Who is Steve McIntyre?
From DeSmogBlog.com:
Stephen McIntyre has been a long-time mining industry executive, mostly working on the “stock market side” of mining exploration deals. He published a blog called Climate Audit where he attempts to analyse in sometimes long and extensive detail the work of climate change scientists where he documents “statistical mistakes” in peer-reviewed scientific literature. …
McIntyre has been described as a “persistent amateur who had no credentials in applied science before stepping into the global warming debate in 2003” and has been a prominent critic of temperature records that suggest increasing global temperatures over the past 1000 years.
As of 2003, McIntyre had worked in the mineral business for 30 years and he has been an officer or director of small public mineral exploration companies for over 16 years…
In February, 2014, he put up four blog posts attacking Dr. Michael Mann in relation to Mann’s defamation suit against Mark Steyn et al, claiming in those posts that Mann had “misrepresented the findings of reports and inquiries into his work and the work of other climate scientists in relation to the so-called “climategate” affair, when the emails of scientists at the Climatic Research Unit at the University of East Anglia were illegally hacked and then published.” (desmogblog.com)
Today, McIntyre’s blog posted a lengthy “fisking” (sort of) written by Jean S, of the climategate emails. The focus is on the famous “Hockey Stick” curve produced by Mann and others just prior to 2000. This graph is an assembly of carefully vetted climate records including proxies and the modern instrumental record. The different sources of information used to show dramatic 20th century warming are well documented in Mann’s publications and further documented in freely distributed supplementary information. Two objections seem to have been raised by climate science denialists such as McIntyre and others. One is that one of the proxies, certain tree ring data, show cooling or at least a lack of warming. The other is that the graphic representation of 20th century warming uncritically combined proxy data and instrumental data.
Denier Complaints About Climate Proxies Are Based On Ignorance Of The Science
A proxy, or proxyindicator, is a natural system that leaves a recoverable residue that varies in some measurable or observable way, such that the variation may correspond to a natural variation happening in the world at the time the proxy was being formed. For example, the ratio of Oxygen-16 and Oxygen-18, two stable isotopes of that element, in Oxygen incorporated in stable form (in biogenic tissues, for example) indicates the ratio of these isotopes in the ambient environment, which in turn, indicates the amount of each element available at the time, which in turn, indicates how much of each type of oxygen is trapped in glacial ice (which tends to have more Oxygen-16 because glaciers are ultimately made of vapor, which is isotopically light). Oxygen isotope ratios of materials recovered from deep sea cores indicate the march of glacial formation and melting over long periods of time.
The first of these two objections to Mann’s work, and other work, relies on naiveté among potential readers about proxyindicators. As is the case with all scientific data, all proxies are suspect, and all proxies have the potential of varying in sensitivity over time. Scientists must always evaluate the quality of the data they use, and not accept it uncritically.
For example, say you wanted to estimate the flow of a major river over time. You could measure how much silt is deposited on the river’s flood plain by taking Carbon-14 samples at numerous depths in the floodplain. A greater depth between samples separated by similar amounts of time might mean more flooding. But, as the flood plain matures and raises in elevation, the frequency of floods year to year may decrease, causing a decline in the rate of siltation, and thus, apparent water flow in relation to the actual water flow. Furthermore, at some point, the flood plain is essentially filled up, and flooding overbank happens elsewhere along the river, and no longer (or infrequently) at your sample site. This is a decline in the usefulness of the proxy to the point that you have to simply stop using it.
Another example. Say you want to use pollen counts form layers found in mud at the bottom of a lake, the layers having been recovered from cores sunk in the lake. Assume your lake is in a region that started out as grassland but slowly became more forested. Trees act as pollen filters. Pollen wafting across the landscape is caught in the trees. In the early part of the lake core data, pollen may be introduced from many kilometers away from wind blown plants (grasses, some trees) and from similar distances from plants that do not distribute pollen via wind, but in small quantities (such plants produce way less pollen than wind pollenated plants) via streams that enter the lake. Over time, however, trees will grow up first around the lake, then over a larger area of the landscape. Windblown pollen from grasses is less likely to get into the pond, and there may be less of that pollen because trees are replacing grassland. Meanwhile, the longer distance stream carried pollen may continue to represent the original catchment of pollen. But, if there are changes in rainfall patterns, that could change too. People looking at pollen in lake cores may use an independent measure, such as the amount of iron in the sediment, to indicate how much water comes in from longer distances via streams vs. how much comes in from groundwater and as direct rain. They will use studies of pollen taphonomy, which look at changes in “pollen rain” as forests develop, to calibrate the effects of trees on the wind blown grass pollen representation. At some point, near the top of the core, the tree pollen may be suddenly and dramatically reduced and the wind blown grass pollen may switch to mostly corn or wheat. This is farmers coming in and completely changing the environment. The core from that point on up may become useless. In sum, the entire core has to be analyzed as a dynamic, changing proxy where some of the changes are important information about the changing environment, while other changes are indicative of an increase or a decline in sensitivity of the proxy as an indicator of what is being studied.
Something similar is going on with the tree ring data Mann used. At around 1960 the ability of the tree ring data to represent regional temperature declines and the tree rings become useless. Prior to that time the data should be used. After that time the data should be discarded.
A proxy is not a pre-calibrated consistent source of information. It is a method that uses measurements of recovered material that allow the reconstruction of an ancient process. But that requires understanding the process well enough to develop a way of determining when the proxy is being helpful and when it is providing noise. A good amount of the research on ancient paleoclimate and paleoecology is about how the proxies work. With this research it is possible in many cases to evaluate the utility of a proxy at a given location, and furthermore, to assess which parts of the proxy can be used, which parts need to be further calibrated, and which parts need to be ignored because of a decline in their usefulness.
We see climate science deniers claiming, for example, that the tree ring proxy used by Mann needs to be used “all or nothing.” This is nothing more than ignorance of how paleoclimatology works.
(See also: Clearing up misconceptions regarding ‘hide the decline’)
Complaints About The Hockey Stick Graph Are Not Valid
McIntyre’s arguments (along with others) about the graph are middle-school level obfuscation of the point. The scientists who published the original Hockey Stick graph went through pains to be clear about what information was going into which part of the overall curve. Subsequent renditions of the same data, or similar sets of data with new information added, range across the board from highly complex constructs showing the different sources of the data, error ranges, etc. to those that simplify by drawing a simple curve of combined information. I wrote about this here, showing how this practice, of sometimes making a very complex thing simpler in a way that makes the point accurately, is done all the time.
The latest post on McIntyre’s site, completely misrepresented what happened with the Hockey Stick curve. Nowhere in the quoted emails is there any suggestion or approval or any indication by Michael Mann of seamlessly merging proxyindicator data and instrumental data. The original documents clearly show that this is not what happened at any stage.
Why Do McIntyre And Others Fabricate These Objections?
If you read JeanS’s post closely, s/he seems to be simultaneously implying that Mann created a falsified representation of how the data come together, while at the same time admitting he did not. This is an increasingly common tactic among climate science denialists. They can no longer totally make up what they are saying because they are too easily called on it, yet want to provide other denialists with fodder, and confuse anyone involved in policy, or who just wants to learn, with more confusion and less clarity.
The only way to accept or even seriously consider the arguments that climate scientists developing the Hockey Stick curve or similar research were involved in inappropriate shenanigans is to anchor oneself deeply in a mire of intentional ignorance. There seems to be only one reason to do that (other than simply being very, very ignorant): a commitment to anti-science activism, with the likely intention of damaging the translation of good science into useful policy.
This is something a mining industry executive might do out of self interest and to represent the interests of that industry. Is that the case here? There is a trick to help determine if that is the case. Follow the money.
The graphic depicted above is from here. It is a 2007 version o the often replicated and used “Hockey Stick.”
There is now a Tumblr for stuff like this.
More on Steve McIntyre:
RealClimate on McIntyre
Skeptical Science on McIntyre
Rational Wiki on McIntyre
Have you read the breakthrough novel of the year? When you are done with that, try:
or Paperback
*Please note:
Links to books and other items on this page and elsewhere on Greg Ladens' blog may send you to Amazon, where I am a registered affiliate. As an Amazon Associate I earn from qualifying purchases, which helps to fund this site.
“McIntyre misrepresents” ho-hum … dog bites mann, or at least tries to gum an ankle, like a toothless Chihuahua, if this is all there down to now.
Remember that McIntyre & McKirtick’s great work in 2005 used a 1:100 cherry-pick to get the results they wanted, Replication and due diligence, Wegman style or later and more graphically, Effect of selection in the Wegman Report
Ah yes, “Jean S” regales the denialati with misinterpretations of issues that have absolutely no bearing on anything, whatsoever.
Greg, if the tree-ring proxies are useless for 1960-2014, why would we trust them for 1000-1200?
At this point I’ve written two blog posts, this one and the one linked to above, that deal with proxies in general that answer your question in a general way. You need to read those posts, not just the titles, not just skimming.
I don’t think I’ve written in detail why tree rings change ca 1960, but you can read about it, called the divergence problem, here:
http://www.skepticalscience.com/Tree-ring-proxies-divergence-problem.htm
Essentially, N. latitude high altitude trees change their growth patterns around 1960 and become lousy proxies. It is not fully understood as to why, but it may ironically be because of global warming. I suspect it could be because of Ozone, too, or at least I wouldn’t rule that out. Not knowing exactly why a proxy goes belly up does not allow a scientist to pretend it has not gone belly up, though. You understand that, right?
Mike… Have you read any of the published research on the divergence problem? Any at all?
Your statement is the equivalent of saying, “I found a fly in my tomato soup, thus all tomato soup should be outlawed!”
Your entire critique is ridiculous. I would give detail but there is no fact which will contradict your apparent belief that climate scientists are by nature….perfect.
Incidentally, this is exactly the same tactic used by creationists. There are known circumstances where carbon dating methods fail, and those are discussed openly in the published literature to communicate to other researchers where and how carbon dating can give wrong results. Creationists use this as an argument to say that “carbon dating doesn’t work,” instead of the more accurate statement of “carbon dating doesn’t work… under certain specific circumstances.”
I have no clue how this post is supposed to be a response to the climateaudit post. I’d suggest people read that post (http://climateaudit.org/2014/09/22/black-tuesday-of-climate-science/) and judge. Kind of mind-boggling.
Wow – you must be betting that your readers have very poor reading comprehension skills. This is a jumbled-up non-sequitor.
Tell you what, if you really want to attack this post, ring up Keith Briffa and ask for his comment. Let’s see what he has to say.
The author evades the fact that if a tree ring series can suddenly become an “unreliable proxy” for temperature in the present, there is absolutely no way to know whether or not that same proxy may have become “unreliable” at some point in the past. This is of crucial importance because the most important part of the hockey stick is not the upward sloping blade but the relatively flat handle — that flat handle wipes out the medieval warm period, said wiping out being absolutely necessary to Michael Mann’s claim of “unprecedented” warming in the twentieth century.
No, that is not true at all. In fact, I deal with that issue as one of the main points in this post and the linked to post wrt proxies in general. The reader (Michael Smith) has apparently evaded reading any of this! 🙂
I am curious as to why you insist on not getting this. Is it that you get it but strategically decide to pretend you don’t?
MikeR… Really? You’re joking, right?
Michael Smith… “…there is absolutely no way to know whether or not that same proxy may have become “unreliable” at some point in the past.”
Perhaps you can explain to us, in detail, how proxy series are validated. It’s a very complicated process but there are no character limits here, so knock yourself out.
@Mike Watson
> Greg, if the tree-ring proxies are useless for 1960-2014, why would we trust them for 1000-1200?
Obviously, a time series doesn’t directly represent temperature — or else it would be a recording thermometer. Tree rings, like all other candidate records, have potential shortcomings. Mike Watson is hardly the first to notice that.
Beyond tree rings, other data series also present challenges. But it’s important to highlight the temperature surge of the past few decades. When it strengthens the validation statistics of “hockey-stick” paleotemperature reconstructions, such proxies should be included.
AMac: Yes. I would simply add that instruments that “directly” measure temperatures are also proxies. We are using the behavior of liquids, metals, or resistors and such as a proxy for temperature!
Greg Laden wrote:
“Not knowing exactly why a proxy goes belly up does not allow a scientist to pretend it has not gone belly up, though. You understand that, right?”
Yes, of course we understand that. What YOU don’t appear to understand is that if proxy behavior can change in the present, without us understanding why, it might well have changed in the past and, thereby, failed to show the medieval warm period. I saw nothing whatever in your post that deals with this point.
All I see you trying to do is bring in a straw man argument, i.e. you go to pains to point out that we must drop a proxy that has become “unreliable”, as if THAT were the issue at hand. Well it isn’t. The issue is what makes anyone believe they have been consistently reliable in the past?
Assertion above:
McIntyre has been described as a “persistent amateur who had no credentials in applied science before stepping into the global warming debate in 2003” and has been a prominent critic of temperature records that suggest increasing global temperatures over the past 1000 years.
Bio of S. McIntyre found on the net includes the following:
I graduated from U.T.S. in 1965. I stood 2nd in Ontario in the then province-wide Grade 13 examinations and was 1st in Ontario (and in Canada) in the high school math contest in 1965. I studied mathematics at University of Toronto, graduating in 1969 with a B.Sc. My focus was on pure mathematics – courses like algebraic topology, group theory and differentiable manifolds. I stood 2nd in my class in 3 of 4 years. As an option, I took several economics courses, including econometrics. I was offered a graduate scholarship to study mathematical economics at MIT, but before doing so, felt that I should have broader social science background and studied PPE (Philosophy, Politics and Economics) at Oxford University on a Commonwealth Scholarship, graduating in 1971. For family reasons, I decided that I should stay in Toronto in 1971 rather than going to MIT and began work.
One could take the position that an undergraduate math degree is not applied science. On the other hand, the oversight of mineral exploration programs, an activity also listed in his bio, certainly encompasses applied science.
He’s a smart guy with a strong math background. Ad hominem attacks on his background distract from the message and put your credibility at risk.
I’m not aware of any reason to think that Professor Mann’s knowledge of math is superior to that of McIntyre.
Don’t change the subject, Rob.
Michael… If you think I’m changing the subject, then you don’t understand the subject.
Michael: ‘Yes, of course we understand that. What YOU don’t appear to understand is that if proxy behavior can change in the present, without us understanding why, it might well have changed in the past and, thereby, failed to show the medieval warm period. I saw nothing whatever in your post that deals with this point.”
We? I thought you were one person.
Anyway, for things like tree rings and temperatures, we know something is a proxy because is demonstrates variation. When treerings stop demonstrating variation they essentially go bad. Also, it is not really true that we simply don’t know why tree rings stop signaling ca 1960. It is not certain, more research is needed, but we are not clueless.
Also important to keep in mind: Tree rings were important in Mann’s pioneering work. Now there are other proxies as well. And as pointed out, the subsequent increase in instrumentally measured temperatures are certainly important as well. In 2000, maybe, it might have been possible to imagine how the Hockey Stick could be invalidated. But it has in fact been confirmed and extended in both directions via multiple approaches.
Micheal again… “The issue is what makes anyone believe they have been consistently reliable in the past?”
How do you think proxies are validated?
Hey AMac, long time no see.
I think Greg Laden September 22, 2014 at #18 did not follow the links provided by AMac September 22, 2014 at #14 🙂
The price of US postage stamps loosely correlated with the US average temperature anomaly; until recently when as we all know the Postal Service introduced the “forever stamp”. Now the average price of stamps is rising, even though the US temperature anomaly has been flat. The recent divergence between price and temperature, though, should in no way prevent us from using stamp prices as a climate proxy in historical reconstruction’s.
Or are stamps somehow intrinsically inferior to greetings and if so why?
You are actually equating the price of postage with climate proxies? That is not what climate proxies are. Now you know!
We are still, to this day, in a situation where no research on global temperature over the past 1000+ years has returned anything other than a hockey stick. It’s a pretty sad thing for the denialati that they’re reduced to continually rehashing such meaningless details as the 1999 WMO cover art, when there have been nearly a dozen new multiproxy reconstructions produced since then.
The thing about science is, it moves forward. Get over it already. We are warming the planet at an unprecedented rate through the emissions of greenhouse gases. We’re on an emissions path that holds the potential for extreme damages for later generations. No detail about the 1999 WMO cover is going to change that.
Greg Laden writes in post #20…”We? I thought you were one person.”
And then Greg Laden continues by writing “we” three separate times in an explanation…WOW GreG Laden, I thought you were one person.
We are not amused.
Where do you find these guys, Greg?
Rob, I thought they came with you!?!
It is very funny that you simply cannot understand what has happened. When someone takes a tree ring proxy and splices it onto the thermometer record (all while trying to “hide” an inconvenient decline) AND then goes on to say that no such thing has ever occurred….AND then the world finds out that this is exactly what did happen….AND the scientists who said they didn’t do it actually wrote emails to each other about doing it……we will have a problem of trust.
So, where is the tropical troposphere heating up? Because it was supposed to, and then it would from the mid-lattitudes reach poleward to melt the ice caps….b-b-b-bbb-but the ice in the Arctic (not close to the poles) was melting….and the ice in the Antarctic is increasing (closer to the equator)!!!! WOW. When you can’t shoot straight, why am I going to believe you when you move the target to where your stray bullet landed?
You are beclowning yourself.
Galen, decline in what? Keep asking yourself that.
Greg Laden wrote:
“When tree rings stop demonstrating variation they essentially go bad.”
Right. Then how, precisely, do you know that the relatively flat — and hence, showing relatively little change — portion of the hockey still called “the handle” doesn’t simply reflect proxies that went bad?
Rob, if you have an answer to this, provide it.
Michael, you are confusing a bad signal with a meaningful signal of cooling. Again.
Greg, it’s almost as though a post like this one sends out a bat signal to the people who not only don’t understand the statistics but want to spread confusion – like a concerted effort on their part.
That couldn’t be the case could it?
Michael… Let me ask you this.
Do you think, for some reason, the only way to calibrate and validate a series is through comparison with the instrumental record elements of a series?
It’s a simple question.
Gaelan… “When someone takes a tree ring proxy and splices it onto the thermometer record…”
Bzzzt!! Right there, you already do not understand anything about this issue.
I can’t remember ever to have read a climate article of such low quality as the article above! Keep up the good work, and the alarmists are destined to lose the battle. Who needs enemies when you have got such friends?
Gaelan… “So, where is the tropical troposphere heating up?”
Ah, jeez. Now you’re going to launch into a full-out Gish gallop?
Real Climate – “Dec 12, 2007 – The basis of the issue is that models produce an enhanced warming in the tropical troposphere”
Skeptical Science-“May 27, 2013 – The IPCC confirms that computer modeling predicts the existence of a tropical, mid-troposphere “hot spot” about 10km above the Earth’s”
Rob Honeycutt…
These are your go-to’s buddy!
If it isn’t the tropical troposphere….itz dem oceans dat ate my warmins, right mate!
“Do you think, for some reason, the only way to calibrate and validate a series is through comparison with the instrumental record elements of a series?”….
I think the way to validate any series is to use it in the same way every time….no upside down Tiljanders and no striking proxies off of the record when the inconveniently diverge from your presupposed plan of continued warming.
Hey Rob Honeycutt…..is this email not really from who it says it is from?
“”””
cc: k.briffa@uea.ac.uk,t.osborn@uea.ac.uk
date: Tue, 16 Nov 1999 13:31:15 +0000
from: Phil Jones
subject: Diagram for WMO Statement
to: ray bradley ,mann@virginia.edu, mhughes@ltrr.arizona.edu
Dear Ray, Mike and Malcolm,
Once Tim’s got a diagram here we’ll send that either later today or
first thing tomorrow.
I’ve just completed Mike’s Nature trick of adding in the real temps
to each series for the last 20 years (ie from 1981 onwards) amd from
1961 for Keith’s to hide the decline.””””
Because surely you have an explanation as to why I can’t understand exactly what they wrote….
Rob Laden…when the going gets tough, erase the comments!!!
WOW, WOW, WOW
Okay, so my browser ate them.
Gaelan, you need to get hold of yourself.
Gaelan @39… We’ll start with this one that your browser didn’t eat.
The tropical hotspot is the expected result of moist adiabatic lapse rate, where heat is released in the mid-troposphere when moist air condenses and releases latent heat.
Got that? It is the result of warming rather than man-made warming. Given that all the data sets show that we’ve had warming, that means we have a couple of choices.
1) All the data sets are wrong, and there has been no warming, or
2) The data sets are correct and the hot spot is difficult to detect.
If you read the published literature on this issue, you find that #2 is correct.
If you bothered to actually READ the SkS and RealClimate posts that you selectively quote from, you find this is what they both say.
Gaelan @ your post #40 that didn’t get eaten by your browser.
“I think the way to validate any series is to use it in the same way every time…”
That’s changing the subject. I’ll ask you my same original question. How do you think tree ring proxies are calibrated and validated?
“Something similar is going on with the tree ring data Mann used. At around 1960 the ability of the tree ring data to represent regional temperature declines and the tree rings become useless. Prior to that time the data should be used. After that time the data should be discarded.”
Greg,
It’s not all tree ring proxies that ceased to respond to temperature around 1960. Many tree ring proxy sets did not respond to the global temperature prior to 1960 at all. These were generally excluded from the tests, again for being poor quality proxies. For normal scientists, at this point alarm bells rightly start ringing. You simply cannot peek at your proxy data like this prior to inclusion, or you are pretty much guaranteed to get a long flat shaft with a blade matching the rising temperature over the callibration period.
If you exclude data because it does not match due to some unknown condition, you still have no idea if the proxies you are keeping are matching the signal merely due to some equally unknown condition.
This is before we even discuss the insane concept of tree rings from local areas being evaluated against a northern hemisphere temperature average, with the presumption of teleconnections to justify the response. It’s not science, it’s wishful thinking.
And Gaelan @ comment #41 that didn’t get eaten by his browser.
“Mike’s Nature trick” refers to Mike Mann’s recently published (at the time) paper that came out in the journal Nature. Notice the capital “N”? The Journal Nature. If you go back and look, what Dr Mann did was just add the instrumental record to the multiproxy record. You can see exactly this at the top of this very page where Greg posted the image.
“Hide the decline” is related to Keith Briffa’s high latitude tree ring series that suffers from “the divergence problem” after 1960. You’ll note that Keith is cc’d in the email. It’s not a problem with all tree ring series. And it’s a problem that is known and widely discussed openly in the scientific literature (and was at the time). Nobody (other than deniers like you) are concerned about this affecting the rest of the proxy record because the rest of the records are calibrated and validated by other means.
“The decline” is just bad data that would confuse the lay audience for whom the 1999 WMO report was intended.
Look, researchers know so much more about what they’re doing than you or McI. You suffer from a chronic disease known as ignorance. You just don’t have the full depth of background to comprehend what you’re looking at. And those that do have the background and HAVE looked at it, they all say there is no issue here.
Remember, this is an issue that has been investigated by, now I think it up to, 10 different independent panels. They all came to the same conclusion. It’s a big nothing sandwich.
None @47… “For normal scientists…”
Really? “Normal” scientists? That phrase sets off alarm bells right there.
Perhaps you mean, “scientists who’ve never studied, even for a day, how dedroclimatology works.”
None, I’ll ask you the same question I’ve been asking others. Do you know anything about how proxy series are calibrated and validated? And, if not, why are you even commenting?
None, yes, the part where I specifically mentioned the tree ring data used in Mann’s work is a signal that I was talking about not all tree ring data. In fact, tree ring data generally indicates growth, primarily, and if it is useful as a temperature proxy depends on geography, timing, and the trees involved.
You are missing the point, and at the same time making it. All proxies have to be interrogated and evaluated whenever and wherever they are used. It is simply not the case that a proxy is thought up then used uncritically whenever it is found.
In my view, the mechanisms behind tree rings as temperature indicators makes them a less than ideal proxy, which is probably why they were not the only proxy used by Mann. But they have some advantages too. Unlike sediments which can mimic each other or change in source material, tree rings are always rings. From trees. So that’s good.
Regarding using tree rings from local areas, that is done because trees are … from local areas. All proxies are collected locally and reflect some range of catchment for the thing being measured (read the post, I discuss that). This, again, is why more than one proxy was used for the Northern Hemisphere.
Meanwhile, more proxies, of different kids, from broader areas have been used for a global verification of the hockey stick. Also, it is confirmed that using a 1960 cutoff for decline in the ability of those particular tree rings to work has been verified as well.
On the ‘Divergence Problem’ in Northern Forests: A review of the tree-ring evidence and possible causes
Rosanne D’Arrigoa, , , Rob Wilsona, b, Beate Lieperta, Paolo Cherubinia, c
Abstract:
An anomalous reduction in forest growth indices and temperature sensitivity has been detected in tree-ring width and density records from many circumpolar northern latitude sites since around the middle 20th century. This phenomenon, also known as the “divergence problem”, is expressed as an offset between warmer instrumental temperatures and their underestimation in reconstruction models based on tree rings. The divergence problem has potentially significant implications for large-scale patterns of forest growth, the development of paleoclimatic reconstructions based on tree-ring records from northern forests, and the global carbon cycle. Herein we review the current literature published on the divergence problem to date, and assess its possible causes and implications. The causes, however, are not well understood and are difficult to test due to the existence of a number of covarying environmental factors that may potentially impact recent tree growth. These possible causes include temperature-induced drought stress, nonlinear thresholds or time-dependent responses to recent warming, delayed snowmelt and related changes in seasonality, and differential growth/climate relationships inferred for maximum, minimum and mean temperatures. Another possible cause of the divergence described briefly herein is ‘global dimming’, a phenomenon that has appeared, in recent decades, to decrease the amount of solar radiation available for photosynthesis and plant growth on a large scale. It is theorized that the dimming phenomenon should have a relatively greater impact on tree growth at higher northern latitudes, consistent with what has been observed from the tree-ring record. Additional potential causes include “end effects” and other methodological issues that can emerge in standardization and chronology development, and biases in instrumental target data and its modeling. Although limited evidence suggests that the divergence may be anthropogenic in nature and restricted to the recent decades of the 20th century, more research is needed to confirm these observations.
Dealing with the problem by carefully examining specific proxies, in “A matter of divergende” Wilson et al. 2007, J of Geophysical Research:
1] No current tree ring (TR) based reconstruction of extratropical Northern Hemisphere (ENH) temperatures that extends into the 1990s captures the full range of late 20th century warming observed in the instrumental record. Over recent decades, a divergence between cooler reconstructed and warmer instrumental large-scale temperatures is observed. We hypothesize that this problem is partly related to the fact that some of the constituent chronologies used for previous reconstructions show divergence against local temperatures in the recent period. In this study, we compiled TR data and published local/regional reconstructions that show no divergence against local temperatures. These data have not been included in other large-scale temperature reconstructions. Utilizing this data set, we developed a new, completely independent reconstruction of ENH annual temperatures (1750–2000). This record is not meant to replace existing reconstructions but allows some degree of independent validation of these earlier studies as well as demonstrating that TR data can better model recent warming at large scales when careful selection of constituent chronologies is made at the local scale. Although the new series tracks the increase in ENH annual temperatures over the last few decades better than any existing reconstruction, it still slightly under predicts values in the post-1988 period. We finally discuss possible reasons why it is so difficult to model post-mid-1980s warming, provide some possible alternative approaches with regards to the instrumental target and detail several recommendations that should be followed in future large-scale reconstruction attempts that may result in more robust temperature estimates.
An example of carefully looking at proxies when reconstructing temperatures from Mann and Jones “Global Surface Temperatures…” 2003, Geophysical Research Letters, to give a flavor of what happens in real science.
[21] Composites were performed separately for both hemispheres, based on the available regional temperature records (8 for the NH, 5 for the SH). Each regional temperature record was standardized by removal of the long-term mean and division by the standard deviation after decadal smoothing (lowpass filtering at f = 0.1 cycle/year). The latter step emphasizes the timescales (decadal and longer) on which the reconstructions are most reliable, given the possible uncertainties in annual dating for certain proxy data used (NH records #5, 7, 8 and SH record #4 and 5).
[22] A flawed recent study [Soon and Baliunas, 2003, hereafter referred to as ‘SB03’] compels us to stress two points which might seem patently obvious: (1) It is essential to assess each proxy series for sensitivity to past temperature variability and not, as in SB03, to equate hydrological influences with temperature influences; (2) It is also essential (e.g., by compositing records) to distinguish between regional anomalies, which often cancel in a hemispheric mean, and not, as in SB03, to equate (e.g., the existence of asynchronous warm anomalies in different regions with a hemispheric mean warm anomaly).
[23] Composite series were formed from weighted combinations of the individual standardized proxy series, employing weights on the individual records that account for the size of the region sampled, and the estimated reliability of the temperature signal as determined by comparison with the instrumental surface temperature record [Jones et al., 1999]. Local (decadal) correlations were calculated between each proxy record and the instrumental grid-box surface temperature records for the regions they represent over the period 1901–1980 (see Figure 1). Proxy records exhibiting negative or approximately zero local correlations (SH record #2 and #3) were eliminated from further consideration in the study. Alternatively, reliability was determined from the correlation of the proxy series against the target (SH or NH) instrumental decadal hemispheric mean series. For reference, an area-weighted average of the instrumental data over the regions sampled by the proxy network (8 regions for NH and 3 for NH) yield extremely high decadal correlations with the associated full hemispheric mean instrumental series [r2 = 0.73 (0.60) for the NH (SH) during the 1901–1980 period].
[24] The hemispheric and global composites were standardized to have the same mean and (decadal) standard deviation as the target instrumental hemispheric mean series over the period of common overlap (1856–1980). While our ‘standard’ reconstruction involved area and local-correlation weighted composites, the sensitivity to the weighting scheme was also examined. Calibration resolved variance (‘?’) [see e.g., Mann et al., 1998] was conservatively estimated from the detrended decadal data variance resolved in the instrumental record. These yield values ? = 0.37 (? = 0.29) for the NH (SH). Rough uncertainty estimates in the hemispheric reconstructions were determined from the magnitude of the unresolved variance during the calibration period, taking into account enhancement of uncertainty at centennial timescales [Mann et al., 1999].
[25] Unlike annually-calibrated reconstructions for which the 20th century instrumental record can be used for calibration, while independent 19th century data is withheld for cross-validation [e.g., Mann et al., 1998], decadal-resolution reconstructions require the use of the extended (1856–1980) instrumental record for an appropriately constrained scaling, leaving no hemispheric-scale instrumental record available for cross-validation. However, previous, more highly (annually) resolved hemispheric temperature reconstructions available back to AD 1000 which have already been successfully cross-validated against the instrumental record [e.g., Mann et al., 1999] provide a means for longer-term cross-validation. We use for cross-validation, in the case of the NH, a simple composite of three previous millennial temperature reconstructions [Jones et al., 1998; Mann et al., 1999; Crowley and Lowery, 2000] and for the SH case, a more tentative composite of the Mann et al. [2000] and Jones et al. [1998] SH reconstructions, which is available back to 1700. This gives a cross-validation of ? = 0.77 (r2 = 0.57) for the NH for the period AD 1600–1855 (? = 0.16 for the full period from AD 1000, likely due to expanded uncertainties in the reconstructions contributing to the composite [e.g., Mann et al., 1999]). As a cross-check, an NH extratropical warm-season temperature reconstruction based on entirely independent (tree ring density) data [Briffa et al., 2001], yielded a highly significant correlation r2 = 0.3 with the NH reconstruction during the overlap interval (1402–1960). For the SH, the verification statistics (available for the interval 1700–1855) are lower (r2 = 0.16), probably due to greater uncertainty in both series being compared.
So eliminating bad data sources from an analysis is bad how? Help me here.
This post and discussion do a spectacular job of missing the point of the article at Climateaudit.
The WMO99 cover was the direct subject of Jones’ “hide the decline” email. In Mann’s latest court submission, he vociferously stated he had no involvement in the WMO99 cover, despite being listed as an author and claiming credit in his CV. He also subsequently claimed at realclimate that no researcher had ever grafted thermometer data onto proxy data.
The CA article clearly shows how Mann commented positively about the cover and made suggestions for improving it. Jones explained to him exactly what he had done. Mann could hardly have missed how there were single lines covering both the proxy period and the subsequent thermometer period.
Greg, you’re going to get yourself into very hot water defending the validity of long term temperature reconstructions from tree rings. This stuff has very major problems, several of them, as some of us have pointed out very clearly. Steve McIntyre’s work is not even chief among them. Paleo T reconstructions from tree rings is probably the single weakest part of climate science, or certainly top three.
Jim, I actually said something similar about tree ring above. They are a useful proxy but for the reasons discussed again and again ignored need to be used carefully. Mann and others used them carefully. You can’t have it both ways.
So, this example you give of real science, Mann and Jones “Global Surface Temperatures…” 2003, Geophysical Research Letters, is that the paper that Ray Bradley thought was “truly pathetic and should never have been published”? And that Ed Cook thought was “probably the worst paper that Phil [Jones] has ever been involved in”?
Just askin’
Igsy, I don’t know, it is a peer reviewed published paper, I don’t make it a practice to check around with people to see if I should use such things. The material I provided above serves the intended purpose well and distracting from that is not an argument.
And yet, Jim… The PAGES2k project, which uses a wide range of proxy data, comes up with very similar results as tree ring reconstructions.
http://www.skepticalscience.com/pages2k-confirms-hockey-stick.html
Are you complaining that tree ring data reconstructions came up with the right answer for the wrong reasons?
Greg, My father was a WWII vet. He fought in the Battle of the Bulge and suffered debilitating injuries for the rest of his life. He also freed the Jewish prisoners from the concentration camps in Germany. The term ‘denier” is intended to mean those that deny the Holocaust. You demean the term when you apply it to the complex problem of Climate. You demean is sacrifice and the sacrifice of thousand of GI who saw the Holocaust first hand. Shame on you.
Bob, no, it isn’t.
Igsy… So, you’re telling me that sometimes members of “The Team” disagree with each other? Even criticize each others’ work? I thought they all had to toe the line or risk losing their grant monies.
(sarc)
Rob, exactly.
Bob Brown… Hmm, how then is it the term dates back to the 1520’s?
Quick and ez summary as to how to deal with “divergent” tree-ring data.
Instrumental data-sets can be used to calibrate/validate the “divergent” tree-ring data-stes from over the pre-divergent time-period for which the instrumental data-sets are available.
For time-periods prior to the availability of reliable instrumental temperature data, the “divergent” tree-ring data-sets can be validated against proxy data-sets that *don’t* suffer from divergence.
The wealth of “non-divergent” tree-ring and other types of proxy data make this possible.
If divergent tree-ring data-sets were the only proxy data available, then they should definitely be considered unreliable.
But with so much other proxy data available, there are ways to cross-check the divergent data for time-periods prior to the availability of instrumental data; as a result, even “divergent” tree-ring data can provide useful information for pre-instrumental-data time-periods.
https://books.google.com/ngrams/graph?content=Denier&year_start=1800&year_end=2008&corpus=15&smoothing=0&share=&direct_url=t1%3B%2CDenier%3B%2Cc0
I do have to smile when I read all the pseudoskeptic arguments over the hockey stick. Everytime I get a denier to actually discuss it they eventually say they’re *not* talking about the blade (we know the blade is accurate because it’s the instrumental record), but the handle (he disappeared the MWP!). Yet the first thing they glom onto is ‘hide the decline’ – absolutely hilarious, but they apparently aren’t in on the joke.
Here’s a clue: MBH98 only went back to 1400. When was the MWP?
As Greg mentions, all of our instrumental measurements are in effect proxy measurements. We are using the known variations in electrical output between two dissimilar metals when we use a Type K thermocouple. These variations are highly correlated to temperature changes.
But a type K thermocouple’s output is only correlated to temperature from -200ºC to 1350ºC. Below or above that we must use a different proxy – so for higher temperatures we might use a Type C thermocouple that has a range from 0ºC to 2300ºC.
If we were to follow the insane arguments put forth by some pseudoskeptics we would have to throw the baby out with the bathwater. You *must* pick a proxy and then use it for all measurements and include all of it’s results. It’s somehow unethical to use a Type K thermocouple for most of your measurement results and a Type C for only those measurements where the Type K is no longer correlated!
Sorry, you use the best proxy you can for the measurement you’re making. Every proxy – whether it be a tree ring series, a voltmeter, a thermocouple, spectrum analyzer, etc. – has a useable range. Using it beyond it’s known range is what will get you in trouble. That’s when you get spurious results. Exactly the opposite of how pseudoskeptics think it should be done.
It’s really a good thing that none of these pseudoskeptics actually *do* any science because if we relied on them we’d be back in the dark ages.
“The WMO99 cover was the direct subject of Jones’ “hide the decline” email.”
Ah, no it wasn’t Andy. The subject was the paper that came out in Nature in 1998, MBH 98. The WMO99 cover came out in 2000; the Jones email was written in 1999. That is why he called it Mike’s Nature trick, not his WMO trick. And what seems to get forgotten is the fact Phil Jones misremembered what Mann et al actually did in the ’98 paper. He says in the email that Mann added the instrumental data from 1981 onward,except that isn’t what Mann did. Mann showed the proxy data on the graph from 1400 to 1981, and on the same graph but with a different line he plotted the instrumental data from 1905 to 1995, Nothing was spliced together. Both lines of evidence were shown together on the graph.
Bob Brown – concern troll. The word ‘denier’ wasn’t retired – like some athlete’s jersey. If you deny the basics of science you’re a science denier. If you deny the attempted extermination of the Jews by Hitler, you’re a holocaust denier.
Go concern troll somewhere else.
By the way, I found another Hockey Stick graph that closely follows Mann’s but with entirely different data. https://books.google.com/ngrams/graph?content=Hockey+stick&year_start=1800&year_end=2008&corpus=15&smoothing=3&share=&direct_url=t1%3B%2CHockey%20stick%3B%2Cc0
I’m not sure if recursion is allowed: the ‘hockey stick’ proxy added in to help produce a hockey stick?
I asked over at HotWhopper if anyone knew of a collection of hockey sticks from phenological studies. We all know that recent temperature reconstructions all show a hockey stick, but many other studies also show them. I just don’t know if anyone has collected them all in one place.
NH lake-ice on/off dates, arctic sea ice volume, and arctic sea ice minimium extent quickly come to mind – but I’ve seen dozens over the past few years – just can’t remember them all 🙁
At #56, Jim Bouldin made the most important point in the thread, and one that is more important than anything asserted in the post.
He knows what he’s talking about — read his blog.
Defending weak science is, ultimately, a bad idea. While usually well-meant, this has damaged climatology more than anything its enemies — real or imagined — have done.
AMac… Then maybe you can answer the same question I asked Jim.
The PAGES2k project comes up with an almost identical reconstruction and relies on a wide range of proxy data. Are you making the complaint somehow that tree ring based multiproxy reconstructions are coming up with the right answer for the wrong reasons?
AMac: I addressed Jim’s comment with a correction. You too shall stand corrected, sir!
@ Rob Honeycutt #74
A light conversational tone can be nice, but there’s a risk of segueing into glibness. “Are you making the complaint somehow that..” may not be the best invitation for a conversation on matters scientific.
I looked at the SkepticalScience page on PAGES2k that you linked in #58. The top graph shows four reconstructions: Moberg, Ljunqvist, Hegerl… and Mann EIV, from his 2008 PNAS paper.
The issue of how to determine which data series to use is a crucial topic, and one that is a challenge for many disciplines. The topic is beyond the scope of this comment — except that this is one of the serious problems that tripped up the authors of Mann08.
Another is that Mann08’s EIV reconstruction depends on the inclusion of the Tiljander proxies, more and more heavily as one goes earlier in time (and tree-ring proxies drop out). Even by Mann et al’s methods, no-Tilj EIV fails to validate prior to 1500. It should give pause that the certainly-incorrect inclusion of these proxies serves to rescue this reconstruction, from ~500 AD to ~1500 AD.
These are well-known and oft-discussed issues (e.g.). And yet, this 2013 SkepticalScience report makes no mention of them. Neither, apparently, does the referenced 2013 Nature Geoscience paper.
Quality control isn’t an optional afterthought; it must be built in at the start. For whatever set of reasons, paleoclimatology is deficient in this regard. Until this is addressed, pace Bouldin, I see little point in diving in to particulars.
AMac… It was a very serious question, and one you completely dodged.
Look, we’re talking about trying to determine past global temperature from proxy data. It’s never going to be absolutely precise, nor does it need to be.
What we do have are multiple complex research projects trying to answer the question about global temperature for the past 1000+ years. So far they have all come up with nearly the same answer. You can dive into niggling points all day long about the precision of the data, but the big picture is very clear.
Over the past 1000 years there has been a slow and gradual cooling trend that ended with the start of the industrial revolution and the emission of large amounts of atmospheric CO2.
…And global temperature today is likely higher than it has been anytime in the past 1000 years.
Wut?
Comparing a “choice” (there is no choice) bertween the major data sets in 2008 with assembling appropriate proxies for a paper doing something entirely different ten years earlier is a “challenge for many disciplines”?
You are trying to sound scientific and all and you are doing a pretty good job but you are not really saying anything.
You might find an answer to what seems like a question you ahve about Tiljander here: http://www.pnas.org/content/106/6/E11.full
Skeptical Science does discuss this issue.
Again, you’ve not really grasped the point of this post (or my earlier one). It is ALL ABOUT quality control. And, many if not most of the denialist objections are about backing off from that (like by treating all proxies that have similar names as equally valid/invalid no matter what).
Rob Honeycutt #77 and Greg Laden #79 —
Thank you both for your prompt responses. I’ve reviewed the issues concerning the Tiljander data series — really, they aren’t temperature proxies, as far as anybody has shown. Neither of you seems very well-informed on this subject, and its implications for the robustness of Mann08’s findings. That’s ok — it’s a wide world, full of interesting topics; nobody can master them all. However, these aren’t the preconditions for a mutually-beneficial conversation. FWIW, I present my perspective in this Yahoo answer (#5) and the links therein.
Greg and Rob,
I’m sure the Skeptical Science Team including Mike Mann are proud of your strong defence! Go on.
My polite and on-topic response to #77 Rob Honeycutt and #79 Greg Laden appears to have failed moderation.
@Kevin O’Neil#72
This might be what you’re looking for
Enough hockey sticks for a team
http://environmentalforest.blogspot.co.uk/2013/10/enough-hockey-sticks-for-team.html
Thanks, johnl. I think BBD might have pointed me there the other day – if you notice I have the last comment on the page 🙂
I’ve got 5 that aren’t listed there, but it’s a good place to start.
Walsh & Chapman Historical Arctic Sea Ice Extent
Hottest Days of Month – Australia
1000-year records of southern hemisphere background concentrations of CO2, N2O, & CH4
Lake Suwa, Japan Ice-On Time Series 1443-1983 (page 24)
AMac… You know, it’s interesting. I was just googling around to find out more about the Tiljander issue and, lo and behold, AMac is there at Stoat’s place discussing the exact same issue five years ago.
http://scienceblogs.com/stoat/2009/10/29/tiljander/
The interesting thing is, you still haven’t managed to respond to anything I’ve said about temperature reconstructions. You’ve just immediately jumped to this Tiljander hobbyhorse.
You’re stuck in a time warp, AMac. Science moves forward. It really wouldn’t matter if Mann has done every single one of his proxies wrong because one single paper does not define the whole of any issue. Not even Einstein’s relativity papers do that. Science is the full body of research, the full body of scientific knowledge.
The point I’m making here is, you and Jim (and McI and others) have made all this noise about tree ring data. And you’ve been stuck on one person’s research for going on years now. You seem completely incapable of letting go and looking at the fact that… science moved on without you. There are a dozen reconstructions all showing the same results as MBH98/99.
And now I’m asking you to stop and look at PAGES2k with it’s broad range of proxy records… again, confirming nearly the exact same result.
You wrote in, the blog post;
“Something similar is going on with the tree ring data Mann used. At around 1960 the ability of the tree ring data to represent regional temperature declines and the tree rings become useless. Prior to that time the data should be used. After that time the data should be discarded.”
You have not show *any* hypothetical basis for what you wrote. That is just a conjecture pulled out of thin air. (Not by you, but by those preceding you)
There is a literature on this you could read, Jim. You’ll find there is a basis.
such as?
What is your hypothesis for the occurrence?
Noting that it occurred, without a falsifiable explanation, only means that your data is unreliable.
“On the ‘Divergence Problem’ in Northern Forests: A review of the tree-ring evidence and possible causes
Rosanne D’Arrigoa, , , Rob Wilsona, b, Beate Lieperta, Paolo Cherubinia, c
Abstract:
An anomalous reduction in forest growth indices and temperature sensitivity has been detected in tree-ring width and density records from many circumpolar northern latitude sites since around the middle 20th century. This phenomenon, also known as the “divergence problem”, is expressed as an offset between warmer instrumental temperatures and their underestimation in reconstruction models based on tree rings. The divergence problem has potentially significant implications for large-scale patterns of forest growth, the development of paleoclimatic reconstructions based on tree-ring records from northern forests, and the global carbon cycle. Herein we review the current literature published on the divergence problem to date, and assess its possible causes and implications. The causes, however, are not well understood and are difficult to test due to the existence of a number of covarying environmental factors that may potentially impact recent tree growth. These possible causes include temperature-induced drought stress, nonlinear thresholds or time-dependent responses to recent warming, delayed snowmelt and related changes in seasonality, and differential growth/climate relationships inferred for maximum, minimum and mean temperatures. Another possible cause of the divergence described briefly herein is ‘global dimming’, a phenomenon that has appeared, in recent decades, to decrease the amount of solar radiation available for photosynthesis and plant growth on a large scale. It is theorized that the dimming phenomenon should have a relatively greater impact on tree growth at higher northern latitudes, consistent with what has been observed from the tree-ring record. Additional potential causes include “end effects” and other methodological issues that can emerge in standardization and chronology development, and biases in instrumental target data and its modeling. Although limited evidence suggests that the divergence may be anthropogenic in nature and restricted to the recent decades of the 20th century, more research is needed to confirm these observations.”
(sorry about the long quote)
There is no falsifiable hypothesis there.
That is only conjecture. Conjecture about why the dendrochronological temp proxy is unreliable now, post 1960.
The same, and more, conjecture can argue why the tree ring temperature reconstruction is unreliable at any time.
What is your falsifiable hypothesis for the divergence?
“Jim, I actually said something similar about tree ring above. They are a useful proxy but for the reasons discussed again and again ignored need to be used carefully. Mann and others used them carefully. You can’t have it both ways.”
Using them “carefully”. That is selection bias. Sometimes it is ok. Many more times it is very bad.
“The causes, however, are not well understood and are difficult to test due to the existence of a number of covarying environmental factors that may potentially impact recent tree growth. These possible causes include… … differential growth/climate relationships inferred for maximum, minimum and mean temperatures.”
That covers it all.
Robert Murphpy:
When Jones wrote “I’ve just completed Mike’s Nature trick of adding in the real temps to each series for the last 20 years” he was in the middle of preparing the WMO99 cover chart. You are right that he mis-remembered the trick, but wrong that it was about something else.
The CA post clearly shows that Mann clearly was consulting durign the WMO cover preparation, despite his denial in his court submission.
Why don’t you read the actual post at CA
http://climateaudit.org/2014/09/22/black-tuesday-of-climate-science/#comment-730970
Rob Hunnicut says; ” And now I’m asking you to stop and look at PAGES2k with it’s broad range of proxy records… again, confirming nearly the exact same result.”
Rob, the result is not the same. The PAGES2k may be good research, but it is *not* the same graph line as the MBH hockey stick.
Show the error bars. Show the SD for the coincidence.
“Additional potential causes include “end effects” and other methodological issues that can emerge in standardization and chronology development, and biases in instrumental target data and its modeling”.
Et tu?
“You are right that he mis-remembered the trick, but wrong that it was about something else.”
No, it was absolutely about MBH 98. That is definite. The email says so explicitly. And the “trick” was no trick at all. Both proxy and instrumental data were shown on separate lines. There was absolutely no splicing. You guys are really making fools out of yourselves.
And yet Huang (et al, 1997) got this confounding paleoclimate result for the world, published in GRL
http://joannenova.com.au/globalwarming/skeptics-handbook-ii/web-pics/boreholes-huang-1997.gif
Robert:
The “I have just completed…” part was written by Jones while he was preparing the WMO graph. The WMO graph clearly splices proxy and thermometer data.
Mann was consulted on the WMO graph preparation and gave suggestions – the timeline for this is in the CA post. Mann then claimed credit for the WMO chart in his CV, yet in his court submission he vociferously denied having any part in its preparation.
“The “I have just completed…” part was written by Jones while he was preparing the WMO graph. The WMO graph clearly splices proxy and thermometer data.”
And MBH ’98, which Jones referenced in the email, absolutely did *not* splice proxy and instrumental data. Read the friggin’ paper. You people are pathetic.
Im interested in the use of proxy data – can you expand upon this statement? – Many thanks.
“Prior to that time the data should be used. After that time the data should be discarded.”
Best
Michael
Robert: I completely agree that Mike’s Nature Trick refers to MBH98.
MBH98 was not the subject of the CA post that Greg Laden criticised. The subject was WMO99, and Mann’s involvement in its preparation. Once again, why not read the actual CA post, that way we can avoid talking at cross-purposes.
Robert: I completely agree that Mike’s Nature Trick refers to MBH98.
MBH98 was not the subject of the CA post that Greg Laden criticised. The subject was WMO99, and Mann’s involvement in its preparation. Once again, why not read the actual CA post, that way we can avoid talking at cross-purposes.
@ Greg Laden,
Thank you for allowing my comment #80 past moderation, making my #82 superfluous.
@ Rob Honeycutt # #85,
Thanks for reading my #80 and responding. I agree, the Stoat thread you link is fascinating.– though we may not agree on the specifics of why that is so! Readers can judge for themselves.
And yet the specific internal guts of the NAS Hockey Stick report (June 2006) accepts M&M’s critique about 95%of the time – how did that happen?
I’m looking over the comments that came through overnight (sorry for some getting stuck in moderation; comment security has been heightened due to some shenanigans that always seem to emerge when I’ve got something of interest to certain groups).
I don’t see anything that can’t be addressed by simply referring back to what I’ve already written. Proxies are not automatically accurate and problem free. They always have to be used carefully, with every attempt to verify that they are representing the expected change in the variable(s) of interest. They can’t be used uncritically.
Using them carefully, using them uncritically, is proper science.
Also, proxies vary in how typically well behaved they are. One could always rank proxies as to how reliable they may be, and in fact, that is part of using them carefully.
The usefulness of the proxies used by Mann et al in the papers that first presented the “Hockey Stick” is something one can ask about, but the fact that the curve showing dramatic increases in surface warming with the increase in CO2 in the atmosphere has been verified multiple times since then indicates clearly that the work they did was on the mark.
I have thus far restrained myself from commenting on this thread. Most of the issues being discussed have been discussed ad infinitum elsewhere. I am a lawyer, who prior to attending law school, had academic training in time-series analysis and who worked professionally for several years as an econometrician in private industry. For what it is worth, IMHO the critics of Michael Mann’s hockey stick graphs have decisively won the debate concerning whether his statistical analysis was flawed, and whether the various “tricks” “bodges” etc. regarding the presentation of hockey sticks graphs were misleading. I’ll resist the temptation to further debate these issues, which in my mind have been settled. YMMV.
I do want to comment, however, on how the main post completely misses the point of the post by Jean S. by taking it out of context. Here is simple explanation of the context.
1) Michael Mann has sued several defendants who asserted, among other things, that Michael Mann hockey stick graph was “fraudulent”.
2) Michael Mann has asserted that the defendants knew or were reckless in not knowing that this claim was fault because numerous government investigations had examined his work and found nothing wrong with it. As one example, Michael Mann’s attorneys have cited the U.K.’s The Muir Russell panel as an investigation that “exonerated” Michael Mann.
3) The defendants have properly pointed out that the Muir Russell panel was never charged with the task of investigating Michael Mann’s work and, therefore, its report cannot be construed as an exoneration of Michael Mann’s work.
4) Moreover, the defendant’s pointed out to the extent that the panel touched upon Michael Mann’s work, it was critical of it, specifically quoting the following conclusion, “In relation to “hide the decline”, we find that, given its subsequent iconic significance (not least the use of a similar figure in the TAR), the figure supplied for the WMO Report was misleading in not describing that one of the series was truncated post 1960 for the figure, and in not being clear on the fact that proxy and instrumental data were spliced together.”
5) Mann’s lawyer then replied that:
“In their brief, the CEI Defendants suggest that the University of East Anglia’s investigation actually found that the hockey stick graph was “misleading” because it did not identify that certain data was “truncated” and that other proxy and instrumental temperature data had been spliced together. See CEI Anti-SLAPP Mem. at 16-17; NRO Mem. at 35. This allegation is yet another example of Defendants’ attempts to obfuscate the evidence in this case. The “misleading” comment made in this report had absolutely nothing to do with Dr. Mann, or with any graph prepared by him. Rather, the report’s comment was directed at an overly simplified and artistic depiction of the hockey stick that was reproduced on the frontispiece of the World Meteorological Organization’s Statement on the Status of the Global Climate in 1999.41 Dr. Mann did not create this depiction, and the attempt to suggest that this report suggested an effort by Dr. Mann to mislead is disingenuous.”
6) The defendants have not yet submitted their reply brief. Jean S., however, has made two posts at climateaudit.org regarding the above quoted material from Mann’s brief. In his first post, Jean S. points out that Michael Mann claims in his CV credit for the WMO cover page graph. Specifically, Mann’s CV contains the following entry: “Jones, P.D., Briffa, K.R., Osborn, T.J., Mann, M.E., Bradley, R.S., Hughes, M.K., Cover Figure for World Meteorological Organization (WMO) 50th Year Anniversary Publication: Temperature changes over the last Millennium, 2000.”
7) Jean S followed up with a post, which is the subject of the main post here. In this follow up post, Jean S. quotes various emails from the climategate emails. The point of the post is to show that Michael Mann, consistent with his CV, actually participated in the discussion of how the “hockey stick” graph should be presented on the WMO cover figure.
8) I am fairly certain that Jean S. would agree with the conclusion of the Muir Russell panel that the graph “was misleading in not describing that one of the series was truncated post 1960 for the figure, and in not being clear on the fact that proxy and instrumental data were spliced together.” The point of Jean S.’s post, however, was not to defend this conclusion, but merely to show that Mann was complicit in the deception identified by the panel.
Rob Honeycutt, when you say non-hockey sticks have not been produced, I disagree. If many of the more recent papers that you refer had been first produced, then they would not be known as hockey sticks. The original MBH papers were much flatter. MBH 98 only goes back to 1400, so no warm period can be expected, though the absence of an extended cool period is surprising there.
http://www.skepticink.com/prussian/2014/09/21/yellow-pages/
Kevin ONeill, you say hide the decline refers to the blade, but the issue is that the decline signals something wrong with the proxies. My theory is that without Mike’s Nature Trick, the chart itself would have been thrown off the front cover, as Phil Jones said there was another candidate(I think Phil wanted to go with the other chart and was outvoted). So if there is a decline now that needs to be hidden, perhaps past declines or non warming are also really increases.
Greg Laden, are these in your opinion the strongest responses to the issue of divergence? I’m surprised there was no Briffa paper. I have read some, but I have yet to see a good explanation. There are two separate issues.
1) Reasons for the divergence. I have seen a number of theories presented, including because of global warming or CO2. However, the issue for skeptics is
2) How do we know the divergence is recent?
Much of what I am seeing is asking about causes while assuming it is recent.
One paper by Briffa says
In the absence of a substantiated explanation for the decline, we make the assumption that it is likely to be a response to some kind of recent anthropogenic forcing. On the basis of this assumption, the pre-twentieth century part of the reconstructions can be considered to be free from similar events and thus accurately represent past temperature variability.
This is an assumption not evidence. I’m wondering if there is evidence. Something like what carebannog wrote above. Has this been done? The different proxies look so different from each other, as well as the reconstructions, so it is not immediately obvious.
So Wilson 2007 and Darrigo et al are the best explanations?
John Mashey, the cherrypicked comment doesn’t apply. When Nick Stokes put up an alternate version, I didn’t realize he had actually taken a random sample out of 10000 to get hockey sticks rather than the top 12 as McIntyre did. The results were that similar. Even DeepClimate’s post still shows hockey sticks when you look at the full random sample, as DC acknowledged. It’s just a few are upside down which is irrelevant in THIS Mann paper.
> One could always rank proxies as to how reliable they may be, and in fact, that is part of using them carefully.
That is a problem with many paleoclimate papers that has been criticized as well. The issue when you do this is that you end up getting hockey sticks even from random data. If you weed out those that show warming in the modern period, to match your known correct data, and then average everything together, then even random data will produce a hockey stick, as the past portions will average out to zero.
Mann 2008 suffers from the same complaints that Mann describes in his paper. Even MBH used some precipitation proxies for which he criticized Soon & Baliunas.
Rob Honeycutt writes: “You’re stuck in a time warp, AMac. Science moves forward. It really wouldn’t matter if Mann has done every single one of his proxies wrong because one single paper does not define the whole of any issue. Not even Einstein’s relativity papers do that. Science is the full body of research, the full body of scientific knowledge.”
Although AMAC is more than capable of defending himself, I suspect he has become weary of doing so. Let me chime in.
AMAC has made two basic criticisms of Michael Mann’s use of the the Tiljander Varve proxies. First, he points out that the scientists who collected the Tiljander Varve samples acknowledged upfront that they were contaminated by land use changes during the modern era when actual temperature records exist. This is not disputed. AMAC then makes the obvious point that it is not possible to properly calibrate the Tailjander proxies to the modern temperature record in a way that would allow them to be used as temperature proxies–any correlation between the Varve thicknesses and the modern temperature record would necessarily be spurious. I have not seen anyone attempt to refute this obvious point.
AMAC’s second point is that things are even worse than they appear at first glance. The spurious correlation between varve thickness and the modern temperature record assigns to varve thickness a sign that is the opposite of physical theory that has been proposed by its collectors to justify its use as a temperature proxy. Again, I have not seen anyone who has attempted to refute this point that was first identified by Steven McIntyre.
Neither point is trivial. Michael Mann used the Tiljander “proxy” in a study that he says is a reconstruction that results in a hockey stick graph without using controversial “bristle cone pine” tree rings. Steven McIntre has emulated Michael Mann’s reconstruction and claims that his reconstruction that includes Tiljander but excludes “bristle cones” fails tests of statistical significance. Michael Mann’s colleague, Gavin Schmidt, has reluctantly conceded this point in a comment buried deep in the comment section of RealClimate.org .
AMAC’s criticism should have required that Mann issue a public acknowledgement of the error or at least provide a demonstration backed up with data and code to show that the error was not significant. This has never occurred. In fact, Mann’s reconstruction continues to be cited. AMAC has not let the error slide by because it is telling about the current state of Paleoclimatology that this error has never been acknowledged or corrected in the scientific literature.
In any other field, Mann’s study would have been widely rejected by the scientific community. The fact that it has not been says much about the field, which is why AMAC has continued to make his points.
I acknowledge not providing specific hyperlinks to back up the above assertions. Anyone who interested can use google to find the appropriate references in AMAC’s previous writings. As far as I am concerned, I have already spent too much valuable time making these points.
Jim said… “…but it is *not* the same graph line as the MBH hockey stick.”
It doesn’t have to be the exact same line, Jim. No one expects that. It only has to support the base conclusions that global temperature slowly fell during the past 1000 years and then rose dramatically when we started burning fossil fuels, and that the current global temperature is likely greater than anytime during the past 1000 years.
MikeN @106… You completely fail to understand the fundamental question at hand. Is the global temperature today anamolous during the past 1000 years?
MBH98 went back to 1400 but MBH99 went back 1000 years, which is why the two papers are generally referred to as MBH98/99.
I’ll make my statement again: No one has produced a global or NH multiproxy temperature reconstruction that has shown anything OTHER than a hockey stick.
If you are defining hockey stick as current temperatures are warmer, that is one thing, and is generally where the dispute lies. I was looking at what the Prussian posted, and pointing out that the newer papers, if Mann in 1998 and 1999 produced something that looked like that, no one would have called it a hockey stick.
MikeN… (sigh)
THAT IS what the entire issue is about. THAT is what the hockey stick is. It is a global or hemispheric multiproxy reconstruction showing that temperatures since the start of the industrial revolution have risen dramatically in response to man-made carbon emissions.
Yes, the various reconstructions have different shapes for the “handle.” That is a meaningless point. You expect that because of the fact that they’re using different proxy sets. But the fundamental conclusion remains the same with every reconstruction that’s been done.
> All proxies are collected locally and reflect some range of catchment for the thing being measured
So you think it would be inappropriate to suggest that proxies respond to global temperature rather than local temperature?
MikeN… All proxies are measuring local and regional temperature based on the type and method. That’s why multiproxies are used to look at global temperature.
Get that? MULTI. They’re taking a large number of individual proxy sets and combining them together to estimate global temperature.
Mike… In fact, go back and look at the PAGES2k paper. They show you the different areas where the proxies came from. And even the regional sets are formed by looking at a variety of individual proxies in order to get a distributed measurement of local and regional temperatures over time.
The thing here is this: Scientists know what they’re doing. The earned their PhD’s doing this work and have a deep basis of knowledge to call upon to perform their work. They know the challenges. They know what works and what doesn’t. They account for all this in order to pull together the best possible representation of global temperature over the past 1000 years.
AMac @103… Yes, and again, there you are five years ago trying to hash out answers to issues you don’t understand relative to a paper that was published 10 years before that. And here you are today, still trying to hash out those exact same issues that you don’t understand.
We’re now 15 years down the road on this one, AMac. We’re now a dozen multiproxy reconstructions past MBH98/99. And we now have a far more comprehensive project before us with PAGES2k, which is supporting Mann’s conclusions from 15 years ago.
That puts us back to the question we started with. Is your complaint that somehow tree ring proxies are giving the right answer for the wrong reasons?
If you’ll pardon the pun, this seems to me to be a massive case of missing the forest for the trees. You can wrestle with Tiljander for the next 20 years and it’s just not going to change the fundamental conclusions of the full body of scientific research on the issue of millennial temperature.
Wow. Reading this comment section is like falling into a sciency trope salad drenched in FUD dressing.
Nothing like the Internet to generate instant scientists spinning out over-generalized hypotheticals and gabbling about their myopic obsessions.
McIntyre: a fancy pants Dr. Oz peddling climate woo.
MikeN writes:”Kevin ONeill, you say hide the decline refers to the blade, but the issue is that the decline signals something wrong with the proxies. My theory is that without Mike’s Nature Trick, the chart itself would have been thrown off the front cover…”
First, who cares what chart was chosen for cover art? Seriously, who cares. We’re talking science, not cover art.
Second, ‘hide the decline’ refers to the divergence problem. A KNOWN problem at the time. The proxies weren’t telling them something they didn’t already know. They already knew the tree ring data past 1960 was no good. The trick was to produce data up to the present day, even though there wasn’t a proxy available with high correlation to temperature past 1960. Obviously the instrumental data filled the void.
Did you read the analogy I posted upthread regarding thermocouples? Is a scientist/technician using a Type C thermocouple at 2000ºC (when he was using a Type K for previous measurements) doing something wrong? Of course not. If he used the Type K at 2000ºC he’d get spurious results by using the Type K outside it’s known range and be reporting bad data. This is what you want MBH to do – use a proxy outside its known range and then report the bad data.
Let me repeat: You want MBH to have used a proxy outside its known range and then report the bad data. Why? Why do you want them to publish bad results when good results (the instrumental record) was available?
See my post above about divergence. The analogy with the thermocouple doesn’t hold up, as the dispute is that the proxies are not valid at older times as well. Started with an analogy to your thermocouple, but it breaks down, because you know your thermocouple is valid at a certain range. I guess the best is what if you were using that thermocouple out of range. How would you know that?
>You want MBH to have used a proxy outside its known range and then report the bad data. Why?
To evaluate the validity of the proxy in past times, it is best to show all data yes. You then have to be able to justify what looks like a bad proxy, as to why it is valid. By ‘hiding the decline’, you are avoiding such questions.
Incidentally, you say it is just cover art, but as shown in the post Laden linked, “will help to bolster the claims to be made in IPCC.” I guess you could say the hockey stick itself is just cover art as well.
Rob, I think we are in agreement as to what is the issue. I was merely referring to the phrase ‘hockey stick’, as the Prussian was doing.
With regards to global and local temperatures, I just threw that in there without describing the issue. Note I said ‘respond.’ Do proxies respond to global temperature, or local ones? Some of the hockey stick papers have proxies responding to global temperature rather than local. See ‘teleconnection’
MikeN… “To evaluate the validity of the proxy in past times, it is best to show all data yes.”
No. That is not how proxy data are validated and calibrated, or when necessary, cross-validated.
You just can’t seem to get past the fact that researchers know what part of the NH proxies are wrong. They can just omit those and proceed to calibrate and validate the remaining data.
Why is that so hard to comprehend?
Obstreperous Applesauce
I agree with your comments about this discussion, but not when you say
“McIntyre: a fancy pants Dr. Oz peddling climate woo.”
This discussion has absolutely nothing at all to do with McIntyre or the post at ClimateAudit, which was solely about WMO99 and Mann’s contribution to it, which Mann claimed on his CV but denied in his court submission.
This is at least the third instance of Mann being, to coin a phrase, economical with the truth in his court submission. Firstly he claimed to be a Nobel prize winner. Secondly he claimed that Oxburgh “exonerated” him when in his own book, Mann said that Oxburgh had not investigated his work. Thirdly he denied any involvement in authorship of WMO99 even though he listed it on his CV and the Climategate emails (analysed at CA in the post under discussion) showed he was consulted and made suggestions to improve it.
MikeN… re: teleconnection
But no one is heavily promoting any single proxy as a definitive global proxy. It’s a footnote at best.
AndyL… “This is at least the third instance of Mann being, to coin a phrase, economical with the truth in his court submission.”
You’re spending waaaay too much time drinking McI’s koolaid.
Just wait and see how the judge rules on McI’s niggling points. Judges have very very little patience for his manner of BS.
The fact with the WMO99 graph is, it was Jones’ project. Not Mike’s. It was a cover graphic. Not a research paper. And that’s about as much as the judge is going to look at.
Rob,
You say: “You just can’t seem to get past the fact that researchers know what part of the NH proxies are wrong. They can just omit those and proceed to calibrate and validate the remaining data.”
The sceptic point of view is to ask in what order those things happen. If scientists could determine which proxies were valid prior to calibration, there would be less concern. However what seems to happen is that the selection happens during the calibration phase, which leads to over-fitting. While sceptics on site such as CA are not all climate scientists, there are some strong statisticians who can analyse the impact of this. Many climate science papers do not have statisticians on their author list, despite the complex analysis involved.
Meanwhile I refer you to what Jim Bouldin said earlier:”Paleo T reconstructions from tree rings is probably the single weakest part of climate science, or certainly top three.”
Rob Honeycutt, why would you state that the shape of the ‘handle’ of a climate reconstruction does not matter? The only reason there is rough coherence in the ‘blade’ is because of arbitrarily offsetting each reconstruction to align with instrumental. There is uncertainty in this arbitrary offset which incoherent ‘handles’ actually demonstrate. One should also consider that many reconstructions are built not just by aligning proxies with instrumental but scaling as well (eg composite plus scale) – an action which amplifies noise and therefore shrinks the scale of the signal (in the handle).
Greg Laden, your suggestion that proxies should be handled carefully (ie cutting off ‘divergence’ or using correlation based proxy selection) makes sense intuitively but, as Jim Bouldin mentioned, causes a selection bias. In such cases the bias is in the composite error term (selection of proxy ‘noise’ which correlates with temperature). IOW the composite error of the proxies is not 0 (not unbiased) which by definition shrinks the scale of the scale of the signal in the reconstruction.
>You just can’t seem to get past the fact that researchers know what part of the NH proxies are wrong. They can just omit those and proceed to calibrate and validate the remaining data.
Why is that so hard to comprehend?
See my post above on divergence. You say researchers know what part of NH proxies are wrong. I have asked for what is the evidence of this knowledge. I will concede that if it is known that the proxies are accurate in earlier times, then the divergence would not be an issue. If ‘Hide the decline’ refers only to the divergence, then that would also not be an issue.
Rob Honeycutt writes: “AMac @103… Yes, and again, there you are five years ago trying to hash out answers to issues you don’t understand relative to a paper that was published 10 years before that. And here you are today, still trying to hash out those exact same issues that you don’t understand.”
Please describe what issues AMAC doesn’t understand. I have read the Stoat post that you reference in comment 85 and it seems to me that AMAC describes the issues accurately and in detail. More importantly, his comments are never effectively refuted.
I have attempted to describe the issues that AMAC raises in my post #112. If I am missing something important, please inform me.
The fact that these issues are somehow the subject of continuing debate is baffling to me and, as I have previously asserted, a sign of the something deeply wrong in this field.
With regards to teleconnections, I threw that in because of something Greg wrote above. The actual impact on hockey sticks papers I think is more in terms of getting more proxies to pass validation. However, would you agree that it is a method error to validate a proxy if it agrees with global temperature instead of local temperature?
AndyL…
A real skeptic would read the research before coming to a conclusion. A real skeptic would not automatically assume that researchers don’t know what they’re doing.
Look, there may be strong statisticians there but that doesn’t mean they understand the nuances of the underlying data or how to process the data without introducing their own errors.
And in reference to Jim Bouldin, I’ve asked this same question about 10 times now: I’m showing you the PAGES2k research, which include a wide range of proxy date other than tree rings, and they’re coming up with the same fundamental conclusions as Mann’s original 98/99 research. So, does that mean your issue is that tree ring reconstructions are turning up the right answer for the wrong reasons?
MikeN… “I have asked for what is the evidence of this knowledge.”
Greg posted two papers on the divergence problem above at 51 and 52. The issue was widely discussed within the field at the time of MBH98/99.
Rob (comment 128)
I don’t know how the judges will react to being misled by one of the parties, but at some point in this case Mann’s credibility will come under question..
Regarding the criticism by Muir Russell of the WMO Cover, which they described as “misleading”, Mann’s submission to court says
“This allegation is yet another example of Defendants’ attempts to obvuscate the evidence in this case. The “misleading” comment made in this report had absolutely nothing to do with Dr Mann or with any graph prepared by him…. Dr Mann did not create this depiction, and to state that this report suggested an effort by Mann to mislead is disingenuous.”
The claim that the WMO cover was “absolutely nothing to do” with Mann is demonstrably false.
MikeN… Here is the key paper that was out on the divergence problem prior to 98/99.
Jacoby, G. C.; d’Arrigo, R. D. (June 1995), “Tree ring width and density evidence of climatic and potential forest change in Alaska”, Global Biogeochemical Cycles 9 (2): 227
And also discussed by Briffa in his 1998 paper in Nature.
Briffa, Keith R.; Schweingruber, F. H.; Jones, Phil D.; Osborn, Tim J.; Shiyatov, S. G.; Vaganov, E. A. (12 February 1998), “Reduced sensitivity of recent tree-growth to temperature at high northern latitudes”, Nature 391
The problem for Warmists like Greg is that Michael Mann has already admitted that his WMO 1999 Hide-the-Decline presentation was misleading. But then he claimed to have nothing to do with it, a point clearly refuted by the timeline and emails in the article by Jean S. Mann is in deep doo-doo.
This also came out in Jan 1998…
http://rstb.royalsocietypublishing.org/content/353/1365/65.short
Rob 132:
you way “Look, there may be strong statisticians there but that doesn’t mean they understand the nuances of the underlying data or how to process the data without introducing their own errors. ”
I belive contributors at CA do understand the nuances of the underlying data at least as well as those compiling multiproxy studies, who did not gather the original data. Two examples, both of which were strong contributors to the “hockey stick” shape:
Tiljander was used by Mann despite the original author stating the data was contaminated in the calibration period.
Braybill, where the CA team found the original trees and re-cored them. Braybill considered the trees to be precipitation proxies, and the NAS panel later said that strip-bark proxies should not be used.
The “misleading” comment was not directed at Dr Mann.
Mann did not create the WMO99 graphic.
Both correct statements.
AndyL… Let’s step back a moment and look at the big picture.
1) Do any of the issues that have been brought up (right or wrongly) change the conclusions of Mann’s research from 15 years ago? No.
2) Has subsequent research shown Mann’s work to be incorrect in it’s conclusions? No.
3) Has subsequent research actually confirmed, repeatedly, the original conclusions of Mann’s research? Yes. Over a dozen times now.
4) Are there any NH or global multiproxy reconstructions that refute Mann’s original conclusions? No.
Are you claiming that Mann mislead people to the correct answer with his original research?
FreedomFan… “The problem for Warmists like Greg is that Michael Mann has already admitted that his WMO 1999 Hide-the-Decline presentation was misleading.”
(sigh)
Let’s go over it again. The “decline” is known to be wrong. We know this because there are thermometer reading that tell us this. The “decline” is only in NH series.
So, are you telling me that it was wrong to remove data that was wrong?
“Not knowing exactly why a proxy goes belly up does not allow a scientist to pretend it has not gone belly up, though. You understand that, right?”
Not knowing exactly why a proxy goes belly up does not allow a scientist to pretend that it hasn’t gone belly up in the past, though. You understand that right?
You are still missing the point entirely. Let me see if I can say it in a way that relates directly to your question.
All proxies have the potential of going belly up for any particular time period or region. All of them. The methodology of paleoclimate reconstruction attempts to address this. If we disregarded a particular category of proxy because it has gone belly up now and then or because we suspect it could, there would be absolutely no paleo reconstructions.
This does mean that if your standard for science is that data are perfect, perfectly reliable, never subject to question, then you can’t accept any of the science done in a very large region of scientific study, including pretty much anything that has to do with natural system. None. It is all invalid on that basis.
And that, Tom, is where you are at. That and that alone is the basis for the argument.
#134 RH
I suppose that you think that the correlation screening used by Mann represents an “understanding” of such “nuances”.
http://climateaudit.org/2012/06/17/screening-proxies-is-it-just-a-lot-of-noise/
Can you spot the green hockey stick in the head post?
This pst at RealClimate may help people understand how proxies generally, and tree-ring data in particular, are addressed, what some of the methodological issues are, and how their role in climate reconstruction has become more useful over the years: http://www.realclimate.org/index.php/archives/2013/06/yamal-and-polar-urals-a-research-update/
“If we disregarded a particular category of proxy because it has gone belly up now and then or because we suspect it could, there would be absolutely no paleo reconstructions.”
Perhaps we should. Perhaps outside of ice cores we really don’t have anything that is very reliable. Of course you don’t like that because the ice cores don’t say what you want them to say.
Tom t, nowhere did I suggest that proxies from ice cores (of which there are several … an “ice core” is not a proxy, actually a tree ring sequence isn’t either btw) is immune to the basic issue that proxies are informants that have to be put under a hot light, their stories verified, and never quite fully trusted.
RomanM… Yes. I believe McI doesn’t comprehend the nuances of tree ring data (explained very well by the RealClimate post Greg just linked to) and Mike Mann does understand them.
Greg ice cores differ from tree rings for one very important reason. They are scientifically verified. We understand the process that causes light and heavy oxygen to precipitate out of water vapor as it makes its way to the poles. We have nothing like that with tree rings.
Tom T:
Ice cores aren’t global proxies.
And the Greenland ice core data so beloved of denialists, is typically charted on a graph that ends in 1855 …
Rob Honeycutt, Mann’s reference to misleading and the panel’s reference are because of Mike’s Nature Trick. Phil Jones did a modified version that has all the proxies go up at the end, despite no proxy data.
Tom T… Ice cores also have their challenges though, with melt water contamination and such. When you get good samples they are very accurate proxies (and yet still frequently misused by “skeptics”). The problem comes with the fact that most ice core data is confined to high latitude and high altitude locations, so their coverage is limited.
No one makes the claim that tree ring temperature proxies are perfect. But they are useful to expand the coverage for millennial reconstructions. And that is why I keep bringing up the PAGES2k project.
The PAGES2k reconstruction is showing a very straight handle and a very definite blade, very similar to the original Mann hockey stick. So, with more coverage and a wider range of proxies, it seems we’re getting something that is lending credence to Mann’s work, perhaps over the other multiproxies done over the past 15 years.
If you’re honest, you have to consider the possibility that Mann, Bradley and Hughes we actually very skillful in their use of the tree ring data they had at their disposal at the time.
MikeN… How many times does this have to be explained?
Mike’s Nature trick is merely the “trick” (as in “clever thing to do”) of using instrumental data at the end of a multiproxy reconstruction.
Phil Jones used that clever application of data for the WMO99 cover graphic. Mike didn’t apply the method to the WMO99 graphic. Phil did it. The panel was saying that the way Phil applied Mike’s trick was misleading.
Do you understand the differences in how the two applied the trick? Or do I need to explain that too?
Thanks for the links on divergence. The last paper you link is evidence in favor of the skeptic viewpoint. Despite the title, it spends more time on other matters. When it gets to the issue in Section 5,
“The implications of this phenomenon are important.
Long-term alteration in the response of tree growth to climate forcing must at least to some extent negate the underlying assumption of uniformitarianism which underlies the use of 20th century derived tree growth climate equations for retrodiction of earlier climates
.
At present, further work is required to explore the detailed nature of this changing growth climate relationship.
It is possible that it has already contributed to some
degree of overestimation in published reconstructed temperature means.”
Note that this divergence is in the opposite direction, helping to create hockey sticks, rather than a decline.
Actually it does need to be explained again. Please explain the difference in how the two used the trick as best you can.
Beyond that, what I need explained again is what is clever about this trick of putting the instrumental data there, at least as Mann used it, the manner in which it would not have been ‘misleading’.
MikeN…
The point is, the issue was known at the time and was openly discussed in the literature and between researchers who work in the field at the time of the WMO99 graphic. They all knew the data after 1960 on Keith’s reconstruction diverged from the instrumental record and was, therefore, wrong.
So, please tell me, what would be the point to including the misleading divergence on a graphic intended for a lay audience?
Ice cores aren’t global proxies
First they are both hemispheric and local. It depends on the method used to analyze the oxygen isotope ratio. Some methods are more representative of the water cycle to the point of condensation others more to the local temperature at point of precipitation. Regardless the general physics of how ice cores work there will be both a local and hemispheric signal in the isotope ratio.
“So, please tell me, what would be the point to including the misleading divergence on a graphic intended for a lay audience?”
To let the audience know that your method isn’t a fool proof as you are trying to claim. If you intent its to mislead your audience through a trick then deleting it is a good choice.
MikeN… There is a difference in how Phil and Mike applied the trick. In Mike’s work the instrumental data is labeled as such. In the graphic that Phil created, since it wasn’t a research paper, wasn’t, so the viewer would not understand what was proxy and what was instrumental. THAT is what they were saying was misleading. Not Mann’s use of the trick, Phil’s reinterpretation of the trick.
None-the-less, it was a graphic for the cover of a report for a lay audience (who could never be adequately informed of what a “divergence problem” is). It was not a graph for a research paper. And, it is still, graphically, consistent with a wide range of millennial reconstructions and doesn’t mislead viewers about the basic fact that global temperatures have risen dramatically over the past 150 years.
Tom T… “To let the audience know that your method isn’t a fool proof as you are trying to claim.”
But that would also be a misleading statement since the divergence problem is limited to the scope (in Briffa’s data set in question) to the period after 1960. The data prior to that is no in question and is, in fact, presented in full detail in Briffa99.
It would be far more misleading to include the divergence.
“MikeN… There is a difference in how Phil and Mike applied the trick. In Mike’s work the instrumental data is labeled as such. In the graphic that Phil created, since it wasn’t a research paper, wasn’t, so the viewer would not understand what was proxy and what was instrumental. THAT is what they were saying was misleading. Not Mann’s use of the trick, Phil’s reinterpretation of the trick.”
Mike didn’t seem to have a problem with it. He even asked or a T-shirt. A classic method of propaganda is to move the ball. Mike labeled the trick. Jones chose not to label it Mike saw no problem with it. Eventually “All animals are equal, but some animals are more equal than others”.
The difference between the nature graph and the WMO graph is just an example of chipping away at the rules over time.
”
Tom T… “To let the audience know that your method isn’t a fool proof as you are trying to claim.”
But that would also be a misleading statement since the divergence problem is limited to the scope (in Briffa’s data set in question) to the period after 1960. The data prior to that is no in question and is, in fact, presented in full detail in Briffa99.
It would be far more misleading to include the divergence.”
No not mentioning it is misleading because you are pretending it doesn’t exit. The correct thing to do is present it and discuss it. Unfortunately, the audience might begin to think you are full of it.
Greg
“You are missing the point, and at the same time making it. All proxies have to be interrogated and evaluated whenever and wherever they are used. It is simply not the case that a proxy is thought up then used uncritically whenever it is found.”
It is simply not rational to pick a proxy group, in this case trees, pick the ones that match a global temperature trend over a short period of their record, say they are good proxies and include them, while dropping proxies which do not match, then say the resulting blend in some way can be used to hindcast temperatures.
A very basic problem with this is that no group can contribute a negative or flat trend (and as we all know, some areas warm, some are roughly flat and some even decline, even as in general temperatures warm). So good proxies could very realistically be dropped for accurately proxying their non rising local temperature merely because they did not follow the global temperature.
A further very basic problem all accept that using tree rings are very noisy proxies, even groups of them. In picking your “good” proxies you do know if it is the noise that is making a bad proxy match. This of course means your hindcasting on the back proxy is poor. Likewise it could mean that noise in the matching period caused what would have been a good proxy for hindcasting to be dropped because it did not match.
A third serious problem is of course that too warm is not good for tree ring growth, especially if there is a lack of precipitation. There is in reality no way to tell the difference between (cold and wet) and (warm and dry). Neither will produce as good growth as (warm and wet). One could even argue that the “divergence” problem was due to increasing temperatures hampering growth, but this of course would render hindcasting impossible as one would never know whether low growth was due to too much warmth or cold.
There are many other problems. I think that in general these types of problems can be summed up by the withdrawn Gergis et al paper. They produced a hockey stick Southern Hemisphere record, attempting to avoid a couple of the above problems (by selecting against local temperatures and detrending the proxy and temperature data prior to evaluation to attempt to minimise the effect of spurious correlations). It was later pointed out that they forgot the detrending step after all, and that including it no longer produced a hockey stick. Rather than republish with the fixed methodology, showing a non hockeystick and if I remember correctly poor verification statistics, the paper was quietly pulled. So they originally published with good verification statistics even though they had failed to follow their detrending step, yet when the detrending step was performed the verification statistics were very poor. Why was it when they failed to follow their planned methodology they got such good verification statistics, yet when they followed it they got such poor statistics ? Did the proxies suddenly go from being good proxies to being bad proxies ? No, the answer is that in reality you do not really know what are good and what are bad proxies because the data is so noisy.
Tom T… I’m also going to keep hammering on this, but if you think that tree ring proxies aren’t as accurate as people are lead to believe… For one, that’s a red herring. Tell me how accurate are people being lead to believe they are? The uncertainty ranges are well represented in the research. And second, look at the PAGES2k data! It’s showing nearly exactly the same thing as Mann’s original work, but is based on a wide range of proxies.
If you believe that tree rings are so problematic, then how the heck did Mann get results, 15 years ago, that so closely match far more extensive reconstructions done today?
Tom… “Mike didn’t seem to have a problem with it.”
It was a cover graphic.
Mikes pat on the back and asking for a T-shit is prima facia evidence that Mike only labeled his nature trick because he had to. I’m sure he felt that when first introducing it he had to label it or the paper would have been rejected. However, its clear in the e-mails that he viewed Jones decision not to label it a good idea.
Tom… “No not mentioning it is misleading because you are pretending it doesn’t exit.”
For a lay audience it would have been misleading because temperatures did NOT fall post 1960. They clearly didn’t pretend it didn’t exist because there are all sorts of research papers on the divergence problem at that time. They were discussing it publicly.
“The uncertainty ranges are well represented in the research.”
But not in how they presented it to the public. In the business world if you leave things of such significance out of your 10k you go to jail.
Tom… “However, its clear in the e-mails that he viewed Jones decision not to label it a good idea.”
It was a cover graphic. (rollseyes)
The presentation of millennial reconstructions showed then, and show today, that modern global temperatures are higher than they have been over the past 1000 years.
That is the subject of the report. That was an accurate statement then. It is an accurate statement today. The cover graphic represents that conclusion.
The report is not about nuances of dedroclimatology. For that one would read the relevant research.
We’ll see, Tom. My guess is Steyn, CEI and the NR don’t have a leg to stand on here. This issue has been investigated numerous times now and NO ONE has come to the conclusion that you’re presenting. Zero. I’d say the judge is going to make a very similar investigation and come to a very similar conclusion.
Tom… “But not in how they presented it to the public. In the business world if you leave things of such significance out of your 10k you go to jail.”
Interesting.
Please explain: How would one go to jail for not showing data that was wrong?
“It was a cover graphic.”
So the trick was just that a trick. An attempt to trick the reader into thinking that the method was more reliable than it actually was.
You do that in any public SEC filing and you go to jail. Why do you think its okay for climate scientists to do it?
The “trick” was a kludge to make the numbers that were scientifically appropriate go into the reconstruction and not the numbers that were not scientifically appropriate. By now, you know this. But you refuse to acknowledge it. Why is that?
@pauldd #112 —
It is gratifying to see that you have grasped the significance of the Tiljander saga that Prof. Mann initiated in 2008. Your comment #112 is also an effective rebuttal to the criticisms offered by Rob Honeycutt at #120. (BTW, that sequence suggests that you have invented a time machine.)
In #132, you note, “The fact that these issues are somehow the subject of continuing debate is… a sign of something deeply wrong in this field.”
Yes. My opinion: every physical science’s primary literature includes papers with bad data, inappropriate analysis, and wrong conclusions. That’s not news. Here, a key opinion leader has doubled down on his mistakes for six years. He’s been explicitly supported in this venture by other opinion leaders, and been accommodated by the silence of journeyman scientists who are in a position to know better.
That’s newsworthy,
Rob
The accuracy or otherwise of Mann’s research is (legally) not the point – either to the court case or to the post at CA.
Steyn et al called Mann’s hockey stick ‘fraudulent’ – which is common usage, just as Mann labels other scientists fruadulent or anti-science. Mann claims that CEI, Steyn etc were speaking with “malice” in the legal sense because they knew he had been “exonerated” by around 8 investigations..
Mann has proven himself an unreliable witness. His claim of having “absolutely nothing to do with” the WMO chart is just one example.
It turns out that of the 8 investigations, only one actually looked into Mann’s work. Mann is on record as knowing that one of the 8 studies (Oxburgh) did not investigate him, but still included it in his court submission. The one study that did investigate Mann was Penn State – and they can hardly be considered the benchmark for thorough investigations.
Steyn, CEI and the NR are not the ones short of legs to stand on.
AMac… “Here, a key opinion leader has doubled down on his mistakes for six years.”
Yet again here, AMac, you’re lost in this time warp. Science moved on. It’s 15 years later! Has subsequent research shown Mann’s work wrong? No.
Essentially, what you seem to be saying is, you think that Mann used bad methods to get the right answers.
Call me unimpressed.
“The “trick” was a kludge to make the numbers that were scientifically appropriate go into the reconstruction and not the numbers that were not scientifically appropriate. By now, you know this. But you refuse to acknowledge it. Why is that?”
But the numbers went into the reconstruction. It was plain as day in Briffa’s paper. The trick was to delete it so the general public who aren’t going to look at the primary source wouldn’t see it.
Most SEC filings are rather complex and intended for public consumption. If you engage such tricks of such significance in your public SEC filings you go to jail. Why is it okay for climate scientists to do it?
The excuse that it was okay because it was intended for the lay public does not fly.
AndyL… We’ll see, Andy. I’m watching this case with bated breath.
Tom… “If you engage such tricks of such significance in your public SEC filings you go to jail.”
Again, so you go to jail for not including incorrect data?
This is a new rule that I’ve not heard of.
Tree Rings and Growth, Fritts (1976 and 2001). Perhaps the most cited reference in dendroclimatology. Explains the climate/tree growth relationship, response functions, reconstructions of climate, and basic tree physiology concerning the formation of annual rings.
A TIME SERIES ANALYSIS APPROACH TO TREE RING STANDARDIZATION, Cook (1985) over 1300 citations.
Fundamentals of Tree Ring Research, Speer, 2010. One of the better textbooks in the field.
Two books and a PhD thesis just to start with. But at least once having read them you’ll have a basic understanding of the common field methods and analytical techniques used.
There are of course hundreds if not thousands of papers relating to individual tree species, regions, and statistical techniques. Apparently some of our pseudoskeptic bretheren think this stuff just pops up out of thin air.
Unfortunately it’s a common tactic of pseudoskeptics to just talk out of their ass without actually knowing anything about the science involved. It’s a major effort for them to scan abstracts and quotemine – much less actually read a paper (god forbid a textbook!!!) and understand it.
And of course the fallback position is someone else must explain it to them in words of no more than two syllables – rather than them actually going out and learning on their own. Ignorance is their position of strength.
AndyL… “It turns out that of the 8 investigations, only one actually looked into Mann’s work. […] The one study that did investigate Mann was Penn State…”
Are you sure about that? This sounds like an investigation into the hockey stick to me:
The NRC Report: “The North Report concluded “with a high level of confidence that global mean surface temperature was higher during the last few decades of the 20th century than during any comparable period during the preceding four centuries”, justified by consistent evidence from a wide variety of geographically diverse proxies, but “Less confidence can be placed in large-scale surface temperature reconstructions for the period from 900 to 1600″. It broadly agreed with the basic findings of the original MBH studies which had subsequently been supported by other reconstructions and proxy records, while emphasising uncertainties over earlier periods. The contested principal component analysis methodology had a small tendency to bias results so was not recommended, but it had little influence on the final reconstructions, and other methods produced similar results.”
Tom t, I was talking about Mann’s paper, as you know. There is no “the reconstruction”. For the last time, you are, well, frankly, lying about the term “trick.”
SEC? I think at this point you have just accused named or unnamed climate scientists do fraud. Try not to do that her, please. And by please I mean just don’t.
Kevin, another good one is Ray Bradley’s Paleoclimatology – Reconstructing Climates of The Quaternary(1999), as it covers various kinds of proxies.
That of course is the one that:
(a) Wegman and Said plagiarized, then distorted conclusions they didn’t like.
(b) And to add insult, mis-referenced the title as Quarternary when there used it for a table, where they also converted phenology as phonology.
Maybe we should be using phrenology.
I hate giving Rob the 14-year-old Hatchet Honeycutt any more attention. But I can’t resist.
1) Do any of the issues that have been brought up (right or wrongly) change the conclusions of Mann’s computer-generated research from 15 years ago? No
2) Have subsequent computer models shown Mann’s work to be incorrect in it’s conclusions? No.
3) Haev subsequent computer models actually confirmed, repeatedly, the original conclusions of Mann’s research? Yes. Over a dozen times now.
4) Are there any computer-generated NH or global multiproxy reconstructions that refute Mann’s original conclusions? No.
5) Will the computer-generated results ever get close to accurate? Or are there still Mountains of aspects to the climate that we simply Do Not Know Enough About?
Greg Laden, simply reading this thread would suggest that that is not what the ‘trick’ is. Now if you think you have it right, then you are in company with John Stewart , ‘trick is just a term for a clever way for scientists to – trick you- into not knowing about the decline.
Rob, you are slightly off in the description of Mike’s Nature Trick, as he did include instrumental data in the actual proxy data, which can be see in the data for the paper. Basically, it’s like he did ‘Phil’s Mike’s Nature Trick’ then cut off the reconstruction.
Still in your version, what is clever about that?
Rob, is there any technical detail where the North Report disagreed with M&M’s critiques? Went before Congress and did not dispute anything in the Wegman Report when specifically asked. Agreed with the tendency to create hockey sticks from random data. Agreed that strip bark should not be used. That they went in and agreed with the general idea of a hockey stick isn’t enough to clear Mann from being called fraudulent by an opponent.
AMAC @ 174 : I haven’t yet invented a time machine, but perhaps we e do occupy the same time warp. I imagine that we are not the only ones..
None1… Aaaah, right. I think you have enough self-inflicted damage on that one to suffice.
MikeN… Interesting.
First you say there was only one actual investigation (Penn State) and now you’re admitting there’s also NSC and Wegman.
None of these, now three, investigations found Mann’s work to be fraudulent.
MikeN – ‘fraud’ doesn’t mean you made an error, mistake, or used a less than perfect method. Nothing in the North Report indicates fraud. Nothing even says Mann was wrong.
BTW, the North Report (which you can read online) isn’t an examination of M&M – so they rarely get mentioned.
phi <- 0.9;
Impressive, no?
MikeN… “Rob, you are slightly off in the description of Mike’s Nature Trick, as he did include instrumental data in the actual proxy data…”
No, Mike. He did not. You’re probably looking at a B&W version of the graph where it’s hard to make out. If you look at a color version it’s very clear.
MikeN – Perhaps you ought to consider what the ‘auditor’ himself said, “Our article did not say that all simulations generated a HS pattern. The point was that MBH operations applied to red noise could generate high HSI-index results.” [emphasis mine] That’s the auditor’s response to M&M’s 100:1 cherry pick based on his HS index. Lame.
Also – note that pseudoskeptics never talk about magnitude. How are they related? Is the noise magnitude sufficient to skew the results outside the stated uncertainties? Besides, didn’t we agree earlier the blade came from the instrumental record? So what relevance does the HS shape have on the reconstruction?
And you don’t want to bring up Wegman – that gets really ugly for pseudoskeptics really fast.
Other than that – it’s 2014, time to find a new horse to beat to death.
Rob and Kevin, you don’t seem to understand that the correlation selection methodology of Mann is indeed “cherry picking” which is no different from what you find unacceptable in the M and M examples.
Were you able to find someone to explain to you what the post that I referenced earlier to you was all about? Rob’s irrelevant response indicates that he either did not read it or did not understand the issues.
Kevin: you might try M&M’s backgrounder, jan 27 2005.
““The main error affects a step called principal component analysis (PCA). We showed that the PCA method as used by Mann et al. effectively mines a data set for hockey stick patterns. Even from meaningless random data (red noise), it nearly always produces a hockey stick.” Page 1.
RomanM… And yet none of these investigations made any mention of “cherry picking.”
John Mashey… I thought the whole thing about random data producing a hockey stick was actually an error that was introduced by McI (the statistical expert that he’s purported to be) adding the residual HS signal into the red noise because of improper filtering. I can’t remember off hand where I read that, though.
Doh!!
Wikipedia, but well cited…
“To demonstrate that that some simulations using their persistent red noise “bore a quite remarkable similarity to the actual MBH98 temperature reconstruction”, McIntyre and McKitrick produced illustrations for comparison.[120] Figure 4.4 of the Wegman Report showed 12 of these pre-selected simulations. It called this “One of the most compelling illustrations that McIntyre and McKitrick have produced”, and said that the “MBH98 algorithm found ‘hockey stick’ trend in each of the independent replications”.[132] McIntyre and McKitrick’s code selected 100 simulations with the highest “hockey stick index” from the 10,000 simulations they had carried out, and their illustrations were taken from this pre-selected 1%.[133]”
Great read, Greg. Hopefully the debate will improve now that someone competent has defined “proxy.”
I just wanted to take a minor issue with a couple of words in your explanation: “… such that the variation may correspond to a natural variation happening in the world…” What climate scientists like Dr Mann have done so much to help us understand is NOT “natural variation” but human-caused climate disruption. So your explanation is right, but fails to credit Dr Mann et al. with their biggest contribution! 🙂
In re the “decline”: what part of the phrase “hidden in plain sight” do deniers not comprehend?
Rob, the investigations concluded that pure red noise time series could indeed produce hockey sticks using the Mann methodology. The correlation selection procedures chose those series which matched the increasing temperatures during recent times, but were basically independent of each other during earlier eras thereby producing the flat hockey stick shaft. the math behind this is not that difficult to understand. They did not need to call it “cherrypicking” because it was obvious that the reconstruction process did exactly that.
What is this nonsense about “adding in the residual HS signal into the red noise through improper filtering”? First time I’m hearing something like that.
You also seem to be avoiding the issues I raised about the fact that Mann’s selection procedure causes serious bias in the reconstruction. I am getting the idea that statistics is not something you are comfortable with.
As far as
Hi RomanM
“Rob, the investigations concluded that pure red noise time series could indeed produce hockey sticks using the Mann methodology. ”
Yeah, so they COULD do it, so what? It didn’t do it EVERY time. The data that Mann had was selected on the POSSIBILITY it could represent temps. The data that M&M used was random. So in effect what they’re trying to sayis that proxy data is no better than random data (using his techniques). Now this COULD be true, I’m not convinced as the data Mann used had been used as temp proxies previously, but it’s irrelevant now as no one has used those techniques for 15 years. All the new techniques support the original work, so it doesn’t even matter if that work didn’t use good techniques.
It simply mad to keep banging on about someones work from 15 years ago, because you disagree with the technique. It’s like complaining that Einstein got his cosmological constant wrong. No one cares, and it doesn’t matter
RomanM
“The correlation selection procedures chose those series which matched the increasing temperatures during recent times, but were basically independent of each other during earlier eras thereby producing the flat hockey stick shaft.”
You need to be more specific here. He chose proxies to construct the series and then compared them to the reference period. And this is entirely appropriate – how else do you test your data? BUT he did not select the series that were ‘independent of each other’. He excluded series based on the reference period alone, not their behavior prior to the reference period. The behavior of the earlier data was emergent as he excluded those series that didn’t correlate with the reference period.
Just posting this, so we have a ref.
http://www.meteo.psu.edu/holocene/public_html/shared/articles/mbh98.pdf
Nathan, great point:
“It simply mad to keep banging on about someones work from 15 years ago, because you disagree with the technique. It’s like complaining that Einstein got his cosmological constant wrong. No one cares, and it doesn’t matter”
Deniers love to focus on the Hockey Stick and not the massive amount of non-Hockey-Stick science surrounding the Hockey Stick.
Deniers are so disproportionate. The Hockey Stick is a footnote, guys (a point Mann devotes several chapters to emphasising in his book “The Hockey Stick and the Climate Wars”).
Rob,
Why do you call the divergent, post-1960 dendro data “misleading”? It’s actually INFORMATIVE. It shows us how well that kind of data works as a proxy, and how much confidence we can put in it in the pre-instrumental period.
You seem to think Jones was showing a graph of temperature vs time. No, that’s only true for the post-1960 interval. The vast majority of the graph is not a graph of temperature over time at all, it’s a graph of tree-ring MXD over time.
You seem to think scientific communication is about conveying the truth. No. It’s about conveying knowledge, which is a bigger payload. As a scientist, one can’t just tell people the “answer” one got while lying about the means by which one got it. That’s (at best) uninformative, at worst disinformative.
I do wish people who don’t know how science works would recuse themselves from the debate. The invalidity of their arguments is going to embarrass their ancestors.
Rob,
There’s a really good deconstruction of what happened in M&M05 and the Wegman report here:
http://deepclimate.org/2010/11/16/replication-and-due-diligence-wegman-style/
McIntyre at CA: “Our article did not say that all simulations generated a HS pattern. The point was that MBH operations applied to red noise could generate high HSI-index results.” [emphasis mine]
Which is quite different from the link John Mashey points to: Backgrounder for McIntyre and McKitrick “Hockey Stick Project” where M&M say: ““The main error affects a step called principal component analysis (PCA). We showed that the PCA method as used by Mann et al. effectively mines a data set for hockey stick patterns. Even from meaningless random data (red noise), it nearly always produces a hockey stick.”
Of course the blade of the reconstruction is the instrumental record. So why any of this matters is beyond me.
Kevin,
Indeed, and what Deep Climate found was that by the words ‘nearly always’ he meant the top 100 of 10,000 simulations.
Greg, your argument on using only the proxies that agree with your premise and dropping them when they do not is appreciated.
Rob Honeycutt. where do you find these guys Greg? Thank you for making this blog so entertaining.You are on a par with Nick Stokes, I feel like asking Mosher to run a word simulation on the two of you to make sure you are not clones.
Re other reconstructs of global temp, Gerghis et al [retracted unfortunately] showed no SH Hockey stick and was the only comprehensive SH study. There was a Chinese study 3 years ago that showed a MWP from tree rings I recall.
There are a couple of other studies and a lot of anecdotal history to support a MWP.
The supporting studies for Mann which would only number about 8, not many really include Mann, Mann , Mann et al, Mann and Jones , Briffa and Jones and Briffa and Jones et al. A real echo chamber of support.
Not a trick, I emphasis.
PS Rob, trick is the wrong word for you to be using with Mann and Jones,machination perhaps or prevarication or anything else really.
“Greg, your argument on using only the proxies that agree with your premise and dropping them when they do not is appreciated.”
That is not what my argument is.
Yes, you can find a few studies that show something totally different from the hockey stick. This is why we do all of these studies, HONESTLY AND SCIENTIFICALLY evaluate them, and move eventually towards one or another consensus. Except the tin hat brigade, the bought and paid for skeptics, and the challenged.
Brad Keyes @205 writes: “Deniers are so disproportionate. The Hockey Stick is a footnote, guys] . . ..”
I would agree that the hockey stick is a footnote in the scientific literature. It is not, however, a footnote in the public debate.
To the laymen, hockey stick graphs supports a simple, easy to understand narrative. They show a relative stable climate until the industrial era and then a dramatic uptick during the period when greenhouse gas emissions increase significantly. I would not doubt that many lay opinions regarding AGW are based solely on their view of the hockey stick graphs. Indeed, that is why it was initially promoted with such enthusiasm.
Kevin O’Neil at 208 writes, “Which is quite different from the link John Mashey points to: Backgrounder for McIntyre and McKitrick “Hockey Stick Project” where M&M say: ‘The main error affects a step called principal component analysis (PCA). We showed that the PCA method as used by Mann et al. effectively mines a data set for hockey stick patterns. Even from meaningless random data (red noise), it nearly always produces a hockey stick.’ Of course the blade of the reconstruction is the instrumental record. So why any of this matters is beyond me.”
Agreed, high-frequency instrumental records dominate the blades of the hockey stick graphs. Moreover, since proxies are chosen or discarded based on how well they correlate with the instrument record, the proxy reconstructions will by logical necessity reproduce the blade.
The terms “hockey stick” pattern, however, refers not just to the blade, but also to the shaft, which shows relatively stable pre-industrial climate. That is where the controversy lies.
” that is why it was initially promoted with such enthusiasm.”
It’s a good thing then that legitimate research has supported the validity of the hockey stick, so that its interpretation and message have been supported.
Kevin ONeill, look at the DeepClimate post that John Mashey linked above. In it, he does a random sample, rather than 100 out of 10000, and acknowledges that there is still a tendency to create hockey sticks. It is clearly different from the proper PCA algorithm. Also note Figure 2 in this paper which shows how the 10000 split. Note almost all hockey sticks.
http://climateaudit.files.wordpress.com/2009/12/mcintyre-grl-2005.pdf
None, you mention how Gergis is withdrawn. Yet the withdrawn work was sent into Pages2K for the Australasia section.
Rob Honeycutt, to expand on what the Prussian has posted about Pages2K, which you seem to think validates the hockey stick with other proxies, let’s look through the different hockey sticks they created.
Antarctica, uses isotope series, no hockey stick.
Asia, tree-rings, no hockey stick
Australia, hockey stick, based on Gergis which was withdrawn.
Europe, hockey stick, treerings
North America pollen, no hockey stick
North America trees, no hockey stick but it’s close and doesn’t go back as far, surprising given they have plenty of overlap with MBH including bristlecones.
South America, lots of proxies no hockey stick but it’s close.
Arctic, hockey stick, also uses upside down data to get the hockey stick. Note that this author fixed Mann’s upside-down usage. http://climateaudit.org/2013/04/28/gifford-miller-vs-upside-down-kaufman/
However, the same author has produced a hockey stick in the past that held up better. Perhaps you should just use that instead of Pages2K, which as rejected by Science, and Nature took as a Progress Article and not a Research Article. Means ‘the field might not yet be mature enough for review’
http://www.nature.com/ngeo/authors/content_types.html
#191-192 Kevin & Rob,
I seem to have jumped into someone else’s discussion as I didn’t bring up the investigations. My point about fraud is that Mann’s brief and complaint says that the bloggers spoke with reckless disregard for the truth and should have known they were lying based on reading these reports. I point out that the North Report added credence to skeptic claims as well as Mann’s and thus a blogger would not be acting in reckless disregard of the truth to say Mann’s work is fraudulent even after reading it AND believing it.
Brad Keyes… “Why do you call the divergent, post-1960 dendro data “misleading”? It’s actually INFORMATIVE.”
Yes! It is! I totally agree… For an audience of researchers in the field. And that’s why there have been a good number of research papers that have discussed that issue.
BUT, you have to look at the audience the graphic is intended for. The WMO report targets a lay audience. So, you throw at them information that is as complex as the divergence problem and what is there initial reaction going to be? They’re going to say to themselves, “Why was the temperature going down in this series?” Then you have to go, “Well, the temps were NOT going down in that series. Temps were going up during that period.”
At that point you’ve clearly mislead them from the core conclusion of the available research that says, quite confidently, global temperature has risen dramatically since the 1850’s and is now higher than any other time in the past 1000 years.
That is the truth you’re trying to tell people. And that IS the truth. You are misleading them if you give them complex information that would cloud that point.
And yes, that is exactly what YOU want. You want to create doubt in people’s minds over a point that is very clearly presented by the available scientific evidence. So, I also clearly see why you would want to see that information included.
MikeN… “Also note Figure 2 in this paper which shows how the 10000 split. Note almost all hockey sticks.”
That’s the whole point, Mike. That figure 2 represents only the most prominent hockey sticks they could create out of 100 runs, and those 100 runs were only a 1% chance out of 10,000.
If Mike Mann did something like that you’d have a flippin’ cow.
MikeN… “My point about fraud is that Mann’s brief and complaint says that the bloggers spoke with reckless disregard for the truth and should have known they were lying based on reading these reports.”
Ah, no. That is incorrect.
Even if he did not read all these reports the statements would be libelous because the information clearly showing that Mann’s work is not libelous is widely available. That is a reckless disregard for facts.
MikeN…
It sounds almost like you’re now convinced that Mann’s work is not fraudulent, since you’re defending Steyn’s ability to come to an incorrect conclusion.
This thread illustrates to me an endless debate that leads to no satisfactory resolution to either side. It might help for me to point out why I find so much of what is said in defense of hockey stick graphs generally to be unpersuasive to me. Each of the point set forth below would require extensive argument to fully support. Such arguments can be found elsewhere. I will not attempt to fully argue the points, but will merely provide a thumbnail sketch of my positions. YMMV
1) When a mistake or flaw is pointed out regarding one hockey stick graph, the defenders say that there are other studies that show the same thing so the mistake is not significant. Nevertheless, the mistakes are never acknowledged, never corrected in the literature and are vigorously defended by supporters who circle the wagon to defend the indefensible. AMAC’s alleged fixation on the Tijlander Varves illustrates the problem well. IMHO his critique is simple and devastating, but it apparently doesn’t matter to the climate science community.
2) Many of other flawed hockey stick graphs pass peer review only to have serious flaws identified after publication, even in prestigious journals. See e.g. http://climateaudit.org/2012/10/30/karoly-and-gergis-vs-journal-of-climate/ . This suggests to skeptics that in this area the desire to create hockey stick graphs lets many studies pass peer review when they should not. In other words, the peer review process is flawed.
3) Many of the hockey stick graphs include proxies that have been highly criticized such as the bristle cone pines and Tijlander Varves. The criticisms of these so-called proxies have been widely documented so one wonders why they are used at all, unless their inclusion is necessary to achieve statistical significance for the study. IMHO this is a reasonable suspicion.
4) Even proxies that are not highly controversial are of poor quality. What I would view as high quality proxies are sufficiently rare that they can only be used for regional reconstructions. I think a reasonable critic could survey the field and find that quality proxies do not exist in sufficient quantity to allow anyone to draw firm conclusions about global temperatures over the past 1000 years.
5) Many of the studies use one version or another of deeply flawed screening criteria that seem to be accepted within the climate science community, but would not be accepted in other disciplines.
6) Many criticisms of the skeptics are frivolous and there are so many that it is difficult to respond to all. Much ink is spilled. See e.g. this recent post http://climateaudit.org/2014/09/24/kevin-oneills-fraud-allegations/#more-19779
I suppose if I gave it more thought, I could add to this list, but it will suffice for now to illustrate my concerns. Maybe other skeptics could add to the list.
Rob, look at the figure. It is not showing hockey sticks but a distribution of the hockey sticks.
#221, I am limiting to the North Report and say that reading that report, it would not be a reckless disregard of the truth to say ‘creator of the fraudulent hockey stick graph’ ‘ringmaster of the treering circus’ etc. Because it supports the skeptics on so many points.
Is it just me or will these skeptics not accept anything unless Steve McIntyre personally approves if it on his blog? Everything comes back to “But McIntyre says…”.
“you can’t use that series because McIntyre”
“that issue hasn’t been resolved because McIntyre”
“those results are questionable because McIntyre”
I’ve never seen the argument from authority fallacy spewed so adamantly. And he’s not even an authority on the matter anyway! Creationists are in envy.
MikeN… The North Report does NOT support anything close to what Steyn said. Not in your wildest dreams, it doesn’t. The report concluded, “with a high level of confidence that global mean surface temperature was higher during the last few decades of the 20th century than during any comparable period during the preceding four centuries” which is absolutely consistent with MBH.
The North Report also concluded the PCA issues brought up by M&M did not sufficient to challenge the conclusions of the research.
You’d have to be completely ideologically blinded to avoid these conclusions and determine that the report supports any notions of fraud.
But the graph isn’t fraudulent! No fraud was committed!
The most comprehensive northern and southern hemisphere reconstructions to date both show hockey sticks and neither show a global MWP nor one that was warmer than present.
There seems to be a fundamental (dare I say willful?) misunderstanding about how science works. Don’t let weird (dare I say obtusely lazy, and conveniently ideological?) ideas about perfection be the enemy of the good.
I suppose you can think wabi sabi, if it helps.
http://www.realclimate.org/index.php/archives/2013/09/on-mismatches-between-models-and-observations/
The science of climate has shown itself to be relentlessly inventive in developing techniques that may seem counter intuitive and tenuous to casual observers invested in conventional wisdom; but then that’s been the history of science in general. Get over it.
Personally I’m bored to distraction by AGW deniers practically peeing themselves with excitement over the notion that they might bring down the pros if they can just tie their brains in the right kind of boy scout knot.
Pauldd (#106) writes in his point 7 that: Jean S followed up with a post, which is the subject of the main post here. In this follow up post, Jean S. quotes various emails from the climategate emails. The point of the post is to show that Michael Mann, consistent with his CV, actually participated in the discussion of how the “hockey stick” graph should be presented on the WMO cover figure.
The fact that Michael Mann participated in the development of the WMO cover figure does not mean he had the final say in how it appeared. He might well have submitted his input at an early stage, then moved on.
I’ve had direct testimony from authors that they do not always get to pick the titles of their books. The same evidently applies to cover art. Example: Climate Crash by John D. Cox (Joseph Henry Press, 2005) has cover art showing a table and chair covered with snow. Through the snow on the table poke a pitcher and a glass. This suggests a sudden freeze like those shown in the movie The Day after Tomorrow — the same movie Mr. Cox describes in the book as a travesty of science.
I submit that no facts in evidence support holding Mr. Cox to account for the accuracy of this cover art.
It seems likely that the same is true WRT Dr. Mann’s responsibility for the WMO cover figure. No doubt this will be decided at trial.
Cristopher Winter is correct. In an ideal world the designer creating a cover will develop an enticing and attractively accurate distillation of the contents. But that’s not always the case. It’s like your mama told you, don’t judge a book by its cover. As an art student, I had the opportunity to provide the artwork for a journal cover that was subsequently accepted and published. Truth be told it, um, bore a somewhat fanciful relation to the contents, as I had scant understanding of the nuclear physics within. None of the authors were involved in the approval process.
Bottom line, the Dunning Kruger effect applies as much to assumptions about how the world of publishing works as it does to other areas of expertise. But hey, don’t let that stop you from making things up and blowing them out of proportion.
Rob Honeycutt, who says they are saying the same thing. Only that it adds some evidence in Steyn’s favor. What did they have to say about PCA? Note that they did the same upside-down usage to which Mashey and others are now saying is something untoward my McIntyre. Tends to create hockey sticks from random data, is not something that Steyn will think is an exoneration.
Rob, what was clever about Mike’s Nature Trick?
MikeN – please quote any sentence from the NRC (North), NAS, Penn St, or Wegman than indicates they believe Mann committed fraud.
Please, one sentence. That’s all. One.
Nothing? Thought so. You may now shut up.
Michael Smith (#15): Yes, of course we understand that. What YOU don’t appear to understand is that if proxy behavior can change in the present, without us understanding why, it might well have changed in the past and, thereby, failed to show the medieval warm period. I saw nothing whatever in your post that deals with this point.
What you’re suggesting seems to be that tree-ring proxies used by MBH98 and subsequent papers might inexplicably have changed their behavior to conform to other proxies like sediments in contributing to the shape of the “hockey stick” curve, which you suspect is misleading.
But in fact your doubts about the accuracy of tree-ring data over the past 1,000 years would apply to all other sorts of proxy data as well. And if you were right, scientists could say nothing at all about the existence of a global temperature trend prior to the start of the instrumental record — and very little about regional trends.
But you’re not right. Or, better, what Caerbannog said in #66.
First of all I was under the impression NRC and NAS were the same thing. Are there two separate reports I should be looking at?
You still miss the point, that there is a difference between stating fraud which the investigations did not, and Steyn’s declaring fraud despite an ‘exoneration’. My comment is limited to NRC/NAS. Steyn is not required to retract an allegation of fraud based on that report.
Suppose someone wishes to accuse someone of robbery.
A report says X is not a crook, then notes that X was seen entering the building.
Again and again throughout these comments we get the assertion that the proxy data used in assembling the “hockey stick” curve (of whatever vintage) must not be valid because if they were valid they would have given the pre-1800 portion of the curve a different shape.
In other words, the assertion is that climate scientists in the mainstream don’t know what they are doing.
Scientists in no other field are charged with such widespread incompetence. It is relevant to consider why climate scientists should be thought so different from all other scientists.
Rob Honeycutt, with regards to Mike’s Nature Trick, I am not referring to the existence of a separate chart for instrumental temperature. Read what I wrote.
Here is the data that was smoothed to make the figure.
http://www.meteo.psu.edu/holocene/public_html/shared/research/MANNETAL98/nhmean.dat
Note how it stretches out to 1995. You can go up one directory and compare to the instrumental data to see the overlap. Indeed, there is even a tiny bit of a dropoff at the end of that chart, so a bit of hide the decline as well.
1) Can anyone who thinks MBH98/99 are wrong explain the general meaning of error bars/confidence intervals in statistical analyses? Do those graphs have such? What do they mean?
How many parts of others’ real constructions have been outside those lines?
2) I suggest that people stop using the unadorned term MWP, which is made for confusion.
a) McIntyre’s post showed an MWP.
let’s call that (accurately) “LambMWP” … long since discarded as a flat-Earth map by real scientists. There are 2 key falsehoods in that post, along with the start of a well-constructed deception based on a quote from Deming. presented as fact, whose source was later given falsely by M&M.
b) On the other hand, we have the “ReconMWP”, which shows up in
MBH99 and most other reasonable reconstructions, especially visible in the 3rd graph Greg showed, IMHO a much better visualization than spaghetti graphs.
The ReconMWP shows a slight warming around 1000CE, with a slow jiggly descent to the LIA.
MBH99 has an explicit regression line, showing slow descent, which Montford misrepresented as horizontal on his book’s cover.
For more commentary, see Strange Scholarship…, pp.136-142, especially 142, where M&M;’s silly curve is superimposed on real reconstructions, and the MBH99 confidence interval included on the spaghetti graph.
Either statistical or science illiteracy can be a problem:
Reconstructions cover different geographies. Those focused on
(land-dominated) NH extratropics are expected to vary more than the entire NH, which in turn varies more than global.
Relative NH areas are:
0.13 60°N (Alaska, N. Canada, Scandinavia, Polar Urals, etc)
Likely to show sharpest swings, ice-albedo feedback. etc
0.50 30°N (to pole, sometimes also called extra-tropic)
0.60 23.5°N (Tropic of Cancer to pole, ~Moberg(2005))
1.00 0°N, NH (MBH98, MBH99, others) smoother curves expected
2.00 pole to pole, NH+SH, smoother yet, given SH ocean.
Of course, LambMWP was Central England sketch, not a reconstruction, but a modest ReconMWP (especially in parts of Europe) is perfectly understandable given the Law Dome CO2 curve + solar + volcanoes or lack thereof. A long slow descent with jiggles is exactly what one would expect from Milankovtich.
Anyway, the unadormed MWP is a useless term.
Rob,
sigh. Science doesn’t work that way. I know you think it does, but you’re ignorant and wrong. Wake up: you don’t have science training so please stop trying to “enlighten” those of us who do. It’s embarrassing.
As a scientist you CAN’T pick and choose which half of your story will be true depending on who you’re telling it to. A graph intended for a readership of muggles is held to exactly the same standards of honesty (viz. ABSOLUTE honesty) as if it were intended for…oh…say, the various *scientists* to whom Jones later emailed the same WMO graph exactly as he’d initially printed it for the muggles!
Science not only doesn’t, it *couldn’t* work under your imaginary ethical rules.
Science no longer works when you can’t look at a graph without wondering: “Hang on, am I the right audience for this? Is this accurate, or am I looking at a version that was dumbed-down and infantilised and simplified for people less sophisticated than me?”
Now you ask, “So, you throw at them information that is as complex as the divergence problem and what is there initial reaction going to be?”
That depends.
If they have half a brain (which scientists now believe a lot of muggles do) their reaction will be:
“What good are these tree rings as indicators of climate change if they go the wrong direction half the time? Why is the y-axis labeled as temperature when the recent data doesn’t remotely track recent climate change?”
And they have every right to ask such questions. And scientists have NO right to stop them by withholding awkward information from them.
When Richard Muller says “You’re not allowed to do this in science!” do you think to yourself, “well, that’s his opinion but what would he know? He’s a scientist. I, Rob Honeycutt, shall go forth and promulgate my contrary, scientifically-ignorant view as Truth on every blog I find!”
You’re entitled to your own opinions, Rob, but stop spreading falsehoods about what is and isn’t ethically acceptable in science, a topic you’re just making wild, and wildly wrong, guesses about.
How much clearer can I be? If you think your apologetics for the WMO graph make sense, though every scientist in the world has been taught NEVER to even think of doing things that way, it’s because you don’t know what you’re talking about. So why are you still talking????
“That is the truth you’re trying to tell people. And that IS the truth. You are misleading them if you give them complex information that would cloud that point.”
Oh lord. You really have no idea how mixed-up you sound to someone who’s been taught how to do science. LISTEN: Scientists are not ALLOWED to choose “which” truth they want to tell in order to get a “point” or “message” or “narrative” across, Rob!
You’re thinking of lawyers, politicians, priests, etc. The standards you hold scientists to are, frankly, insulting.
“And yes, that is exactly what YOU want. You want to create doubt in people’s minds over a point that is very clearly presented by the available scientific evidence.”
Wow. Just when I thought there was no topic you knew LESS about than scientific ethics, you’ve now started expounding on a subject you’re absolutely, encyclopedically ignorant of: my thoughts.
Congratulations, Rob, you get the lowest possible score for mind-reading. Nobody could have guessed my intentions LESS accurately.
The only DOUBT I’m advocating the promotion of is doubt in the validity of the dendro proxies, because anyone with half a brain would have serious doubts about that once they knew the truth about what happens after 1960.
But that completely rational “doubt” is incidental.
The real merchants of doubt are scientists who get caught hiding things but don’t even have the grace to apologize; and of course the useful idiots in the non-scientist population who then defend their academic misconduct to the bitter end.
When the general public doubts—distrusts—every word climate scientists tell them, that’s not on me. That’s on Phil Jones and Rob Honeycutt.
“So, I also clearly see why you would want to see that information included.”
Nope, you’re just seeing things.
Christopher Winter@2311
You have one of the very rare posts on this thread that is actually relevant to Jean S ‘s original post at climate audit that the main post purports to criticize. I am not saying I agree but it is on point. I think you could have an interesting discussion by cross posting it on climateaudit.org
It appears this thread has gotten long enough to annoy the server.
Brad… You know what’s really embarrassing? That you just stammered on for that long to say, essentially nothing.
It was a graphic for the cover of a report, Brad. It was not a scientific research paper.
What’s also embarrassing is that you’re sent into such a tizzy over something that is now 15 years old. You know what science does do, Brad? It moves on.
We now have some two dozen multiproxy temperature reconstructions that all agree with each other, many utilizing proxies other than tree rings. But you’re all over giving Phil Jones a spanking because you (and Muller) think he made a boo-boo.
Grow up.
Your writing is even childish. “Muggles?” Come on. (rollseyes)
Nathan Tetlaw@203 writes, “He chose proxies to construct the series and then compared them to the reference period. And this is entirely appropriate – how else do you test your data? ”
In climate science this seems to be entirely appropriate, but in other disciplines it is frowned upon. A fairly simple explanation of the problem can be found here http://noconsensus.wordpress.com/2008/09/04/how-to-make-a-hockey-stick-paleoclimatology-what-they-dont-want-you-to-know/
MikeN: The server is having a bad day, we’ve had a number of those lately. It should be able to handle 500 comments without flinching; something else is going on.
Pauldd – there’s enough insane nonsense at the Air vent that I’d be ashamed to link there except as an example *of* nonsense. An actual scientific paper might be a better reference.
Jeff(Patrick) Id (Congdon) posts thoughts like this:”Anyone with any background in climate change science knows full well (or should) that the human component of observed warming is completely 100% unknown.”
Now that’s not a source with much credibility. Apparently he’s never heard of attribution studies.
Brad… I’ll make one more point.
If Jones had done exactly the same thing he did on the WMO cover IN a research paper, I’d be right there with you. And, a lot of other academics would as well. And besides, it’s not even something a reviewer would have let past peer review. And in a paper Jones would have been careful to explain the details of the divergence in the first place.
But… it was a cover graphic.
From… 15 years ago.
A nice little walkthru for those interested in what’s involved in a proxy: Calibration, validation and application of foraminiferal carbonate based proxies: reconstructing temperature, salinity and seawater Mg/Ca, PhD thesis by J.C. Wit.
Hmmmm …. you’d think from reading pseudoskeptic you could just throw some numbers into Excel or R and do a regression analysis. Instead we find:
Contents:
Chapter 1. General Introduction
Chapter 2. Unmixing of stable isotope signals using single specimen ?18O analyses
Chapter 3. Approaches to unravel seasonality in sea surface temperatures using paired single specimen foraminiferal ?18O and Mg/Ca analyses
Chapter 4. A reappraisal of the vital effect in cultured benthic foraminifer Bulimina marginata on Mg/C a values: assessing temperature uncertainty relationships
Chapter 5. Reconstructing ancient seawater Mg/Ca by combining porcelaneous and hyaline foraminiferal Mg/Ca-temperature calibrations
Chapter 6. Development and Application of a Novel, Direct Proxy for Paleo-Salinity
Chapter 7. Sea surface conditions during deposition of Mediterranean sapropel S5
References
Appendices
I’m sure all those outraged by the WMO report will demand retraction of The Hockey Stick Illusion, whose cover misrepresents the handle as horizontal, although oddly evaporating at the left where the discrepancy would have been most obvious.
Of course, this could be one of those little oversights 🙂 although it certainly does fit key messages of the book.
That’s only 4 years old.
Again, MBH99 had a ReconMWP with a descending regression line.
John… That is exactly what bugs me. These guys go absolutely haywire over the most nuanced and inconsequential issues in the real science, but give a blank check when M&M muck up and misrepresent their red noise tests to such an absurd degree.
JM #240
I assume that you are aware that if they were done correctly, there are still a couple of interpretations of the error bars depending on how they were calculated. As far as I know, the programming for their calculation has never been made available nor was any other explanation given by Mann as to their origin.
If you can indicate which of these apply, I would be happy to give you a statistical answer.
@ 249: “”Pauldd – there’s enough insane nonsense at the Air vent that I’d be ashamed to link there except as an example *of* nonsense. An actual scientific paper might be a better reference.”
Sure a scientific paper would be better. I don’t like spending too much time from my day digging up reference for this type of discussion–for the most part pointless.
So I recalled Jeff I’D had once written a simple explanation of the point I wanted to make. I am not appealing to authority–just to his explanation, which the reader can evaluate. I hope curious readers will ignore your ad hom argument.
Roman… 15 years ago.
Let it go.
Move on.
RomanM – what magnitude error did this “serious bias” add to the reconstruction?
And given that the blade of the hockey stick was taken from /and or corroborated by the instrumental data – did it have any effect on that at all?
Pauldd – Your link states, “I then explained that I could see no basis for elimination of any of the data sets by correlation to measured temperature…”
In other words, to build a temperature reconstruction you should use proxies that may or may not be correlated to temperature.
Damn, I never thought of that. Heck, we could just use any old random datasets lying around if there’s no need for them to be correlated with temperature. What a brilliant idea !? My god these climate scientists are dumb. Wasting all this time trying to find proxies that are correlated for temperature when the Air Vent has already shown that’s a mistake.
My mistake, I thought the guy was a true idiot. Now I know better.
…. hey, wait a second …. that doesn’t make an freaking sense …
RomanM writes: “As far as I know, the programming for their calculation has never been made available nor was any other explanation given by Mann as to their origin.”
Perhaps you’ve forgotten: “Several checks were performed to ensure a reasonably unbiased calibration procedure. The histograms of calibration residuals were examined for possible heteroscedasticity, but were found to pass a x2 test for gaussian characteristics at reasonably high levels of significance (NH, 95% level; NINO3, 99% level). The spectra of the calibration residuals for these quantities were, furthermore, found to be approximately ‘white’, showing little evidence for preferred or deficiently resolved timescales in the calibration process. Having established reasonably unbiased calibration residuals, we were able to calculate uncertainties in the reconstructions by assuming that the unresolved variance is gaussian distributed over time. This variance increases back in time (the increasingly sparse multiproxy network calibrates smaller fractions of variance), yielding error bars which expand back in time. Page 786
and
“The results of all calibration and verification experiments are available; see Supplementary Information. Page 787
Kevin, here’s what DOES make sense:
1. If the data, the science, and the work being done by the climate scientists demonstrate AGW, then we’re causing a catastrophic change to the Earth’s ability to support life as we know it.
2. Given the above, I will be required to change my lifestyle (and consequently, my politics), lest I succumb.
3. I don’t wanna change my treasured lifestyle!!!!!
4. Conclusion: The data is bad, the science is bad, and the climate scientists are all a bunch of bad guys who are only trying to sabotage my beloved lifestyle. Therefore I don’t have to acknowledge it or act on it.
5. Outcome: I win because I want to, and I don’t have to change anything I’m doing. (Back to sleep…)
You’re dealing with that, not with real skeptics who are looking for truth — whichever direction it happens to blow.
A while back, I wrote a couple of comments on the problems of retrospective analysis and the Divergence Problem (these aren’t new topics). I think they’re relevant to the discussion here.
Kevin O’Neal: Actually what you should do is create a priori a set of criteria that defines your proxy based on a physical theory that justifies its use as a proxy.
Next, you include every collected sample that meets your a priori criteria. You can then preform a regression of the entire sample against the temperature record. If the regression results do not meet a predefined standard, the entire sample set should be discarded. If the regression meets a predefined requirement the entire sample can be used as a proxy.
What you cannot do is perform a regression of each individual sample against the temperature record and accept or reject the sample based on the regression result rather than the a priori defined physical criteria. I realize that this result may seem counterintuitive, especially to climate scientists, but it is an essential rule for the reasons explained in the link I previously provided.
Brainstorm @258: I am curious as to what lifestyle changes you advocate to address global warming? Have you adopted your recommended changes?
Just curious after reading this article http://freebeacon.com/issues/study-environmentalists-have-substantially-worse-than-average-carbon-footprints/
The train wreck that is Mann’s “Hockey Stick” can be seen when the NAS panel failed to resolve disputes over its soundness in their report released in June, 2006.
For an overview, SEE Roger Pielke, Jrs Quick Take from June 22nd – with comments from some major climatology players as well as the policy wonk set running a further two days.
At the time, Pielke thought it did decisively affirm Mann. But over the years since then, one sees that he’s come closer to Eduardo Zorita or Hans Von Storch’s skepticism – save for adopting a “what does [climate history] matter?” “let’s set this aside” tone. And we see the pushback against his uncritical reading in many paragraphs from a variety of critics, therein.
And even then in 2006, even before the climategate scandal, many knew that it would be a recurring and revisited theme. For example:
Did AR4 (2007) reject it? Demoting it from Mann’s self-promoted prominence in the prior report? Did the AR4 spaghetti graph in it supplant it? Or merely recycle many of its fundamental errors and paper-over the flaws by turning the spot-light off MBH?
As I linked above, Huang’s (1997) borehole paper – affirming the MWP as a half degree warmer and the LIA as a half degree colder than our recent century – got no mention in the NAS panel’s work. Nor did it get consideration in the successive IPCC reports. Perhaps the methodology was too new for the senior scientists who directed the NAS panel that, in Gerald North’s infamous phrase, “just winged it.”
But the successive IPCC groups have no such excuse – why did they neglect this new state-of-the-art technique of measuring the temperatures of soil horizons?
To me, this oversight, considering how much more comprehensive and numerous its time series are, either suggests a science in the midst of change, resulting in gross neglect of contradictory data – or willful neglect.
Therefore, I affirm the polarization evident in the many comments in Pielke’s early report on the NAS. And like Pielke, I’ve moved against supplicating to Mann’s enormous, unjustified egocentric self-righteous claims, because the email documented record – again, SEE “JeanS” – contradicts him.
In this, I have also been joined by Matt Ridley, persuaded of Mann’s hokum by Andrew Monfort’s book “The Hocker Stick Illusion.”
LINK to Pielke, Jun 22-24, 2006 – his early “take” on the Hockey Stick report by the NAS, http://cstpr.colorado.edu/prometheus/archives/climate_change/000859quick_reaction_to_th.html
@pauldd,
#260 — Yes, exactly, on the pitfalls of post hoc analysis. See the comments I linked in #259 for a step-by-step exposition of a closely related topic.
#261 — Clear thinking about global warming can lead to painful insights — very uncomfortable, if one has come to the issue from a hipster (SWPL) perspective. E.g The Acid Test of the Moral Seriousness of Global Warming Activists Is Immigration Policy.
pauldd – “Actually what you should do is create a priori a set of criteria that defines your proxy based on a physical theory that justifies its use as a proxy. Next, you include every collected sample that meets your a priori criteria.”
Assume a spherical cow …..
Yeah, works great in theory – unfortunately we don’t live in a theoretical world. We have to deal with the one we live in.
Do you seriously think scientists are unaware of the degrees of freedom and the problem selection poses? Do you bother to read the literature at all?
Kevin O’Neal @265: Perhaps we are not understanding each other,but I don’t see the problem as one involving degrees of freedom. The problem is guarding against spurious correlations that can occur when a researcher mines data.
Kevin O’Neal@265: By the way,I do agree that existing data is noisy and scarce such that it likely not possible to create a proxy reconstruction using a rigorous methodology. That is why all existing paleontologist-climate reconstructions need to viewed with caution.
@Kevin O’Neill #265 —
> Do you seriously think scientists are unaware of the degrees of freedom and the problem selection poses?
The problem isn’t that scientists are unaware. For example, every clinical trial that is designed to be used in a drug or device approval in front of the FDA or EMA is designed with these considerations in mind. Another example — for the past half-decade or so, all published GWAS’s (studies connecting genes with phenotypes such as disease state) have a discovery cohort followed by a validation cohort (see Wikipedia entry).
Nor is the issue that all climate scientists are unaware. Jim Bouldin is a tree-ring PhD; he contributed upthread. Quiz him, or read the linked posts on his blog.
The problem can be stated this way: “Best practices” in climatology haven’t changed to insure that the pitfalls of post hoc analysis are addressed.
In other words, a manuscript that ignores this issue will fail the peer review process of most journals in most other fields. In paleoclimatogy, the erroneous statistical findings that result are not grounds for rejection.
Science is a tough career, so in one sense, I don’t blame paleoclimatologists for working with the “Least Publishable Unit” (see Wikipedia) that they can utilize. But the unfortunate overall result is that the field has steered itself down a dead-end street, when it comes to paleotemperature reconstructions.
.
I recognize that most readers of this thread will vehemently disagree with what I (and pauldd, and Jim Bouldin, and others) have written on this subject.
You should recognize that your comrades-in-arms are the shills working for sleazy drug companies, pushing to get ineffective treatments cleared by regulatory efforts. “It [all this pointy-headed statistical stuff] doesn’t matter” is the cri de coeur.
There is a contradiction, in saying that you shouldn’t use series that diverge because they might not be accurate in the past, and you shouldn’t use series that correlate because you could be getting them by luck.
Nevertheless, both criticisms are valid.
The second one regarding correlation is a reference to Mann 08 and other papers as well.
Kevin writes:
Damn, I never thought of that. Heck, we could just use any old random datasets lying around if there’s no need for them to be correlated with temperature. What a brilliant idea !? My god these climate scientists are dumb. Wasting all this time trying to find proxies that are correlated for temperature when the Air Vent has already shown that’s a mistake.
The issue is Mann took 1200 proxies, and then filtered out the ones that were correlated to temperature and used those. If he had thrown out just 100, that would be one thing. Instead he ended up with a smaller number, around 400, while using a low correlation number .10 as the filter.
So the effect of having proxies appear by chance and creating a hockey stick from random data is likely, just as Jeff Id showed.
Orson… But you know what has repeatedly confirmed Mann’s original research over the past 15 years? Subsequent research.
THAT is how science works. The test of time.
Prior to MBH98/99 we had, for the most part, just one graph that everyone was referring to as “global temperature” over the previous millennia. H. H. Lamb. And that was only an anecdotal construction, with no actual Y-axis data, and limited to northern England.
MBH98/99 collected all the available dendro data and collected it into a northern hemispheric MULTIproxy reconstruction of global temperature. A brand new approach.
They easily could have been wrong. The North Reports stated that MBH were clearly attempting to do the best leading edge research they could. Nothing more, nothing less. But it’s the test of time that shows us whether leading edge research got it right or wrong. THAT really is how science works!!
How has MBH98/99 stood the test of time? Well, 15 years later we have about two dozen multiproxy reconstruction, both NH and global. Every… single… one of these has confirmed the conclusions of MBH98/99, that modern global temperature is very likely higher today than at anytime in the past 1000 years, and the global temperature started rising in the mid-1800 when humans started emitting CO2, and has risen rapidly since then.
There have been no multiproxy reconstructions that show different.
So, I will repeat my question that I’ve made repeatedly on this thread. Are you folks claiming that Dr Mann got the RIGHT answer for the wrong reasons? Really?
MikeN… “The issue is Mann took 1200 proxies, and then filtered out the ones….” and on and on.
Okay, Mike. Riddle me this.
How is it that if, as you (and others) claim, Mann mucked up his research so terribly, how is it every subsequent multiproxy reconstruction has turned up the same results?
Think about for just two seconds.
If MBH had done such an awful job, there is no way other research would have turned up the same result. The odds of that happening are astronomical!
You really should stop to consider the possibility that, maybe, just maybe, you’re not fully comprehending the methods applied by MBH, and that, maybe, just maybe, the researchers actually knew what they were doing.
Their results have stood the test of time.
In fact, their results have stood the test of time while under serious fire from factions who find their results undesirable. (Hm, much like the Galileo’s of the past.)
AMac… “Jim Bouldin is a tree-ring PhD; he contributed upthread. Quiz him,…”
Tried that. No response.
I’ll try you, again, though.
Okay, we’ll accept for a moment that the uncertainty range for tree ring data is very high, perhaps higher than we may expect. (These are temperature proxies after all.)
We now have the PAGES2k project going. This uses a very broad range of proxies, many that have much lower uncertainty ranges. Proxies that can be calibrated to 18O isotopes. These can be cross-verified with other proxies in order to constrain the uncertainty ranges.
What does the PAGES2k data show? Well… Pretty much what we saw with Mike Mann’s original research 15 years ago.
What irks all of us here, on my side of things, is the fact that we have such overwhelming evidence that the results are actually highly likely to be correct, but the “other side” keeps bleating on about meaningless nuances of 15 year old research.
Step back and look at the whole picture. The research is looking highly conclusive at this point in time.
Rob,
this is an innovative attempt at an excuse for scientific malpractice, I must say:
“It was a graphic for the cover of a report, Brad. It was not a scientific research paper.”
I never knew the ethics of science varied page by page. Tell me, Rob, since you’re just pulling ideas ex posteriori, what are the guidelines for the *back* cover? Are there any, or is it just an ethical free-for-all where rampant confabulation is encouraged?
“What’s also embarrassing is that you’re sent into such a tizzy over something that is now 15 years old.”
I’m glad it embarrasses you that a guy who conceals awkward information from the public has been allowed to keep calling himself a scientist (or at least a climate scientist) for 15 years, Rob. It should. There’s hope for you yet.
“You know what science does do, Brad?”
It terminates the employment of admitted frauds.
“We now have some two dozen multiproxy temperature reconstructions that all agree with each other,”
LOL. What kind of innumerate hand-waving is this? “Agree with each other”? Uh, how vaguely are we talking here, Rob?
“But you’re all over giving Phil Jones a spanking because you (and Muller)”
And Jonathan Jones. And Paul Dennis. And every other self-respecting scientist acquainted with the trick.
” think he made a boo-boo.”
He DID make a “boo-boo”, or as grownups would say “hanging offense against scientific morality,” “grounds for seppuku,” etc.
“Your writing is even childish. “Muggles?” Come on. (rollseyes)”
Sorry, I was just trying to communicate at the level of someone who says “boo-boo” and thinks he can bail out of arguments by pretending there’s a 15-year statute of limitation on crimes against science.
However, since your pleas to “move on” are an admission of defeat, I should congratulate you for knowing when you’re wrong.
Brad… Look, you and I going back and forth decides exactly nothing on this issue. In fact, the issue has been thoroughly investigated and, you know what? They’ve come down saying essentially what I’ve been telling you.
Sorry if their determinations don’t conform to your ideologies.
Yet again, none of this changes anything relative to the overwhelming body of scientific research that shows that humans are warming the planet through emissions of CO2 with potentially severe effects.
Brad… “I should congratulate you for knowing when you’re wrong.”
No. In fact, I’m correct. And I’m giving you and your buddies ample evidence that shows Mann’s original research has stood the test of time.
You can rehash the past and the inconsequential nuances of how research was done, but the ultimate test, as I’ve been stating, is the test of time.
There are no multiproxy NH or global temperature reconstructions that contradict Mann’s research. There are now about two dozen that actually confirm his original conclusions.
THAT is how science works.
@ Rob Honeycutt #270
> Are you folks claiming that Dr Mann got the RIGHT answer for the wrong reasons?
I. Math jokes: Right Answer, Wrong Reason, January 6, 2011
II. MikeN’s explained what he’s saying, as has pauldd. Paraphrasing people that one disagrees with turns out to be a difficult skill to master. “Are you folks claiming that Dr Mann got the RIGHT answer for the wrong reasons?” might have more to do with sharing your views, than with understanding theirs.
.
III. In my view, Mann08’s methods are fatally flawed, akin to the arithmetic solution in (I.) above. For example: it’s remarkable that Mann and co-authors improved their “wrong” No-dendro/No-Tilj EIV reconstruction by adding uncalibratable and upside-down data series (Tiljander) to come up with the “right” No-dendro/Yes-Tilj EIV reconstruction. (link).
In bottoms-up science, this sort of result sets alarm bells ringing. In policy-driven top-down science, it’s met with a shrug.
My view is that the past two decades have been among the warmest that the earth has experienced for 2,000 years — very possibly the warmest. In that regard, the “consensus” of multi-century reconstructions is likely “right”.
On the other hand, my view is that “consensus” reconstructions systematically suppress climate variability on a multi-decade scale — thus, they are likely “wrong” in over-flattening the blade of the hockey stick.
Many contributors to this thread haven’t seemed to grasp the severity of the problems posed by the methods of “consensus” paleoclimatologists — post-hoc analysis is one example. At this level of scientific literacy, it’s beyond my abilities (and patience) to discuss why the blade can get flattened.
A bit off topic but the cei has filed in reply brief in the COA regarding the Mann libel case.It is well-done and an absolute scorcher. It requests sanctions against Mann’s attorneys for misrepresenting the record, a charge that attorneys do not lightly make. The link is over at watts up with that.For some reason I can’t copy it.
Brad et al… For fun, I’m going to post the conclusions as written in MBH99 for you:
“Although NH reconstructions prior to ab out AD 1400
exhibit expanded uncertainties, several imp ortant conclu-
sions are possible, notwithstanding certain caveats. While
warmth early in the millennium approaches mean 20th cen-
tury levels, the late 20th century still appears anomalous:
the 1990’s are likely the warmest decade, and 1998 the
warmest year, in at least a millennium. More widespread
high-resolution data which can resolve millennia-scale vari-
ability are needed before more confident conclusions can be
reached with regard to the spatial and temporal details of
climate change in the past millennium and beyond.”
Paul… Ah, “scorching?” Really?
Reading the very first paragraph of the brief I can tell the judge is going to reject this out of hand because it’s completely false.
AMac… The science we’re discussing is a little more complex than “cancel the sixes.”
You’re looking at a decade and a half of research, all coming to the same conclusion, and saying it’s the results of poor methods.
In fact, the repeatedly replicated results are fully consistent with what is expected from a wide range of research in other areas. It would be highly unusual if we did NOT see a hockey stick, since we know that we’ve dramatically altered the radiative balance of the earth’s atmosphere!
AMac… “On the other hand, my view is that “consensus” reconstructions systematically suppress climate variability on a multi-decade scale — thus, they are likely “wrong” in over-flattening the blade of the hockey stick.”
And yet, the more this area of research advances, and the more proxies that we add to the mix, and the more global in scope the data become, the flatter the handle gets.
That is also an expected result as well. (Whether you realize it or not.)
> It would be highly unusual if we did NOT see a hockey stick, since we know that we’ve dramatically altered the radiative balance of the earth’s atmosphere!
That is silly. Why didn’t the first IPCC report notice anything amiss when a non-hockey stick was put forward?
Poeple keep asking why a cover from 15 years ago is still under discussion
The reason is that Mann so strongly denied “absolutely” any involvement in its creation. Whatever some people on this thread say, Mann clearly believes that the cover is embarrasing and not something he wants to be associated with. Sadly for him, he’s been busted.
Rob Honeycutt, that is a strange argument. I suppose if you are incapable of evaluating the arguments against Mann’s methods, then that is a reasonable position to take. It could be that the other papers are flawed as well, perhaps because they are using similar methods, or perhaps because they are reusing the same flawed proxies. A number of hockey stick papers use PC1 from MBH. Now perhaps they have their own separate flaws, and critics are rebutted with the arguments that the results are similar to others.
Perhaps Mann’s answer is correct with the wrong methods and the whole thing doesn’t matter. However, when scientists defend what looks to be incorrect, it leads to less credibility. As an example, I’ve convinced many people of flaws in climate science just by loading the Excel spreadsheet from the Kaufman Arctic warming paper. Rather than claiming that nothing was used upside-down, or that the results ‘don’t matter’, or that other papers show similar things, or that CO2 warms the planet as basic physics, a correction was issued. Yet Mann has not issued any corrections in his own upside-down usage. Instead he claimed such an accusation is ‘bizarre’, and that McIntyre didn’t understand what he was doing.
No Excel spreadsheet, but it is still not too hard to explain how things were used uspide-down.
#257 KO
I was thinking of MBH99 when I mentioned that there was no indication of the way the error bars were being calculated. That said, the quote you gave from the paper is by no means sufficiently descriptive to indicate what was done in that paper as well. If you look at the post:
http://climateaudit.org/2008/04/05/squared-weights-in-mbh98/
you will see the math involved. You will note that if he used the residuals from the regression where the proxies were being predicted from the ever-changing sets of PCs, then turning the residuals into confidence intervals for the temperature reconstruction would require transformation by a matrix inversion (involving unannounced weighting of proxies). Such an inversion would likely produce bounds which are asymmetric unlike the ones in the paper. There are other possibilities which I won’t go into.
To answer the original question regarding the interpretation of error bars, there are two possibilities. If the bars are calculated at a 95% level for each year, then (assuming no bias) the methodology produces bounds which will generally include the true target temperature for about 95% of the years with values outside the bars for about 5% of the years. We can also calculate error bars such that the true temperature is inside the the bounds for every year for about 95% of the occasions on which the procedure is applied to such data. The latter produces larger bounds than the former. It is almost certain that Mann would have calculated bounds of the first type.
#265
Pauldd is correct. Look at the scatter plot on my post:
http://climateaudit.org/2012/06/17/screening-proxies-is-it-just-a-lot-of-noise/
The simulated proxies used in the post contain genuine information about the model temperature output. These proxies were created by climate scientist for testing reconstruction methodology. These is a plot showing correlation between proxies and temperature during a “calibration” period against the correlation outside that period. You will notice that higher calibration correlation is not an indicator of usefulness of the process and the point of the post is that using it to select proxies has an extrmely negative side effect.
MikeN… “That is silly. Why didn’t the first IPCC report notice anything amiss when a non-hockey stick was put forward?”
Mike, what is the purpose of the IPCC? It is to present the most current research on climate change. The FAR presented the H. H. Lamb graph because that was the the most current research at the time.
@ Dean #284 —
> [referring to AMac #276] translates to “Since the original and reconstructions give the same result, one I don’t like, I will conclude the people behind them are all liars.”
I’ll share with you what I said to Rob Honeycutt: paraphrasing people with whom you disagree is a difficult skill to master.
As an alternative, here is a link to Where’s the fraud?, an on-point discussion that Consensus supporter Arthur Smith hosted at his blog. That’s from 2010, by the way.
MikeN… “I suppose if you are incapable of evaluating the arguments against Mann’s methods…”
It just doesn’t matter, Mike! The fundamental of the scientific process says that, you’re allowed to get things wrong! If you try to put forth what you think is your best estimate of what the data is telling you, then it doesn’t really matter if you mucked up the details.
Conclusions that are wrong or done in error will slowly fade from view. There’s lots and lots of bad research that’s been published. But the bad research just ends up not agreeing with good research. It’s not the fact that every T is crossed and I is dotted that makes research important. It’s when it stands the test of time.
Arguing nuances of inconsequential details that do not change the basic conclusions of a piece of research is just idiotic. If YOU are genuinely interested it THE TRUTH, the honest scientific truth about the issue in question, then you have to step back and look at what the full body of research tells you.
That full body of research is telling us, very very conclusively, that global temperature is very likely higher than at any time in the past 1000 years, and the modern trend in warming began when humans started emitting CO2 into the atmosphere.
I want to try to understand this. The Hockey Stick conclusion published in 1999 is wrong because the data it used were bad data, data that showed no warming was left out, and the methods were wrong. But the temperature since 1999 has gone up significantly. If the Hockey Stick was wrong, why did that happen?
A second question is this: All tree ring data is bad. What papers are there that I can read that show temperatures over the last hundreds of years to the present that use no tree ring data at all, and what do they show? I think I only want to know about published articles in scientific journals.
MikeN… “I’ve convinced many people of flaws in climate science just by…”
Then you are doing humanity a grave disservice.
Have I missed an important announcement? Have scientists created a hockey stick without using bristlecones or Tiljander?
MikeN… “Yet Mann has not issued any corrections in his own upside-down usage.”
Yes they did.
“The claim that “upside down” data were used is bizarre. Multivariate regression methods are insensitive to the sign of predictors. Screening, when used, employed one-sided tests only when a definite sign could be a priori reasoned on physical grounds. Potential nonclimatic influences on the Tiljander and other proxies were discussed in the SI, which showed that none of our central conclusions relied on their use.”
http://www.pnas.org/content/106/6/E11.full
Brad… See PAGES2k.
AndyL @283… It’s a 15 year old cover graphic that has exactly zero bearing on scientific research.
Should Phil Jones have done the graph differently? Sure. Grow up. Get over it. Move on.
AlisonT,
“But the temperature since 1999 has gone up significantly.”
Where?
“If the Hockey Stick was wrong, why did that happen?”
1. The Hockey Stick papers were bad science, or rather pseudoscience (because they couldn’t be replicated using the information provided, and Mann obstructed all efforts to do so). But that doesn’t mean the Hockey Stick is not “true”.
2. The Hockey Stick didn’t predict post-1999 temps, so the question is a bit of a non-sequitur.
3. “Why” has to do with physical causes—nature doesn’t read scientific articles, let alone obey them, it does its own thing.
4. Why did *what* happen?
Rob,
“Should Phil Jones have done the graph differently? Sure.”
Right. So we agree.
Why did you waste all that time defending him? You looked silly, and you did the image of cli sci no favors either.
I have my own takeaway from all of this. The warming we’ve seen over the past century is so pronounced, robust, and unprecedented that even the first attempt at a large scale reconstruction, with its less than perfect methods (dare i say poor) of analyzing the data still produces the same general result as the much more accepted methods of the present used in the most current reconstructions.
I look at it like the instrumental record -the warming is so strong that you can get results that match the global average, calculated from thousands of stations, with just a few handfulls of stations.
@ Alison #290
> What papers are there that I can read that show temperatures over the last hundreds of years to the present that use no tree ring data at all, and what do they show? I think I only want to know about published articles in scientific journals.
Actually, answering your question was the primary motivation for the ill-starred Mann et al., PNAS (2008). That paper is available at PNAS.org, IIRC (or via my blog, search “AMac Tiljander”).
Lundqvist (2010) was very aware of the problems posed by faulty proxies. And as Rob Honeycutt said in #294, there’s the PAGES 2k 2013 publication.
MikeN and RomanM know the current literature much better than I do, perhaps they’ll weigh in.
AMac, thanks. So those papers show cooling or no warming? And do not use tree rings?
That would make me worry less about global warming.
@AlisonT #300
> That would make me worry less about global warming.
Well.. I dunno about that. I self-identify as a “lukewarmer”, which the “alarmists” view as ill-founded and anti-science. Example Lukewarmerism, a.k.a. Ignoring Inconvenient Evidence. IMO the two places to go to get a balanced view of the alarmist/lukewarmer dispute are the blogs “Lucia’s Blackboard” and Judy Curry’s “Climate, Etc.” See what you think, and happy reading. 🙂
Fine, I’ll change my sentence: your arguments and those of the people you reference have been shown to be wrong.
Brad… “Why did you waste all that time defending him?”
I’m not defending him, per se. I’m saying that what he did on the cover of the WMO99 report has no bearing on anything whatsoever.
Look, you guys keep presenting this as if it were some sort of academic malfeasance, when in fact it was a graphic. Like, you’d want to sue Picasso for not putting the eyes in their proper places.
Jones knew the difference between a graphic and a piece of research. The investigators knew the difference between a graphic and a piece of research. Apparently it’s only the fake skeptics who don’t.
Brad… “The Hockey Stick papers were bad science…”
(sigh)
So, you’re saying all the two dozen multiproxy reconstructions that have been done over the past 15 years are bad science?
That’s an astoundingly silly claim.
Brad… Did you read what I posted just a few comments earlier?
People are allowed to publish bad science. Peer review is there to filter out clearly bad stuff, but bad stuff gets through all the time. And that’s perfectly fine.
Over time bad research falls out of favor because new science, work that better explains observations, will take it’s place. Good science gets bolstered when subsequent research confirms those findings.
What we have with MBH98/99 is early, leading edge research. Their work was really the first large scale attempt to reconstruct millennial global temperature. The conclusions of their research was very carefully worded and modest (read the conclusions of MBH99 posted above). Their conclusions very easily could have been found to be wrong and been supplanted by better research that showed different conclusions.
What happened though? Well, scientists did what scientists do. They argued. They argued the methods. They argued the data. They challenged each other to do better. And genuine scientists looking for genuine answers did their own research. They created their own multiproxy reconstructions. They tested better statistical methods, and better data sources.
The result?
Well, every single multiproxy reconstruction looking at NH or global temperature has confirmed the conclusions of MBH98/99.
That is science doing what it is supposed to do. It is science being skeptical. It’s science testing ideas. It’s science looking for ways to expand our understanding of this particular question.
BUT… you guys are stuck at step 1 and just can’t get past that. It’s like you want to throw people in jail for not doing something perfectly exactly as you think it should have been done. And that’s just complete insanity.
Brad, ALL science is bad science by your definition, because there is always better science that follows it. There is no science that can stand up to the scrutiny that you seem to think it requires.
Audits are great for financial transactions, but they just don’t work for complex science issues like this. Science isn’t about audits, it’s about replication.
Rob, you are completely ruining it for everyone.
It is much easier to deny climate change by assuming that one bad apple spoils the whole bunch. Pick a specific climate scientist, or a particular paper, or a particular model, or a particular proxy, or a particular use of a particular proxy using a particular model in a particular paper. Then spend all your efforts crafting your critique of that one thing. Doesn’t matter if that one thing is actually good or bad or problematic or not. If you spend enough time on it, and keep saying it again and again and again, it will eventually become established that that one thing is bad. Even if it isn’t. That is what the deniers (and the lukers) are saying here.
But that only works if it can be the one bad faux apple that spoils the whole bunch!
@ Greg Laden #306 —
😉
Greg, whew! I was afraid maybe you’d abandoned this thread in disgust!
Yeah, you could nitpick Newton’s theory of gravity all you want and claim orbits are really just a hoax and impossible, but without an alternate testable model of how everything works you’ve got absolutely nothing (and that includes general relativity).
Likewise a lot of what deniers are saying boils down to a pseudo alternate theory in the form of natural variation, which in other words is simply longing for the days when we were ignorant of how climate works; as AGW is inextricably bound up with everything we understand about climate today.
Yes, we have published materials from Johannes Kepler that clearly show that he had fumbled his physics and published that planets have ovoid orbits.
True, he eventually suggested elliptical orbits, and yes, further research by others did result in a consensus that planets travel in elliptical orbits, but…
We MUST return to Kepler’s original publications — THAT’s what matters, after all. His math was wrong, he didn’t treat things in a statistically correct fashion, and HORRORS! Kepler’s early publications did NOT explain how elliptical orbits could be derived from observational data!
This is a crime against science! The flaws of this early work of his thoroughly discredits the entire field of astrometry and astrophysics, and we can no longer trust the results of astronomers.
It does not matter that Kepler eventually got it right (it may have been he accidentally got the right answer for the wrong reasons), nor does it matter at all that his results have been confirmed time & again for years afterward.
He got it wrong at the beginning, he published incorrect conclusions, he destroyed the trust of the public in astronomy, physics, and science, discredited himself, and …
…inspired centuries of self-righteous know-it-all physicists who keep arguing that elliptical orbits can be demonstrated from the wealth of observational and proxy data when we KNOW (even though we are not experts ourselves, but we do know some math and some logic that sounds good to us) that it’s all made up. These astronomers are conspiring to mislead us.
If you don’t agree with me, I’ll start citing web blogs written by well-known (and therefore credentialed) Flat Earth Society members who will set you STRAIGHT!
Brainstorms, if scientists are insisting that Kepler’s early work was correct(assuming what you say), then I would be suspicious of the whole field.
Rob, perhaps you should follow some of AMac’s links. That is not a correction. Indeed, it is a sign of either a deliberate lie or incompetence.
>Mike, what is the purpose of the IPCC? It is to present the most current research on climate change. The FAR presented the H. H. Lamb graph because that was the the most current research at the time.
Your statement was that anything other than a hockey stick would be surprising given the changes to the radiative imbalance. So did they not know of this radiative imbalance in 1990? Why weren’t they shocked by this chart?
Mike, you may be on to something…
“He got it incorrect for the right reasons.” ;^)
Rob Honeycutt @ 305 we sees things very differently, which is ok. If I believed that the process of advancement you described was an actual description of what is occurring with climate reconstructions I would have absolutely no problem.
I do have a problem, however, because what I see is:
1) A field where major mistakes are made, not acknowledged and therefore repeated.
2) Studies are completed with very noisy, poor quality proxies and less than rigorous methodology. The results are then claimed to be much more certain than is reasonably warranted because other studies that suffer from the same shortcomings reach the same results.
It is very possible that many scientists in good faith are doing the best they can with what they have to work with. I just do not believe the field has reached sufficient maturity that it can say anything about temperatures over the past 500 to 2,000 year with much confidence.
If people would actually read the FAR, IPCC(1990), pp.199-203, they might learn
a) That version was done in a rush.
b) There weren’t any serious reconstructions, so they grabbed as sketch.
c) That was identified as needing serious research, which they started doing,
FAR Supplement(1992), with Briffa et al, regional studies
1993: Bradley and Jones paper, alter adapted by:
IPCC(1995)
19954: Hughes and Diaz: Was there a Medieval Warm Period, and IF SO, Where and When
“These warmer regional episodes were not strongly synchronous”
“p.136 ‘In particular, the simplified representations of the course of global
temperature variation over the last thousand years reproduced in various
technical and popular publications (for example, Eddy et al., 1991; Firor, 1990;
Houghton et al., 1990; Mayewski et al., 1993) should be disregarded, since” (Houghton et al 1990 = IPCC FAR)
IPCC(1995/(1996))
p.174-175 ‘Bradley and Jones (1993, 1995) developed a Northern Hemisphere summer temperature reconstruction since 1400 from 16 palaeoclimatic records
(Figure 3.20). Recent decades (the recent instrumental record is also plotted on this figure) appear to be warmer than any extended period since 1400, and the warming since the late 19th century is unprecedented in this record.’
Wow, adding the modern instrumental record was not invented by Mann, the IPCC was already doing it years earlier, because any sane person would want to see how reconstructions compared to the period of overlap with modern instruments.
Normal science, like this. Of course, those dedicated to flat-earth maps need not bother.
By the way, McIntyre wrote falsely when he claimed the Lamb(1965) (IPCC(1990) Fig.7.1(c)) came from IPCC(1995), which showed he actually didn’t have either one. In academe, that’s called false citation. In this case, the 19890s to 1995 shift made his story slightly less absurd. of course, the image didn’t actually come from IPCC(1990), but when asked later, he couldn’t recall where he got it. So, we have an unsourced image (that most likely came from a Western Fuels Association “science advisor”) that is key to all this nonsense.
Indeed, the unravelling of the LIA, MWP, and Post Industrial climate change mystery and the development of the methods and data sets went hand in hand, outside of instrumental and delta 18.