Update See my review of the BEST methodology here.
Physicist Richard Muller has a piece in today’s Wall Street Journal that should be read by everyone (The Case Against Global-Warming Skepticism).
Muller concedes—in public—what many skeptics have claimed for years: that our temperature record is poor, especially over the oceans, that it is limited, filled with errors and biases, and when used as a basis for judgment, leads to over-certainty.
The temperature-station quality is largely awful. The most important stations in the U.S. are included in the Department of Energy’s Historical Climatology Network. A careful survey of these stations by a team led by meteorologist Anthony Watts showed that 70% of these stations have such poor siting that, by the U.S. government’s own measure, they result in temperature uncertainties of between two and five degrees Celsius or more. We do not know how much worse are the stations in the developing world.
Using data from all these poor stations, the U.N.’s Intergovernmental Panel on Climate Change estimates an average global 0.64oC temperature rise in the past 50 years, “most” of which the IPCC says is due to humans. Yet the margin of error for the stations is at least three times larger than the estimated warming. [link mine]
Even more delightfully, Muller admits that it has not been growing stormier (sorry, Big Al):
The number of named hurricanes has been on the rise for years, but that’s in part a result of better detection technologies (satellites and buoys) that find storms in remote regions. The number of hurricanes hitting the U.S., even more intense Category 4 and 5 storms, has been gradually decreasing since 1850. The number of detected tornadoes has been increasing, possibly because radar technology has improved, but the number that touch down and cause damage has been decreasing. Meanwhile, the short-term variability in U.S. surface temperatures has been decreasing since 1800, suggesting a more stable climate.
Nevertheless, the Berkeley project he led—which brought together physicists and (finally!) statisticians—were able to perform a complete re-analysis of all the temperature data, this time taking the main statistical statistical criticisms into account. Let’s leave aside whether these analyses were complete, rigorous, or recommended. Assume that they were. The findings?
We discovered that about one-third of the world’s temperature stations have recorded cooling temperatures, and about two-thirds have recorded warming. The two-to-one ratio reflects global warming. The changes at the locations that showed warming were typically between 1-2oC, much greater than the IPCC’s average of 0.64oC.
His conclusion is that “Global warming is real.” He hopes that “Perhaps our results will help cool this portion of the climate debate.”
But this blog, and all of the scientists who are critics, have agreed with this conclusion since this beginning. There simply is no debate on this question. There are no tempers to cool.
There has been, and still is, a vigorous disputation on the size of the warming and our confidence on this magnitude of warming. And Muller forgets that there has been a much more contentious argumentation about why the temperature has increased (in some places and cooled in others).
Just one thing about the first point of contention. If you look, say, at the year 1945 and compare it to the year 2010, you find warming of a certain size. But if you begin at 1940, just five years earlier, you find much less warming. Temperature increases (or decreases) are always relative to something (this is a point of logic, not physics). The choice of the comparator is arbitrary and subjective. Because of this, it is possible, and it has oft occurred, that someone wanting to stress the size of the increase will choose a comparator that best makes his case. Muller doesn’t state in his editorial what his comparator is; or why he has chosen just one.
However, we can afford to be as generous as Muller when he invited skepticism and allow that his statistical results are far more certain than any prior analysis. This merely brings us to the big question. As to that, Muller admits:
How much of the warming is due to humans and what will be the likely effects? We made no independent assessment of that.
From that he insists, “you should not be a skeptic, at least not any longer.” Somebody has to remind Mr Muller that skeptics aren’t skeptical of that some warming (and some cooling) has occurred. We are skeptics about our ability to explain this warming (and cooling), and to predict skillfully future warming (and cooling).
The fallacy—and it is a fallacy Mr Muller commits—is to suppose that because many climatologists have offered one theory for the observed warming (and cooling), and that, at least for the moment, they cannot think of one better, that therefore their theory is true. Thus, I remain skeptical.
Update Analysis of DJ Keenan’s BEST analysis coming tomorrow.
Matt
You might well be interested in – and I’d certainly appreciate your thoughts on – Doug Keenan’s correspondence with Richard Muller, recently posted up at Bishop Hill. Doug quotes you both liberally and favourably!
RichieRich,
Coming tomorrow. Thanks.
This certainly gives me confidence that we can accuratly measure the global temperature.
http://blogs.dailymail.com/donsurber/archives/44605#more-44605
I hope you also will read, and comment on, the technical paper on BEST’s statistical methods. Link is at Keenan’s headpost, and Muller touts the”sophistication and care that we took in the [statistical] analysis” in his email that Keenan quotes.
TIA & cheers,
Pete Tillman
I am flummoxed by the idea of averaging error ridden data. Doesn’t that make the mean noisier? I do not understand how we can give any credence to global means based on pre-satellite data.
Walter Sobchak,
I don’t mean this to apply to the BEST analysis, but what you suggest is possible and even useful. If you suppose you measured some thing with error, and you had a idea of the structure of that error, the average of the things can be improved by properly accounting for the error structure. We call these measurement error models. I have a paper (under Resume) for measuring skill in the presence of measurement error that is somewhat relevant.
Could you kindly explain the following:
if this is true: “Yet the margin of error for the stations is at least three times larger than the estimated warming”, then how can anyone claim to have detected a trend? Is such a claim statistically valid ever?
What Paul said.
“I do not understand how we can give any credence to global means based on pre-satellite data.”
Exactly, I even have doubts about satellite data as well. It amazes me that we believe we can measure the temperature of the earth. Measure it where?
I can’t even figure out a way to accuratly measure the temperature of my car. Where do I measure? On the top, sides, underneath, under the hood, in the trunk, in the passenger compartment, the tires, where? What time of day do I measure it and for how many years?
How can we accuratly measure the temp of the entire earth? Especially if we are using proxy data like tree rings or ancient lake slime. C’mon now.
I guess I should wait for the analysis but a couple of Muller’s statements are curious.
1) How does he jump from 2:1 to “reflects global warming”? How did he derive cause from that?
2) What exactly does “splice” mean? Is it:
a) ignore the missing data splice or
b) station moved next to blast furnace before/after record splice or
c) treat each string on its own (how is that a “splice”?)
The likelihood of global warming being real is very high but I’m not so sure it can be separated out of the temperature record without using independent sources. I’m working on something with discontinuities and I’m interested in his approach.
Walter Sobchak,
presumably averaging will reduce error. The assumption is that the errors will sum to zero. You should see the data coming off a thermocouple. The readings are all over the place. On spacecraft we always averaged the readings. The thermocouple readings are such because they have high impedance and are subject to external noise.
What Paul said is a different problem. Averaging may reduce the error. It doesn’t necessarily make it go away.
I don’t understand the last part of:
“We discovered that about one-third of the world’s temperature stations have recorded cooling temperatures, and about two-thirds have recorded warming. The two-to-one ratio reflects global warming. The changes at the locations that showed warming were typically between 1-2ºC, much greater than the IPCC’s average of 0.64ºC.”
If 2/3 report warming between 1C and 2C, doesn’t that make the average at most 2/3 to 4/3C assuming those that showed cooling actually showed 0 change? That’s a lot closer to being in line with the IPCC’s point estimate of 0.64C than Muller implies. Also, does the IPCC’s estimate include maritime temperatures? If only one study included them, wouldn’t that bias the other study’s results up?
Pingback: Global Warming Controversy Heats Up « Bohls Neighborhood Talk
I feel uncomfortable when a scientist, talking about science, uses words like “truly” or “very”.
For example:
(Oh yes. By the way, I thought that all scientists were supposed to be skeptics … but that’s by the way).
Or “much” as in:
These words have the same effect on me as someone shouting or armwaving – the louder the argument, the weaker the case being presented. Often, the ‘speaker’ knows it either conciously or subconciously and is trying to compensate.
“Truly”, “very” and “much” mean nothing – except as warnings that the argument is so weak that it cannot stand assessment without them. I will judge whether something is truthful or not based on the evidence.
(In a previous ‘life’ I occasionally wrote nuclear safety case documentation where such empty words were anathema.)
Pingback: William M. Briggs, Statistician » Do not smooth times series, you hockey puck!
Readers may enjoy this podcast from Cato.
Skepticism and Climate Science by Swaminathan S. Anklesaria Aiyar
http://www.cato.org/multimedia/daily-podcast/skepticism-climate-science
Just five minutes.
Pingback: It’s all about Muller…BEST it is not | pindanpost
I took the time to read Muller’s article carefully. It is worse than I imagined.
His intro is great. he concedes that we have good reason to have been skeptical. Many of the things we have harped on have been true. The record is spotty, the data is of very low quality. And some data comes from stations subject to UHI.
But, he says, with my magic machine, I can turn the sows ears into silk purses. Then he takes off his top hat, pulls out a bunny, and says: “behold — A Rabbit”
His first sally is great: He says, by way of concession that “Our work covers only land temperature—not the oceans—but that’s where warming appears to be the greatest”
A report from the guy who lost his keys and searched for them under the streetlamp. He left out 70% of the world’s surface — maybe more. Antarctica counts as land. So does the Sahara.
But the part that really slays me is the last clause. How can we possibly know whether the air over the land has gotten warmer than the air over the water, unless we measure it?
Then he makes his UHI move: “Only 0.5% of the globe is urbanized, so it makes sense that even a 2ºC rise in urban regions would contribute negligibly to the global average.”
The question is not how much UHI contributes to global warming, it is how much it biases the record. What percentage of the stations are urban, what percentage of the readings are subject to UHI?
His last move, to whitewash the station quality issue has been analyzed by Anthony Watt.
His conclusion “Global warming is real” avoids the questions of how much and what is causing it.
My conclusion is that the argument that the globe is warming is based on noisy data. Data that is far nosier than the signal extracted from it. All the warmists have is GIGO. 0.64 degrees my foot. I say they have proven nothing so far.
If one were really interested in a warming/cooling, then one would measure the enthalpy (heat content) of the whole “climate system”.
As it stands, the surface temperature analysis is based on the “average” temperature of each day, computed as the temperature half-way between maximum and minimum… that’s only valid arithmetically. The average heat content at that location for the day is not necessarily close to that. e.g. it may only be maximum temperature for an hour a day, but (near) minimum for 20 hours.
The heat content of the climate system isn’t just in the dry air over time. One has to measure moisture content and soil-/water-surface temperature for a start. Then, for each component, calculate enthalpy over each area (specifically, the thermal mass of each component). That gives the “instantaneous” heat content for the measured region.
Do that for the whole globe. Then sum for the global total at that instant.
It’s that simple. And it could be done if less money were spent on climate models and more on reliable measurements. One doesn’t get a staistically-valid representation of global weather if one only measures where it’s convenient.
Pingback: William M. Briggs, Statistician » D.J. Keenan’s And My BEST Methodology Reviews
You write: “skeptics aren’t skeptical of that some warming (and some cooling) has occurred. We are skeptics about our ability to explain this warming (and cooling), and to predict skillfully future warming (and cooling).”
And your commentators demonstrate that this is not true. eg https://www.wmbriggs.com/blog/?p=4525&cpage=1#comment-50277
I suspect your argument about management of the statistics is as perceptive as that. By which I mean wrong.
Mr Briggs,
I’m sure your blog has never contained any explicit denial of the reality of global warming, one can certainly check this anyway if inclined to. But your assertion that “all of the scientists who are critics have agreed with this conclusion since this beginning”, by being a perfect example of a No True Scostman, excludes quite a number of people from being a scientist and/or a critic. It’s heartening to see that at least on your blog science is the ultimate argument and Wattses, Pielkes and McIntyres of this world are not regarded as scientists and/or critics. Together with those dozens of thousands of scientists who tour the world collecting signatures uder petitions “against the global warming”. That’s a very positive advancement, Mr. Briggs. May their partisan trash journalism rest in peace, now let’s concentrate on the science next time, please.
Nice to see you in good form again, sir. There’s even a “fallacy” of a new sort. Well, “No True Scotsman” is a fallacy itself, isn’t it?
Adrian, you’re a ninny. One random comment does not a serious argument make. Yes, there are people who are skeptical that any warming has occurred, some of whom are skeptical because they claim the Second Law of Thermodynamics don’t allow, or because they think the notion of a global average surface temperature isn’t well defined, or half a dozen other reasons.
Scientifically, those aren’t serious.
If you dig through enough comments, you can undoubtedly find people who claim there’s not been average warming because before the Fall, the Earth was at a perfect temperature world-wide; on the other side, you can find people who think ET is going to come wipe out humanity for the crime of allowing global warming.
Those aren’t serious scientific notions either.
The real problem with this whole debate is that people aren’t precise in their definitions or careful in their arguments. One that’s suddenly become common in the last week is to say “BEST showed warming, so the CO2-forced AGW hypothesis is confirmed.”
Scientifically, that argument is no more serious than the other examples.
I agree with Charles Martin about one thing: “Adrian, you’re a ninny.”
I also agree with Charles when he says: “people aren’t precise in their definitions or careful in their arguments.”
What I said in the offending comment (and I truly care not a fig if you are offended) is that:
“I say they have proven nothing so far.”
There is a difference between believing something, and proving it. I believe that on June 12, 1994, O.J. Simpson killed his ex-wife Nicole and a young man named Ronald Goldman’ with “malice aforethought”. But the Prosecutor was not able to prove that proposition to the jury “beyond a reasonable doubt”.
With that in mind, please return to what I said. I found nothing probative of Muller’s proposition “global warming is real”, in his article.
His assertion that he proved that the temperature record over land shows an increase of less than one degree C, while admitting that few, if any, of the instruments used could determine temperature with an accuracy of less than one degree, does not persuade me of anything.
Of course, the real issue is whether a record that was derived from a few places on the earth’s surface, all of them on land, and most in a few prosperous, temperate zone countries, proves anything about the rest of the world. This is especially true of the 70% of the globe that is covered with water.
I have been told that some 20,000 years ago, the spot where I sit as I write this was covered by a mile of ice, and that the world was a colder place then. I have been shown a great deal of evidence of that assertion, such as river valleys without rivers, and I accept it. Clearly therefor, the globe is warmer now than it was then.
Further, I am a student of history, and I have made some study of the Norsemen and their travels in the Middle Ages, which were nothing short of amazing. That they settled in a place they named Greenland shows that the North Atlantic was warmer then. But those settlements died out before the European colonization of the Americas. In the middle of century XIX, seamen would sing that: “Greenland is a dreadful place, a land that is never ever green”.
In Early Modern Europe, the Thames often froze as did the canals of the Low Countries. So I am willing to accept the notion that the globe is warmer now than it was then.
What I am not willing to accept is the idea that we can derive meaningful global data from the instrumental record before the era of satellite surveillance that began ~30 years ago.
@Walter Sobchak
What I am not willing to accept is the idea that we can derive meaningful global data from the instrumental record before the era of satellite surveillance that began ~30 years ago.
Please explain why temperature anomaly data should not be “meaningful”. You’re aware that temperature measurements are not used to construct a “global temperature”, right?
Pingback: Monday Highlights | Pseudo-Polymath
Pingback: Stones Cry Out - If they keep silent… » Things Heard: e195v1
Grzegorz Staniak,
Foolish me, I assumed that the global temperature anomaly was crafted from temperature measurements. What, pray tell, do they use instead?
@Earle Williams
Stop building that straw-man. “Global temperature” =/= “global temperature anomaly”.
“Please explain why temperature anomaly data should not be “meaningfulâ€. ”
Because it is GIGO. According to Muller:
“The temperature-station quality is largely awful. … A careful survey of these stations by a team led by meteorologist Anthony Watts showed that 70% of these stations have such poor siting that, by the U.S. government’s own measure, they result in temperature uncertainties of between two and five degrees Celsius or more. …”
And that is in the 21st century. As we go back in time, the people get better, but the instrumentation gets worse. Imagine, if you will, a station attendant obtaining a reading on a moonless night in century XIX. He will have to hold a burning candle or a hot oil lamp next to the thermometer to obtain his reading. How accurate could that reading possibly be?
“You’re aware that temperature measurements are not used to construct a “global temperatureâ€, right?”
Everything I know, I learned from Richard Muller. He said:
“Using data from all these poor stations, the [IPCC] estimates an average global 0.64ºC temperature rise in the past 50 years, … Yet the margin of error for the stations is at least three times larger than the estimated warming.”
Did he give us bad information?
@Walter Sobchak
Did he give us bad information?
Most probably: no. He just drew erroneous conclusions.
If muller did not give us bad information, I can make no sense out of your statement:
“You’re aware that temperature measurements are not used to construct a “global temperatureâ€, right?â€
@Walter Sobchak
He implied a false conclusion. There wouldn’t be anything strange about a trend value of 0.64ºC even if all thermometers in the world were scaled in discrete 1ºC intervals and were a few degrees off. It’s not a trend of the average global temperature, but a global average of trends.
“It’s not a trend of the average global temperature, but a global average of trends.”
They take sloppy data, with unaccounted for sources of error, analyze that data, come up with a “trend”, as if it were meaningful, and run analysis on the trend numbers they derived.
It doesn’t make any difference, the in put is garbage, and I can only presume, the out put is garbage.
GIGO.
P.S. if they don’t know what the global temperature is, how can they continue to make dire predictions about the damage a rise would do?
Today’s news:
LONDON (Reuters) – Global temperature rise could exceed “safe” levels of two degrees Celsius in some parts of the world in many of our lifetimes if greenhouse gas emissions continue to increase, two research papers published in the journal Nature warned.
“Certain levels of climate change are very likely within the lifetimes of many people living now … unless emissions of greenhouse gases are substantially reduced in the coming decades,” said a study on Sunday by academics at the English universities of Reading and Oxford, the UK’s Met Office Hadley Center and the Victoria University of Wellington, New Zealand.
“Large parts of Eurasia, North Africa and Canada could potentially experience individual five-year average temperatures that exceed the 2 degree Celsius threshold by 2030 — a timescale that is not so distant,” the paper said.
I apologize for omitting the link to the Reuters item. Here it is:
http://old.news.yahoo.com/s/nm/20111023/ts_nm/us_warming
Re Sobchak’s post:
I guess these scientists published in Nature don’t know that the BEST long-term records, maintained in the US, state record high temperatures, do not show any trend.
Surely this (Reuters story) at least reads like voodoo “science.” (That is, they didn’t check their results against obvious available comparisons to check face validity.)
The Trend Is Your Friend.
The scientific finding that does not settle the climate-change debate
By S. Fred Singer
Letter to WashPost Oct 25, 2011**
Before you write off Bachmann, Cain, and Perry as cynical diehards, deniers, idiots, or whatever, [WashPost Oct 24] consider this:
Why are you surprised by the results of the Berkeley Climate Project? They used data from the same weather stations as the Climategate people, but reported that one-third showed cooling — not warming.
They covered the same land area – less than 30% of the Earth’s surface – with recording stations that are poorly distributed, mainly in US and Western Europe. They state that 70% of US stations are badly sited and don’t meet the standards set by government; the rest of the world is likely worse.
Unlike the land surface, the atmosphere showed no warming trend, either over land or over ocean — according to satellites and independent data from weather balloons. This indicates to me that there is something very wrong with the land surface data. And did you know that the climate models, run on super-computers, all show that the atmosphere must warm faster than the surface. What does this tell you?
And finally, we have non-thermometer temperature data from so-called “proxiesâ€: tree rings, ice cores, ocean sediments, stalagmites. They don’t show any global warming since 1940!
The Berkeley results in no way confirm the scientifically discredited Hockeystick graph, which had been so eagerly adopted by climate alarmists. In fact, the Hockeystick authors have never published their temperature results after 1978. The reason for hiding them? It’s likely that their proxy data show no warming either.
One last word: In their scientific paper, submitted for peer review, the Berkeley scientists disclaim knowing the cause of the temperature increase reported by their project. However, their research paper comments: “The human component of global warming may be somewhat overestimated.†I commend them for their honesty and skepticism.
********************************************************************
S. Fred Singer is professor emeritus at the University of Virginia and director of the Science & Environmental Policy Project. He is a Senior Fellow of the Heartland Institute and of the Independent Institute. His specialty is atmospheric and space physics. An expert in remote sensing and satellites, he served as the founding director of the US Weather Satellite Service and, more recently, as vice chair of the US National Advisory Committee on Oceans & Atmosphere. He is co-author of Climate Change Reconsidered [2009 and 2011] and of Unstoppable Global Warming [2007]
**Responding to: http://www.washingtonpost.com/opinions/the-scientific-finding-that-settles-the-climate-change-debate/2011/03/01/gIQAd6QfDM_story.html
@Walter Sobchak
However you paint it, it suddenly looks like everybody, even yesterday’s “skeptics”, have to reach the same conclusions faced with the same data. Yesterday Muller smeared climatologists, Curry agreed with some clearly erroneous “skeptical” claims, Brillinger was “another affable Canadian […] more like Steve McIntyre” (http://wattsupwiththat.com/2011/03/06/briggs-on-berkeleys-best-plus-my-thoughts-from-my-visit-there/) etc. etc. — like a team of Skeptical Superheros expected to bring the people The Truth. Today it’s suddenly GIGO, same sloppy science as always, just a PR stunt (even though physicist have been publishing pre-prints in arXiv.org for years and decades now), and you can find the word “WORST” used to refer to the project in a lot of places: just because the results are not “skeptical” enough for the denialist taste.
Sorry, but that’s just whining because scientists failed to pander to the prejudiced. And ignorant, too. The amount of nonsense and folly in the comments above alone is a nice proof that the “skeptics” are actually such a gullible bunch that sell them anything just as long as it sounds “skeptical” enough.
Orson: “I guess these scientists published in Nature don’t know that the BEST long-term records, maintained in the US, state record high temperatures, do not show any trend.”
State record high temperatures are the best? Have you demanded a station-quality analysis on the sites producing these records, some dating back a hundred years or more? No, because confirmation bias grabbed your brain when it saw what it wanted. Nice skepticism.
As for Dr. Singer, the response from realclimate.org at http://www.realclimate.org/index.php/archives/2004/12/myths-vs-fact-regarding-the-hockey-stick/ would be first, that it’s not one set of “Hockeystick authors”, it is several different sets of paleoclimatologists.
1) “There is no uniquely defined ‘hockey stick’ reconstruction. This term, as defined in our glossary, refers to the general shape, common to a number of independent reconstructions (see this comparison of various proxy- and model-based estimates), characterized by 20th century large-scale warming that exceeds the bounds of the more moderate temperature variations characterizing preceding centuries.
and secondly, that the proxy data cutoff is driven by the quality and quantity of available proxy data rather than by a desired outcome.
2) “Most reconstructions only extend through about 1980 because the vast majority of tree-ring, coral, and ice core records currently available in the public domain do not extend into the most recent decades. While paleoclimatologists are attempting to update many important proxy records to the present, this is a costly, and labor-intensive activity, often requiring expensive field campaigns that involve traveling with heavy equipment to difficult-to-reach locations (such as high-elevation or remote polar sites). For historical reasons, many of the important records were obtained in the 1970s and 1980s and have yet to be updated.”
Using proxy data has its undeniable problems, but it’s one thread in a blanket. If you argue that skepticism is warranted, shouldn’t we apply that same skepticism to the idea of allowing the unlimited release of greenhouse gases into the atmosphere without actively PROVING there is no harm from unlimited AGHG release? Yet it seems as though the burden of proof is entirely borne by the side which ultimately advocates the more sensible approach to risk management!
I will take Mr. Staniak’s decent into personal abuse as a declaration of intellectual bankruptcy.
Mike: He who asserts a proposition must prove it. Skeptics have the luxury of playing defense. None of us has asserted anything like what you claim. We have nothing to prove, we just have fun watching people like you and Staniak flopping around looking for a cogent argument.
@Walter Sobchak
Have you seen “personal abuse” in your life? I’m not the one calling other people “clowns” here. And the bit about “cogent argument” is quite funny, considering your own confusion and ignorance (“if they don’t know what the global temperature is, how can they continue to make dire predictions about the damage a rise would do”, repeating Singers lies etc). You’re continously proving you don’t know what you’re talking about.
Pingback: William M. Briggs, Statistician » BEST’s Worst Work
One can rightly say that the mean temperatures of land-based thermometers have increased. Beyond that, one can argue from physics there is no such thing as a “global mean temperature” (see Gerlich and Tscheuschner for example). We know that if we wants to measure an average speed, we do not average the “speeds” measured but take the total distance divided by total time. In the same sense, one should think of total heat distributed over total mass when thinking about some sort of “mean” temperature. Sadly, global writing on climate ignores heat and mass and is therefore weakly founded in basic physics. For an alternative kind of view, see Bo Nordell on global warming.