it available to scientists to their delight. Unanticipated goal accomplished.
5) Learned standard ins and outs of crossdating. Data not fully crossdated yet. In process.
Additional goals are being developed that will take more time to unfold. I’m playing with 3D trunk/bark models in my copious (hah) spare time. We hope our (dumb?) fresh look at methods and process will eventually bring clarity of understanding, whether publishable or not. Might even produce a real innovation or two.
Bottom line: Original #1 and #2 goals 100% complete. Added goals #3 and #4 also 100% complete. Added goal #5 in process. Additional objectives under development.
If you see ways to improve on that record, I’m all ears. Honest! E.g. one bit to ponder: appropriate method to make core micro-scans publicly available with no overhead, so more people can usefully participate in manual crossdating. From what I hear, it’s a LOT of data (just a guess: 600-1000MB).
For an off-hours unfunded citizen science project, I’m perfectly satisfied. If you want to see more, better or quicker results, feel free to jump in and help. Nothing we’ve done is proprietary. So yes, you can shoot it full of holes :).
MrPete // August 17, 2008 at 9:44 pm
dhogaza: I’m not saying anything new. I’m putting it in layman’s terms, that’s all. The technical discussion is complete, AFAIK. See the multiyear argument between W&A and M&M, which has perhaps reached a final conclusion. (I’m not holding my breath, of course.)
My desire for climate science is nothing more nor less than taking it up a few notches to match what other sciences expect. OK, and if it’s to be influential for policy, to the level of sciences that influence costly life/death/$$$ decisions. Yes, that’s a lot to chew off. Having been through this (slow/widespread change) process before, I have to afford to be patient. Zero control.
Gavin's Pussycat // August 17, 2008 at 11:22 pm
Study climate field reconstruction MrPete. To the point where you understand it. Lazar did, so could you. Nobody needs those functions.
dhogaza // August 18, 2008 at 12:16 am
Which doesn’t make it correct, of course …
MrPete // August 18, 2008 at 12:48 am
(Let’s move this to Open Thread. This is barely related to the original topic. See you there!)
[Response: Good idea.]
Ray Ladbury // August 18, 2008 at 1:10 am
Actually, MrPete, until it is published, we cannot shoot it full of holes. But then, until it is published in a peer-reviewed scientific journal, it means bupkis.
tamino // August 18, 2008 at 1:37 am
The decidedly off-topic discussion should move to the open thread.
Like gas stations in rural Texas after 10 pm, comments are closed.