by Judith Curry
The new surface temperature dataset developed by the Berkeley group is now available, along with four manuscripts that have been submitted for publication.
For background on the Berkeley Earth Surface Temperature project, see this previous post. Richard Muller’s congressional testimony was discussed here.
Information on the new data set can be found on the Resources page at the Berkeley Earth Surface Temperature web site.
The papers
The group has submitted four papers for publication:
- Berkeley Earth Temperature Averaging Process
- Influence of Urban Heating on the Global Temperature Land Average
- Earth Atmospheric Land Surface Temperature and Station Quality in the United States
- Decadal Variations in the Global Atmospheric Land Temperatures
The fact that my name appears as second author on some of these papers is attributable to my last name starting with the letter “C”. The group has taken a “team” approach to authorship on this set of papers. My contribution to these papers has been in the writing stage and suggesting analyses. I have not had “hands on” the data, one of the reasons being that I do not have funding to do any analysis.
Here are my summary comments on the papers. I think the most significant paper is the Berkeley Earth Temperature Averaging Process, where Rohde et al. have introduced a new method for analyzing the data that appears to me to be a substantial improvement over existing methods. That said, I suspect there is still considerable research still to be done in this area. The Urban Heat Island effect paper is a nice study, and based upon the reporters’ questions, seems to be the primary paper of interest. Is this paper the final word on the UHI issue? Probably not. I will be interested in the reactions to this paper, since UHI is a contentious issue. The decadal variability study is very provocative (this is the paper for which I had the greatest input), but IMO the study needs to be extended back further in time. The Surface Station quality paper shows nothing surprising, but I understand that Anthony Watts is raising a concern over the paper.
In summary, there are no particularly surprising results here. The papers are initial steps in analyzing the data set, and the verdict on these particular papers will be given by others who do subsequent analyses.
The BEST PR strategy
In my relatively minor role in all this, I have had virtually no input into the BEST PR strategy. I have encouraged making the data set available as soon as possible. They were reluctant to do this before papers had been submitted for publication, and cited the problems that Anthony Watts had with releasing his surfacestations.org dataset before papers were accepted for publication. IMO, two of the papers (decadal and surface station quality) should have been extended and further analyzed before submitting (but that very well may be the response of the reviewers/editors.) I agree that it is important to get the papers out there and not be scooped on this by others, especially since Muller and other team members have been giving presentations on this. I have no problem with posting the papers before they are accepted for publication, in fact I encourage people to post their papers before publication.
In terms of how effective the team’s overall PR strategy has been, that is a subject that is certainly open to debate. But my impression is that the group has been honest brokers in all this in terms of trying to improve our understanding of the surface temperature data, while maximizing the impact of the data set and their research.
The spin
The press embargo on this lifts today at noon Pacific time. I suspect there will be pretty widespread media coverage on this, with both sides of the debate spinning this to suit their purposes. I have had queries from several journalists, to whom I probably did not provide any usable sound bites. Lets see how it plays out.
Note, the Economist published its article at noon EST, the link is here.
Anthony Watts has a post here.
Given that these two posts are already up and I quote nothing from the press release, I will go ahead and post this now.
JC summary:
Although the results of the analysis aren’t particularly surprising relative to previous analyses, I think the BEST project is very important given the importance of the surface temperature data set and the problems that have been associated with the CRU and NASA data sets, not to mention their disagreement. The BEST group is comprised of some extremely distinguished scientists (including Nobel Prize awardee Saul Perlmutter), and this topic has benefitted greatly from the examination of this problem by physicists and statisticians who were prepared to take a fresh look at this problem.
I am honored to have been invited to participate in this study, which I think was conducted very well.
In concluding, I will remind everyone that the REAL problem with the surface temperature data set lies with the ocean data. I hope that the Berkeley group will be able to extend their efforts to include ocean data.
I will provide updates with MSM and blog articles on this as I spot them.
Updates:
Thank you Judy.
We already knew temperatures have been rising since the last Little Ice Age (~1300 to 1870).
The real question now is whether or not:
a.) CO2 caused the recent part of this general trend [Consensus government science, Al Gore, the UN’s IPCC, etc.], or
b.) World leaders used global climate change to take control [1], remove the right to self-government [2], and rule through tyranny [3].
Near the end of the Little Ice Age, Thomas Jefferson recognized how very fragile self-government is:
“I regret that I am now to die in the belief, that the useless sacrifice of themselves by the generation of 1776, to acquire self-government and happiness to their country, is to be thrown away by the unwise and unworthy passions of their sons, and that my only consolation is to be, that I live not to weep over it.” [2]
1. “Deep roots of the global climate scandal (1971-2011)”
http://dl.dropbox.com/u/10640850/20110722_Climategate_Roots.pdf
2. “Thomas Jefferson’s letter to John Holmes (22 April 1880)”
http://teachingamericanhistory.org/library/index.asp?document=461
3. George Orwell’s 1949 book on tyrannical government, “1984”
http://www.online-literature.com/orwell/1984/
Oh definitely (b). HAHAHA!
Correction:
2. “Thomas Jefferson’s letter to John Holmes (22 April 1820)”
http://teachingamericanhistory.org/library/index.asp?document=461
RICHARD A. MULLER booms: “The Case Against Global-Warming Skepticism: There were good reasons for doubt, until now.
However not until the very end does he admit:
I find it Muller’s article repugnant in that he raises the straw man that “skeptics” say long term global temperature has not been warming. No “‘skeptic” scientist that I know of says that global temperature has been not been rising since the Little Ice Age. Muller builds a lot of publicity over the equivocation that “global warming” referring to temperature only versus “anthropogenic warming”.
Denis Ables responds
James Dillingpole rebuts: “Global Warming is Real”
Marc Morano summarizes critiques of Muller
Morano further summarizes how all the major evidence is turning against the warmists.
Lets get back to objective evaluation of the evidence, rather than “science by opinion article.”
With all BEST’s emphasis on improving uncertainty, I am surprised that BEST’s papers do NOT follow the international standards for reporting uncertainty. i.e. where the Type A and Type B uncertainties are reported separately, as well as reporting the combined “expanded uncertainty”, including the coverage factor k.
NIST Technical Note 1297 1994 Edition, Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results, Barry N. Taylor and Chris E. Kuyatt
Read what you write Judith! A PR strategy! Did Bohr have a PR strategy, or Maxwell, or Dirac.
The BEST PR strategy is not the best PR strategy because it became so important as to become visible. It’s THE news, as you can read at WUWT. And a total failure: science takes once again the back seat, and who cares if BEST does it for visibility rather than politics?
Your results and your work have just been buried by your team. Congratulations! /sarc
It seems to me that at this point, getting any controversial paper published is a political struggle. BEST (heh!) if it be in the full media spotlight. This may be unfortunate, but the political aspects have been true for years, it’s only the media spotlight that’s new.
Editors of whatever journal(s) these papers have been submitted to will know that their deliberations and decisions, no matter how secret right now, will eventually be exposed to the full glare of public scrutiny. IMO it’s not very good, but better than the available alternative.
I don’t think this is supposed to be new science; it’s supposed to be cleaning up old science. It’s actually less exciting than that; it’s dotting the ‘i’s and crossing the ‘t’s. It’s more like an accounting report. I can’t think of an historical parallel in science.
@P.E….
Something of an indictment of itself: Science ought to be full of peer-reviewed reports of replication of results. In the case of models, both other teams with the same models, and other teams with different models.
I’m not talking about just climate science here, All scientific results should be replicated one way or another.
Yes, et cetera.
Did Newton have a PR strategy?
Galileo?
Copernicus?
Archimedes?
And what about the first guy who discovered fire?
Did Newton have a load of blowhards on blogs accusing him of fraud?
Galileo?
Copernicus?
Archimedes?
if WUWT & Co (“Surface Temperature Records: Policy Driven Deception?”) hadn’t played on smearing the surface records BEST wouldn’t even exist.
And why would you say that so much time and energy is being expended by so many people in trying to discredit some unimportant sceptic blowhard?
Because “skeptics” overplayed it, even deluding themselves in many cases. If Muller hadn’t fallen for the propaganda too and so decided to investigate the matter all those “skeptic” smears would have been left standing.
And why would you say that some unimportant skeptik blowhard, whoever that is, expends so much time and energy discrediting so many people?
Why would you worry?
Since when has some insignificant blowhard’s opinion counted for jack-schi**??
And much more to the point, why should Muller & co worry about some insignificant blowhard’s opinion? I mean, it’s not as if policymakers and people that matter pay any attention to them, is it?
Since when feigned humility and rhetorical questions and mindreading counted for (insert your favorite dysphemism)?
Why are you losing sleep over it? What do you care?
Peter317,
Why do you think I’m losing sleep over this?
Willard, if you have no dog in the fight then why do you bother commenting here?
Peter317,
I opened my browser this morning to let you know I was going to think about that question:
> Why do you care?
I believe that this is a good question and is worth reflecting. But I am not sure what you have in mind. There are lots of things to care about, as should be obvious by reading the comment threads.
Your last question seems to presume that one must “have a dog in the fight” to care about these debates. I fail to see why this must be so, but even supposing you’re right it depends on whose dogs and which fights.
I do have dogs and I do have fights. You can click on my name and look around my tumblog to see for yourself. There is a FAQ where I compiled some answers.
I’ll think about your question. I’ll try to come up with an answer that I hope will satisfy us. And I’ll post it in my FAQ.
That being said, suppose I simply reply that I can afford to have an eccentric hobby. How would you react to that answer? Please be as honest as you see fit.
Best,
w
As usual you make idiotic analogies. Blogs did not show up casting sceptical eyes on climate science as soon as it came into existence. If Man, Hansen and his ilk were not that ready to go policy hunting with half baked ideas the science would not have moved into the realm of politics..
I wonder if the timing of this pre-publication media blitz is directed at the release of Donna Laframboise’s book discussed on the prior thread?
Ever hear the term “the medium is the message?” Here is how it is described on Wkipedia:
“Hence in Understanding Media, McLuhan describes the “content” of a medium as a juicy piece of meat carried by the burglar to distract the watchdog of the mind. This means that people tend to focus on the obvious, which is the content, to provide us valuable information, but in the process, we largely miss the structural changes in our affairs that are introduced subtly, or over long periods of time.”
Is there a better description of the CAGW movement’s use of mass media to effect policy?
Indeed – whenever information conflicts with my viewpoint, I also always turn to conspiracy and corruption as the most likely explanation. I mean, of course, there’s abundant evidence that explains the “timing” of this pre-publication – but why let facts get in the way of a good conspiracy theory?
You may want to check a dictionary on the actual meaning of “conspiracy and corruption.” But given the lack of effort (other than lots of typing) that you put into most of your comments, I will save you some time. Neither one is synonymous with public relations and the use of public media, which is what McLuhan was describing.
But thanks for reading anyway.
Gary –
The notion that all those people involved in the pre-publication of this report– in numbers such as would be required to result in coinciding with the release of Donna’s book, and with none of them raising the issue publicly that such a conspiracy or corruption was at play – is completely consistent with my understanding of the term conspiracy.
If your definition of conspiracy differs – so be it.
Joshua,
You clearly do not understand what you are talking about, which is not suprising.
Hunter,
I think Joshua knows exactly what he is talking about, and that you don’t.
Joris,
I will take your opinion in the spirit in which it is offered and give it all due consideration.
No need for a dictionary for the word “corruption”, Joshua, you can read all there is needed therein:
http://www.collide-a-scape.com/2010/04/23/an-inconvenient-provocateur/
Well, I DO think that whoever released the Climategate files and emails did it just before Copenhagen for the same reason you imply. And it did its job well. It has been nearly two years of the warmists being back on their heels. Now it is our turn to be back on our heels a bit, while they thumb their noses at us.
To quote Churchill, “This is not the end. This is not even the beginning of the end. But it IS the end of the beginning.”
The war goes on…
GaryM
http://xkcd.com/966/
I wonder if GaryM’s wondering has anything to do with BEST’s results.
I wonder if the timing of this is directed at
I find the timing and the full-court press – along with Muller’s “no need to be a skeptic anymore” op ed in the WSJ – just a little too coincidental for my comfort level. And (I could retrace my steps to find link if required, but I think it was the Press Release on their site) one of the pieces I came across indicated that BEST is sufficiently confident of the outcome of the journal peer-review that they anticipate inclusion in AR5!
How convenient and presumptuous is that, eh?!
If BEST is truly committed to “transparency” and input from the wider community, why on earth did they choose to take this particular route – rather than follow Judith’s example of circulating a draft and requesting input – before submitting for publication, not to mention going to the MSM!?
Very decent of BEST to make data etc. available for wider scrutiny, supposedly in parallel with “peer-review” process. But (pardon my skepticism), is this not a perfect set-up for dismissing any wider community criticisms if the “official” peer reviewers give the papers their blessings?!
And since you mentioned The Delinquent Teenager…, Gary, maybe someone could point Mandia and Gleick to Canada’s National Post, where Peter Foster could show them how a real and credible book review:is written:
Oh, and this edition of the NP also includes the first of two in a series of excerpts from The Delinquent Teenager…
..maybe someone could point Mandia and Gleick to Canada’s National Post..
Thereby adding 10% to NP’s readership. ;)
Oops – posted on the wrong thread earlier:
Too funny: From Anthony’s post:
Apparently, like Judith, Anthony has some difficulty in reading the comments posted by “climate skeptics” at his website.
Jesus, Joshua, you really are this dense.
Really?
Doubt that “the world is warming” aren’t posted in numerous comments here and at WUWT on a daily basis?
Yeah – I guess I am dense, because I could swear I’ve seen such comments often. I guess I’m either mistaken and actually haven’t read them, or those comments are actually false-flag comments posted by “warmists.” Sorry for my confusion, hunter.
I’m sure that we will see post after post by “skeptics” in complete agreement with the BEST findings that the “world is warming.”
Joshua, Anthony is right. Magnitude and causes. And I would add time scale and significance. I see post after post by skeptics about it.
Edim –
I see many posts from skeptics that focus on magnitude and cause also – but I will repost from the previous thread:
I read numerous posts: (1) Questioning that there is a valid mathematical trend of warming over the past 50 years, and (2) that consequently, the estimates of recent warming by the “climate community” are incorrect.
The BEST report indicates both a valid mathematical trend of warming and (essentially) agreement with the estimates of the “climate community.”
Posts in disagreement with the BEST findings are constant and ongoing – and have been for years. To suggest otherwise is disingenuous.
No doubt, there are some “skeptics” who accept the scientific estimates that are consistent with those (essentially) confirmed by BEST.
But to characterize “climate skeptics,” as a group, to be in agreement with the idea that the “world is warming” can only be done through a process of highly motivated reasoning.
BEST hasn’t been around for years.
P.E. –
In disagreement with BEST’s findings.
Joshua,
(1) Questioning that there is a valid mathematical trend of warming over the past 50 years
I question magnitude (and causes) of the warming trend over the past 50 years. I am still not convinced that the trend is not somewhat overestimated (UHI or human local warming). Even if we accept that real, non-local warming is for example 0.6 °C (over the past 50 years), that’s all causes (net). We don’t know how much of that is caused by anthropogenic CO2. It could be 0.5, 1.0 or 0 °C. Yes it could be zero.
Climate skeptics as a group? I don’t think so. By the way, the world is cooling over the past 8000 years.
Joshua, you have your issues reversed. What the trend has been is a big question. Whether there has any warming is a different issue.
Edim I cannot count the number of times I have had to argue with skeptics about whether it has warmed or not. Only to see them turn around and talk about the LIA.
“Edim I cannot count the number of times I have had to argue with skeptics about whether it has warmed or not. Only to see them turn around and talk about the LIA.”
I think most think that the climate has warmed since the Little Ice Age.
And there were glaciers which were formed during the LIA- hence it’s name. Not all of those glaciers which formed during during LIA have melted.
There some articles out there which say within 50 year or so a glacier which formed during LIA will have completely melted. Though because variation, no doubt glaciers formed from the last Ice age or maybe the one before that will also melt. As will some glaciers form during the LIA won’t completely melt within 50 year or 100 years.
We are in a warming period. You are either in warming period or a cooling period. Cooling are marked by most glaciers advancing. The LIA ended when most glacier stopped advancing and began retreating.
In terms of the interglacial period we been in for 10,000+ years we at around the average temperature or below average temperature- the warming we are measuring is a recovery from unusual cold period called the Little Ice Age.
Steven Mosher,
Whether it has warmed or not (or cooled) depends on time scale, magnitude and significance. It has cooled over the last 10 ka, 8 ka, 5 ka, 4 ka, 2 ka and probably over the last 1000 years. The coldest point of LIA was very likely the coldest point for the last ~10 ka. It has warmed since that point and nothing unusual has happened over the last 50 years. Solar activity points to an upcoming period of cooling (sc 23, 24 and maybe 25 are all weak cycles). After that who knows? Maybe Kim?
Heh, rising slowly toward the light. The globe is more likely to cool than to warm. If we can change the odds to warmer, why don’t we? Blow CO2 on dem Bones. Shake, rattle that coal.
==============
Joshua
You said it was off topic in a previous thread but its certainly not in this one.
Question-During which decade did the Earth start warming (around 1900 or so according to Giss. but you give your own answer)
Question -During which decade did this rise turn from natural to man made.
thanks
tonyb
tony –
I answered you previously. (because I wouldn’t want you to think me obstinate). But no probs, I’ll repeat.
As for a valid warming trend, I’ll go with what Judith has indicated
http://judithcurry.com/2011/05/24/what-we-agree-on/
As for when warming turned from natural to man-made, I’m a bit more agnostic, but I am not inclined to dismiss the opinion of the vast majority of experts in the field: controlling for natural variables, the warming trend over the last few decades is 90% likely to be more than 50% anthropogenic. I do remain somewhat “skeptical,” however – that the uncertainty might not be greater than captured in that statement.
TonyB,
Yes the warming started around 1900, or so. That’s a fair enough statement. That’s what we see in Mann’s hockey stick too. Mann’s graph isn’t just about the existence or otherwise of the MWP.
You ask when the warming turned from natural to man made. The warming of the early 20th century is usually assumed to be mainly natural – meaning that the ‘turn’ was mid 20th century. That is an assumption which may not be correct. Besides CO2 concentrations starting to rise, there were large scale changes in land usage at that time which had their effect on climate too.
Incidentally, any chance of the Berkeley group taking a look at the longer term temperature record? It would be good if we could finally get agreement on the temperature record of the last millenium too.
Nah, Joshua, you are putting forth a straw man argument.
The question is not has it warmed up since the little ice age. The question is whether it warming due to mankind and the method that is causing the warming.
Do some skeptics say it isn’t warming? Yes. Do some warmist (actually a lot ) say it is not only warming but the warming is increasing based on computer simulation? Has there been an appreciable warming in the last 12 years?
Nope, playing the straw man card doesn’t work.
goodspkr –
What “straw man?”
You say that there is one question (through your use of the definite article). I say that there is more than one question.
The problem with some “skeptical” argument, IMO, is that they lack clarity. On the one hand some “skeptics question the validity of assessments that show an increased rate of warming over the past 50 years, [so that’s question #1], and that assign a value to that warming [that’s question #2] that seems significantly greater than previous warming under similar conditions [that’s question #3] under similar conditions save the influence of anthropogenic CO2.
Yet on the other hand, some of those same “skeptics” say that they don’t question whether the globe is warming. I think if you consider that a bit, you’ll see a logical problem: If you don’t think the assessments that show warming are valid, how can you not question whether the world is warming?
BESTs work addressed some of the issues that some “skeptics” (who obviously doubted whether the world is warming – despite whether they might proclaim otherwise) had with the previous assessments of warming.
It is what it is. All-in-all, a relatively minor part of the big picture. As such, the question of attribution [that’s question #4] was not intended to be addressed by the BEST analysis – as the scientists involved have been quoted. But for those “skeptics” who based their “skepticism” on question #1, #2, or #3 (and I have seen many), BESTs work should be relevant.
Assuming you accept BEST’s analysis as valid, then what remains is the question of attribution – but as such, then what has to be answered is what is the variable that has increased the recent rate of warming to the extent shown in the BEST analysis. The vast majority of climate scientists think that the most likely explanation is the correlated increase in anthropogenic CO2 emissions – based to a large degree on a sophisticated analysis of the physics of how CO2 in the atmosphere interacts with sunlight. If you have another differential variable to explain the recent rise in the rate of warming, I’d love to hear about it.
But if you reject BEST’s analysis – then it stands to reason that you are not one of those “skeptics” that agrees that the world is warming.
The problem with some “skeptics” is that they don’t seem to realize that you can’t have your cake and eat it too.
Joshua,
While there are a few people who think that ALL data sets are so fudged up that “all warming has been purely fabricated” the vast majority of posters at WUWT, here, and elsewhere all pretty much agree that the Earth has warmed something on the order of 0.8 to 1.0 degrees C since 1850. Skeptics question how much of this has to do with the emergence from the little ice age vs. how much of this has to do with human emissions of carbon dioxide. If you could swear that you have seen comments “often” that indicate that no warming whatsoever has occurred since 1850, I don’t really know what to tell you other than (???).
Even skeptics realize that had there been no warming whatsoever since the tail end of the little ice age, we would all still be really uncomfortable right now, and still be suffering from famines and disease. So, in short, I suggest you do a LOT more reading of what the VERY common skeptical position actually is.
And yes, in anticipation, I am absolutely certain that you can find, copy, and paste at least several examples of comments which are in favor of the “no warming has actually happened at all, it is all merely data manipulation” theory. As I said earlier, such posts do exist, and if your bias is to look for them and single them out, then yes, you may be under the impression that they reflect the majority of skeptic claims.
Here is a brief summary of what most skeptics (especially those who understand the scientific method) believe:
1. There has been on the order of 0.8 to 1.0 degrees of warming since 1850.
2. Bad data quality and bad “handling” have probably caused some error in the exact calculated rise, and in the slope of the rise.
3. Calculating a “global average temperature” is probably a futile (and possibly a meaningless) exercise, but for now it’s what we have.
4. Some portion of the global temperature rise is completely natural emergence from the little ice age, and some may be due to increasing greenhouse gas emissions, but we haven’t seen any real nuts and bolts analysis attempting to show what percentage is due to what. We have merely been subjected to claims that “virtually 100% of the warming is unexplainable without the increase in CO2 emissions; therefore virtually 100% of the warming MUST be due to the increase in CO2 emissions.” This is clearly nonsense, because in order to believe that we would be forced to believe that we never would have emerged from the little ice age without the increase in CO2 emissions.
5. Depending on just how skeptical the commentor is, he/she may believe that nearly 0% of the warming is due to CO2, and nearly 100% is due to natural causes. Most skeptics do actually believe that at least some percentage of the warming probably IS due to CO2 emissions (though not all)
You can see MANY similar posts on virtually any “luke-warmist” or skeptical site. Unfortunately you make the same error that the majority of the BEST press releases are going to make. You assume that the majority of skeptics believe “there has been no warming at all since 1850”. Let me assure you, the majority of skeptics are damn glad we are not currently as cold as we were at the tail end of the little ice age, and you should be glad of it as well.
Peter –
I have no idea where this comes from. I never speculated on what “skeptics” think about temperature trends since the little ice age.
I have made no such assumption, and I can assure you that no matter how hard you try, you will find no comment of mine that rests on such an assumption.
My point is that Anthony’s statement, that “skeptics” don’t doubt that the “world is warming” is not consistent with a significant number of comments I read in the “skeptical” blogosphere.
There are two kinds of comments that would apply. One would be of the type you address in your post – that no warming has occurred since 1850. I don’t know how many such posts one might find, but I have no doubt that they would be fewer in number than two other kinds of posts that I read often.
The fist of those two other kinds of posts refers to a recent more time reference. For example, we can find quite a few posts that say something very much to the effect of: “The warming has stopped.” or “There is global cooling, not global warming.” These types of statements would be in contradiction to Anthony’s statement, and those of many other skeptics, when they say that “Skeptics don’t doubt that the world is warming.” Our illustrious host, herself, as said that she doesn’t know any “skeptics” who doubt the sign of recent temperature trends (it is positive over the past 50 years) – so your reference to data from 1850 is largely irrelevant.
The second of those two other kinds of posts are those which say that the analyses of temperature trends from the “climate community” are invalid. If someone believes that to be true, then there would be no basis for not doubting that the “world is warming,” as a belief otherwise would be founding on no evidence.
The qualified language you use here “probably” is to be commended – but I think that if you read many comments in the “skeptical” blogosphere, you will find that such qualified language is definitely different than the language, if not most opinions expressed. It isn’t valid to characterize the opinions of “skeptics” as a group using language that is decidedly different than the language “skeptics” use to express their opinions.
You will note that the IPCC says that it is 90% likely that more than 50% of recent warming is attributable to anthropogenic CO2. That language is distinctly different than the claims you are attributing, as a group, to those you disagree with in respect to climate change.
Read David W’s post above regarding discrepancies between land and satellite data. There you will find a “skeptical” opinion that is not internally inconsistent, and that clearly lies outside of Anthony’s characterization (rather than, as we see “skeptics” doing often – hiding behind a trivial statement such as “The world is warming AND cooling”).
As much as I appreciate you sharing your perspective, your description about what “skeptics,” or “most skeptics” believe does not account for the evidence I see when I read sites such as WUWT or this site. What people say is more instructive in that regard than how you describe what they say.
There are clearly some “skeptics” who have this opinion, and who post comments that are entirely consistent with that opinion. This is the opinion that Anthony suggested that this viewpoint is uniformly characteristic of “skeptics.” My experience shows me that assessment is inaccurate at two levels. First, I read many posts that are outright in contradiction to that viewpoint. Second, I read many posts that state such viewpoint as an underlying belief – but where the accompanying argument is inconsistent with that underlying belief. Again, if you do consider the “climate community’s” analyses of temperature data to be invalid, then you have no reason to believe that: there is a trend of warming (although limited), and consequently, there is no reason to believe that anthropogenic CO2 is explanatory for any % of recent warming (because you have no valid analysis to prove any recent warming has taken place).
Great work Joshua, I admire your patience and energy.
So if one loony commenter says something on Anthony’s site, that has to be his position? What does that say about some of the comments in alarmosphere sites? Or OWS?
Maybe Joshua is thinking of one of those sites where the moderators delete all comments they don’t agree with.
“One loony commenter.”
Too funny.
And allow me to ask, when Anthony posts breathlessly about a snowstorm in the Sierras, or a seal showing up in Boston Harbor, is he doing so to underline his belief that “the world is warming,” but he only questions the attribution and extent?
Of course, that last question asks for speculation – and reasonable disagreement is understandable.
But the notion that the phenomenon can fairly be described as “one” commenter is clearly false.
And I’ll further add that I don’t judge such comments to be “loony.” That is entirely your assessment, not mine.
I judge them (usually) to be a product of motivated reasoning. I have seen many logically constructed arguments that conclude the world isn’t warming. Not “lunatic” in the least. But the conclusions seem (often) to clearly be derived from a motivated process of reasoning.
You’ll have to ask Anthony about what his intent is with some of his threads. He goes off on a lot of non-climate related tangents. A few days ago, Willis had a thread about hitchhiking to Grants Pass. It was Willis’ version the Hitchhiker’s guide to the Galaxy or Zen and the Art of Motorcycle Maintenance, but it had nothing directly to do with the climate. I think one of the reasons why his site is #1 in readership is that in some ways, it’s more like a nerd magazine than a climate blog.
Sometimes an article is just an article.
Joshua,
How do you feel about the constant stream of marginal papers in marginal journals and the fear-mongering based on them that you see in the news on a daily basis?
Stories that state that a freak winter storm are fully consistent with global warming and that this or that disease “could be” worse in 100 years due to global warming or insisting that tropical storm frequency is worse and is definitely due to global warming? Don’t you ever just roll your eyes and say “Oh brother”? I agree there are a few people on here with no scientific background and some will say there is no warming at all. But this may be more a reaction to the constant hype in the other direction. I think when looking at the data, most people agree the planet has been warming for hundreds of years and don’t really dispute warming in general but rather its causes and the extent they will go up in the future and whether this will be a good thing or a bad thing (depends on how much they go up) and how to address the issue. The stories Anthony posts about harbor seals or what-not are simply his rebuttal to what we see in the news every day. He does not think the world is drastically cooling.
Bill –
Polls show that the majority of Tea Partiers – the majority of whom believe that they are very well-informed about climate science, to the point where they say they don’t need any more information to inform their opinion (and I’ll note that David H. has said their superior knowledge about climate change is “documented”) – say that the globe isn’t warming.
Now – that doesn’t speak to the readership of this site or WUWT per se, but there is at least some degree of crossover (the exact amount is unknown, but many have expressed allegiance with Tea Party-like ideology), and certainly, if you include Tea Partiers in your categorization of “skeptic,” then it is fair to say, I think that in fact, a reasonably significant % of “skeptics” doubt that the globe is warming (according to a recent poll, @50+% Tea Partiers, @30% Republicans, @15% independents, and @8% Dems).
Now – I think there are some problems with the polls, and their results by shouldn’t be considered as precisely accurate. But I do think that there is a substantial amount of evidence that: (1) a significant % of “skeptics” think that the globe isn’t warming and, (2) a not negligible % of commenters at sites like Climate Etc., and WUWT do not think that the globe is warming.
In sum, I think that statements like Anthony’s and similar statements by Judith that skeptics as a group don’t doubt that the world is warming are demonstrably false.
Isn’t this Watts guy a TV weatherman, and don’t TV weathermen live off of anecdotes? Long term climate change is the opposite of anecdotal information.
Don’t Malthusians always dress up their doom in fancy formulas that always work, right up until reality happens?
WHT, you believers are the ones confusing anecdotes about every snow storm, every drought, every rain every hurricane as *proof* of AGW.
That’s what I mean. Between the fundies who believe in “God’s will” to explain everything and the rest of the populace lacking intuition on statistics and thus easily enthralled by a hole-in-one, we often have a tough job explaining the obvious.
“The issue of “the world is warming” is not one that climate skeptics question, it is the magnitude and causes.”
A. Watts 2011
“Instrumental temperature data for the pre-satellite era (1850-1980) have been so widely, systematically, and uni-directionally tampered with that it cannot be credibly asserted there has been any significant “global warming” in the 20th century.”
A. Watts 2010.
http://scienceandpublicpolicy.org/images/stories/papers/originals/surface_temp.pdf
First line of the summary. What a difference a year makes! I wonder what the Watts 2012 position is going to be?
Joshua is a self-appointed hall monitor.
He decides what is right and wrong, and issues firm notes to be taken home to those he flatters himself sa busting.
If they do not bring back signed notes, he calls his mommy.
It’s clear that Joshua is as transparant, precise and complete in his reasoning and writing as anyone who ever was! It’s just your awful bad luck that you appear to be utterly blind to this. A clear case of ‘pearls for the swine’ it seems…
Joris,
Oddly but not surprisingly, you sound like a sock puppet for our dear Joshua.
I will gladly stick with the opinion shared by many here, and let Joshua and his sock puppets play with themselves.
I’m rent with jowls of jocularity.
==============
You do seem to appeal to the opinion shared by many, hunter.
Would that be a consensus, by any chance?
Joshua
Thanks-it’s sometimes difficult to follow a thread because of the nesting. I agree with the 50 years but what about the hundreds of years of warming before that?
tonyb
tony –
Well – since we do seem to agree about what the data show over the last 50 years, and accordingly, since we apparently agree about the mathematical validity of the analysis of those data as represented by BESTs analysis – then your question is an important one, now isn’t it?
Take a look at this video.
http://berkeleyearth.org/video/
That video shows that the nature of the warming over the last 50 years is dramatically different than the warming for hundreds of years prior. My assumption is that the forces at play that increased temps in the hundreds of years prior to the last 50 have still been at play during the last 50 years, and that in the last 50 years the increasing rate of rise is attributable to some other forcing. Unless some other factor is offered to explain that difference, then I think the explanation that it is 90% probable that more than 50% of the recent warming can be attributed to anthropogenic CO2 emissions seems logical (although I’m not entirely convinced about the level of certainty that statement implies).
As far as that goes, there are occasional “stone cold” comments here. Does that make Judith a “stone cold climate denier”? Are you participating on a “stone cold” Tyndall-denying website?
Don’t worry josh, you seldom post/troll-bait on topic anyhow.
Oh no! Josh has resurfaced on another thread with the same post that he used in a lame attempt to derail the last thread. Pathetic. You got issues, Josh-ua.
Hi Judy – I encourage you to document how much overlap there is in Muller’s analysis with the locations used by GISS, NCDC and CRU. In our paper
Pielke Sr., R.A., C. Davey, D. Niyogi, S. Fall, J. Steinweg-Woods, K. Hubbard, X. Lin, M. Cai, Y.-K. Lim, H. Li, J. Nielsen-Gammon, K. Gallo, R. Hale, R. Mahmood, S. Foster, R.T. McNider, and P. Blanken, 2007: Unresolved issues with the assessment of multi-decadal global land surface temperature trends. J. Geophys. Res., 112, D24S08, doi:10.1029/2006JD008229. http://pielkeclimatesci.files.wordpress.com/2009/10/r-321.pdf
we reported that
“The raw surface temperature data from which all of
the different global surface temperature trend analyses are
derived are essentially the same. The best estimate that has
been reported is that 90–95% of the raw data in each of the
analyses is the same (P. Jones, personal communication,
2003).”
Unless, Muller pulls from a significanty different set of raw data, it is no surprise that his trends are the same.
Same crap, different day(SSDD).
actually they did…..why not read the paper first?
“We obtained this estimate by choosing 2000 sites randomly from a list ofapproximately 30,964 temperature recording stations world-wide that had not been used by NOAA, GISS, or HadCRU.”
Yes, and they got a different result. Look at fig 1 (we are talking about paper 4, “Decadal variations”) and there is a clear cooling over the last decade in the black BEST graph whereas the other 3 just show a levelling off.
Unless, Muller pulls from a significanty different set of raw data, it is no surprise that his trends are the same.
Assuming they used the same data (which apparently they didn’t, at least for some results), it would indeed not be surprising, if we assume that urbanization or siting quality do not have a large effect on trends.
Which I think was precisely the point of contention.
if we assume that urbanization or siting quality do not have a large effect on trends.
sometimes for amusement I go to the Weather Underground site and pull up all the weather stations in my area, there are 22 of them reporting currently. As I write this post the lowest is 55.7F out in the boonies at a State Agriculture research center and the highest is 61.3 which is located in the middle of town near the maritime center. There are about 4 miles from each other as the crow flies. Other stations near the water vary from 57.3 – 61. Two of the stations located on different barrier islands read in the 57 degree range so it is not just the ocean effecting the Maritime Center Station. All are within 5 miles of the center.
So please tell me how you derive a .1 or .2 degree per decade warming using these again. Maybe I am just missing something in the way you guys do trends. Since what is in the country now will be urban 5 to 10 years from now, the UHI effect is constantly increasing, at least where I live.
What they want is to eliminate that 5.6 degree difference from the trend. The problem isn’t urban heat, it’s the effect of urban heat in distorting the overall trend data.
. I read the BEST Station Quality tentative paper from the link over at Watts site. I suppose I was unclear as to my doubts so let me rephrase my concern on station readings.
lets say I take data from Kennedy Airport in 1970 and again in 2011 it shows a warming trend. I take data from John Browns farm in 1970 and again in 2011 and it shows the same warming trend. That I can buy and acknowledge as good data.
If I took data from a station in 1970 when it was in the middle of bean fields then took data form the same or nearby station 40 years later and it was now a urban shopping area and residential neighborhoods and it shows a warming trend how much is natural and how much is from the UHI ? A question totaly unrelated to this thread would be how models take into consideration the sheer amount of roofs that have been built, Air conditioners that have been installed and asphalt that has been poured int he last 40 years.
I do not doubt the Earth is warming, UHI data shows me that. It is the land and ocean data I have a issue with.
Apologies to Dr Curry and all the rest of the posters for the slightly off topic questions but this is a fascinating and complex field.
@Jim in SC…
Well, once we have a more general trend, you might subtract it from that. But you have to understand: they’re looking for a global trend. They included your Kennedy Airport station and your urban shopping area stations in the total average, but left them out of the “very-rural” average. Of the stations you mentioned, only John Brown’s farm was included in the “very-rural” average. Then the compared the two averages.
steven mosher is correct that the farm may have an acre of asphault around the station today when it didn’t 40 years ago, but unless there’s a general pattern of that sort of thing happening, it probably won’t be significant. IMO.
Thanks for the reply AK, so why not just eliminate all urban data from the reports,that way any UHI effect would certainly be eliminated.
In my mind at least though I don’t think I could ever have any confidence in report using surface temperatures. Too many factors involved that allow for error, including the political bias of those who are compiling and “homogenizing” the data. That does not mean that I question the warming, it ‘s just the degree of warming that I have questions on. That and what if anything we can do to change it and for that matter whether it is even a crisis
Anyway when looking at temperatures we still cannot explain past temperature changes and until we can do that we cannot determine how much of present day warming is due to anthropological influences.
Steven Goddard makes the interesting observation that the BEST temperature matches or is below Hansen’s 1988 Scenario C forecasts with NO CO2 emissions after 2000. i.e. Muller supports the case for Global warming skepticism.
I like the Economist magazine but the chart they have only goes to 2006, what’s up with that? One would think in this age that missing 5 years of data is odd. Plus the graphic shows a temp increase since 1800 anyhow. I’ll admit the video time lapse temp graphic of the temp data at dotearth is entertaining even though the colours are silly and misleading – the planet is not a melting hot iron, lol. They don’t show ocean temps on it either.
If the temp has gone up 1 degree, so what? Most of it looks like it was going up anyhow. We have 7 billion people now but back when it was cooler it was much less population so we need the extra heat and co2 and we will need more. We are lucky temperature is not going down.
Its a 10-year moving average, with the median year plotted, so the 2006 datapoint is the mean 2001-2011 temperature.
Thanks Zeke, I didn’t notice that, is it a simple moving average or exponential? And do you know why they use 10 years and not 5 or 8 or 20? My first experience with moving averages was in the stock market. The problem with moving averages is that all points are usually weighted equally but the last points are actually the more important as you probably know. Also it adds another factor that the average time could have been cherry picked to suit a trader’s message. For example for the stock, if choose a bigger number you flatten the line out. Traders often used both exponential in combination with simple along with the real numbers because both have criticisms. Also in the stock market people that use technical analysis like that are usually no better off than other types of traders. The trend is your friend (but all trends must end).
Kermit
Good points.
10 is always a suspect number. It’s unlikely to be the fittest number, if it was just picked for convenience; it’s unlikely a number picked based on the properties of the data would come out to exactly 10; even if it was picked to be fittest, the question is always fittest for what purpose.
And I don’t recall seeing the rationale discussed anywhere in their documents. If anyone knows, I’d be obliged.
Judith wrote: “I am honored to have been invited to participate in this study, which I think was conducted very well.”
Hmmm are you sure? Even though I sent a correction a correction to Dr. Richard Muller on October 6th, and got a reply, they still can’t fix the 6 occurrences of getting the name of our paper right. Look right under Figure 1 here:
http://berkeleyearth.org/Resources/Berkeley_Earth_Station_Quality
—–Original Message—–
From: Anthony Watts- ItWorks
Date: Thursday, October 06, 2011 3:25 PM
To: Richard A Muller
Subject: Re: Our paper is attached
Dear Richard,
Thank you for the courtesy, correction: Fell et al needs to be corrected to
Fall et al in several occurrences.
When we complete GHCN (which we are starting on now) we’ll have a greater
insight globally.
Best Regards,
Anthony Watts
And Dr. Muller’s reply:
—–Original Message—–
From: Richard A Muller
Date: Friday, October 14, 2011 3:35 PM
To: Anthony Watts- ItWorks
Subject: Re: Our paper is attached
Anthony,
We sent a copy to only one media person, from The Economist, whom we trust to keep it confidential. I sent a copy to you because I knew you would also keep it confidential.
I apologize for not having gotten back to you about your comments. I particularly like your suggestion about the title; that is an improvement.
Rich
On Oct 14, 2011, at 3:04 PM, Anthony Watts- ItWorks wrote:
> Dear Richard,
>
> I sent a reply with some suggested corrections. But I have not heard back
> from you.
>
> Does the preprints peer review you speak of for this paper include sending
> copies to media?
>
> Best Regards,
>
>
> Anthony Watts
Maybe nitpicking here, but Revkin is with the NYT, does that count as a copy to the media? He seemed to have a post ready. Doesn’t really matter but still.
Anthony Watts,
In the first comment to your post where you put the topic of Public Relations before the results of Best Temperature Project, we read this comment from JeffC:
> I warned you that this would happen when they asked you to assist … I hope you really aren’t suprised to have gotten fleas from these dogs …
Here is a Voice of God reply:
> REPLY: I took the BEST team at their word, without ascribing motives. Perhaps they started out with the best of intentions, but got paved along the way. – Anthony
http://wattsupwiththat.com/2011/10/20/the-berkeley-earth-surface-temperature-project-puts-pr-before-peer-review/#comment-772832
Your reply does seem to imply that the Best Team went to Hell with the process as you first envisioned it.
To clarify, here are two questions:
First, could we presume that you started with the best of intentions, and then legitimately speculate about anything we fancy and make some wink wink, say about your morality, once we said that we won’t ascribe motive to you?
Second, do you imagine Hell as cold or hot? There are conflicting views regarding this symbol. In some circles, it’s very hot. In others, it’s glacial.
Many thanks!
To true climate scientists this is a wonderful new dataset that might provide insite into real climate studies.
To the believers on AGW orthodoxy it is another black eye because this dataset still shows warming from 1910 to 1942 with an increase in emissions from 3.5Gt/y to 4Gt/y and cooling from 1942 to 1975 when emissions increased from 4gt/y to 20gt/y disproving any possible correlation between CO2 emissions and global warming which is the entire foundation for the case in support of AGW
honest science 1 global warming fanatics zero
“…warming from 1910 to 1942 with an increase in emissions from … cooling from 1942 to 1975 … disproving any possible correlation between CO2 emissions and global warming…”
It doesn’t disprove anything of the sort. Do you *really* need this explained to you?
Yes, Norm. Just because there is a lack of correlation, doesn’t mean there’s no link to cause. You must BELIEVE!
Hi Judith. Strange that such an important paper would be released before peer review. Is that a recognition that review is not important?
It’s fairly common for papers to be made public once submitted. It’s less common to have a media hoopla to celebrate the fact.
From what I’ve read it sounds like the papers have been submitted. I wonder if any reviews have come back yet?
Yes it is common. Do you think it is good practice?
It gives the authors a greater chance to avoid committing a flawed paper to the permanent record, perhaps. It gets more eyes on the analysis at an earlier stage. Some papers get mired in peer review and don’t see the light of day for months, or years. I, for one, am pretty glad the papers are out there now rather than a year from now. I’ll be interested to see what, if any, changes get made if, or when, the papers get published.
No, it seems to be a compromise between making the data available as soon as possible, providing documentation, not getting scooped in terms of initial publications.
But why would Muller make his dumb statement about convincing skeptics? The work is related to a very small piece of a tiny part of the CAGW pie. Does he have an inflated view of his own importance? Or a very flawed understanding of what the argument is about?
I guess since the “sceptics” were trumpeting the “unreliable temperature record” so loudly, the fact that Muller has proven this to not be the case should give many of them pause for thought, meaning this paper is significant to “sceptics”.
What remaining justification is there for denying what climate scientists have been telling us?
“the fact that Muller has proven this to not be the case should give many of them pause for thought,”
How has Muller proven that it is not so?
“What remaining justification is there for denying what climate scientists have been telling us?”
Climate science is telling us a lot of stuff, most of which has nothing to do with the temperature record, so what are you talking about?
“I guess since the “sceptics” were trumpeting the “unreliable temperature record” so loudly, the fact that Muller has proven this to not be the case should give many of them pause for thought, meaning this paper is significant to “sceptics”.”
No one is disputing that the temperature record was taken and is being taken inaccurately.
Meaning that the procedure one is suppose follow to take temperatures readings is not widespread and commonly followed. Resulting in getting measurements which can be 5 C in error.
The merit of such “unreliable temperature record” hasn’t being proven, rather instead, it’s considered adequate to provide accurate average temperature.
Just because you drive a backfiring car with poor brakes, doesn’t mean it’s unproven to be good car.
Since this has not been examined as all science has to be examined, there no reason to assume the driver hasn’t wrapped the car around some telephone pole.
“What remaining justification is there for denying what climate scientists have been telling us?”
There is plenty of justification, the temperature record was (and to an extent still is) about data quality and transparency. It was claimed/suspected that data was being manipulated in order to advocate a theory – a perfectly reasonable theory – but one in which the manner of scientific conduct opened itself to a great deal of suspicion, the infamous hockey-stick is but one example.
Since I have been following the debate, very few skeptics have been concerned by the voracity of the temperature record – in fact they use it to make the case about the doubtfulness of the CAGW proposition. (It’s worth noting that most skeptics also accept much of the AGW idea and science behind it). What most skeptics object to is overstating the attribution of late 20th century warming to GHGs and anthropogenic influences. And they point to the existing temperature record to make the case.
You responded to Norm earlier in the thread but I suspect you may have missed his point. He was pointing out that the correlation with CO2 emissions and temperature change is poor. It maybe that it has some effect but it is completely swamped by natural variability. The last decade is showing no warming at all despite 30% of all the GHGs we have ever emitted entering the atmosphere.
That does not mean we won’t ultimately have a problem, but it is safe to assume that the problem is not as large or as urgent as we first feared.
It seems to me the ‘voracity of the temperature record’ done et Dr. Muller up.
=========
Agnostic says:
“it was claimed/suspected that data was being manipulated in order to advocate a theory – a perfectly reasonable theory”
I’m sorry. Nutty conspiracy theories such as the one you describe are not reasonable. At all. Now proven to have been completely wrong, of course.
“the correlation with CO2 emissions and temperature change is poor”
Which part of climate science suggests there is such a correlation? How on earth could anybody who has spent even the most cursory amount of time on this issue even make such a statement?
What is see here is *not* any sceptical analysis of available information. What I see are facts, logic, and history being tortured to fit a pre-conceived notion.
BEST has put to rest some of that torturing (the silly conspiracy narrative Agnostic describes).
Judith
Is this posting before peer review
A rush to meet the deadline of IPCC’s AR5 cutoff? Or
Seeking a widespread public/blogsphere “peer review”?
Trying to make the data set available as soon as possible with some documentation
Judith
I’ve read all the way through but can’t see any of the nuances I expected-just a ‘global’ increase. Where is the mention of the 28% or so of stations that Dr Muller said were cooling, how many are static and how many are situated in uhi affected environments? Are there other papers I haven’t seen?
tonyb
Judith
I found the other papers and read this;
“We compare the distribution of linear temperature trends for these sites to the distribution for a rural subset of 16,132 sites chosen to be distant from all MODISidentified urban areas. While the trend distributions are broad, with one-third of the stations in the US and worldwide having a negative trend, both distributions show significant warming.”
AND
” A histogram of the station trends is shown in Error! Reference source not found.a, categorized by station record length. The distribution is broad with a width substantially larger than the mean; 67% of the slopes are positive, i.e. there are about twice as many warming stations as cooling stations.”
There seem to be several places with the message
‘Error reference source not found’
Consequently I’m not sure if the histogram showing the cooling stations is actually present in the paper. Do you know?
Poor piece of research claiming London is the first example of Urban heating. It was Rome, as various articles attest, with Nero being perfectly aware of the effect and the causes.
tonyb
If roughly one third show cooling and two thirds show warming then, given the huge uncertainties inherent in the averaging, it seems we do not know if it is warming or not. Note too that the UAH readings show no warming in the 1978-97 period when the surface statistical models show the most warming. The surface models are probably falsified by this difference.
Steve McIntyre said:
“I commented on one replication issue – the BEST assertion that, using the Anthony Watts classification, the trends at “good” USHCN stations was higher than the trends at “bad” USHCN stations. I attempted to replicate this aspect of the analysis and got an opposite result.”
If true, this demolishes BEST and those who helped them.
http://wattsupwiththat.com/2011/10/21/sceptical-berkeley-scientists-say-human-component-of-global-warming-may-be-somewhat-overstated/#comment-774358
It will probably end up demolishing McIntyre. He says that he hasn’t looked at the BEST methodology yet in his comment. So whatever he did, he did it wrong.
I’ve got popcorn and look forward to more Mannian-type “innovative” methods being exposed.
Rattus Norvegicus. Truly and admirable character. Never let’s preconceptions get in the way of his rock solid objectivity.
David,
As long as the paper is submitted by mid-Nov and made available to the authors of the appropriate chapter then its eligible to go in, far as I know. You don’t need to make it public in any way.
I’m intrigued by the statement that they wanted to avoid being scooped. Scooped by whom?
That caught my eye, too. I wasn’t aware that this was the newspaper business.
I expect we should stay tuned…
“I’m intrigued by the statement that they wanted to avoid being scooped. Scooped by whom?”
Maybe they were afraid Anthony Watts would be called to testify before congress and he would make a statement that he had seen the preliminary BEST data and it coincided with his views on UHI, and undermined Muller’s.
Nah, he’d never do a classless thing like that.
The cut off for submissions for AR5 WG1 is July 31, 2012. The cut off date does not seem to be driving this.
My email to BEST:
I was really looking forward to the BEST project results.
Then you go and make a dumb-a$$ed move like this.
Nothing yet peer-reviewed, simple typo mistakes uncorrected after advance notification, and different time series comparisons for the UHI paper.
Not to mention the full-court PR campaign. Just what are you trying to accomplish here?
It may be salvagable if you ever get it right – we’ll see – certainly off to a really bad start.
Way to blow a great opportunity – unless of course, this was the plan all along!
I’ve just taken a quick look at the funding information for the BEST team, which is:
Major Donors include:
– The Lee and Juliet Folger Fund ($20,000)
– William K. Bowes, Jr. Foundation ($100,000)
– Fund for Innovative Climate and Energy Research (created by Bill Gates) ($100,000)
– Charles G. Koch Charitable Foundation ($150,000)
– The Ann & Gordon Getty Foundation ($50,000)
In addition to donations:
So now (pending peer-review and publication) we have the interesting situation of a Koch institution, a left-wing boogy-man, funding an unbiased study that confirms the previous temperature estimates, “consistent with global land-surface warming results previously reported, but with reduced
uncertainty.”
The identities of the people involved with these two organizations can be found on their websites. Let the smirching begin.
Not sure where you are going with this one either.
Bill Gates is rich, but not much of a right winger or oil industry insider. I would tend to think his donation counter acts any Koch money.
I’m thinking maybe this is actually what it claims to be: an objective look at the data without reference to any political agendas. People like the Koch’s and others who’ve been pointed at by their enemies have a decent incentive to sign on, both to clear their names and to get in on the ground floor of what might actually replace the IPCC.
Wouldn’t be any surprise to me if the Open Society Institute also offered a chunk of money, and it was accepted. Presumably without strings, although that’s something that the public needs to keep an eye on. (In both cases.)
The big question regards the people running it: the executive board and the Science Advisory Board. I’m not as familiar as I should be with the names, but given our host’s apparent approval, and a very brief scanning, I’d guess this is a very good start.
Kermit,
Perhaps you missed the most important operative word-of-the-day: MAYBE.
I’ll try to underline them as we go along in the thread, for our own edification.
The Koch family, in the real world, funds a great deal of science research and journalism.
In the AGW fever swamp, they are funding satan’s child eating followers.
Well, I’d guess the same is true of the Open Society Institute, although I don’t have any proof, and I have to admit moveon.org isn’t a very good example.
Dr. Curry
Just finished reading the paper:
Decadal Variations in the Global Atmospheric Land Temperatures
I am still rewriting my rudimentary reflections on the AMO, so when completed it might be an entertaining short ‘read’ for the co-authors of the above paper.
On warming, my view is that the first two IPCC propositions, (i) that the earth has warmed over the 20th century, and (ii) that this warming is historically unprecedented, cannot be supported strongly by observational data.
IMO the data vary so much in quality over time that we do not possess valid or reliable data for the 20th century at all. We may have a rough picture of what has happened, but not something that warrants findings expressed to decimal places. To compare observational data for the 20th century with proxies for preceding centuries seems pretty hairy to me but, again, it may give us a rough picture.
The BEST analysis is of land data, which covers less than half of the planet’s surface. Everything I have read about SST gives me much less confidence that we know anything about temperatures in and over the oceans even now, and hardly at all for the 20th century.
What happens when you put land and sea data together? You get a dreadful mess of incompatible results, from which someone points out that year X was warmer than year Y, because of a ‘finding’ expressed to three decimal places. What has happened to common sense, to ordinary scepticism?
It seems to me that people must be almost desperate to be analysing and extrapolating stuff that is so poor in origin. Why do we do it? Because, again IMO, some people are simply wedded to the threat, and we have all been caught up in that mood. It’s so important. Humanity is at stake. Catastrophy looms. Etc.
I don’t think any believes catastrophe looms because everyone knows nothing will change. Even if with four more years of Obama. Why? Because the rest of the world intuitively understands that the science authoritarians of Western civilization are a big joke on the US.
Don,
“Everything I have read about SST gives me much less confidence that we know anything about temperatures in and over the oceans even now, and hardly at all for the 20th century.”
—————————————————————
From BEST FAQ site:-
http://berkeleyearth.org/FAQ.php
“What is next for Berkeley Earth?”
“As a next step, Berkeley Earth plans to address the total warming of the oceans, with a view to obtaining a more accurate figure for the total amount of global warming observable.”
Last I heard from JC was that funding was not yet in place for “next”. Maybe that has changed?
Green sand
Following my article on SSTs carried here
http://judithcurry.com/2011/06/27/unknown-and-uncertain-sea-surface-temperatures/
I have been out in my rowing boat taking SSts. Can I get some funding please?
tonyb
Tony, many thanks for the reminder, well worth a re-visit.
Re funding, I might stretch to an outboard:-)
Interesting that the BEST “plans to address the total warming of the oceans” rather than assess the accuracy/uncertanties in the historic SST data sets, which what I thought was the BEST objective?
Hay, ho, round and round we go.
tonyb, Your article was very clear as to how some of these T’s were collected and raised many questions about the validity of the readings as well as the problems with global coverage over time. Brief, as it has been. Even the blind could read the size of the actual error bars that came with most of the early data that had been collected.
We don’t have all, and it is not the best…NOW blame the Weatherman
For me, you did an outstanding job of putting light on our subject here.
Thank you, Tom
Tom
Thanks for your kind comments. I have had several email exchanges with John Kennedyof the Met office (a fine scientist) who wrote an article about SSts immediately following mine. The uncertainties are to me so great that it is impossible to quantfy the corrections needed to make them useable. The SST variability is also much greater than is claimed and the time window in which anyone could put a thermometer in the bucket and get an accurate picture of the internal temperature is as little as three minutes, depending on ambient temperatures. This information was derived through actual physical observation. In my opinion SSTs prior to 1950 or so should not be used. The jury is out on more modern readings.
tonyb
tonyb, If you would please ask your friend at the Met office, to give you ‘the rest of this story’, from late 2009. Just what were all of the ‘originals’, that were put down on paper… recorded and then ‘dumped’? What did Phil Jones mean exactly, when he wrote those words? Also, what is the percentage of historic weather records of the B.E., that are still extent? If your friend could provide you with a link to the manifest of what was collected from around the world and over what time frame does this useable record cover? This has been an obvious question that has yet to be addressed by MSM. We have come along weigh, from the Dooey Decimal System II.
“Scientists at the University of East Anglia (UEA) have admitted throwing away much of the raw temperature data on which their predictions of global warming are based.
It means that other academics are not able to check basic calculations said to show a long-term rise in temperature over the past 150 years.
The UEA’s Climatic Research Unit (CRU) was forced to reveal the loss following requests for the data under Freedom of Information legislation.
The data were gathered from weather stations around the world and then adjusted to take account of variables in the way they were collected. The revised figures were kept, but the originals — stored on paper and magnetic tape — were dumped to save space when the CRU moved to a new building.”
I just have to wonder what price a collector / Universal Investor, would have paid to keep it out of the land fill / energy to waste plant…?
Whooo No’s
Tom,
according to the Climategate e-mails, Phil Jones sent the data to at least two academics for research. I wonder if anyone has requested copies from them????
I am increasingly leaning towards the conclusion that they flew out the window when “post-normal science” walked through the door.
“My hope is that this will win over those people who are properly sceptical,” Richard Muller, a physicist and head of the Berkeley Earth project, told the Guardian.
“Some people lump the properly sceptical in with the deniers and that makes it easy to dismiss them, because the deniers pay no attention to science. But there have been people out there who have raised legitimate issues,” he said.
He can’t possibly be that stupid. His work doesn’t even begin to address causation. But I think he makes his agenda pretty clear.
I’m also curious. I wonder how he knows whether deniers pay attention to science. Dr. Deming is a scientist who proudly claims the title of denier.
It’s really amazing how often scientists open their mouths and manage to reduce the public’s respect for science with every word they utter. One is forced to wonder how good the science can be when rational thought seems so difficult.
Having read his WSJ piece I can only conclude that it’s worse than we thought.
I’ve always felt that from a strategic POV, the constant fighting about relatively small differences in recent temperature records is not smart. Which is not to say it’s not legitimate.
Already the MSM is misinterpreting this just as you’d expect: Skeptics Can No Longer Doubt That the World is Warming!!
As if it were some great AGW triumph. That is of course, not the issue.
I have this vague feeling that Anthony got himself snookered on this one.
“Already the MSM is misinterpreting this just as you’d expect: Skeptics Can No Longer Doubt That the World is Warming!!”
Actually, they seem to be following the lead of the scientist in charge of the project.
“I’ve always felt that from a strategic POV”
And that’s what’s important, that our strategies to defeat belief in CAGW be successful. Facts and science is not important, validating our prior held position is.
Absolutely.
Add to that the fact that land station temperatures were never ever the issue in the first place – it’s *sea* temperatures that are going to prove that CAGW is all a hoax.
“I will remind everyone that the REAL problem with the surface temperature data set lies with the ocean data. I hope that the Berkeley group will be able to extend their efforts to include ocean data.”
I think that deserves emphasis. If we look at global warming trends, they closely parallel ocean warming, and the land warming data have less impact on mean values. This is not surprising given that the oceans cover about 70% of global surface.
On the other hand, if the BEST land data are unsurprising, the ocean data will probably not be either to a large extent. Over the long term, there is a relationship between land and ocean temperature change, such that the ocean warms at a rate somewhat lower than the land rate but not dramatically lower. The relationship is determined largely by the greater heat capacity of the ocean and the temperature-moderating effect of evaporative cooling over the ocean. In addition, the ocean itself has a significant impact on land temperature change through advective patterns. Although controversial, some land/ocean temperature model-based studies have suggested that land warming is more a consequence of ocean warming rather than direct radiative forcing over land (this is possible despite the higher land rate because the smaller land heat capacity permits greater warming for the same amount of heat shared by land and oceans).
Keeping the ball rolling…nuts! How long will be before we see ads from class action attorneys on TV: “Are you suffering from meso-globalwarmia caused by big oil?
Funding five more years of more filing cabinets full of AGW junk science should pretty much do US in.
Question: Greece is currently suffering from a man-made disaster (the we’re creating here at home). If we want to make a donation to help them out, do we send the money to Greece or Germany?
put it to your nearest bank. it’ll find it’s way to greece somehow
holds head in hands…again Muller manages to grasp defeat from the jaws of victory…he could have published a definitive set of agreed temperature records….and the idiot thatch-gobbler blew it.
It looks like you science rejectionists are now denying you’d ever questioned the temperature record in the first place. You’d never raised the issue of the “urban heat island effect”. You’d never argued that the recent warming was due to various natural cycles and wasn’t at all unusual.
And there was I thinking that was exactly what you’d said! I’ll have to take a look at some old postings to refresh my memory, perhaps.
Never mind. You’ll still be able to say that “correlation doesn’t prove causation”. How about the “global warming is a good thing” argument? There must be a fair bit of mileage left in that one. We all like warmer weather, don’t we?
Tempterrain – nice to see you’ve solved the UHI problem. I’m sure Judith will gladly invite you to join BEST, since you’re so much ahead of them on the topic.
It fascinating to see how Ms. Curry avoids actually saying anything about the result … namely, that there is no significant UHI effect, in agreement with previous scientific studies. And it’s fascinating to see the lengths to which the deniers go to discount this work promoted by and funded by deniers.
Clearly a tribal alarmist.
No MAYBE now, but a CLEARLY.
tempt. writes: “It looks like you science rejectionists are now denying you’d ever questioned the temperature record in the first place. You’d never raised the issue of the “urban heat island effect”. You’d never argued that the recent warming was due to various natural cycles and wasn’t at all unusual.”
There’s a lot wrong with the above. Your first few sentences are simply not true. Who is denying those things? And your last sentence would lead one to believe that this project in some ways demonstrates causation. Show me where it does that so I can go ahead and eat my hat (my favorite Key West baseball cap in fact).
BEST seems like it took a fair amount time and effort.
1/2 million dollars donated and unknown [to me] amount manhours
spent on project.
What aspect of project took the most effort?
Anyone know?
Developing the Muller & Associates web site probably took a lot of it.
What did they do about polar coverage?
Why do their charts leave out comparison with the satellites?
Tilo… The data and papers released today are about land based data. The data they are looking at are surface stations. They relate to the land data from the other three land based data sets. Next they are going to produce ocean data to complete the global data set. The satellite data will relate more to the total global data when it comes out.
The issue with the papers released today have to do with the reliability of the surface station data. All the papers released today are available to download at the BEST website. They are taking public comments on them prior to submission for peer review.
Buzz. GISS polar data is also land based from surface stations around the rim of the Arctic. And then it is simply extrapolated across the poles. This causes huge errors.
As far as I can tell, Berkley also does a form of statistical extrapolation and interpolation that is called Kriging. If they do this across the poles like GISS does, then I see no reason to think that it doesn’t have the same problems as GISS.
Tilo… But that’s not what they’re looking at in these papers. They’re looking at land data. You can pull up the exact same land data for GISS and HadCRU.
I suggest reading the papers.
“The issue with the papers released today have to do with the reliability of the surface station data. All the papers released today are available to download at the BEST website. They are taking public comments on them prior to submission for peer review.”
Could it be said their greatest contribution is providing an accessible data base?
A summary of the UHI pre-pre-pre-pre-paper:
1. The UHI exists in named places (eg Tokyo)
2. Within the same paper, Tokyo’s UHI varies from 2C/century (introduction) to 3C/century (discussion)
3. By shooting in the dark (classifying in a rough fashion a large number of unnamed places), the UHI doesn’t exist any longer
4. We don’t have the time and/or the money to verify if the UHI remains disappeared when using named places, individually assessed and perhaps even (shock! horror!) assigned a degree of urbanization
5. A back of the envelope calculation suggests that if 27% of the Global Historical Climatology Network Monthly (GHCN-M) stations are located in cities with a population greater than 50,000, then assuming a UHI of 3C/century the the contribution BY UHI to the world MEASURED average is 3*27%=0.81C
6. Nevermind, there’s always space for pure speculation
No effect for UHI simply doesn’t pass the smell test. But let’s try a thought experiment. Let’s say that in 1900 you have a thermometer that is surrounded by 5 miles of concrete and asphalt in all directions. By 2000 it has 10 miles of concrete and asphalt in all directions. We’ll call this the Urban thermometer.
Then, let’s take a thermometer that is sitting in the middle of a cow pasture in 1900 and then in 2000 it is surrounded by concrete and asphalt for 500 yards. We’ll call this the rural thermometer.
Which of the two thermometers will have increased the most in temperature?
Seems to me that using the 16,000 least urban out of the 39,000 and then comparing them against the 39,000 doesn’t prove a thing. Better to use 100 thermometers that are known to have no change at all in site conditions than 16,000 that could still be effected by construction. Their method seems to suffer from the old brain dead mechanical approach combined with unverifiable assumptions that all the other sources suffer from.
And to simply conclude that UHI makes no difference? Come on – get out of my face.
Tilo… You’re missing the whole point. Rather than doing thought experiments and performing smell tests why not just read the paper?
Buzz – you’ve missed the whole paper.
Buzz:
I looked through the paper. I don’t see how it answers any of the questions. Comparing the 16000 least urban against the entire 39000 doesn’t answer the questions that I asked above. If you think it does, tell me how.
If find it strange that when one drops large quantities of napalm on a city like Tokyo, then the temperature drops.
I also like the idea that the ‘urban’ heat effect is directly coupled to lights at night. Do not bother actually visiting the site, use a satellite. Sure, you will miss the sewage farm, the rural airport, the warehouse, the quarry and so on.
Why bother going to all the trouble of monitoring what actually happened in an area, say from grassland, to dust bowl, to soy, to rice, to soy again, to corn to scrub land if you can uses a satellite to measure night time light pollution.
Pity you didn’t use a satellite to monitor the night time IR emissions and white light sources.
Nor does there appear to be any attempt to actually look where and why some stations appear to be cooling and other heating. Why not have a look at the top/bottom 15%. I know that it may be novel in this field, but in most other fields of scientific study there is an attempt to know where variance comes from.
UHI part discusses how much warmer urban areas can heat up.
With UHI one has the aspect is getting higher daytime temperatures and
one has aspect of increased night time temperature.
In Preliminary Findings regarding UHI
http://www.berkeleyearth.org/resources.php
They mention that Tokyo as compared to global land temperature is about 2 C higher. That doesn’t mean much as regional areas can vary a fair amount.
UHI seems to me to have greatest increase in night time temperatures, but for daytime increase in temperature the most important heating factors seems to darker colored surfaces and reduction in heat convection.
With night time temperatures it seems the greatest affect is the increase of heat capacity of man made structures- mostly paving.
Is UHI considered a “greenhouse affect”?
Okay, so we know that Berkely uses much the same station data as the other sources. And it looks like they use a method of statistical interpolation and extrapolation called kriging. Is there any reason to think that this might have produced a record different from what GISS got with normal extrapolation and interpolation?
Also, GISS extrapolates from polar shore stations across the ice. Does Berkely then “krig” from polar shore stations across the ice?
Tilo. I show 17272 stations in the inventory.
Thats twice GISS.
WRT polar stations, you best understand kridging before you spout off.
If you look at the time series they will show how every area is handled.
To show that there is a problem with the pole you have to show that. not merely assert it. at less than 2% of the total surface area you wont find much effect from treating it differently
You are welcome to download any data and prove that. not assert, prove. with your code and data made available
“Thats twice GISS.”
Uuuuuhhhh. And that’s important why?
Steven: “WRT polar stations, you best understand kridging before you spout off.”
Why do I need to understand kriging? It is a mathematical method. How is a mathematical method going to deal with going from water warmed shore stations to ice bound areas away from shore. They still have no thermometers across the ice. And in order for the method to work they would still have to consider transitions from melted shore stations to ice covered arctic. I doubt that they can do that, since the ice line information isn’t even available before the satellite era. But the question remains, are the “kriging” across the arctic. Maybe we should get the answer to that question that I asked before you spout off.
“To show that there is a problem with the pole you have to show that. not merely assert it.”
Wrong again Mosher. If they are going to use a statistical form of extrapolation then it is their job to show that it works in every place where they use it – not just assert it.
“at less than 2% of the total surface area you wont find much effect from treating it differently”
Wrong again Mosher. Hansen did a HadCrut mask of the GISS data and showed that there was no divergence when the poles were left out. Whereas there was significant divergence when they were included.
Mosher: “You are welcome to download any data and prove that. not assert, prove. with your code and data made available”
There you go blathering on again about code and data. Just because you found a new love that you think makes you a magician you now want to impose that on the rest of us. But it is the methods that matter, and they can be described in simple English. And if the methods are inadequate, then the code and data are also going to be inadequate. It’s the old GIGO rule that every programmer knows.
Berkeley, AFAIK, is land stations only. They probably treated the Arctic ocean as not land.
Does GISS treat the Arctic ocean as “not land”? If so, why do they give us coverage by extrapolation from land?
I don’t pretend to know the answer to what Berkely is doing. That is why I’m asking. So thanks Nick, but I don’t need the guesses. I can do guessing for myself.
The Kridging is going to use a correlation length and correlation scale. Its in the paper. go read it.
or watch how the data is calculated.. see something interesting?
http://berkeleyearth.org/movies.php
The question was, “do they do their kriging over the Arctic”? Now, “how is kriging done”? Did you get that?
You obviously dont get kridging Tilo. go read the paper and get the code.
you already think its extrapolation and interpolation. hopeless for you.
When are you getting your nobel?
Uh, not sure about ‘Kridging’, but ‘Kriging’ seems to be interpolation.
http://www.amazon.com/Interpolation-Spatial-Data-Springer-Statistics/dp/0387986294
Mosher: “you already think its extrapolation and interpolation. hopeless for you.”
Yeah, that’s what I think. From wiki:
“The theory behind interpolation and extrapolation by kriging was developed by the French mathematician Georges Matheron based on the Master’s thesis of Daniel Gerhardus Krige, the pioneering plotter of distance-weighted average gold grades at the Witwatersrand reef complex in South Africa.”
The question is “what do you think it is”.
Mosher’s definition of kriging. “A magical process for getting the right answer.”
“So thanks Nick, but I don’t need the guesses.”
No need to guess. At the bottom of p 30 of the Averaging paper they say:
“Our analysis and that produced by
NOAA explicitly construct an average that only considers temperature values over land..”
That rules out the Arctic Ocean.
Nick: “That rules out the Arctic Ocean.”
Oh, I forgot about the word “ocean”. So they are going to give us the Arctic Ocean as an SST? Eh, right Nick?
No, they have done a land only analysis, and have not mentioned any plan to include SST’s for anywhere.
Well, I’ve read the Urban Heating paper, (Wickham et al. 2011), and it looks like they’ve done a pretty good job. Basically what they did was “global land temperature record reconstruction using the Berkeley Earth Surface Temperature averaging methodology (Rohde et al., 2011).” for the entire dataset, 39028 stations, and a dataset of 16,132 “very-rural” stations that were at least 1/10 of a degree in latitude and longitude away from any urban regions shown on the “MODIS 500m Global Urban Extent classification map (MOD500) of Schneider et al. (2009, 2010).” (see Wickham et al. 2011 for quoted ref’s.)
They then compared these two reconstructions (see figure 5 on page 11), finding little evidence of significant Urban Heat Island effect, and indeed after about 1950 the rural reconstruction actually trends very slightly upwards compared to the total.
This is a perfectly valid way of doing it (IMO, but theirs also), and certainly settles any questions about UHI in my mind. (As if that matters, but let’s see what experts say.)
Ref:
Wickham, C., Curry, J., Groom, D., Jacobsen, R., Muller, R., Perlmutter, S., Rohde, R., Rosenfeld, A., Wurtele, J. (2011) Influence of Urban Heating on the Global Temperature Land Average Using Rural Sites Identified from MODIS Classifications Submitted
Devil is in the details.
1. Modis500 was calibrated on a training dataset that used cities in excess of 100K.
2. Modis picks out “built” not urban. A golf course or city park will register as Rural, however their .1 degree slop factor should get that.
3. .1 degrees is roughly 10km at the equator but less as you go toward the poles. They need to fix that in the final paper and use a constant radius.
That work is a bitch,
4. Station location errors average over 2km, last I looked. This BEYOND reporting precision. So they often report the Wrong location to .1 degrees
5. The urban/rural differential depends upon the actual character of the rural land before it was urbanized. see Imhoff and Oke.
6. They didnt control for de urbanization. Some rural sites used to be urban, so you need historical population to eliminate that.
That’s off the top of my head
If I’m going to second-guess the experts, I have to use my own intuitions based on their data, methods, intermediates, and conclusions. I’ll admit it isn’t “pound on the table” proven, but when, for instance, MOD500 is replaced with something more detailed, the whole thing can be re-run. If they did it right, it probably wouldn’t take more than a day or two.
You know more about this than I do, why don’t you go right to the paper and see what (if anything) they’ve done about the issues you see? That’s part of why they made it available now, rather than at publication.
Their station data set only has 17K stations in it, so Im trying to figure that out as we speak. Some of the files they speak about in the readme are absent from the download
The whole UHI question turns on the categorization question and its not that easy. Its certainly not as easy as using on indicator even with a buffer.
I read the paper already. that why I can refer to their .1 degree buffer.
I know thats an potential issue since I tried the same approach and then realized that the .1 degree buffer changes as a function of latitude. Thats why a more accurate approach is to use a distance buffer rather than a “degrees” buffer.
ask yourself
1. How far is .1 degree latitude and longitude at the equator?
2. How far is .1 degree latitude at 45 degrees latitude?
Ask yourself, where are over 50% of the stations located..
it aint the equator.
So basically by using a fixed degree buffer they have a distance buffer that varies with latitude. Small point, but those are the things I would start to look at
But it’s only a problem if the buffer is to small close to the pole. Else it just excludes a few stations that might have good data.
“If they did it right, it probably wouldn’t take more than a day or two.”
That doesn’t help. The categorization of rural would probably not change if you use a more detailed source. But “rural” can still allow for a lot of building around the thermometer. You simply can’t resolve the issue through mechanical means. You have to know what happened to the surroundings of every individual thermometer. That said, I find any conclusion that says there is no UHI effect absurd. The problem in part lies in the name. You don’t have to be urban to get a UHI effect. Take a thermometer 50 miles from anything and build a parking lot next to it and it will show a UHI effect. But the thermometer will not be reclassified as urban.
Agree Tilo. UHI is a misnomer. A better name would be Anthropogenic Local Warming (ALW) or ALCC. Some “rural” stations might have more warming trend than some “urban” ones if they developed more over the time period. Land-use change is important.
The first useful thing to do would be to use only the best stations with known stations history and as little change as possible.
Ok, AK that was pretty easy.
using the 17000 stations in their site_detail.tx file, I did the following
Using the modis data they used and several other datasets I looked for a rural station according to their rules. Then I pulled up other data to show you what you get
Step1: take the station locations and record the Modis value at the purported station location.
Step2. Take the average of all pixels within 5 minutes ( .0833 degree)
Step3. Take the average of nightlights within 5 minutes
step 4 Take the average of Impervious surface within 5 minutes
Step5 get the percentage of land within that 5 minute grid that is
designated at uban land
step 6: Look for a “built” modis pixel ( urban) within 10000 meters
Then I select all stations that have NO “built” modis pixels within 5 minutes.
Then, I look through all of those to find stations that “look” rural according to Modis ( at 5 minutes) but may be urban. I stress the “maybe” but this is the kind of double checking you have to do when using Modis or other products. Thats because Modis is built using supervised classification trained on a dataset where the “built” pixels came from heavily urban areas
Found 97 stations that would be classified as rural by Modis ( .08 degree buffer) but that MAY not be rural
Here is the first one, Lon Lat
110673 NORILSK 69.333, 88.125
According to Modis there are no built pixels within 5 minutes .0833 degrees
But go to google earth and have a look. about 12% of the surface is impervious, 3% of the land is urban according to other measures of urbanity
Here’s another one
20.3500,-74.5000
Modis thinks that this is unbuilt, BUT the impervious surface database says otherwise
If I get time I’ll post a google earth tour of the stations, but I really would like to understand why the file only has 17K stations in it first
It would also help if they publish the station list
If it has 17K stations, It’s probably the one that’s already passed their filter. These would be the ones they’re treating as “very-rural”.
So if I understand correctly, you found 97 stations out of 17K stations that were “iffy”. If so, my next step (if I had the resources) would be to replicate their run with and without those 97 stations and see what difference it makes.
It just occurred to me, Steven…
You’re looking for built-up areas encroaching on the station, but isn’t the Urban Heat Island effect dependent on lots of buildings? We’re not talking about the station being next to a sun-warmed building, we’re talking about a bubble of warm air a kilometer or so high, and several times that in diameter. Putting a parking lot on a farm isn’t going to create an UHI.
AK: “You’re looking for built-up areas encroaching on the station, but isn’t the Urban Heat Island effect dependent on lots of buildings?”
No, you are confusing classifications with reality. Certainly it is true that the more built up an area is, the more UHI effect you would expect there. But all we are talking about with UHI is the effect on instruments that results from the addition of man made structures around the instrument. So, indeed, if you have an instrument on a farm and if you put a concrete parking lot next to it, the instrument is going to start yielding higher temperatures. And that is what we care about – not the fact that the name has “urban” in it. On the other hand, if you have a thermometer in the middle of a city and it’s already built up, you won’t see much change if you put up another building or parking lot a mile away. Of course total city build up will effect the temperature over time because the larger the warmed area is the harder for the heat to dissipate. But when you start to build in a rural area, immediately around the instrument, the effect can increase even faster than it can in an already built up urban area. And that rural area won’t change its classification status for a very long time and until the building gets really dense.
I’m only pulling this out of my backside, but I would guess that out of 39,000 thermometers in their study they would be lucky to have a thousand with no UHI effect. These things are almost never put in national wilderness areas. They are near roads, close to where someone is able to read them and maintain them. And such areas are naturally subject to population growth.
But really, do you need to know any more than that they came up with a conclusion of “none, or negative, UHI effect” to know that they are doing it wrong. Come on – a little common sense please.
There is more to the problem than that AK
When you classify urban and rural as a binary distinction and say “rural is X” and throw all stations that dont met that standard into a urban bin you may be putting rural into the urban bin. you have to assess the false negative and false positive.
That is one of the problems with a binary distinction. when you get it wrong its a double whammy
Suppose there are 1000 stations.
Your classifier puts 300 into rural and assumes the other 700 are urban
The first problem is the false positive rate. so 97 out of 17000. Thats good. Next is the false negative rate. how many of those 700 are really URBAN?
Lets say that without more detail like a station list I’m not convinced they get the categorization done right.
Thats NOT saying the did it wrong. I’m not seeing evidence that they tried to assess their false negative and false positve rate. So, throw a few urban in the rural bin and some rural in the urban bin and IF uhi is small then you’ll show no difference or a negative difference.
There is another way to skin this cat. more later
Mosher, that makes no sense at all. BEST didn’t assume the non-rural stations were urban. They specifically sought to avoid that problem by not doing any comparisons involving urban only stations. Instead, they compared the stations “known” to be rural to all stations, including those that are “known” to be rural.
There couldn’t possibly be false negatives with that.
Steve, I have always thought that places that have been known to have had changes in the period of interest should be used as a positive control. USAAF/USAF bases that swelled in WWII, then had runway extensions in the 50’s, then were run down, increased in 80’s and then closed post-Berlin War would be a good start.
On the other hand, I am sure there mus be land grant agricultural universities, like MSU, that have farming research stations that have records of crop changes in different areas would also be useful.
The lack of positive/negative controls sickens me.
As for scouting out locations, hell the Boy Scouts of America have 4.5 million youth members, they all want badges.
You could do a deal with the BSA and get every location scouted, along with precision GPS coordinates and photographs inside a year.
Now there’s a good idea.
WMO has instituted a program to have all member nations upgrade the quality of their location data. It is now reported down to the second
which is what you really need. Issues
1. many reporting countries just put in 00 for the seconds
2. Historical records of prior locations
3. Metadata for prior locations.
That said, if you have a really TIGHT filter for rural and its rural now
they you can probably assume it was rural then. I can even check that with historical population figures for the entire planet. Question can you determine the location for the last 50 years.
For those where we can… more on that
Ya know, if you could replicate their work in a decent run-time, you could try it with some rural stations deliberately modified, see how many and how much error you could introduce before it starts to show. That would give you a limit of sorts on how many stations could be bad without affecting the data.
Course, they could have done/do that as well, it would make for a tighter paper. Or did they and it somehow didn’t register with me?
I think you have to have a record of changes in the immediate vicinity of the instrument – regardless of rural or urban classification.
But since you have been looking at this stuff, maybe you can tell me about time of day data. Given that time of day readings have an effect, why is there a bias introduced by the time of day changes when they are taken all together?
Tilo TOBS in the US is well documented. ROW? who knows.
But you need to either
1. do an explicit TOBS adjustment
2. “break” the record and let the kridger do its magic
Mosher. Again, not what I asked. The question is, why should time of day adjustments produce a positive temperature bias when all of the adjustments are taken together. And the last time I saw the chart, the bias was big.
Tilo. You dont know the first thing about TOBS if you have to ask that question. go read Karl. or read the report that jerry B did on John Daly’s old site. Or download jerry’s data. Or do your own homework
Mosher: “You dont know the first thing about TOBS if you have to ask that question.”
Well, Mosher, if you know the answer, then why are you blathering like an idiot without giving the answer? I’m not going to follow your “go read” recommendations because they never answer the questions that I’m asking. Your recomendations simply waste my time. Most of the time you don’t even seem to understand the question. You simply look at it as another chance to rant.
Tilo Reber, for what it’s worth, a couple days ago I was curious about an issue Mosher claimed had been disproven by Lucia. When I tried to verify this claim, I found nothing of the sort. Instead, all I found is she had addressed a different, far more minor issue which was somewhat related to the issue being discussed. I asked Mosher for clarification, accepting that I could have simply not found whatever he was referring to, but Mosher never responded. The only response I got was from another person, who got the same results as I did.
Given that (and other, similar experiences), I support your decision not to follow his “go read” recommendations. I know I won’t.
Brandon: Mosher loves to hear himself talk. He’s so anxious to take off and show everyone how insightful he is that he misses the question or the subject much of the time.
I can’t speak to that Tilo Reber, but I can say, to me at least, he’s a questionable source. There is no doubt he can provide useful information and insight, but I won’t trust what he says unless I’ve verified it. That isn’t horrible, but it does mean I question a lot of what he says simply because I don’t have the interest it would take to check for myself.
I’ll listen to somebody who tells me to “look it up” if they’ve earned my “trust,” but otherwise, it just rings hollow.
Tilo. do your homework.
Start here
http://www.john-daly.com/tob/TOBSUM.HTM
http://journals.ametsoc.org/doi/abs/10.1175/1520-0450%281986%29025%3C0145%3AAMTETT%3E2.0.CO%3B2
prior to WWII most stations used to measure in the evening, but over time they shifted to the morning ( 7AM) the empirical studies show that moving from 5pm as a TOB to 7AM introduces a cooling bias of -.9C
Approximately 30% of all stations ( as I recall) had to be adjusted for this change in TOB.
At some point you will do your own reading
Okay, let me repeat what I just said to Omnologos:
No effect for UHI simply doesn’t pass the smell test. But let’s try a thought experiment. Let’s say that in 1900 you have a thermometer that is surrounded by 5 miles of concrete and asphalt in all directions. By 2000 it has 10 miles of concrete and asphalt in all directions. We’ll call this the Urban thermometer.
Then, let’s take a thermometer that is sitting in the middle of a cow pasture in 1900 and then in 2000 it is surrounded by concrete and asphalt for 500 yards. We’ll call this the rural thermometer.
Which of the two thermometers will have increased the most in temperature?
Seems to me that using the 16,000 least urban out of the 39,000 and then comparing them against the 39,000 doesn’t prove a thing. Better to use 100 thermometers that are known to have no change at all in site conditions than 16,000 that could still be effected by construction. Their method seems to suffer from the old brain dead mechanical approach combined with unverifiable assumptions that all the other sources suffer from.
And to simply conclude that UHI makes no difference? Come on – get out of my face.
Judith, quick question. It looks like the AMO index increase of 0.55K from 1980 to 2000 roughly is not outside the natural variability as shown for the previous 100 years. Would not this be prima facia evidence that a lot of the land temp increase since 1980 is due to internal variability?
Thanks
Very interesting. Obviously it looks at the urban heat island effect in some detail, does the study also look at the effects of anthropogenic heat flux outside of built up areas e.g. freeways through largely rural areas etc?
Actually, it’s doing the exact opposite of looking “at the urban heat island effect in some detail“, rather it’s just using a very sophisticated global averaging system to compare the average for all the stations with those known to be not urban.
How about comparing the URBAN stations to the NOT-URBAN ones? And are you really unable to see how many questions are left open because of the way a station is classified as “urban”?
Well, you see, because of reporting problem with too many of the stations, they don’t know the locations of many stations that closely. Urban areas comprise only about half a per cent of the Earth’s land surface. Sure, if there was good data, they could do anything they wanted with it. The challenge here is to extract useful knowledge from really noisy data.
Point is, when it comes to the Urban Heat Island effect, if they have a known non-urban population to compare with, they can eliminate the effect, or at least put a good rough size to it.
OTOH there was no obligation to stick to all that “noisy data”. And given what is known about the well-known places (such as Tokyo) one would have expected a little more humility and questioning of their own methodology, in the discussion. If they could speculate about asphalt, why couldn’t they speculate about the apparent counterintuitiveness of their results?
Urban areas comprise more than a quarter of the measurement sites. Don’t forget that. As far as uncertainties go, a great deal of the measured warming could still be due to UHI.
“it’s just using a very sophisticated global averaging system to compare the average for all the stations with those known to be not urban.”
The results don’t mean anything because “not urban” doesn’t mean anything. You can have all kinds of construction happen around a thermometer and still have it classified as “not urban”. No UHI effect defies common sense.
I expect plenty of people will be looking for the reason. After all, it’s not that there’s none, it’s that it’s negative.
Wrong. Modis is 500 meter data. Not urban by their definition means that there are no built structures within roughly 10,000 meters of the station location.. .1degrees.
If you want Modis data you have to write the PI.
No Mosher, that is not their definition. I thought you read the paper. Their definition is that urban environments are those that are “dominated” by a built environment. Nothing says that you couldn’t have buildings and parking lots right next to the instrument. In fact, later discussion would indicate that any place smaller than 100,000 would be rural. Check page 5 under “Data”. And the 10,000 meters is relative to places that are classified by a map to be urban. It has nothing to do with there being structures present near the instrument. ROFL. Did you actually think that one third of their instruments were in completely unbuilt environments?
Tilo. You are confusing two things. here is how BEST used MODIS to classify
“An alternative, which we adopt here, is to analyze the urban-rural split in a
different way. Rather than compare urban sites to non-urban, thereby explicitly estimating
UHI effects, we split sites into very-rural and not very-rural. We defined a site as “veryrural” if the MOD500 map showed no urban regions within one tenth of a degree in latitude
or longitude of the site”
Like I said
:Not urban by their definition means that there are no built structures within roughly 10,000 meters of the station location.. .1degrees.
Perhaps you thought I was talking about the Modis definition of a built pixel which is 50% of the pixel being “built” and a continous area of 1km.
Still wrong Mosher. There is no such thing in the Modis definition or the BEST definition. You are purely pulling that out of your backside. Very rural simply means that it’s 10K from an area classified as urban. But it can be 49.9% built around the thermometer – as long as it’s 10K from an urban location.
Somebody needs to get their feet wet!
The oceans have the data, always had, always will. But we do not know about the historic, so quite frankly stop guessing about the past and use resources to ensure present day integrity.
Quite frankly, any statement or claim about whether this planet has warmed, is warming, is not warming, solely on the basis of land based sensors is incomplete and misleading.
I suggested the BEST team just graph just the rural unadjusted data and was told it was done but I can’t find it on the link they gave me.
I’d be more inclined to trust this data than the adjusted data.
I surmise that the four pre-pre-pre-papers will get torn to pieces in the next few days. The quality of the BEST work will be measurable in the way they will react to that (making the necessary adjustments, ignoring the lot, or circling the wagons).
Best part of the papers. its all about the uncertainty
Follands estimate of measurement noise has always bothered a nunch of us..
Here is BEST:
We note that the correlation in the limit of zero distance, � 0 = 0.8802, has a natural and
important physical interpretation. It is an estimate of the correlation that one expects to see
between two typical weather monitors placed at the same location. By extension, if we assume
such stations would report the same temperature except that each is subject to random and
uncorrelated error, then it follows that 1 − � 0 = 12.0% of the non-seasonal variation in the
typical station record is caused by measurement noise that is unrelated to the variation in the
underlying temperature field. Since the average root-mean-square non-seasonal variability is
~2.0 C, it follows that an estimate of the short-term instrumental noise for the typical month at a
typical station is ~0.47 C at 95% confidence. This estimate is much larger than the
approximately 0.06 C typically used for the random monthly measurement error (Folland et al.
2001).
I must admit I didn’t notice that. I suggested a while ago that one places a grip of temperature reading stations a couple of hundred yards apart and actually measure the error.
Did they actually do a control experiment?
Its based on the correlation length . Probably more comprehensive than a controlled test and derived from the actual data
It’s why you have a professional statistican on the job
NIce.
I noticed the 12% thing, but I thought it was a bit odd how they explained it. They appear to be looking at the correlation r and ascribing proportions of it to different sources – but one normally does that with r^2, and ascribes contributions to the variance. That would mean 22% of the differences is due to independent measurement errors. However, since there are two errors involved – one for each station being compared – the independent error in each station is presumably half that, and then you have to take the square root to get an SD. I’m not sure if they simply skipped over some intermediate steps or if they meant something else.
But besides the number itself, I found the chart it came from (fig 2 in the averaging paper) to be fascinating. Even at near zero range, there are correlations ranging from 1 right down to zero – 0.88 is just the average. The scatter then drops and widens down to 3000 km, and then it rises and narrows. What is the cause of this behaviour? Why the dip? What is causing so many stations to still be correlated (and anti-correlated) at extreme distances? What sort of spread ought to be expected?
The following charts add to the mystery. Figure 3 shows that equatorial stations show an even lower zero-range correlation, but a much higher positive long-range correlation. What causes that? They say that it is that tropical weather is less variable, so other influences are proportionately bigger, but it seems to me there’s a lot left unexplained here. If you interpolate between those stations using the mean correlation for the whole dataset, or if you interpolate between stations with poor local correlation generally, you would base your interpolation on assumptions of far more consistency than is the case, which I’d expect to underestimate error.
I can’t say what the consequences would be, and they might not be significant, but it seems like an area worth looking at in more detail. As this is work in progress, I don’t see that as fatal – presumably they plan to look at it in more detail. But I think the confidence they express in their conclusions is overblown when they have this and many other features, assumptions, and approximations down in the detail yet to explain/resolve. It’s a set of tentative early observations from work in progress, not settled science.
Figure 2 is of course a radial correlation function so that there are many more data points that are spatially separated at a distance than at close range. You can see that in that the sparse density of points near the origin. I am not sure how they are incorporating the radial density factor, but certainly the counting statistics are poorer near the origin.
Correlation functions of energy-conserved systems can show all sorts of wiggles. If some region is too hot, another spot will likely compensate by turning cooler. These are the long-range correlations that scientists are always on the look-out for.
The main reason for doing this correlation is to provide a good model for the kriging algorithm used elsewhere. The empirical correlation function is the piece that provides the interpolation for the missing points in the grid.
Scientists and engineers have been doing this kind of spatial regression under different names for years. I just picked up a book yesterday by Mumford called Pattern Theory and once again see how image reconstruction algorithms borrow a lot of these optimal prediction ideas.
“Figure 2 is of course a radial correlation function so that there are many more data points that are spatially separated at a distance than at close range.”
They haven’t sampled it uniformly on range? Oh dear!
The left hand end of the chart is the bit we’re most interested in. Why would they undersample it?
So that would mean the spread at short range is even broader than the visual appearance would indicate, and the narrowing of the spread at the other end is even more pronounced. Rather than any long range correlations, I would suspect in this case that it is because the radial correlation changes with latitude (and probably other factors). There isn’t any single correlation function, there’s a mixture; and there’s more diversity at short range than long.
Nothing they can do about that because it is just a geometrical effect. Do one of those Google shopping exercises and ask how many restaurants are within a certain radius of your location, and then start shrinking it and the number of course drops. That is basically the idea and it is not a matter of undersampling as it is of oversampling all the distant ones.
As far as the damping of the waveform with distance, there is indeed disorder associated with this. One of the classic models of uniform disorder on signals is what is called the Sine Integral and it gives an idea of how fast you can lose coherence with a mix of frequencies. On of the benefits of hands-on doing statistical signal processing is that you can look at these kinds of plots as illustrated by Figure 2 and immediately tell what is going on.
“Nothing they can do about that because it is just a geometrical effect.”
Of course they can do something about that! There’s a limited supply of point-pairs for the very shortest distances, but over most of the range you can set the sample density to suit. They just picked a method for choosing points that weights wider separations more heavily.
Not sure what you mean by uniform disorder on signals. I’m guessing you mean that if you have a uniformly distributed power spectrum of spatial frequencies over some fixed width, then you get a Sinc-shaped autocorrelation, or something related. Sure, but why would spatial frequencies be sharply limited that way? And given that the next figure shows that we do have a mixture of different correlation functions at different latitudes at least, why that explanation in particular?
Yes, the shape is reminiscent of a damped wave, and there may be something to that, but it would take more than just a similar shape to justify it to me. Why is it happening?
Thanks, but I’m having difficulty understanding.
Because of the disorder, this is run of the mill stuff. I will give you a simple derivation.
I’m glad you noticed that.
Thank God effects from UHI have been halted — that is what was leaving polar bears adrift on small chunks of ice to starve and die.
He’s even dumber than I thought. http://online.wsj.com/article/SB10001424052970204422404576594872796327348.html
The Case Against Global-Warming Skepticism
There were good reasons for doubt, until now..
So he’s just talking about past warming.
That won’t stop plenty of people from using it to browbeat skeptics about the whole AGW schtick. And did you notice how that disclaimer was just at the bottom of the story? It’s probably syndicated, and anybody needing to remove a couple of column-inches will leave it out.
Deliberate?
The the only thing I would argue with in that op-ed is the characterization of the surfacestations project as careful.
The dividing line between CRN3 and CRN4 is the 10 meter distance to heat sources. They didn’t go out there with measuring tape for the most part and just eyeballed the station. Given that many of the surveyors probably had a motivation to put the worst spin possible on station siting. How good are you at determining the difference between 32 feet and 33 feet? This is the difference between CRN3 and CRN4.(10 meters is 32′ 9.7″). I race 30′ sailboats, where the 2 boat length rule comes into effect ofen and I always had problems getting that eyeball guess right, even though I had a good sense of what 30′ looks like. My guess is than many of the CRN4 stations are actually CRN3 stations and this would account for the vast bulge of CRN4 stations in the dataset.
I don’t think that in the end the misclassification of stations would make much difference in the conclusions, I am just arguing with the characterization of the surfacestations survey as careful.
It is disconcerting to read that someone width such vast resources of knowledge and thoughtfulness about the problem could say something this fundamentally stupid.
Apparently he seems to think that climate skepticism could only be based on doubt about the temperature record, and now that he’s settled his doubts about the temperature record he thinks the case against all climate skepticism is settled. There is a name for that category of error.
===========================
Apparently he seems to think that climate skepticism could only be based on doubt about the temperature record [snip]
Actually rather than settling anything at all this begs the question of how anyone seems to think we can determine the global temp (whatever that really is) to within a fraction of a degree when the resolution of the instrumentation is +/- a whole degree (or more!) This of course is exacerbated by worse resolution in past records.
Moreover, that the world has warmed since the little ice age isn’t being questioned anyway. We have plenty of historic documentation regarding the freezing of dutch canals, the hudson, thames, etc during the little ice age, so observing that this no longer happens isn’t fungible with “buying in” to CO2 alarmism and/or other anthropogenic attribution.
So certainly a great deal of skepticism *IS* based on the temperature records, and rightly so.
As near as I can determine the world has been warming naturally for the past 175 years and climate science has yet to quantify what the human contribution is. It seems to me that land use etc contributes the majority of record skew and CO2 is but a minor component. This due to known physics may change if the present understanding of things is correct, but I think from a systems standpoint it seems unlikley that our understanding of the “radiative balance” vis a vis the effects of convection is well understood at all.
In sum BEST contributes very little here. I’d like to see Dr Curry’s contribution to their knowledge base.
Bob Ward wastes no occasion to say something stupid: “So-called ‘sceptics’ should now drop their thoroughly discredited claims that the increase in global average temperature could be attributed to the impact of growing cities”.
Will the BEST team ask for a retraction?
Same comment I just left over at Lucia’s:
Based solely on a quick glance, (all I have time for), the paper “Decadal Variations in the Global Atmospheric Land Temperatures” appears to be a statistical analysis, and fails to consider the processes. It also seems to fail to acknowledge that:
1. the PDO does not represent the detrended SST anomalies of the North Pacific north of 20N,
2. the AMO does represent the detrended SST anomalies of the North Atlantic,
3. and that ENSO is represented by NINO3.4 SST anomalies (not detrended)
4. that their analyses are comparing apples to oranges to pineapples.
Since they were just trying to construct the evolution of LST’s why would they try to do attribution?
Retract that last comment, I had just looked at the press releases and comments from Rohde about attribution w/o looking at the decadal variation paper yet.
“The group has taken a “team” approach to authorship on this set of papers.”
Judith, I was encouraged when I learnt that David Brillinger was on the Team, as I expected some interesting statistical advances. And indeed the main statistical paper, BET Averaging, is advanced and interesting, but he’s not listed as an author, though warmly acknowledged. What happened there?
Didn’t Charlotte Wickham handle the stats?
Nick,
I asked Judith in another thread:
judithcurry.com/2011/10/22/changing-minds/#comment-125855
Here is her answer:
> I don’t know anything about this[.]
I met both Charlotte Wickham and David Brillinger while I was there visiting in March. Charlotte showed me work in progress, Brillinger did not. I think then that “thingsbreak” is correct.
Charlotte is a grad student. Nice lady, listened to what I had to say carefully. Brillinger mostly talked about himself and his work.
I think Brillinger, like Saul Perlmutter http://berkeleyearth.org/saulperlmutter.php was added for “star power”. I didn’t see any evidence of output from either. That’s not to say they did nothing, but I got the impression that the core workers were Liz Muller, Robert Rhode, Charlotte Wickham and Don Groom. http://berkeleyearth.org/dongroom.php
Groom was brought in because he was good at handling large datasets like is found when looking at scads of accelerator particle data.
From: Richard Muller
To: James Astill
Cc: Elizabeth Muller
Sent: 17 October 2011, 23:33
Subject: Re: BEST papers
…………..
‘Keenan states that he had not yet read our long paper on statistical methods. I think if he reads this he is more likely to appreciate the sophistication and care that we took in the analysis. David Brillinger, our chief advisor on statistics, warned us that by avoiding the jargon of statistics, we would mislead statisticians to think we had a naive approach. But we decided to write in a more casual style, specifically to be able to reach the wider world of geophysicists and climate scientists who don’t understand the jargon. Again, if Keenan reads the methods paper, he will have a deeper appreciation of what we have done.’
http://www.bishop-hill.net/blog/2011/10/21/keenans-response-to-the-best-paper.html
Does that mean the Brillinger who is not mentioned on the list of authors?
I’m confused. I’ve seen a number of people make comments similar to Judith Curry’s:
However, the UHI paper says:
How is this not surprising? BEST says their conclusions not only show UHI isn’t introducing a warming bias in the record, but instead, urban sites are introducing a cooling bias. They have no explanation as to why this would be, and it contradicts everything we know about UHI.
Surely I’m not the only one who was surprised to hear UHI isn’t a problem because the cooling bias it introduces is small (not that 10% seems small to me).
We’ll have to wait for the final paper release. THis is certainly one i’ll be looking at and having a go at crunching myself.
It’s a very suprising result to me, not to say it couldn’t be correct, i just can’t see how at present.
I believe that their result is correct in a technical sense. It’s just that what they were checking had no correlation to checking for the effect of build on thermometers.
They have no explanation as to why this would be, and it contradicts everything we know about UHI.
Ummm… no.
UHI may make an area warmer but once it hits a threshold level increase of population etc doesn’t change much annually. Dr Roy Spencer posted something at WUWT to this effect by saying that it appeared the major change in temp increase took place going from uninhabited land to a farming community level and then the rate of change slowed down (think of an exponential curve here.)
As such what Spencer says and what BEST claims is consistent and I reckon this to be accurate.
randomengineer, your response makes no sense. You say there is a threshold level for UHI, which is obviously true (though what that level is may be debatable). However, there is no logical connection between that point and what I said. That UHI in an area will eventually plateau could in no way cause the UHI effect to be negative. It could cause UHI to be smaller over time if those plateaus were not offset by newer areas of urbanization, but that can only reduce it toward zero. It cannot make it negative.
You have just argued the warming effect from UHI becomes smaller over time, therefore the warming effect becomes negative…
Does this mean that the UHI effect is actually negative or just that it is less positive than it was?
tt
Yes.
Yes to “just that it is less positive than it was” ?
It wouldn’t be possible to measure anything but a relative change would it?
tt
Sorry, was treating your question in a strictly Boolean sense.
I’m not sure how to interpret it otherwise.
First impressions.
The papers themselves, hardly important or unpredictable, but pending peer review and audit, seem final enough a slamming of the lid on UHI and of the various other commonplace attempts to impeach the terrestrial surface temperature record’s fitness for at least some climatology.
Jawdropping the thoroughness of the work and especially the improvement in the dataset by methodology, for the purpose of converting what had been mere meteorological data into climatology evidence.
And moving the record itself back 50 years (albeit with ridiculously large uncertainty).. useful for impeaching all the LIA garbage so many spout so willfully, and putting into perspective the uncertainty associated with the paleo derivations.
0.911C +/- etc. at 95% and so forth, back over 50 years? That’s.. just unbelievably improved.
How improved? Well, for a quarter century, I’ve been holding fast to the position that +/-AGW was probably unprovable on the surface temperature record within our lifetimes, that it was only the best candidate by something like a 10:1 ratio.
Now, it’s pretty much a done deal, barring something on the scale of a major meteor striking the Earth, and I have to change my stance. Unreservedly, AGW is certain, based on BEST.
That’s what’s come out of BEST.
One last impression. Sad commentary that an obvious and sloppy hatchet-job based on blog posts and one-sided research like the IPCC piece of the prior thread is going to make its author money through sales, while such a worthy effort of science as BEST is underfunded to achieve even a fraction of what it might do for policy and science.
while such a worthy effort of science as BEST is underfunded to achieve even a fraction of what it might do for policy and science.
Did you read the acknowledgements to see who funded the reearch?
Rick
Yeah.
And I didn’t see nearly enough names in the list, other than Koch, of free riders who benefit disproportionately from the subsidized and unpriced use of the scarce carbon cycle resource.
If they’re going to get so much for free, they ought at least have the decency to better fund scientific inquiry into the harms they do.
If they’re going to get so much for free, they ought at least have the decency to better fund scientific inquiry into the harms they do.
WTF? Harms?
If anyone else said this I’d expect it as tongue in cheek but I suspect that you actually believe this rubbish.
randomengineer
So we’ll put you down on the list of people who believe there has been no harm from any fossil fuel industry ever, then?
“So we’ll put you down on the list of people who believe there has been no harm from any fossil fuel industry ever, then?”
I can think nothing more “good” than extracting oil from the ground.
There are numerous consequences which aren’t good. To focus only what isn’t good is foolish.
If free people are given the choice: mine oil or not mine oil, the overwhelming majority choice will be to mine oil.
To think what the overwhelming majority would want as being “bad” is
fundamentally immoral. And simply very silly elitism.
gbaikie
“I can think nothing more “good” than extracting oil from the ground.”
Oil pumps produce the ultimate good.
I’m sure the theologists will be grateful they can stop arguing about that question now that it’s been settled. ;)
You’re missing the part about ‘free riders’.
That is to say, those who extract excess benefit without due compensation to those who ought benefit because of the cost imposed on them.
The carbon cycle is a scarce resource, a system that can refresh the carbon balance at a rate far lower than we impose on it through fossil combustion.
Who pays for is deficit?
The future, in unacknowledged carbon cycle debt.
Why do we leave this burden to the future?
Because free riders have persuaded us by touting ‘cheap energy’ arguments (that have no actual benefit to the economy, but only distort prices and reduce the democracy of the Market) to government to let them get away with it.
It’s past time to put an end to that.
Oh. You thought I meant oil spills and dead sea birds? What sort of nut do you think I am?
The correct term is not “mine oil” but “drill oil” as mines are not typically needed to extract oil.
But then again you have a point for the future, as the only way we will get any future gains is to lay waste to the landscape and actually mine for all this low-grade-oil-containing “substance”. This would be funny were it not so depressing.
Webby,
the correct term is to drill oil? Is that what they do for Canadian oil sands, a rather significant supply??
http://www.thefreedictionary.com/mine
It may not be the most common term, but, it is perfectly acceptable.
Hey Kumkwat, Lower grades of oil are of course mined and that’s why I say it will “lay waste to the landscape”. Look up EROEI and “tar sands” if you don’t understand the concept of “grade”.
Webby,
lay waste to the landscape?
Into propaganda much??
You might try something other than VideoGames for science some day.
Kumquat,
You haven’t seen the aerial photographs I assume. The Canadian government has regulations in place that the excavators have to replant all the vegetation and recreate the landscape and drainage once they are done scraping the earth, but not clear how comprehensive that will be in practice. Only time will tell. It turns out that the remediation takes a lot of energy and that is subtracting from the already meager EROEI from the tar sands itself.
Webby,
your problem is your assumption that everyone is as ignorant as you.
Yes I have seen the pictures of the mining areas. It would be irresponsible to talk about the issue without a minimum of familiarity.
Have YOU seen the pictures of restored areas???
http://www.capp.ca/environmentCommunity/land/Pages/RestoringLand.aspx#wdVL13PgVg4P
Hey Kumkwat, It figures that you would link to an industry-sponsored web page. Your fundamental problem is that you have this naive belief that everyone around you was born yesterday.
Webby,
yup I am naieve.
Should be easy for you to show they are LYING then!!
The naievete is electing alledged environmentalist Politicians who promise they will oversee these activities when all they do is extort money from the companies and then let them go.
Every volume is removed from the ground, transported, processed, returned to the location, and then the floral landscape is restored after the soil is returned in place (being so north the thickness isn’t that great and a lot may be peat). Thicknesses are between 30 and 100 feet, so it’s reminiscent of open pit mining . The USA doesn’t restore the West Virginia mountain-tops after the coal is removed (obviously, since all the energy used for this would result in a net negative energy capture), so we will see how the Canadians do.
How do you conclude it seems a “final enough a slamming of the lid on UHI”? BEST’s conclusions about the UHI (or at least urban/rural site differences) are in direct contradiction to practically everything which has come before it. If anything, it should open a whole new can of worms.
Also, how do you conclude AGW is certain based on BEST when BEST is just a temperature record? You can say it proves warming, but not the source of that warming.
Finally, if you’re going to talk about how great BEST’s uncertainty calculations are, you shouldn’t limit your comment to it’s spatial and statistical uncertainties. Acting as though there are no other uncertainties involved is grossly misleading.
Brandon Shollenberger
Good questions!
How do I repeat the conclusion BEST draws about the UHI, in plainer language? Well, not plainer, really. Perhaps I ought have disambiguated, by clarifying that I meant only the argument at WUWT that UHI could contribute significantly to the shift in global surface temperature was finally put to rest and buried. UHI itself is real and worth studying. At a couple orders of magnitude too small to affect AGW.
BEST’s conclusions on UHI are hardly new.
The minorly intriguing puzzle that some cities are heat islands some of the time, sometimes even significantly for the local region, and still might lead to cooler than otherwise wider area temperatures downwind some of the time (due likely particulates mostly) has been commented and speculated on before.
If you’ve missed the entire body of literature studying the field of urban cooling (here’s just one reference, http://www.hindawi.com/journals/amet/2010/681587/) and this is so novel to you as that you find it contradictory of all your one-sided sources, how is BEST to be held to account for that?
How do I conclude AGW is certain based on BEST?
The thing that most substantially held back the certainty of the AGW argument was the poverty of the datasets it was using as evidence. They were a juryrig of meteorological records slapped together with improvised methods that gained most of their credibility from the convergences that three somewhat separate groups obtained from largely overlapping raw data, and the somewhat consistent results over time.
Now, pending audit and review, AGW has a purpose-built dataset fit for use in climatology for the major AGW hypotheses, to sufficient precision and with sufficient accuracy. I cannot say it proves any of the hypotheses, but I can say it makes a much better foundation for the AGW hypothesis, when combined with other data, than the prior datasets, enough to half and half again the observational uncertainties.
Finally, while I’m a huge supporter of Chaos arguments and acknowledge this leads to important spatiotemporal questions, there are good methods for determining the scale and scope on which such considerations dominate.
AGW is chiefly a global multidecadal-to-millennial phenomenon — the observed magnitude of sources of uncertainty in climate over the globe on that timescale due Chaos is smaller than the margin of error in BEST.
Occam’s Razor tells us not to include more assumptions in our considerations than are necessary. If you’re assuming unknown unknowns, then you’re just being irrational.
I hope you find in my explanations sufficient justification for the shift in opinion I have expressed. If not, by all means, I remain open to convincing presentations of fact and method.
Bart R, thank you for responding, but I’m afraid what you’ve said is extremely difficult to understand. You claim BEST’s results are not new, then go on to say:
First, BEST didn’t cite any sources discussing “urban cooling,” so it seems cheeky to act as though I am just ignorant. This is especially true when the paper you linked to doesn’t even support your claim. That paper says:
The only example from the “entire body of literature” you claim I “missed” explicitly states the urban cooling effect cannot even offset, much less overwhelm, the urban heating effect. The only source you provide to dispute and criticize my points is a source which directly contradicts you. This contradiction isn’t subtle or hard to find. It’s clearly stated at the very beginning of the paper’s conclusions. I cannot imagine how you could possibly read your own source and not see it.
I can accept what you’ve said about why you changed your opinion on AGW, but I cannot think of any justification for what you’ve said about urban cooling.
Brandon Shollenberger
Thank you for your thoughtful correspondence.
It appears my tone came across poorly, and that it’s sometimes difficult for me to communicate even simple ideas without complicating them needlessly, for which I apologize.
I don’t claim to know certainly if urban cooling counters entirely or overwhelms UHI. The cite I provided was meant to produce a toe-hold into the literature, not a weighty argument. And yes, the UHI cooling shadow, though probably greater in area by some multiple than the direct UHI zones, does not typically overwhelm the geographic spike of urban heating upwind of that shadow, as you point out.
There is a large body of research looking into that subject, which Google is infinitely wiser about than am I; it appears to come down largely on the net cooling side, from what I can gather out of the democracy of the internet and the little reading on these matters that I have from time to time done.
You may wish to scan for original source material from reliable publications within this corpus and inform your own views.
That BEST restricted its efforts to its own clearly stated goals, and its references to only the references necessary to the goals it stated, ought not be surprising and ought not lift from either you nor I the obligation to inform ourselves independently on the topic.
That was the whole of my intention in providing that cite.
However, what is abundantly clear from BEST about the 2% of land surface area that experiences this UHI, and the somewhat larger areas included in the cooling shadows, is that 98% remains forty nine times larger, and that 98% is dominated by a heating trend that is larger than the UHI can account for alone.
A back-of-the-envelope maybe 2x-to-5x the size of GHG temperature difference, divided by 49 for the area, and compared to the 0.911 figure of the past half-century, amounts to little more than the size of the error bar contributed by UHI.
Cities might be pilot lights of AGW, but it is amply clear they cannot be the furnace, not by a longshot, given the scale and scope involved.
Bart R, I don’t think there was a problem with your tone. The problem is I’m somewhat well acquainted with the literature regarding UHI, and I’ve seen nothing of what you describe. For example, while I hadn’t seen the source you provided (I imagine due to it being newer) I had read the Jin 2005 paper cited in it. The results of both papers indicate there is an urban cooling effect, and both agree it is small in comparison to the urban heating effect. As such, the source you provided changes nothing about the situation as I understand it, and it certainly doesn’t indicate urban stations might have a cooler trend than rural stations due to their urban nature.
I’ll freely admit I don’t follow the studies of UHI as closely as I could, and I appreciate the source you provided, but as far as I have seen, there is nothing to support your point. To my knowledge, there is no work which indicates urban stations should be expected to have cooler trends than rural stations. I’d be happy to read more if I’m wrong, but I’d need help finding what I’ve missed.
With that said, I don’t think you and I disagree that much about UHI. As I’ve said before, my expectation has been the effect of UHI is 10-20% of the total magnitude of the trend. To me, the reason UHI is important isn’t that it would invalidate AGW. Instead, I’ve always seen it as mattering for how it affects our views on sensitivities, especially since it would be expected to have an uneven temporal distribution.
In any event, BEST acknowledges it has no real idea as to why the trend came out the opposite of what was expected. This is a trend which is 10% of the magnitude of the warming being observed. If the trend had been the same size, but in the expected direction, that would make for a 20% difference in results. To me, simply shrugging one’s shoulders in the face of such is highly peculiar.
I thought BEST was seeking to make a temperature record which was as accurate as possible. Their interest in UHI doesn’t seem to indicate such. They apparently made no effort to understand their results, choosing instead to simply dismiss the error as “very small.” They didn’t even discuss how it impacts the total uncertainty of their results, and it isn’t included in the uncertainty in their graphs. Their results may offer support to the idea UHI isn’t inordinately large (50%+), but they certainly don’t resolve my concerns about UHI.
Brandon Shollenberger
Yes, we somewhat agree, thank you for your clarifications.
BEST slightly amplifies some of the puzzles of the previous literature, with data that clarifies aspects of evidence for cooling.
Overall, I doubt anyone expected BEST to settle more questions than it in turn raises.
Now, we have a dataset more fit to build questions from.
“Now, pending audit and review, AGW has a purpose-built dataset fit for use in climatology for the major AGW hypotheses, to sufficient precision and with sufficient accuracy.”
LOL. There is no new data set. It’s the same old stuff that has been around forever manipulated in new ways.
“by clarifying that I meant only the argument at WUWT that UHI could contribute significantly to the shift in global surface temperature was finally put to rest and buried. ”
That may have been the case if the BEST UHI test actually checked for UHI. Unfortunately, it did no such thing.
Tilo Reber
I take a somewhat different view.
UHI is, while a valid topic of its own in climatology, a completely spent and irrelevant consideration in the big AGW picture, and especially no more can be claimed to disprove AGW by GHE or even to significantly reduce impact.
There may well be no (or little) new data per se though there is a substantial increase in the amount of data considered in constructing the dataset.
All datasets are subject to manipulations, or if no explicit manipulations are done, that in itself represents a type of manipulation.
The fitness of a dataset for various uses depends largely on the type of manipulation.
The meteorological datasets that have been slapped together for use in climatology are about as suitable as using milkshakes for motor oil.
True, they contain many of the same elements, but it just doesn’t work very well.
BEST hasn’t resolved all the problems of data for climatology from the 1800-present day timespan. Nothing can.
The opportunity to very meaningfully collect such information is truly gone.
The improvisation BEST came up with, however, allows some analysis that has some validity and value, and helps point the way to the type of information we ought be collecting, and yet still fail to.
Bart R: if you’re familiar with the previous thread you should know one has to.read and understand the actual text before commenting on it. There’s something jawdropping about the uhi paper for.example, but it’s not thoroughness.
omnologos (I still love the irony of the name. It was intentional, right?)
I’m familiar with the previous thread, and so I know how much one has to read of that text before understanding its agenda, method and most glaring weaknesses.
It’s a slam of the IPCC by stringing together a bunch of nasty things people have said.
So what? Where is there any value in that?
The number of nasty things I’ve said myself about just about everything would fill whole volumes on any one topic.
It doesn’t mean I’d encourage anyone to read the accumulated trash and jetsam of my own intemperate and one-sided slantedly hostile screeds, much less if intermixed with the bad-mouthing one could collect just by waiting for a blog to fill up with flames and trolling.
And here’s where BEST and the delinquent thing justify some comparative literature. In the case of BEST, skeptical but earnest investigators applied the best thinking they could to all the data they could collect, with remarkable transparency and intellectual honesty.. and surprised themselves by coming to conclusions that they had not expected, fairly representing both sides of argument, and including a fair selection of analyses and criticism from acknowledged experts coming from diverse positions. In the other case.. it’s a litany of faults chained together to come to the foregone conclusion, lacking representative evidence from the many actual stakeholders who the IPCC serves.
So, no, I don’t need to read a Harlequin Romance to know the type of fiction I’m missing, and I don’t need more than the ample evidence presented by the bookseller and commentators and what I’ve previously seen of the author’s work to dismiss her diatribe.
If there were a balanced discussion with details of who has used the IPCC’s work, and for what, insider information of the inner workings of their (ie the IPCC’s clients’) operations that no one else knows very much about, an in-depth cost/benefit analysis, a comparison of the cost of the IPCC to the cost, say, of one presidential nomination campaign, maybe I might be tempted to read more.
As it is, I have a stack of Mad Magazines I haven’t gotten to yet, that’ll be higher on my reading list.
Somebody please check if London is classified as “rural” for UHI purposes? :)
The real “clincher” is this. Imagine if a worldwide study on smoking or asbestos were suddenly to find that not only those won’t kill you prematurely, they actually increase your lifespan by 10%.
Now of course science can always be falsified. However when one’s study’s conclusions run against something quite well-known (be it tobacco smoke, asbestos, UHI, or slower-than-light neutrinos) then one should recognize the potential momentousness of the “discovery” and focus on finding out if there’s been errors about it, rather than casually mention it as a “confirmation” (and of what?).
London is “built” according to Modis. Wibleton is ‘unbuilt’
Steven Mosher is putting up a valiant defence, not so sure about effectiveness of it.
Oh, Mosh will be the first one to admit that UHI is likely small; while the BEST report is far from the final word on the subject, they do narrow the window of expectations quite a bit. But there is still work to be done, so we will see what we end up with.
To me, this paper increases, rather than decreases, the window of expectations in regard to UHI. I’ve said a few times my best guess for UHI was it would have a magnitude of 10-20% of the observed warming. This agrees with BEST’s conclusions, except it never occurred to me the bias would be negative. That means BEST has doubled my expected range, unless I just dismiss their results out of hand.
Then again, I’m one of those people who say 10% is not “very small” or “almost insignificant,”
I should probably point out the 10% figure I mentioned doesn’t include uncertainties. If you include those, you get a range from 0-20%. Personally, I cannot see how one would conclude it is almost not worth considering the possibility of a 20% error.
Zeke: Why would you think that the test that BEST ran has anything to do with checking for the effect of build on thermometers?
I have now read the first paper on the averaging procedures. It is well done (and data and code are available, so if I want to pick nits, I can). I have considerably more confidence in these data than in the other three records.
Why do they use a 10 year moving average? not 5 years or 20 or something else? It seems to me it hides the recent temp stability since 1998 even though we have record high 390ppm co2 in 2011 and the most recent years are more important.
I have a simple question. Does anyone know where I can find the data graphed in BEST’s temperature record? I know there is a directory with lots of data on their site, and that may have what I’m looking for, but it’s 174MB. That’s more than I can reasonably download with my internet connection, especially when I just one a single file.
How slow is your connection? 174Mb is relatively small. I do download/upload of files of that size pretty often on my 7Mb/S down / 1Mb/S up DSL connection. A download of that size only takes about 5 minutes or so in the real world. Uploads are quite a bit longer, but tolerable.
I have to say I’m rather amused at your comment about “the real world.” A lot of my neighbors would take issue with being told they aren’t living there.
As it happens, I currently live in an area where you cannot get a good, high-speed internet connection. The two main options here are satellite and wireless. I have a wireless connection, and it is limited to one gigabyte a month. This means downloading that would consume almost 20% of the bandwidth I can use each month. Obviously, I don’t want to do that just to get single file which is probably less than 100kb.
As a side note, there’s a difference between bandwidth (size of the data flow) and speed (quickness of the data flow). The wireless internet connection I have is quite good about speed, but bandwidth is far more limited (at any given time, and in total). The satellite options I looked into were the exact opposite, giving me much better bandwidth, but far worse speed. I know this is irrelevant to what you asked, but it’s always irked me that people use “fast” and “slow” in regard to bandwidth as though the actual speed is irrelevant. To a gamer, bandwidth is good, but speed is vital (it’s what controls your lag).
The terminology is bandwidth vs. latency. And you are correct on it. I got in on the Verizon data plan before they limited it, so I pay $30/mo for an unlimited data plan. Ain’t gonna change it until they make me! I use it to stream music from a favorite radio station a long, long way from here, although generally only in my car, and eat up a lot of bandwidth that way.
If you are looking for the data for the whole record though, the text dataset is actually 203Mb — I’m downloading it right now. It is a zip file, so I don’t think it will get any smaller. The Matlab dataset is 103Mb. If you want an copy, send me your address at johnsully at hotmal dot com and I’ll send it to you on a CD.
First, you have no idea how jealous I am of your data plan. Second, latency is a measure of how much the signal slows down, so it doesn’t directly describe the speed of the signal. The difference doesn’t change much, but it does make the comparison between bandwidth and speed more direct than the one between bandwidth and latency. Of course, we all know what each other is talking about, so it doesn’t matter.
Anyway, thanks for your offer to send me the data on a CD, but it’s really not necessary. I wasn’t looking for the entire dataset. I just want the final results. There is a graph (well, a couple graphs) on the BEST website which shows their temperature record, and I just want the numbers for it because eyeballing graphs is fraught with difficulties. As Kermit suggests, I can always use a different connection to download the full dataset if I want, but in the meantime, it would be handy to have the values for that one graph. It seems strange to think one would have to download over 100MB just to be able to graph their results, but as far as I can tell, the zip file you mentioned is the only spot you can find them.
With that said, if you do have a file with the data used for the graph(s) on that page, I’d love it if you could e-mail it to me. It’d be far smaller than the entire dataset, so internet issues wouldn’t matter, and it would take a lot less work for you. My e-mail is just a gmail account with my name separated by a period.
Maybe go to a Starbucks and download it there the next time you are near civilization? They and most other cafes have good speeds.
I’m not a fan of coffee, but it has been a while since I spent time in a Barnes & Nobles. Maybe I ought to visit one sometime next week.
I think there is a lot more in the Muller report that meets the eye
1) It notes great variability in climate prior to 1950-this directly contradicts the view of the Met office and IPCC who say the climate in the past was much more stable. I have been pointing this out for years and asked the Met office where their claim comes from. The great variability in the climate can be seen on Dartmoor, an upland area not 10 miles from their base where evidence of warmer and cooler times than today have been written about at grreat length and the physical evidence remains to this day.
2) I have also been pointing out for years that one third of the planet is COOLING-this was again confirmed in Mullers report but very little made of it. The Ipcc say that other than South Greenland and a few places in the tropics everywhere is warming. Here is an article on the cooling written by Verity Jones and myself whttp://diggingintheclay.wordpress.com/2010/09/01/in-search-of-cooling-trends/
3) The uhi effect and station move have created entirely different micro climates over the years-we have known about UHI since Roman times-it was NOT first noted in 1833. It is counter intuitive to me that we should take known uhi affected temperature stations and spread that amongst areas that don’t even have stations. I guess that is an artifice of the grid system
4) As regards the warming since 1950, I don’t disagree, but we once again see someone measuring from a trough. The period 1950 -1970 was cold.
http://data.giss.nasa.gov/gistemp/graphs/
5) As Judith Curry notes there is great uncertainty over SSTs.
tonyb
At first glance, two of the papers appear to contradict each other.
Paper 1 says that there is no evidence that global warming has abated since 1998.
Paper 4 fig 1 shows cooling of at least 0.2C since 1998.
I noticed the same issue, but then I looked closer. The figure you refer to from Paper 4 isn’t BEST’s temperature record. It’s a record they generated by using only a fraction of the stations they had (2,000 stations not used by the other groups releasing temperature records). There will obviously be differences between that and the record they get when using all the stations. To get an idea of the difference, look at Figure 8 of their first paper.
That said, I really have no idea why they used a subset of their data in that figure in the fourth paper. I can’t find any explanation in the text, and none come to mind. It seems completely random, so it’s hardly surprising people might not catch it.
Yes, that’s right, I think they wanted to use data that was independent of the other groups (see Roger Pielke’s comment earlier). So as you say it is not a direct contradiction, but still an interesting anomaly.
It is not clear what the source of these 2000 records is, in paper 4. This is something that certainly needs to be fixed in the text.
The paper does call the independence a “virtue,” which makes a certain amount of sense, but it still leaves the question of why only use 2,000 stations? All it seems to do is raise additional issues. For example, the paper says the four lines being compared are broadly consistent, but just look at the 1960s. Ultimately, they could have done the same comparison with 4,000, or even 20,000 stations without changing anything in their paper except for making the line more accurate. Given that, I can’t see why they made the figure they made.
That said, it is certainly interesting to see such large differences between that subset of their data and the full data set. It makes me wonder just how many stations are needed to actually get the “right” results. I’ve heard a number of people, such as Mosher, say relatively few stations are necessary for coming up with the temperature record, but this seems to be pretty clear evidence you need a lot of them.
As a side note, I find the description of how they generated that line both amusing and confusing. I’m curious just what level of precision they were trying to convey:
It may every now ant then be good to think again, why the temperature time series are of interest an what kind of time series provide most useful information. Often it’s useful to consider several complementary time series, which each provide or some additional relevant information to supplement the others.
The temperatures at all points of the Earth surface do not have equal significance. The temperature variations are exceptionally large at some locations at some particular times of year and day (like Arctic or near Arctic winter nights), but all have the same weight for the same surface area in the global average surface temperature. In particular the polar areas and other Arctic areas in Siberia, Canada and Greenland may have exceptionally large temperature variations, but rather little overall importance. That’s just an example on the fact, that the global AST does not necessarily represent best either the most important real trends nor the significance of these trends.
The global AST is also difficult to determine from the existing data. The oceans may be largest problem, but also the wide Arctic areas influence the outcome and are a major factor in the differences between existing time series.
The global temperature is the most natural concept to define, but some more limited average temperatures may actually reflect better the significance of the temperature changes and be easier to estimate. The longer time series are much more reliable and accurate, when the area covered is reduced.
One of the more meaningful variables is the total heat content of the Earth system, but it’s difficult to determine and long time series are impossible to obtain. That raises the question of the best proxies for that. Are the proxies improved, when areal coverage is increased with sparse and problematic data which is given a great weight based on the wide areas for which no better data exists? I don’t know, but I doubt.
The choice is often between data of better accuracy with lesser coverage and of essentially lower quality with wider coverage. When there are good reasons to conclude that, what we want to know is the full coverage, there isn’t much choice, but with the temperature data the full global average isn’t necessarily even the most relevant data series.
The oceans are an exceptionally interesting case from the scientific point of view due to the nature of available data. I can believe that there’s still much to do in extracting useful knowledge from the historical data. Again it’s not likely that any single time series is able to summarize optimally the information, but there may be several alternative approaches with their strong and weak points. That’s an area, where a cooperation of statisticians and substance experts may be very valuable. They should really search for best ways of summarizing the information without preset restrictions on what the time series should represent.
Bart R: you haven’t read a thing about omnology, Donna’s book, my reply to you, or the BEST UHI paper. What’s left? The yellow pages?
Omnologos — My word! You weren’t being intentionally ironic!
Nomology is a real word; “omnology”, just a spoof of Scientology.
And while your list of reading materials may be nearly exhaustive to your mind, one remarks the yellow pages (which from time to time I do read through), the BEST UHI pre-release (which I’ve gone over a couple or three times), your replies and the book in question don’t really add up to much of a foundation for drawing logical conclusions about anything.
How could anyone who values logic or pretends to “omnology” find a one-sided and narrow treatment of anything, much less of an enterprise that is dedicated to casting a wide net of scientific reasoning and presenting the most up-to-date and coherent reports on a broad field as possible, anything but antithetical?
Oh, the hypocrisy.
Judith wrote:
In concluding, I will remind everyone that the REAL problem with the surface temperature data set lies with the ocean data.
Yes and to emphasize and generalize this comment I would say that the REAL problem is with the spatial distributions.
I have always thought that this work was a very minor and largely insignificant contribution to the understanding of the system’s dynamics.
Getting a “new”, more “reliable” way to compute a spatial average of a finite number of data randomly distributed in space?
An interesting statistical exercice but largely irrelevant.
Finding that 1/3 of the points have cooled while 2/3 have warmed, the total giving (obviously) a warming?
L.Motl has found that first already 3 months ago (http://motls.blogspot.com/2011/07/hadcrut3-30-of-stations-recorded.html) even without BEST methodology but he asked the right question : why is there a large number of points which cool ?
Where is a convincing argument that a spatial global temperature average is in ANY way relevant to the dynamics? Where are the spatially localised oscillations at llow frequency inthe global averaging?
As this number (spatially averaged temperature) has no relevance for the dynamics and for the accurate predictions at long time scales then this study is an interesting statistical exercice with no practical interest.
I have never been interested by those hundreds of “studies” that all just take the data and then basically only do averages.
I have always known that this exercice will always give a single number which will be the more or less the same for all methods because they simply must all use the same data.
Well yes, there is certainly an UHI but, again, it is uninteresting to know what is some second order effect on an irrelevant parameter.
Now if they used the data to spatially identify and characterise multi decadal oscillations, for instance by testing the Tsonis hypothesis, then it would be a bit more interesting. Or if they looked in depth on spatial correlations.
If its is only an Nth attempt to say that GlobalT(t2) – GlobalT(t1) = 0,x °C, and pretend to be enthousiastic because we got the sign of an irrelevant number with a 3 sigma confidence, color me unimpressed.
Mr. Milanovic, could not agree more. (your final paragraph made my day!)
Tomas Milanovic
One must admit, the standard of “impresses Tomas” is too strict by many times a metric.
How many times in a century would such a rarefied measure be met?
It would be awe-inspiring were the dataset you seek available, and yet no one had made every effort to assemble it, for the two century span of time BEST covers.
What was done instead, BEST assembled a fit-to-(some)-purpose dataset out of what actually was available.
While it can be (maybe) determined what the minimum number of dynamically significant datapoints for the purposes you describe, I’m dubious even current satellites and methods could produce the beginning of such an information collection.
Why talk about what we cannot have, when if we’re clever we can find some advantage in what we do have?
Besides, you have yet to demonstrate how the considerations you mention are likely to apply on the timescales that matter to AGW.
Thomas Milanovic: As this number (spatially averaged temperature) has no relevance for the dynamics and for the accurate predictions at long time scales then this study is an interesting statistical exercice with no practical interest.
The location of the center of gravity of a passenger aircraft also gives no information about its dynamics, but a sequence of such locations can tell you its destination and time of arrival. It is possible that a sequence of mean earth surface temperatures might tell us what to expect globally in the next few decades. Mean earth surface temperature, and the related climate sensitivity to a doubling of CO2 are not intrinsically meaningless; not any more than the location and fuel of the aircraft.
I believe it has not been adequately demonstrated that the mean temperature record and CO2 sensitivity are as useful as aircraft location and fuel, but the ideas are not intrinsically meaningless.
why is there a large number of points which cool ?
I believe that this question must be investigated. Are there any consistent differences between warming and cooling locations? Can we account for heat transfers? Do the confirmable mechanisms have implications for mechanisms of CO2 effects (in different places or at different altitudes.) If everyone agrees that BEST is the best effort of its kind to date, and especially if AGW advocates call it good work, then BEST will focus attention on that question. It’s intuitively unlikely (for what that’s worth) that the differences between warming locations and cooling locations are either random or irrelevant.
Allmost everybody believes that there has been warming, but what causes it is diffrent issue. Nobody knows for shure about Sun’s affection to climate still IPCC says it’s not the sun, it’s CO2. I believe Spencer is right when he says that climate sensitivity to CO2 is low. I still believe i’ts the Sun after all.
Light’s beginning to dawn. Somewhere on earth.
====================
Yes, Kim, our thinking about the Sun has been “directed” since 1971. Belatedly space scientists realize they were used. – O Manuel
Sorry for the bad use of the arrow brackets.
I just noticed another strange thing about the UHI paper, though this one isn’t as important. Table 1 in the paper says there are 38,898 total sites used, 16,068 of which are “very rural.” Earlier in the paper it gave those values as 39,028 and 16,132.
I can’t find any explanation as to why 130 and 64 sites were excluded from the calculations.
Relevant to this point is the abstract, which explicitly states:
Make yourselves ready! Bishop Hill has the following:-
http://www.bishop-hill.net/blog/2011/10/21/keenans-response-to-the-best-paper.html
“Keenan’s response to the BEST paper”
“Doug Keenan has posted up his correspondence with the Economist and Richard Muller about the BEST paper. I reproduce it here with permission.”
A lot to go through and comprehend
I’m amazed and thunderstruck at the absurdity of some correspondents to try to convey that this study somehow trumps skepticism. This is madness, plain and simple.
Judy, you are going to have to talk to the good Doctor about his loose lips. We’ve got a ship to keep afloat for the voyage.
================
Muller’s found something right because he sneers at finding something wrong.
===============
I probably shouldn’t want to intrude into private grief – but I do!
This graph says it all really.
http://news.bbcimg.co.uk/media/images/56197000/gif/_56197115_climate_change_624gr.gif
TT
Any comments on the one third of stations that are cooling?
You also might like to extend that graph to 1607/8 from which the upwards trend seems to spring.
tonyb
Sounds like situation normal to me.
How many would you expect to be warming? All of them?
Judith Curry has her name to this paper. Why challenge me? Why don’t you ask her about the details?
Those GISS crazies.
TT –
Check out this version of that graph:
http://berkeleyearth.org/video/
(it will take a little while to download)
Devastating.
No wonder Watts has been hyperventilating.
So what have we learnt?
Well our themometers, the same thermometers as before, show that the world has warmed over the past 200 years.
Take a bow Muller & Curry for stating the obvious. Nobel prize in the post.
Now what of attribution?
After some length Muller eventually stated, “How much of the warming is due to humans and what will be the likely effects? We made no independent assessment of that.”
My good lord!
Right around the planet Muller, Curry et al, have given the impression to journalists, to the politcians and to the public that AGW has been cuplable in excessively warming the planet, but by their own belated confession now admit “We made no independent assessment of that.”
Now that is devious.
Is this BEST’s version of Manns-Jones trick? They got caught hiding contrary data from scrutiny, whilst BEST appears to have attempted to hide a very important caveat that somehow got lost in this PR exercise.
This is not science, but maybe that was the intention.
The jury is still out on attribution, but who is now listening to those diliberations?
Muller – FAIL
Curry – FAIL
BEST used something like thirty thousand stations the other groups didn’t use. If you cannot get basic facts straight, it’s hard to expect anyone to take you seriously.
Same data source, not an independent source, so it makes no difference how many stations were used or not – the planet has warmed – which is something we are now being told we didn’t know or didn’t accept before.
Now we are either dealing with a straw-man arguement or revisionism,or both. Either ways it is devious to imply that is due to AGW when Muller was forced to admit, ““How much of the warming is due to humans and what will be the likely effects? We made no independent assessment of that.”
So you’ve changed your story from “same thermometers” to “same source,” and yet, you’re still just making things up. Tell me, just what is this “same source” being used?
I half suspect Muller deceived himself. Reading the ClimateGate emails made him suspect the temperature record and he’s settled his doubts on that score. What perhaps deceived him is an industrial strength belief in the correctness of the IPCC’s understanding of CO2’s radiative effect in climate.
Otherwise, I’m bamboozled, and just as outraged as you are at the straw man/revisionism on display.
Keep an eye on future temps, and wonder what Fig. 1 in Paper 4 means to Muller.
=============
I need to post a partial retraction of what I said above. It turns out one of BEST’s papers does use GHCN entirely for its data. I didn’t realize this before as I had focused on the other papers which used BEST’s new dataset. This doesn’t make Mac right, especially in regards to his tone, but it does make the tone of my comments unjustified. For that, I apologize.
Of course, it’s clear Mac didn’t know what he was talking about, otherwise he would have pointed out my mistake in his first response to me. I was right to think he wasn’t bothering to learn the facts; I was just wrong to think those facts were so simple.
My tone is bang on.
Did you believe that the world had warmed prior to BEST?
Do you still believe that the world has warmed after BEST reported?
Did you doubt?
Now here is the Economist headline;
“The heat is on – A new analysis of the temperature record leaves little room for the doubters. The world is warming”
Who were these doubters????????????????
BEST did not feature two very important caveats;
1. This was not a truly independent verification that the world has warmed – it used the same data and came up with the same answer.
2. This was not a test of attribution.
However from reading blurb in the mediayou would have thought that BEST was saying the opposite. We are left in no ‘doubt’ that humanity is to blame for the current warming.
Now that is devious.
Surely the point that Keenan is trying to make, over at Bishop Hill, is that the linear models are expecting TSI to have a linear trend so the satellite data is adjusted to fit that trend.
In reality the Sun is cyclical so it’s effects must be cyclical, which differs from what the models are expecting because they require a linear trend.
The data stipulates cyclical the models stiputate linear so what happens, the data is adjusted to fit the model requirements instead of the the models are adjusted to fit the cyclical data.
A quick search gives some detail:
http://icecap.us/images/uploads/Lockwood_and_Frolich_Review.pdf
Or am I missing something?
Mac –
There has been much criticism among “skeptics” of previous assessment made WRT recent warming on the basis of failure to account for the potentially corrupting influence of an UHI effect, or the statistical methodology used by the “climate community” to assess trends in land temperatures. These criticisms asserted that for those reasons, the existing analyses of the recorded data did not validly show recent warming, let alone the degree of recent warming. Hence, the criticisms were rooted in a view that questioned recent warming – and on that basis justified “skepticism.”
From reading comments from “skeptics” at various websites, I have seen exactly that basis for skepticism expressed many, many times.
The BEST analysis controlled for the UHI effect, and used a different statistical methodology that addressed (at least some of) what some critics felt were methodological flaws in previous analyses.
Assuming that you accept the BEST methodology as valid – and you were a “skeptic” for the reasons I described above, then you would, theoretically, no longer be skeptical. If you were a “skeptic” who doubted AGW, not because of questions about the “climate community’s” analyses showing warming, or about the “climate community’s” assessments of the degree of recent warming, but simply on a questioning of whether the recent warming is attributable to anthropogenic CO2 emissions – then obviously, BEST’s work would not change your “skepticism.”
I’m a little curious how you expect someone who sneers at finding something wrong, to find something wrong.
===============
“why is there a large number of points which cool ?”
Simply because perceived climate change is a consequence of a redistribution of the climate zones as they jostle about in the process of speeding up or slowing down the rate of energy loss to space in a powerfully negative system response to ANY forcing of temperature from ANY internal or external source.
With that process going on some places warm up and other places cool down as they change their positions relative to the nearest climate zones. More or less cloudiness, more or less sunshine, more or less wind from oceans, more or less wind from land masses etc. etc.
If the globe is trying to accumulate more energy for whatever reason the climate zones realign to get rid of it primarily by altering the speed and/or volume of the water cycle. There might be a net globally averaged overall surface warming for locations in the troposphere but not necessarily any significant warming (or total energy content) for the system as a whole which must include the oceans from surface to depths.
The data shown just provides evidence for a slight long term surface warming trend in the regions (mostly northern hemisphere, mid latitude continental) where any changes would be most apparent from surface observation recorded by thermometers. That trend clearly going right back to the LIA perhaps supplemented a bit in recent years by the UHI effect.
It does not follow trhat there is necessarily any significant increase in total system energy content.
I could imagine that global warming could melt polar ice, thus “disappearing” in the latent heat of melting, increased cold water would arive in the temperate regions, and so lower temperatures in the local climate.
Whether this is true or not, I’ve no idea, but I imagine that the relationship between a global positive radiation balance and regional climate is going to be complex.
Just to clarify then – assuming that you are a “skeptic,” that means that you doubt whether the world is warming.
Correct? Because I’ve been told that “skeptics” don’t doubt that the world is warming.
I don’t doubt that some portions of the surface have warmed but there is no evidence that total system energy content has increased significantly taking into account the entire body of the oceans.
From the point of the entire system we could just be seeing faster energy flow from negative system responses such as the faster or larger water cycle through a small portion of the system offsetting the slower flow from more GHGs.
Just to clarify then – assuming that you are a “skeptic,” that means that you doubt whether the world is warming.
Depends on your definitions.Your terms are too imprecise to allow for a definitive response.
A tiny amount of energy is retained in the system for a fraction longer by GHGs but the amount due to humans is infinitesimal as against naturally varying CO2 flows plus solar and oceanic effects.
Any warming there has been is therefore more likely sun and ocean induced than a consequence of human emissions.
And warming of one part of the system is not necessarily significant if it fails to measurably warm up the system as a whole due to negative system responses.
Lots of misconceptions there.
GHG will contribute immediately as the energy is transferred quickly via statistical physics of gas particles.
Bogus. Uptake and release are slow diffusional processes when it comes to excess concentrations. CO2 diffusing to sequestering sites and excess thermal energy diffusing show similar random walk dynamics, the first random step is what fools everyone. Do you actually study or have any experience with the laws of mass balance and conservation?
The world is warming AND cooling, depending on time scale. Temperatures are used as the measure of this warming/cooling. Change in temperatures is not necessarily proportional to the change in energy content of the system.
Edim –
I’m confused. Just recently you told me that “Anthony was right” when he said that skeptics don’t question whether the “world is warming.”
Now here you indicate that you believe that the world is “warming AND cooling.”
I don’t think that I’m being overly semantical here by seeing a conflict in your most recent statement and your previous statement.
Anthony’s statement clearly meant that “skeptics” don’t question whether or not in balance -small, or regional variations notwithstanding – there is a mathematically valid trend of warming. That would imply that “skeptics” would not answer a question of “Is the world warming,” with an answer of “It is warming AND cooling.” But an answer of “Yes, the world is warming.”
Joshua,
An important problem with the skeptikal position (with two K’s, pretty please with some sugar on it) is that we rarely take the time to identify what one is skeptikal about. For you see, there are an indefinite amount of subject matter about which one can be skeptikal. Here are some examples:
One can be skeptikal about the very existence of a temperature average.
One can be skeptikal about the warming of surface-land in recent centuries.
One can be skeptikal about the quality of the land-surface temperature data we have.
One can be skeptikal about the importance of such questions when compared to the Very Real Green Menace.
There are lots of positions about which one can be skeptikal.
* * *
So we now know that Anthony Watts never doubted that there was a surface-land warming. That is, we now know that Anthony Watts position is compatible with the proposition according to which there MAY have been a warming of 0.0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000001
Celcius during the recent century or so.
Anthony Watt’s skeptikism is about HOW MUCH warming there really were.
* * *
What we do not know from Anthony’s last post, is where he stands regarding his previous skeptikism regarding the quality of land-surface measurement data we have.
Anthony Watts does seem skeptikal about the priority of such questions when compared to the Very Bad PR Strategies BEST is showing.
* * *
What can be noted is that all the skeptikal position are quite independent from each other. Not only are there indefinite skeptikal positions, there are indefinite combinations of skeptikal positions. Some Mainstream Mathematicians believe there are more combinations than there are positions, but they’re all Krooks bought by the Realist Church.
That these positions are all independent from each other is quite obvious from the fact that we do observe an indefinite combination of skeptikism in blogland. IF, and that is a big IF, there are logical impediments prohibiting for a skeptikal agent to entertain two skeptikal positions, there is nothing that prevents to share the skeptikal work among many skeptiks.
I hope this MAY BE of help to understand your problem,
w
> The world is warming AND cooling, depending on time scale.
A lukewarm answer.
Sure – the world is warming AND cooling in the sense that on various time scales, at various places on the globe, even as I type, temperatures are rising and falling.
The statement that the world is warming and cooling depending on time scale, is essentially trivial.
The non-trivial questions at hand are: what has been happening, in balance, over the past 50 years, how does that differ from previous trends (by sign and degree), what are the variables that might explain any differences that might exist, and is anthropogenic CO2 the most likely (or perhaps the only possible) explanatory variable for differences that might exist?
Anthony’s statement is that “skeptics” don’t question whether the “world is warming.” That would clearly imply that “skeptics” don’t doubt that the recent trend, in balance, is one of warming.
If your answer to that simple question is in any way ambiguous, and you are a “skeptic,” then Anthony’s statement is wrong with respect to your opinion.
Anthony did not say that “skeptics” don’t doubt that the world is warming and cooling.
If your answer to that simple question is that you don’t think that the methodology of the analyses that have been conducted are valid, then there again, if your are a “skeptic,” then Anthony’s statement is wrong with respect to your opinion (as you would have no basis for belief, one way or the other, as to whether the “world is warming,” although you could point to evidence of past climate change to say that it is likely that the world is warming AND cooling.
Look – what “skeptics” as a group do or don’t believe isn’t really that important with respect to the science. And Anthony’s opinion is just one person’s opinion, and as such it is largely irrelevant. But in terms of the climate debate, it is relevant if some significant number of “skeptics” regularly mischaracterize other “skeptics.”
The conflict between Anthony’s statement and what I see written in comments from “skeptics” all over the blogosphere becomes significant because I see that same inconsistency all the time.
Time after time I am told that “skeptics” don’t question whether the world is warming – they only question the degree and the attribution. Our very gracious host has, herself, endorsed that characterization.
Polls show that a very significant % of Republicans and Tea Partiers – who presumably could be referred to as “skeptics,” don’t think that the world is warming.
What do you suppose is the explanation for such a widespread and repeated mischaracterization of the viewpoints of “skeptics?”
Joshua,
I’ve replied to this in the wrong sub-sub-sub-sub-thread:
http://judithcurry.com/2011/10/20/berkeley-surface-temperatures-released/#comment-125465
Thanks, willard –
For the benefit of Richard Saumarez, I’ll respond later. I wouldn’t want the poor fella to be forced to endure another of my posts without a break.
It’s always been about the attribution. Is that just now dawning on them?
Frankly unbelievable. So why the perception that they’ve been so clueless?
It’s like Muller is writing a review of the book ‘Climate Skepticism’, but fails to buy, download, or read the book.
He’s documented that the typescript is the same as previous editions, but otherwise leaves the book unexamined.
The work is not a waste of time and money, but the jumped up spin, the pumped up kick, is lethally self-destructive.
============
Attribution matters only if there’s something to be attributed.
Okay here are two questions.
How many people believed that the world has warmed over the past 200 years before BEST published their results?
How many people now believe now that the world has warmed over the past 200 years after BEST published their results?
If your answer is Yes-Yes take a bow.
If your answer is No-Yes you won’t be able to take a bow because you are made of straw.
That is the basis of the BEST project.
How many people believe that it isn’t possible to answer Yes or No to a question beginning by “How many…?” ?
(Yes)
How many no’s does it take before you realize I mean yes?
==============
In answer to your question: maybe
That lacks nuance, Joshua. Try this:
IMO, from my standpoint, perhaps we could envision the possibility that maybe.
Perhaps I stand corrected.
I don’t know, but the answer is caused by global warming
And we’re talking about papers that are not even peer reviewed yet – so applying “the team” standard for truth – I can safely claim that the conclusions of these four papers are false – without even reading them. And obviously “the team” doesn’t have any pull at WSJ or else the op-ed would not have been published at all.
I’ve been looking at the GHCN data from a different perspective, some might say from a rather limited signal processing approach (see post), so I wasn’t looking at the data from a global trend standpoint. Yesterday, I did a quality control exercise. By selecting 15 stations at random, 4 had series that lasted nearly a century, but contained minimal data, 3 had unexplained “jumps” in the data of well over 2oC ,two showed negative trends and the rest showed positive trends in region of 1.2 C/century. Were one asked to make a judgement on data of this quality in process control or medicine, one would might question the quality of the data.
The primary data appears to me to be somewhat uncertain. I am well aware that statistics is the art of making judgements in the light of uncertainty, but I do worry about the quality of some the primary data.
Having said that I am in no doubt that the “global mean temperature” is rising and has been doing so for some time.
I think the historical record is in interesting as history. I don’t imagine that anyone with any sense is going to dispute the broad brush findings of these studies. What appears to me to be far more important is that the climate community is starting to gather far higher quality data with satellites that can address the problems of radiation balance, clouds, regional atmospheric temperature profiles, humidity. Given detailed geographic data from satellites and, I presume, increasing resolution of satellite temperature measurement, I guess that in 10-20 years this will give a much clearer picture of what the climate is doing than historical instrumental temperatures.
One of the most important parts of BEST’s work, to me at least, was it was going to include a large amount of data GHCN didn’t include. After all, a 400% increase in data is great one is looking for a signal in “noisy” data.
Unfortunately, they apparently aren’t publishing the results they got with that new data just yet. Or rather, they’re publishing partial results, but not the full ones.
Richard, I would like to sample some stations data as you did, how do I get it?
I have looked at WolframAlpha’s historical weather data for specific cities and stations and it often jumps and often has data gaps too.
http://climate.usurf.usu.edu/products/data.php?tab=coop
This is one place to start.
I have another issue with the UHI paper. It’s second section begins by saying:
However, the paper they cite says this:
It seems the citation is bogus. As far as I can tell, there is nothing in the paper they cite which discusses the formation of their dataset. At best, all the paper they cite does is make vague references to to the material it is supposed to be providing.
Hi Judy
I wrote this comment on Watts Up With That in response to comments from Zeke Hausfather-
“Zeke Hausfather – I realize that he has a much larger set of locations, but many of them are very short term in duration (as I read from his write up). Moreover, if they are in nearly the same geographic location as the GHCN sites, they are not providing much independent new information.
What I would like to see is a map with the GHCN sites and with the added sites from Muller et al. The years of record for his added sites should be given. Also, were the GHCN sites excluded when they did their trend assessment? If not, and the results are weighted by the years of record, this would bias the results towards the GHCN trends.
The evaluation of the degree of indepenence of the Muller et al sites from the GHCN needs to be quantified.
Perhaps they have done these evaluations. However, from my reading of their work, I have not yet seen it.”
I have made a KMZ file of the 36700 odd BEST sites. This is a Google Earth file – if you click on it’s name, it will bring up GE with all the sites marked. There are flags with info, got by clicking on the pushpin markers.
It is accessible from here.
.When I get further into the data, I’ll make one with the years of data indicated, with folders and colors, and also the GHCN stations. I could do GHCN now, but it isn’t very meaningful without the duration of data info.
Nick did you find 36K in the best file?
Steven, I found 36736 stations in the inventory, and I’m now in the data file data.txt. 35121 stations there have at least one month with more than 10 days data.
Steven,
I found the 10 days in a month test is no good, because the number of days is unavailable for a lot of stations. Without it, in the latest data, almost all stations have at least some data, 39026 out of 39056. I’ve updated the KMZ file.
cool
I’ve put up a new KMZ file which has more of what Dr Pielke is looking for. It has the years of data info, and the four data sets BEST, GHCN, GSOD and CRUTEM3. The markers are colored, sized and foldered. The folders let you look at many different subsets.
I think that one has to be very careful about the implications of the temperature history of Earth, when thinking about this data.
The error limits on the data in say 1850, are considerably wider than they are today. If one were ask, what was the change in mean surface temperature between 1830 and 1880, and is this comparable to today, could one answer this question? I think not, because there isn’t nearly enough data to enable one to answer it.
I would suspect that the temperature record in this study is interpretable in mean global terms for about a century and probably only “hard” for a shorter period. Therefore I would be reluctant to make arguments about the implications of the recent temperature excursion, which I am sure is rea,l in terms of whether it is unusual in terms of a longer time frame.
This is not meant as a criticism of these studies which are designed to synthesise conclusions from the existing data as accurately as possible.
Richard, what recent “excursion” are you talking about? Do you mean the 1978-97 warming? The satellites do not show it and they are instruments, not arcane statistical models. I doubt it is real.
http://www.mediafire.com/file/a9tv9tad9e6216p/UAH_2011_06_19_two_regressions.pdf
See my post at 8.52 re data quality and satellites.
I am prepared to believe that there is warming in the instrumental temperature record. How this relate to the overall land temperature, I am less certain.
I am shocked by your statement (well actually not all!). In my last post, I got dumped on by a couple of trolls who disrupted the entire thread, so I am being careful not to make deliberate troll-baiting statements. Not that I think their opinions are particularly interesting or correct. If you haven’t discovered it, you should go to “The Idiot Tracker” home website of “Robert” and have a look at the psychopathology on display. From a purely psychiatric point of view, one should avoid becoming a victim of these individuals’ paranoia.
What shocked you? I think the satellite record falsifies the surface statistical models, that is all. These Jones type area averaging models are mathematically strange to begin with, so I fail to see the shock value in my statement.
I was being purely ironic.
I should add that I think that the divergence of satellite data with the thermometer data is of great importance. I am certainly not a standard bearer for the somewhat questionable approaches of dubious interpolation based on correlation, the methods of data reduction used (see my thread on Aliasing) and I am somewhat sceptical about Kriging, having used it on some data sets.
As I said in a previous post – I believe satellite data will be vastly more important and physically important in determining major issues such as overall radiation balance.
Thanks Richard. As a Spock-like logician I have no ear for irony. It often escapes me.
The hero and mineral exploration veteran McIntyre will come to the rescue and teach everyone about kriging. It is widely used in that field because mineral core samples are very expensive measurements and they have to get by with strategically placed samples. Of course there are well known predictive disasters but this is a lot different than the gampling involved in prospecting. In the kind of kriging used for temperature estimation, they are not betting the farm on where to place a mine but more like performing optimal extrapolation for lots of missing spatial data points. Those will regress to the mean as the law of large numbers takes effect.
Richard,
Thank you for taking the time to comment here.
I have two questions for you from your experience in measuring sensitive biological systems and applying sophisticated statistical tools:
1-How signficant is 0.45C* over >20 year period of time in a dynamic system that has daily fluctuations often in the 10 X degree range, and inter-day changes that are frequently multiples of a degree range?
*granting a recent claim made on this blog as fully correct.
2- The second question may be out of an area you wish to consider, but I think your opinion would be valuable:
How meaningful and useful is the concept of ‘global temperature’ or ‘global climate’?
It is not uncommonn for there to exist simulataneous areas of Earth experiencing record heat and cold. It seems that regions of the Earth is where people and life exist, not globally per se. I may not be expressing this clearly, and if not please advise.
Regards,
David,
Do you really think there’s a big difference between satellite TLT trends and those from surface records? It seems to me they are very similar and any differences between the featured four do not seem dependent on a satellite/surface divide.
Thermometers are instruments too. Producing a global temperature map from a series of moving measurement platforms is a complex process, particularly so for the lower troposphere because you have to look through the rest of the atmosphere to get at it. This is a process that will necessarily involve statistical models, whether they are arcane or not.
Paul, the surface models show steady warming from 1978 to 1997, while the UAH shows none. That is a big difference in my book. Especially since this supposed 20 year warming spurt is the only warming in the last 70 years in the surface models.
The issues with the satellite readings pale to insignificance compared to the issues with the surface models, beginning with the fact that this is a convenience sample. It is fundamental to statistical sampling theory that you can draw no firm conclusions fro a convenience sample.
Paul, the surface models show steady warming from 1978 to 1997, while the UAH shows none.
But RSS does show warming, about as much as the surface records. As I said, there are short term differences between them all but no strong satellite/surface divide.
Especially since this supposed 20 year warming spurt is the only warming in the last 70 years in the surface models.
Huh? How about the warming from 1984-2003, from 1989-2008, if we stick to 20 year periods.
The issues with the satellite readings pale to insignificance compared to the issues with the surface models
I don’t think so. There are potential issues with both, but they seem to have been largely overcome. If you think the surface record is more problematic why do you think it was that the UAH record was reporting erroneous trends for several years?
I have no use for RSS. It was cooked up by people who did not like the UAH readings. And HadCRU shows no warming trend from 1984-2003, nor from 1989-2008. The only warming period it shows in the last 70 years is roughly 1978-1998, where UAH shows no warming at all.
Of course you could try to eliminate the contradiction by saying both UAH and HadCRU are no good. That is basically what the NAS/NRC did. But then it means we have no idea if it is warming or not.
I have no use for RSS. It was cooked up by people who did not like the UAH readings.
The UAH record was incontrovertibly wrong at the time RSS came along with their TLT record. Christy and Spencer have admitted as much. My question is how could they have made such a mistake if satellite measurements are so plain and simple compared to the arcane ‘surface models’.
HAdCRUT3v trends from 1984-2003 and 1989-2008: http://www.woodfortrees.org/plot/hadcrut3vgl/from:1984/to:2009/plot/hadcrut3vgl/from:1984/to:2004/trend/plot/hadcrut3vgl/from:1989/to:2009/trend. There seems to plenty of warming to go around there.
David Wojick writes: HadCRU shows no warming trend from 1984-2003, nor from 1989-2008.
HADCRU shows a warming of approximately 0.45C from 1984-2003 (by OLS). This is a slightly larger warming than GISSTEMP has for the same time period.
It shows a warming of about 0.36C from 1989-2008. This is just slightly less warming than GISSTEMP shows for the same years.
Whoops, it looks like Paul S beat me to it there. Sorry for the redundancy.
david do you accept the PHYSICS models used to create the satellite temperature products
The satellites do not show it and they are instruments, not arcane statistical models.
My pool was registering 71 °F this morning, and 73 °F later in the day. An arcane statistical model claimed that its average temperature over that period was 72 °F. Next time I’ll use instruments, not arcane statistical models.
I love the logic skeptics use.
Better yet, they dont realize that satellite temperature products RELY on radiative transfer equations
Synthesizing bad data as accurately as possible is an oxymoron.
David –
Do you think that the “world is warming?”
Joshua, the satellite measurements indicate a step function warming coincident with the 1998-2001 ENSO cycle. See http://www.mediafire.com/file/a9tv9tad9e6216p/UAH_2011_06_19_two_regressions.pdf That single event of warming is probably real, so in this limited sense the world is indeed warming a bit.
Prior to 1978 we do not know, but we know that the supposed warming from 1978-97 did not in fact occur. This is the supposed warming that AGW is now based on. So I neither think the world is warming over the long run nor do I think it is not, but I do think AGW has been falsified by the satellite data.
I also believe that we have an entire scientific community trying to explain a statistical fantasy, in order to support a political movement. That is the real problem, not whether or not it is warming a bit.
David –
So – as I understand it, you trust the satellite data as valid, but you do not trust surface temperature data – and thus, you would question any assessment that the “world is warming.”
I commend you for having an opinion that remains consistent. However, your opinion would suggest that it is not only the “scientific community trying to explain a statistical fantasy in order to support a political movement” that is mistaken in their opinions about global temperatures.
It would also be Anthony, and the many “skeptics” who say that there is no doubt that the world is warming, who you also think are trying to explain a statistical fantasy (although presumably you’d have to find another explanation for their fantasy – as their political orientation would, no doubt, be inconsistent with your thinking).
Obviously, those other folks are, as well, either misreading the satellite data, or are trusting the surface temperature data which, no doubt, you consider to be invalid.
Is my understanding correct?
Yes Joshua, you have it about right. What you are calling the surface temperature data is in fact the output of a complex computer model. It is not data at all, in the sense that data usually refers to measurements.
I spent several years studying these models, in the context of statistical sampling theory, and I came to the conclusion that they are inaccurate at best. Far worse than the climate models. This is a math issue, not a physics issue. Most people do not even realize that these models exist. They think these so-called global average temperatures are statistical sampling averages, but they are far from that.
However, I also participate in those debates that assume the surface models are correct, just to keep my hand in the big game. My sole goal is to stop the greens, so I have multiple lines of defense.
Going down this road, before the first prediction, the GCM model-makers already have become, “convinced disciples of the mathematical reductionism.” They also wish to be counted among the converted who eschew the concept that an average global temperature is meaningless because temperature is an intensive variable.
Moreover, the AGW model-makers’ GCMs fail grandly compared to reality on either a zonal or seasonal scale. That is why the AGW model-makers indulge themselves in the fiction that their GCMs will be reliable when predicting long range climate instead of short range climate–which is nothing more than a preconception–i.e., a simple-minded belief that has never been validated, nor can it be becasue the AGW null hypothesis has never been rejected nor can it be.
This is probably a waste of time, considering Judy routinely ignores this question but I shall try once more. Given that the earth is below average temperature and below average co2 levels in the atmosphere, how can you attribute a warming of .8 kelvin to mankind? It just doesn’t make sense. It is very similar to Lucia claiming global warming is melting ice at the North Pole, without posting the temperatures. It is no wonder you 2 do most of your work behind the computer, there is a serious lack of critical thought.
ah your over here. the nurse was wondering where you had run off to. Time for your meds.
Jay, you need to unpack you argument for it to make sense, speaking of making sense. What average temperatures and CO2 levels are you referring to? One million years, one hundred million, or one billion?
I do not buy AGW but it certainly makes sense as an hypothesis. If you accept that it has warmed over the last 30, or 50, or 100, or 150 years, and that the CO2 levels have risen over the same period then it is a straightforward hypothesis that the latter caused the former. It may not be true but it is scientifically sensible. In short this is a genuine scientific debate.
Finally, OK, are daily TMax and TMin available for download? I have done that for a few places, specifically Canada. Summer TMax in Canada has been falling. In spite of all the hype from AGW faithful, summer heat wave days has been falling here. What has warmed, and is driving up the average, is winter temps have been getting milder. Maybe some of the AGW faithful here can explain how this is bad.
yes. depnds on the data source. Some sources have tmin/tmax some only tave. some daily. some hourly. some only monthly
***** Message to Joshua *****
Joshua.
Could I make a suggestion? You make a long and detailed post outlining your position. Having done that we can all read it. Some of us may agree with you and others may not. That is the way of the world.
After doing that could you then stop introducing flatulent and totally boring posts arguing whether people are deniers, sceptcs or true believers?
If you were to do that , the quality of the discussion would rise immensely and be considerably more interesting.
**** Message to Richard ****
Thanks for reading.
There are some folks in these here parts that have a program you can download to enable you to easily discard the comments by specific people at this blog.
If you don’t like my comments, then I suggest one of two approaches:
(1) ask for the download – I don’t doubt that someone will provide it.
(2) in lieu of that approach – don’t read my comments – or the myriad comments by those who (even though they may claim otherwise) think that my comments are interesting enough for them to respond.
(3) endure the pain of reading my comments (honestly, I don’t force you to do so) and simply not respond.
I would suggest to you those approaches, also, are the way of the world.
And Richard:
I’ll point you to this comment that appeared immediately below mine in response to you:
The questions I address WRT the beliefs about “skeptics” are relevant to a great number of the posts you might read at Climate Etc. – both in the sense that my posts discuss the reasoning they contain, and in the sense that even more specifically, many other posts contain opinions on the topic of the beliefs of “skeptics” and the beliefs of those on the other side of the climate debate.
Perhaps you should take up your issues with them? If they didn’t write what they write, then I wouldn’t write what I write discussing what they write.
Whew!
***** Message to Richard *****
***** And to Richard Only *****
***** No, not you, Joshua! *****
***** Everybody else, stay away! *****
***** Pretty Please *****
***** with Some *****
***** Sugar on it *****
http://www.greasespot.net/
You didn’t ask for my advice but I’ll give it anyway. Do not punch the tar baby!
Has anyone mentioned Keenan’s post on Bishop Hill?
http://www.bishop-hill.net/blog/2011/10/21/keenans-response-to-the-best-paper.html
I have great respect for Anthony Watts, and put up a short post of appreciation on his blog this morning. But I don’t think I’m speaking out of both sides of my mouth when I say that I don’t believe he’s not being smart in this case. The NYT’s made him look foolish today in their short piece (I expect something much larger there soon as they see this as a great victory)…I actually cringed in sympathy for him.
Of course the NYT’s is going to do everything they can to disparage skeptics, that’s not new. But in this case Mr. Watts is handing them the ammunition by giving the impression, or at least by giving them the chances to create the impression, of intellectual intransigence (otherwise known as denialism). Just as bad, he continues to allow them to make the case that skeptics do not accept the warming trend over the last few hundred years, which of course is laughable.
He’s just one guy doing the best he can, and as I said on his blog I think one day he’ll be recognized as a hero in this fight to expose perhaps the greatest science fraud in history, but there are times when I think he would benefit from some expert PR advice. Of course such advice would probably have saved him from getting sucked into this mess in the first place.
“According to the UN’s Intergovernmental Panel on Climate Change, the greenhouse forcing from man made greenhouse gases is already about 86% of what one expects from a doubling of CO2 (with about half coming from methane, nitrous oxide, freons and ozone), and alarming predictions depend on models for which the sensitivity to a doubling for CO2 is greater than 2C which implies that we should already have seen much more warming than we have seen thus far, even if all the warming we have seen so far were due to man. This contradiction is rendered more acute by the fact that there has been no statistically significant net global warming for the last fourteen years.” ~Richard Lindzen
Sorry, above should read : “I don’t believe he’s being smart.”
At the moment accurate temperatures using satellites was possible, amazingly, global warming stopped.
At the moment accurate temperatures using satellites was possible, amazingly, global warming stopped.
Both satellite temperature records show warming, as long as you don’t cherry-pick start and end dates.
J, so you admit that the alarmist just cherry pick the start and end dates?
Er, what? Who’s the “alarmist” cherry-picking start and end dates? And where exactly do I “admit” anything?
Your comment makes no sense.
Q1. Is BEST doing independent science regarding global warming by using GHCN data?
A1. No!
Q2. Does BEST prove attribution?
A2. No!
Where the data reflects an increase in the growth of atmospheric CO2 levels, we see decreases in global temperatures, as follows:
“… since 1940 and from 1940 until 1980, even the surface record shows cooling. The argument is that there has been warming since then but, in fact, almost all of that is due to what is called the ‘urban heat island’ effect — that is, that the weather stations are around the edge of cities and the cities expanded out and distorted the record. When you look at rural stations — if you look at the Antarctic, for example – the South Pole shows cooling since 1957 and the satellite data which has been up since 1978 shows a slight cooling trend as well.” –Dr. Tim Ball (Nov. 2004)
Let’s get real. Scientists in China, Japan, Russia — they’re all getting a good chuckle at the boy-yoy-yoing! science found in secular, socialist Western academia.
Jim Cripwell raised the issue of the UHI distortion to the surface temperature record.
The BEST UHI study group reported:
http://www.berkeleyearth.org/Resources/Berkeley_Earth_UHI
But, as Judith stated, there is more work to be done on this topic.
But let’s take a quick look at the data.
Since January 1979, the GISS surface record shows a linear warming trend of +0.17°C per decade, while the UAH tropospheric record shows a warming rate of +0.14°C per decade over the same period.
The troposphere is supposed to warm faster than the surface, according to GH theory, but the record shows that it appears to have warmed more slowly.
The accuracy of the ocean record prior to around 2000 has been questioned (see Judith’s comment above).
If one ignores these issues for simplification, one could postulate that the difference between the two records is due to distortions in the surface record, which can be attributed to the UHI effect, poor station coverage, loss of stations, etc. (all issues, which the satellite record does not have).
This would represent +0.03°C per decade (a bit higher than the BEST estimate of +0.019°C per decade) or 0.3°C out of the 0.7°C warming reported over the 20th century.
So even at only two-thirds this value, the UHI effect is not insignificant over a longer time period, as it represents 2/7 or 29% of the observed long-term warming.
Max
PS Just see that BEST found that urbanization results in cooling instead of warming.
Ouch!
Let’s do some rework here, folks!
There are so many effects involved that one shouldn’t really be surprised on the nominal sign of the signal, which is anyway not statistically significant. The question is not, whether there are UHI effects, the issue is the net influence of all changes that have affected the time series. The presence of spurious effects in both direction has been known for years. Similar conclusions have been presented also previously (I cannot give references).
Various tests give fair support to the conclusion that the changes that affect individual time series add up to a very small bias. There may be more systematic bias in the earliest data. It may be very difficult to estimate the size of such potential bias.
Pekka
Would you agree that this is only a step in the overall process? The overall process could be described as:
1st trying to get a better estimation of how much the earth is actually warming,
2nd- then trying to model what that temperature increase will do relative to general atmospheric circulation (what the impact will be on precipitation in specific regions…GCMs or regional CM’s), and then:
3rd- trying to determine what the impact the changes in temperature and rainfall will have upon humanity in the different regions.
yes.
For some time I’ve argued that the UHI bias, while large in some isolated cases, is actually rather smallish when you consider a wider range of circumstances. That small effect size makes categorization of rural and urban critical. If urban warms slightly more than rural, then your categorization better not have any cross contamination: a urban mis classified as a rural, will have a double whammy on your power. A rural classified as urban will likewise have a double whammy effect.. IF the effect is small. If the effect is large then miscategorization isnt that much of an issue
“For some time I’ve argued that the UHI bias, while large in some isolated cases, is actually rather smallish when you consider a wider range of circumstances. ”
If it’s raining I doubt there is any UHI.
Also it’s cloudy. I doubt it affects very much average temperature of Seattle or any place in Washington state.
Also if it’s windy and/or it’s area that receives little sunlight- or most land areas on earth.
UHI should most affect in Southern parts of US or during summer at higher latitudes.
Though location temperature reading site and environment they in are different aspect issue than UHI. Having temperature reading near pavement or near an air conditioner exhaust will increase the temperature the “micro environment”.
The misclassification problem is relevant for all large-scale urbanisation studies I’ve seen. An alternative way of looking at it is to find the lowest and highest trending subset of the data while still maintaining a regular coverage and it looks like the BEST analysis can provide some measure of that. One can they say if urbanisation (or greenification) affects X% of stations then this is the largest that effect can be.
Pekka
Gimme a break!
Surely you are not going to try to rationalize that, despite the many studies that have shown the opposite in real-life examples, the BEST sub-group finding that URBANIZATION HAS CAUSED A SLIGHT COOLING EFFECT ON GLOBAL TEMPERATURES has any validity whatsoever – are you?
Max
I very much doubt these “many studies that have shown the opposite in real-life examples” were looking at exactly the same thing as the BEST study was. And exactly matters at this small level of significance.
AK
Are you trying to rationalize that urbanization has caused cooling?
Please explain.
Max
@manacker…
First let’s be precise here: we’re not talking about cooling, we’re talking about slightly smaller levels of warming. If you want to subtract out the overall “global warming” trend for the total dataset, the remaining trend for the “very-rural” dataset is actually slightly warmer. This leads to a possible “cooling” Urban Heat Island effect.
Now, one possibility is that urban areas aren’t as much warmer today than they were in the 1950’s, and there are many possible explanations for that. Alternatively, there are all sorts of other possible reasons for this effect. But note: in terms of “Global Warming” the UHI is and always has been an artifact of the measurement system. The claimed effect is thought to be caused by the effect of urban heat islands on thermometers, causing a false signal of “Global Warming”.
Max,
that must be more of that Goreball Warming causes Goreball Cooling science!!!
That a spurious warming signal from UHI might be “statistically insignificant” is a point that could be debated.
The comparison of the satellite and surface rate of warming indicates a difference of 0.03C per decade, and one could also say that this difference is “statistically insignificant”.
But saying that urbanization causes cooling is downright nonsense, whether the number is “statistically insignificant” or not.
That was my point, Pekka.
Max
Max,
since they have no statistically significant group of stations that are proveably unaffected by UHI the results are absolutely useless to determine the amount of UHI affecting the record. The theory used to be that Urban areas had UHI and Rural had little to none so the differencing was thought to have some usefulness. Dr. Spencer’s work and now this statistical video game show that we didn’t know much then and still don’t.
Without serious baselining work where stations are set up to compare PRISTINE areas with Rural and Urban this is one more issue that makes the GMT pretty much useless no matter whose it is.
I have a possible solution. Let’s remove a few large cities and rural developments and return the areas to pristine condition and see how the temps develop!! (I get to pick)
It would be impossible to find any stations that would be proveably “unaffected by UHI.” Of course, you must realize that your statement undermines the “skeptical” argument that UHI distorts findings of trends in warming (which is rests on the distinction between “good” stations and “bad” stations). You’re asking to prove that any given line can’t be bisected.
Joshua,
“You’re asking to prove that any given line can’t be bisected.”
Uhh, don’t really know what to make of this response. I am saying that they HAVEN’T identifed amount of the various contributions to the warming or lack thereof or even cooling. As I posted elsewhere, we have several contributors to the possible warming of any station. As all of these can vary what would YOU do to show us how each item, Solar, GHG, UHI, Cloud… contributed to the warming or lack?
Muller found that a subset of stations warmed at ~.2c/C faster than the other stations. This is an average of all stations so does NOT show the individual variance which is probably larger in some cases plus and minus. The problem is that none of us have no way of know how much of the actual signal for each station is UHI so this number is meaningless whether for the station set or for individual pairs. We do not know if one station is cloudier than another. We do not know if one station is windier or higher humidity or more precipitation or has frequent inversions…….
I am beginning to wonder if this is a fault in the thinking process. Socialists think in terms of the collective and discount individual contribution!!
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
Joshua,
“If you don’t think the assessments that show warming are valid, how can you not question whether the world is warming?”
Straw Man strikes again. It is PERFECTLY possible to assert the following:
1. The temperature datasets (all of them) especially prior to 1979, are all suspect to one degree or another for several reasons (I can go into these if you like).
2. The way the data have been manipulated and homogenized has been shown to be suspect in many cases, although BEST did its “best” to eliminate these homogenization “issues” and still shows a similar trend.
3. There have been several instances in the past where the slope of the warming curve has been greater than or equal to the current slope of the warming curve, so the claim that the rate of warming is unprecedented and can only be explained by CO2 is spurious. For example the slope of the warming curve coming out of the Younger Dryas was remarkably more steep than the current warming curve in every study I have seen relating to that time period.
4. In spite of 1,2 and 3, we were at the end of the Little Ice Age in 1850, and we clearly are not STILL in the Little Ice Age NOW, so obviously there has been warming.
Any more questions?
Peter –
You are still going back to 1850. What happened over 50 years ago is largely irrelevant to evaluating the impact of CO2 emissions. We could STILL not be in an ice age, but not be warming. We could be not be an ice age and be cooling. We could not be in an ice age with there being no temperature trend. There is no reason to assume that “the world is warming” based on the trivial observation that we are not in an ice age.
Once again – our illustrious host has said (paraphrasing) that she doesn’t know any “skeptics” that doubt that there is a warming trend over the past 50 years. Now maybe many “skeptics” do agree with that – but to agree with that would imply an opinion that the temperature record data is valid. To say that you believe that the world has been warming over the past 50 years, but at the same time to say that the data don’t validate such an opinion is internally inconsistent.
Again – read David W.’s post above. It is internally consistent – and it also lies outside of the categorization that “skeptics” don’t doubt that the world is warming.
“Once again – our illustrious host has said (paraphrasing) that she doesn’t know any “skeptics” that doubt that there is a warming trend over the past 50 years.”
Now you seem to understand. You specified the time scale. Fifty years, OK that’s 1960 -2010. Now, I don’t know anybody who doubts that there is a warming trend over that period. Of, course it depends on definition of warming (temperature or heat/energy), but ignoring that and talking strictly temperatures, I have never seen a post doubting that “global temperature” hasn’t increased over the last 50 years. What most sceptics are critical of is magnitude and causes.
Yes – she specified whereas Anthony didn’t. That doesn’t address whether his statement (that you agreed with) is accurate.
But:
Really? You haven’t seen comments about “global cooling?” About “What warming?” That the “Warming has stopped?” That “There is no warming?” All those opinions would be based on the opinion that there is no valid trend of warming over the past 50 years. You can’t have a valid trend of warming over the past 50 years if the “warming has stopped.”
And even if you haven’t (which I find hard to believe) – have you never read posts that doubt the validity of the methods that establish a trend of warming over the past 50 years? How can one agree that there is a trend of warming over the past 50 years if one feels that none of the methodology used to quantify the data are valid?
I’ll refer you also to David W.’s post. At least it is internally consistent – although it lies outside Anthony’s characterization of “skeptics.”
Josh-ua,
I have asked you a similar question before, with no reply: Why don’t you enumerate those skeptical commenters on this board, who have said that there has been no warming over the last 50 years?
Let me do an “independent assessment of how much of the observed warming is due to human actions”
1) We have learned that Berkely Earth’s global mean temperature anomaly data are similar to hadcrut3vgl and gistemp
2) Oscillation in global mean temperature is due to thermohiline circulation cycles as described in the following paper:
http://bit.ly/nfQr92
3) There was a long-term global warming of 0.06 deg C per decade since temperature record begun in the 1850, one hundred years before the start of the exponential growth in human emission of CO2 in the 1950s, as shown in the following graph:
http://bit.ly/pJhIaw
4) As there was no change in the global warming rate of 0.06 deg C per decade since temperature record begun in the 1850s, the contribution of human actions on the observed warming is nil, naught, zilch.
except girma now you gotta draw your lines back to 1800. have fun
steven
I am dumbfounded with the IPCC’s failure to subtract the warming rate due to ocean cycles from the recent global warming rate of 0.16 deg C per decade for the period from 1970s to 2000s.
Warming rate due to ocean cycles = Actual recent global warming rate – Long-term warming rate
Warming rate due to ocean cycles = 0.16 – 0.06 = 0.1 deg C per decade.
Girma
see:
http://www.vukcevic.talktalk.net/NA-SST.htm
vukcevic
What is “North Atlantic SST precursor”?
Why not add a caption for the graph?
get busy redrawing your lines. dont cherry pick the start date
steven
I think girma’s figures (like HadCRUT3’s) going back to 1850 are good enough for some things.
They simply show the observed multi-decadal cycles of warming and slight cooling, the last two of which were declared as “statistically indistinguishable” by Phi Jones.
These represent a bit of a dilemma for those who firmly believe that CO2 has been the main climate “:control knob”, as there are no such cycles in the CO2 record. As a result, they require separate rationalizations that represent a departure from the Occam’s Razor principle.
I know that it is not part of BEST’s brief to try to attribute a CO2 causation for the temperature record, but simply to establish if the surface record, itself, is meaningful at all.
The preliminary BEST report has been informative, but I am looking forward to its final report – so far there isn’t much “meat on the bone” IMO.
As our host here has put it:
Let’s see what these “subsequent analyses” tell us.
Max
It seems to me that there are some things in the Berkeley report that are very interesting.
When I look at Fig. 1 in the report, available here:
http://www.berkeleyearth.org/Resources/Berkeley_Earth_Decadal_Variations
you will see that the 2010 temperature is considerably lower than in the other three records. In the El Nino year of 1998, three of the four records (including Berkeley) were at an anamoly of almost plus 1.0 degrees. At the end point, 2010, three of the records were between plus 0.85 and 0.95, but Berkeley was under 0.6. It is literally the last year where Berkeley breaks away, downward, from the pack. Could that possibly have to do with how Berkeley handled the dropping of so many surface stations? I’d look into that, ASAP.
Eyeballing the change from the beginning of the satellite temperature era, 1979, it seems to me that Berkeley’s trends through 2010 are the closest in temperature trends to the satellite records, which have lower trends than the land based records (0.140 degrees per decade for UAH, 0.148 for RSS, through Jan. 2011:
http://en.wikipedia.org/wiki/Satellite_temperature_measurements#Trends_from_the_record
If I were you, I’d concentrate really hard on the Berkeley 2010 temperature, which goes strongly in the opposite direction than the other three records.
The other thing that really stood out in the Berkeley work is that the land surface temperature record is much more closely related to the Atlantic Multidecadal Oscillation (AMO) than to ENSO. The PDO and AMO appear to be closely correlated, with PDO changes leading AMO changes by about 2.5 years (Fig. 5).
We’ve all read the articles by “mainstream” researchers suggesting that temperature might not rise for a couple of decades because the PDO is in its declining 25 to 30 year cycle. If this is true, then the AMO may not be far behind, and if Berkeley is correct that the AMO and surface temperatures are tightly linked, then we may indeed not see much warming for a while.
I’m wondering if anyone has looked at the actual data yet – and I don’t mean with a meat grinder. In BEST, ID #141457 is listed as ‘BUFFALO GREATER BUFFALO INT’L’. There are data from 1873 to 2010 in the file ‘data.txt’. First, there was no Buffalo Int’l Airport in 1873, so either the data are mislocated or they are a combination of separate stations. I thought eliminating these problems was a feature of BEST. Second, for 1873, the 12 monthly values for temperature are:
12.782, 11.615, 12.295, 12.208, 11.568, 11.546, 11.403, 11.300, 10.938, 10.632, 7.576, 13.909
Besides being unrealistically warm, it seems unreasonable that January and December would be the warmest months of the year in Buffalo. I hope someone can point me to where I am misinterpreting the data.
bad conclusions. You can’t merely look at input data. you have to look at intermediate steps, also the quality flags, also the station combining.
The station name is just the most recent name. So before the airport was located there it would be named differently. You’d have to look at station histories to get that.
I havent followed the source all the way through and dont know if they dump intermediate files ( as GISS does) so that you can track what happens at various datasteps.
Wonder Under what circumstances January and December have ever been the warmest months in Buffalo, NY?
Again, until you trace those values through the outlier detection and correction part of the code you dont have a clue whether that raw data made its way into the final result. hence the need for source code and intermediate data steps. I should not have to explain that.
They took raw data in. They applied there own quality tests.
Looking at the raw data tells you nothing we dont already know. raw data sucks
OK, Mosh … I’m not a climate scientist, and I strongly suspect that the only reason I got a passing grade in the stats 101 course I was obliged to take during the process of acquiring my Psych/English degree was that I was successful in using pre-post-modernist logic (and a tiny bit of good old-fashioned rhetoric).
In the intervening years (which are more than I care to reveal!), I have acquired a number of proficiencies in working with data.
As a big picture pattern picker-outer person, I’ve always found that working with/examining source (i.e. raw) data has enabled me to solve more problems than data which has been filtered and/or otherwise distanced from the source/raw data..
So please tell me, why is climate-related raw data different from all other raw data? Or should I, uh, pass over this question?!
So you have not actually looked at the data – thanks.
BEST has since replaced ‘data.txt’ with a new version. The new version has Tavg for Buffalo back to 1831, which matches GHCN, but the temperatures are still screwy values. What a disaster.
After reading the post, the comments to date, and the Wickham et al preprint on the UHI effect, here are some tentative impressions on a subject I haven’t studied in detail. They are based on my previous understanding aided by the Wickham paper and some insightful comments in the thread.
The paper finds a “slight” negative slope – 0.19 C/100 years – in comparing the warming of all stations with those judged “very rural”, suggesting what one might call a “reverse UHI” that differs from previous estimates, most of which have suggested a positive but usually very small effect. The cause of the discrepancy is unknown. It may be artefactual, but adds to evidence that tends to exclude a climatically important UHI effect that exaggerates warming trends.
It is not clear to me why there should be a significant trend in either direction a priori. There is no question that urbanized areas tend to be warmer than rural ones but that does not inevitably require them to warm significantly faster over the long term on a global average, and so the latter is a question that must be settled empirically. The new analysis adds to that evidence.
A UHI effect must be distinguished from inappropriate siting of instruments – e.g., next to an air-conditioning vent that might blow warm air onto the measuring devices. The UHI effect refers to the real – not spurious – finding that urban areas tend to be warmer than the rural surroundings due to changes in albedo, reduced heat loss because of the obstruction from buildings, reduced evapotranspiration, etc. In general, these phenomena require fairly substantial urban buildup – a few buildings and a roadway should have very little effect. For those areas that are clearly urban, I’m not sure we know enough about the UHI phenomenon to predict how the growth of big cities into even bigger ones will affect their warming rates – in some cases, the rates might increase while other cities might show a reduction due to urban planning that creates more open areas.
Legitimate questions are raised about misclassification of “very rural” stations. We might therefore ask how much misclassification would be required to convert the rural vs total station difference to zero, how much to reverse it, and how much to reverse it sufficiently to create an important positive effect. This would be worth modeling to some extent, but I’m inclined to believe that misclassification sufficient to miss an important positive effect is very unlikely. Perhaps that can be tested. At this point, the new information, in concert with existing data, fails to support a UHI effect as a source of significant distortion in the interpretation of global warming over land.
AFAIK the cause of the hypothesized spurious warming is the expansion of urban area to include stations that previously were rural.
Hear Hear!
Not just urban planning. AFAIK most models of UHI are a sort of “bubble” of warm air over the urban area. However, if such a “bubble” gets big enough, convection will set in, producing something like a mushroom. On some days, that mushroom might actually act to draw surrounding air in, cooling all but the most central stations.
A good idea.
my comment below, thx
Josh-ua,
Why don’t you emulate Fred. His droning defense of the dogma is at least based on somewhat plausible, although arguable, interpretations of the alleged climate science consensus. And Fred seems to be a decent and honest guy, who is just misguided. He is clearly not here only to be an antagonist. In other words Josh, you could be tedious in a much nicer way.
Fred, I mostly agree. The only thing I would change is rural vs. urban. Rural stations can have high UHI growths, higher than urban and therefore higher trends. The proper distinction should be unchanged vs. changed (or wild/built or similar). So, absolutely best stations please (few hundreds is enough) and it doesn’t even have to be 100 years or more. The last 50 years would be something. Let’s see what 200 unchanged stations show over the last 50 years (or more if available).
Fred
There are scores of studies out there, which clearly show, based on individual locations all over the world, that urbanization has caused warming at a local level.
Therefore, a preliminary conclusion, which shows that the impact of urbanization on the global temperature record has been one of slight cooling is not logical. Whether or not it should be discarded a priori as an error, is another point.
But is surely does not make any sense, so should not be used as a rationalization that the UHI impact is negligible, as you are apparently doing.
Max
“There is no question that urbanized areas tend to be warmer than rural ones but that does not inevitably require them to warm significantly faster over the long term on a global average, and so the latter is a question that must be settled empirically.”
Isn’t there also no question that as time goes on and more places became urbanized? And so the temperature should go up. The point would be that warming is still due to man but not co2. Or do I misunderstand you?
There could also be an affect of wilderness to agricultural or other vegetation changes. Even a distinction and change in urban residential, versus urban industrial or commercial, high density versus low density. I think this is the areas of interest of r peilke sr.
That was in reponse to fred above
I think the point is whether urban areas warm faster than non-urban areas.
In terms of evaluating trends, you can still find a valid trend of warming if both types of areas warm at a similar rate. You are determining trend by comparing the stats from the same locations over a period of time. You are not determining trends by simply by overall warming – in which case a warming trend might be only an artifact of urbanization.
Or do I misunderstand you?
Yes
Isn’t there also no question that as time goes on and more places became urbanized? And so the temperature should go up.
Kermit – In response to you and relevant to points made by AK above, the assertion that a UHI effect is a distorting influence on temperature interpretation, as I understand it, involves the fact that urban stations are greatly over-represented in temperature records and so the effect of their temperature change on global temperature is over-represented. Urban areas in the record represent a very small percentage of the actual land area of the globe. The claim is not that because cities are growing, the heat they exude is making the world warmer (urban heat is probably insufficient to do much of that, although Pat Michaels did imply otherwise), but rather that they are making it appear warmer than it is. As long as locations are classified correctly in both current and past records, it should be possible to ascertain whether urban locations are warming faster than rural ones. If they are not, then changes in global land temperature can probably not be attributed to urban heat to any large extent. Whether there is any significant contribution is a topic that will probably continued to be argued.
The two concepts to be distinguished are (a) cities are expanding and making the world warmer. and (b) cities are warming faster than rural areas and that causes the record to show more warming than is actually occurring. It is item (b) that is the “UHI effect”.
There are some other concepts or subconcepts that you touch on and others I think could be important.
– cities are expanding and (maybe) are making the world only appear warmer.
– cities are where people live so measurements may be more accurate, and/or maybe more important to us. I assume data is easier to get from urban areas and at earlier periods in cities, perhaps science stations also. Here is why I say this, only land areas were used in this study. I assume it was for these same reasons, people live on land and so it is both easier to get, and more relevant because that is where we live. Take Greenland, in the video by BEST they have data points back to 1850 yet I would assume that data would be more extensive in the UK at that time. Yet the video shows it of equal value. It’s likely that the UK data is even more accurate, even though the temperature in Greenland is probably of more life-and-death value.
There are many other considerations, climate (on a tropical island there is less variation then an arctic one), cultural (in the 19th century UK they had a culture that put importance on measurement), purpose of the measurement (for science or for navigation or for agriculture), time period, methods. Perhaps these all even out in the end.
So to me this means that some data should be weighted accordingly, maybe they do, but then someone has to judge what is the weight, then the analysis becomes subjective to become of value. There is where bias comes in. These are the tricks referred to in ClimateGate. If you take your car into your mechanic and he says there is nothing wrong with your car you don’t pay him. The mechanic will try very hard to find a problem with your car.
IMO
Fred Moolten:
I disagree with the whole concept of urban versus rural when one is trying to judge the accuracy of a global temperature record. The classification of rural has nothing to do with construction that can effect rural records.
But let’s look at what can happen. If you put a thermometer in a city that is already built up you may get very little additional build in the immediate area of the thermometer. But the city will continue to grow at the edges, and to a much smaller extent in already built areas. This will certainly give an increasing UHI effect for the thermometer.
Now let’s consider the rural areas. Notice that virtually the entire east coast and west coast of the US are dotted solid with black dots indicating that there are thermometers there that are considered rural. These areas are densly populated and getting more so all the time. So a designation of rural should not make one think of a cow pasture in the country side. Rural can still mean significant build and it can still mean that significant building is happening. In fact, the availability of open land means that construction could happen very close to the place where the thermometer is placed, without any change in the thermometers status. This construction close to the thermometer would have a more direct effect that construction in a city that is happening at the edges, far from the thermometer. So putting a city of, say, 20,000 around a thermometer will still leave that as a rural thermometer, but the effect on the thermometer can be greater than what is happening in a city where the thermometer is in an area that is built and not changing much.
In any case, the question that BEST asked, which is “how do areas that are classified as rural compare with areas that are classified as urban”, completely fails to answer the question, “how has construction of roads, parking lots, buildings, etc. effected temperature trends”. In fact, the answer that they have produced is simply stupid. I can’t believe that they accepted it without having it occur to them that they may have done something wrong. It seems like we should be able to get some indication of what is happening to global temperature trends that are uneffected by UHI by looking at SSTs.
Douglas Keenan was a reviewer of four of the BEST papers. He has made his review comments available at Bishop Hill, http://www.bishop-hill.net/blog/2011/10/21/keenans-response-to-the-best-paper.html
They are worth reading.
Bravo to Keenan. Maybe BEST can change its name to SAME.
Thanks for a very interesting post and for a link to the Economist article on the BEST project. As much as I like the Economist (subscriber for more than 15 years) I find its reporting on climate issues to be irritating for a newspaper that believes in economic rationalism and free enterprise. And this this article misses the mark entirely with its silly, triumphalist spin. The main question remains whether an upward trend in temperatures observed for the last century or so is part of a low-frequency cycle that is a component of its “normal” stochastic behavior or if it signals an aberation due to human activity. An improved 150-year series of temperatures cannot answer this question or reduce valid skepticism about the methods used to backcast prior temperatures using proxy variables. Having better data, though, is always a welcome thing and makes for better science.
The UAH satellite record show substantial cooling from the endo of 2009 through the end of 2010, of roughly 6/10ths of a degree:
http://img856.imageshack.us/img856/2403/dailyuahtempsmar92010.png
This big drop in in agreement with an even larger drop found by the Berkeley group, but opposite in sign and magnitude to the three other land based records (NOAA, NASA-GISS, HadCRU). See Fig. 1 — zoom in — in the Berkeley report:
http://www.berkeleyearth.org/Resources/Berkeley_Earth_Decadal_Variations
This divergence between Berkeley and the other three records might be very important. Suppose that Berkeley has found an anomaly that has escaped the other three “official” records. In theory, that would be the type of thing that Berkeley set out to do. If this 2010 temperature finding is correct, then the last year of the series is between 0.15 and 0.2 degrees lower than the other three, “official” series. That would mean that the trend for the Berkeley series is about 15% to 20% lower than for the other three.
So it seems to me that we have to understand why the Berkeley 2010 temperature figure is so much lower than the other three land records, but in agreement with UAH.
The graph shows 12-month moving averages with the final data point appearing to show the average of July 2009 to June 2010, so in any case it doesn’t show a comparison of 2010 anomalies.
The most likely reason for the discrepancy is that BEST don’t have 2010 data finalised yet so in their comparison setup January to June 2010 monthly anomalies are set to zero, creating a downward diagonal in the moving average.
Also, the 2010 anomaly in UAH is in good agreement with the surface records.
“Our biggest surprise was that the new results agreed so closely with the warming values published previously by other teams in the US and the UK,” said Prof Muller.
Now that it has been adequately confirmed that the analysis of global temperature data by NOAA, GISS, and HadCRU (showing that the global temperature has been increasing by 0.2 °C per decades since 1970) has been correct after all, perhaps Prof Muller and associates will now also take a look and verify that the radiative forcing due to the increase in greenhouse gases has indeed been responsible for this observed increase in the global surface temperature.
Perhaps one should answer the concerns from Doug Keenan before assuming the BEST data and analysis are correct.
Also, when do we anticipate that an appropriate and accurate description of the results will be released.
As I read what BEST has done, it is to:
– analyze data from a large number of stations that still represent a very small percentage of the land mass of the globe, which in itself represents only 30% of the total area of the globe.
– average the Tmin and Tmax, which most certainly does not represent the average temeperature for a day, which would be the integral over 24 hours of all temperature function at that station for that day.
– no addressed whether the temperature variations seen are statistically significant or just random variations in climate.
So at best (pun intended) they’ve shown that some temperature metric at a small fraction of the globe’s surface is now higher than it was at some time ago, but we don’t know if that is due to random variations or as part of a warming trend, and if part of a trend, we don’t know the causation of that trend.
Now that’s not much we can say after half a million and a lot of work!
How one gets from this to the headline in a press release about non-peer reviewed papers is, in a nutshell as they say, the problem that many skeptics (including me) have with climate “science”!
IMHO, the climate “scientists” have shot themselves in the foot yet again.
Now that there is agreement with observations ,this tells us that the physics are indeed wrong ie the Krakatoa problem eg Stenchnikov 2006
The observed value no doubt has substantial temporal sampling errors, and the observations for the earliest eruptions considered may suffer from inadequate geographical sampling or other data quality issues. It
is noteworthy in this respect that the global-mean surface temperature record used in this study surprisingly shows no global cooling following the very large 1883 Krakatau eruption [Jones et al., 2003]. However, the surface air temperature reconstructed by Hansen and Lebedeff [1988]
shows a more sizable cooling effect of Krakatau, which indicates the level of uncertainty in the observations themselves, especially for earlier volcanic events
The interesting thing will be seeing how and if reconstructions change.
You have reconstructions that are calibrated to HADCRU back to 1850.
Looking at the differences between HADCRU and BEST you’ll see that HADCRU lies outside the boundaries of BESTs CI.
The HADCRU methodology results in using a much smaller number of stations to estimate the mean. This is especially important in early years where you have records which stop before the HADCRU common anomaly period.
I think the next course of action would be to see how this better more statistically sound series changes the envelop of reconstructions. Since, as Hansen argues, Paleo gives us the best look at the ECR it would seem that the next course of action would be to recompile reconstruction results using the best temperature series. If everything turns out the same, all the better.
Thats another brick in the wall.
So, given the talents in Mullers group and given the importance of the instrumented record to reconstructions and given the importance of recons to ECR.. the logical thing for them to do is to have a look at recons.
Mosher: “So, given the talents in Mullers group”
What talent? They decided that there is no UHI effect. You can hardly get dumber than that.
If you think that there is hude UHI effect you are welcomed to do the following
1. Download their data
2. Download their code
3. Define your own criteria for urban and rural
4. Prove that the trends in urban are substantially greater.
This problem is not easy. I suspect you won’t do a lick of work
Mosher: “You have reconstructions that are calibrated to HADCRU back to 1850.
Looking at the differences between HADCRU and BEST you’ll see that HADCRU lies outside the boundaries of BESTs CI.”
LOL. If you are talking about proxy reconstructions, they can’t even find proxies that warm as much as HadCrut. How much more cherry picking are they going to try to do to match BEST?
Nothing has been confirmed. First of all, they diverge from HadCrut. Second of all, they have concluded that there is no UHI effect using a methodology that is incapable of drawing such a conclusion. A brain dead conclusion if there ever was one. Third of all, their source data sets are drawn from the same place as the previous sets and are subject to the same potential problems. Fourth of all, they still diverge from the satellite records. Fifth of all, they still outrun the proxy data, just like all the other sets.
Dr. Lacis
In the paper
Decadal Variations in the Global Atmospheric Land Temperatures
It is also said:
“On the other hand, some of the long-term change in the AMO could be driven by natural variability, e.g. fluctuations in thermohaline flow.”
I happen to be reasonably familiar with the AMO (or the N. Atlantic SST) and the most climate related events (both causes and consequences) in the North Atlantic.
I have identified process which is advanced to the AMO and the NAO by some years, which may or may not be at the root of the SST’s upward trend.
http://www.vukcevic.talktalk.net/NA-SST.htm
In my view it is not possible to determine degree of the CO2 contribution to any degree of certainty until all possible natural causes are eliminated.
We should not be guided by our beliefs, but with our knowledge, and as far as the AMO is concerned our knowledge is far from complete or even adequate to be able to make scientifically valid judgment about its contribution to the global temperature rise or fall.
What is an SST precursor?
I am currently putting together short article, so all necessary information will be there. Data I use are from NOAA, NCAR etc.
Sure thing. We should not be relying on beliefs when we have real knowledge of facts and physics to go on. We must of course keep in mind that the best that our current ocean models can do, is to generate natural variability in ocean response that has some resemblance to the El Nino, La Nina variability. I am not aware that the current ocean models can do anything reasonable with respect to modeling AMO climate variability.
Accordingly, the best that we can do with El Nino, La Nina, AMO, and PDO variability is to try identifying their spatial and temporal patterns in their early stages so that their short term climate impacts can to some extent be extrapolated into the future.
As for the climate impacts of CO2 and the other greenhouse gases, we have the advantage that here we have facts and physics to bring to bear. The time trends for the GHGs are being accurately monitored. We also have available accurate HITRAN absorption coefficients for all the relevant GHGs. And we have available reliable radiative transfer models to calculate the radiative forcings due to these GHG increases, and reliable models to calculate the climate impact of these radiative forcings including the time delay of the climate system due to the large ocean heat capacity.
We should not be relying on beliefs when we have real knowledge of facts and physics to go on.
I disagree, it is all in the data, physics of it is basic, far simpler than one may suspect.
Correction: It should be ‘I agree’.
Hello Andy, I have a question that you can probably answer for me. I look at Figure 2 in the decadal paper and I see the AMO index. It looks to me like there is no upward trend. Has this data been detrended?
Thanks,
David,
I don’t see any reason why they would try to ‘de-trend’ the AMO index. Moreover, to first order, I would not expect there to be any long-term trend in the AMO index. I tend to think that the AMO and PDO, and the shorter period El Nino and La Nina variability represent natural (unforced) fluctuations and oscillations that the climate system undergoes about a zero reference point – hence, no long-term trend.
It is possible that global warming might induce some change in the amplitude, frequency, or pattern of these unforced fluctuations, but I have seen no evidence that that might be happening.
A Lacis
“fluctuations and oscillations”=> http://bit.ly/p1VPcC
“long-term trend” => http://bit.ly/pJhIaw
(a warming rate of only 0.06 deg C per decade, not IPCC’s 0.16 deg C per decade; therefore, true climate sensitivity = 3 *0.06/0.16 = 1.1 deg C)
AMO
Also see Is The AMO A Statistical Phantom?
The AMO is treated as some kind of taboo by the climate scientists. The AMO is simply N.A. SST with the trend taken out, so AMO by definition is the DETRENDED N.A. sea surface temperature.
Many suspected that this oscillation may be do with TSI, with some solar papers (Reading University) trying to establish the link. I do not think that to be the case, the change (according to my findings, which will be fiercely disputed) is directly related to the events in the Arctic Circle.
It is wrong to assume that the AMO (or the SST) has a regular cycle, as it is often suggested.
A Lacis
What caused the earlier similar warming both in magnitude and duration from 1910 to 1940, when human emission of CO2 was 1/5th of that during the recent warming period, shown below?
http://bit.ly/bx3Ebh
The recent warming is identical to the previous one, and it is not unprecedented. By Ockham’s razor principle you assume it is natural.
In addition, the globe has started its cooling phase as shown in the following data.
http://bit.ly/nz6PFx
AGW is not supported by the data!
Wow,
Leslie Kaufman at the NYT really butchered by comments. I specifically told her that my issue on this wasnt transparency. specifically told her that I thought the results were as expected.
Then again, she thought I was steve mcintyre till I corrected her.
maybe I should post our email exchange? it’s hilarious..
Votes?
Yeah, let’s see the exchange, thanks.
Steven, I would ask her/him permission and give her a chance to defend herself.
I agree with MattStat, but it’s a judgement call
I’ve asked her to make a correction.
No. I don’t see a purpose to be served. It’s as though you were to purposely introduce a troll. Your comments are good. Don’t dilute them.
that’s my vote.
I’ll take your suggestion. Thanks
Mosh
I always ask permission before publishing private emails , It’s why I get sent so many interesting bits of information, but also means a good proportion of it I can’t pass on. :)
tonyb
On the other hand, if publishing the email exchange made her angry, she might not misquote Mosh ever again – because she would not contact him again. :)
There’s enough evidence anyway, people seem to have joined the journalistic profession because not only they weren’t able to do a thing, they couldn’t even teach it.
Yes, post it.
In my experience, journalists get away with very poor work queslity- for whatever reason- far too often.
…..sigh……
quality
ironic, no?
I would ask her/him permission and give her a chance to defend herself.
Figure 1 in the Berkeley Earth Decadal Variations shows intervals where the BEST temperature series departs dramatically from all 3 of the other series: 1954, 1956, 1968, 1993, 2010 (years approximate.) Have the authors addressed such seemingly bizarre discrepancies? I can imagine explanations, but have the authors explicated some of these?
UHI and erroneous ‘averaging of data’ actually hid a cooling trend taking place in France since 1871. The land-based thermometers were located at airports were snow is always immediately cleared, resulting in the recording of anomalously higher winter temperatures compared to the surrounding countryside.
Added to that are other factors. For example, airports get hotter in the summer and many use de-icing sprays or burn tons of kerosene to keep August and Christmas flights humming.
These average-skewing UHI effects actually masked a decline that everyone really knew had actually occurred because reality even overwhelmed the erroneous data collection methods and termperatures still dropped.
What we learn from the lesson of France is that we are not dealing with an ‘average temperature’ — what we really have is an ‘AVERAGE OF TEMPERATURES.’ And, worse yet – in France – we’ve got erroneously structured data and results obtained from areas of airport tarmac instead of French countryside.
I suspect this is a naive question but why is BEST comparing the set of “very rural” to the set of “all stations including very rural”?
I would have thought that the proper comparison would be “very rural” to “all stations without very rural”. Comparing something to itself seems odd.
Hi Jay – I think the distinction may not be very important because we could calculate one from the other. To my mind, the method used by BEST offers an advantage in that it’s easier to visualize what the results mean. The overall average temperature trend is something whose meaning we can visualize, and so is the trend of very rural stations. On the other hand, what to make of “non-very rural” stations doesn’t lend itself to easy interpretation because they are so heterogeneous. If the two concepts were not intercovertible, I would think it might be wise to present both datasets, but as it is, there is probably no need..
Fred,
you are just bloviating now. Comparing a set of stations against a set of stations that INCLUDES that set of stations REDUCES any possible differences.
So we’re all, or nearly all, in agreement now that the temperature record is reliable, are we? Not that anyone has ever said otherwise of course!
http://www.skepticalscience.com/surface-temperature-measurements-advanced.htm
http://www.skepticalscience.com/the-best-kind-of-skepticism.html
HAHAHAHAHAHAHAHAHAHAHAHAHA
Independent verification is always nice. How much discussion was there on V&V procedures on this site in the past, and now it is marginalized by all sorts of excuses. What many people do not seem to understand is that verification is also the process of gaining confidence in the models and results.
Its a semi-effect PR stunt, while their not in competition with any one else so had no need to rush it out ‘to beat’ someone , the need to rush it out was political one hence all the effort gone into the press and marketing which lets remembers has nothing to do and does not add to the science .
And why semi-effect, while although they know that should these papers fall to bits in peer review this will not make the headlines, as by they this will be ‘old news’ Its not actual getting the coverage or impact it was designed to . I would take a good bet Bob Ward and his fellow AGW spinners are spitting feathers over what happened to Gaddafi and the way its dominating the headlines. While given the economic situation people attentions are rather fixed on bigger concerns and I cannot see the return of the amount of political will needed to ensure a return of the ‘good times ‘ for the Team the and friends on the back of BEST work .
Maybe a good way to get some feeling for the effects of UHI is to compare land based trends to SST trends. We know that there is no UHI effect to the SST trends. Here is a NOAA chart.
http://processtrends.com/images/RClimate_SST_A_latest.png
Looks like about 0.5 C since 1880. I’m not sure that it’s a fair comparison due to ocean circulation. But the current SST trend looks to be flat (last dozen years). And I can’t see how it could be flat if global temperatures are, in fact, rising.
Don’t worry the SST will catch up to the land trend, as soon as that heat hiding down in the depths resurfaces. I used to think there was a lot of cold hiding down there, but now I know better.
Actually, the land temperatures have been going down for the last 10 years, more so in BEST than in the other datasets.
Hmm. I must have had my head tilted when I looked at their chart.
Sorry, They are going down in the BEST and are nearly flat in the others.
David:
Just to make sure we are not talking past each other. I’m looking at this chart. Are you looking at the same thing? I know that it’s a compressed scale for looking at the recent trend. And it’s kind of burried with the others. But I don’t see a decline. Maybe you have a more isolated chart that only covers the last dozen years or so and that has a trend line.
http://www.berkeleyearth.org/analysis.php
Yes, I’m looking at Figure 1 in the Decadal paper. It shows more detail in the last 15 years. If you look very hard you will see that in 2010 Best goes down quite a bit and is about 0.1 C lower than in 1998.
David Young, it’s important to remember Figure 1 in that paper was made by using data from only 2,000 stations. It will be different than what is gotten when all stations are used.
Brandon, I see your point. Thanks
Not a problem David Young. I was confused about the same thing. It’s especially confusing because none of the four papers actually publish the BEST temperature record, so it’s not like you’d see the “right” graph in addition to that one.
For what it’s worth, you can find the record posted on their site, here.
You know, in the time it took you to make all these grumpy comments, I’m pretty sure you could have answered your own questions three times over. As in, getting some station data from some locations *near*, but not *in* urban centers (the kind that were pretty much rural in the early 20th century, but are suburban or fully urban by now), and see what happens over the 20th century as urbanization extends.
.
Then you could actually try to estimate just how much of the global land surface is composed of such areas, which might give you an idea of how much you should weigh the resulting value to estimate its actual effect on global land temps. Looking out the window next time you take a plane might provide some interesting clues.
.
If you find a huge number, you can book your trip to Stockholm! (And when you get there, be sure to tell them about your theory that UHI effects can be estimated by looking at the difference between land and sea warming rates. I’m sure they’ll be quite impressed :) )
toto: “Then you could actually try to estimate just how much of the global land surface is composed of such areas, which might give you an idea of how much you should weigh the resulting value to estimate its actual effect on global land temps.”
It’s not about how much land is urban and how much is rural. It’s about where the thermometers are. You won’t find many of them in national parks or on farms. In other words, the distribution of the thermometers doesn’t come close to represtenting the percentage of true rural area versus the percentage of built area.
“(And when you get there, be sure to tell them about your theory that UHI effects can be estimated by looking at the difference between land and sea warming rates. I’m sure they’ll be quite impressed ”
Not as impressed as by your theory that SSTs can be going down while land based temps are going up.
No I’m not making this up. Anthony Watts really did say:
“And, I’m prepared to accept whatever result they [the BEST group] produce, even if it proves my premise wrong. I’m taking this bold step because the method has promise………… but the method isn’t the madness that we’ve seen from NOAA, NCDC, GISS, and CRU, and, there aren’t any monetary strings attached to the result that I can tell.
http://wattsupwiththat.com/2011/03/06/briggs-on-berkeleys-best-plus-my-thoughts-from-my-visit-there/
Guess that Anthony, in his wildest dreams, never anticipated that this group would be so stupid as to declare that there was no UHI.
I’ll have to take a look at what the BEST group have said about the UHI effect in more detail but its my understanding that groups like GISS have never claimed that it didn’t exist and they have corrected their data for this effect as described by James Hansen.
http://pubs.giss.nasa.gov/docs/2001/2001_Hansen_etal.pdf
Maybe Judith could clarify if the BEST group disagree with Hansen on this? If they do, it can’t be by much. The BEST and GISS results are very close.
It doesn’t really matter about the comparisons between BEST and GISS regarding UHI. The simple matter is that the BEST UHI test does not come close to testing what they claim to be testing.
So you say. The BEST papers aren’t yet peer reviewed and the idea of making them publicly available at this stage is to give everyone an opportunity to correct any faults. There can be no charges of “Pal review”.
So we’ll have to see what objections are made and whether they are valid.
Reading between the lines, Judith doesn’t sound too happy about the section on the UHI. However, she’s the one with her name on the four papers. It’s a bit lame to say she is second author for alphabetical reasons only. It doesn’t matter if she’s first or last, she has to agree with their contents or ask for her name to be removed completely.
What exactly is ‘the premise’ referring to in that quotation? I can’t see from that webpage what Watts means. It can’t be that no warming is to be found, or that poorly sited stations exaggerate warming, since he has a paper out claiming the opposite: http://wattsupwiththat.com/2011/05/11/the-long-awaited-surfacestations-paper/.
I take Watts’ premise to be that the surface temperature record, (GISS , Hadcrut etc) is/was unreliable. Now BEST has come up with a very close match. So it will be interesting see if Anthony Watts will be as good as his word and “accept” the BEST findings. I doubt it somehow.
Temp,
statistics cannot fix bad input and insufficient coverage. Poorly posed adjustments only add to the issue the same as the original. Mistakes like adjusting the recent temp up after a step decrease to homogenize the series must be researched to determine if it is appropriate, not simply applied because it MIGHT be the right thing to do. Could be the series was biased HOT before the step decrease and the older temps should have been adjusted DOWN!! Video Game studies simply cannot add certainty to the mess we have.
Here is the first one, Lon Lat
110673 NORILSK 69.333, 88.125
According to Modis there are no built pixels within 5 minutes .0833 degrees
But go to google earth and have a look. about 12% of the surface is impervious, 3% of the land is urban according to other measures of urbanity
Nor’ilsk is either the largest or 2nd largest nickel smelting operation in the world.
While NORILSK is still not hightly developed, the actual build change that happened there over the last 100 years could be greater than the change experienced by an urban thermometer that was already surrounded by build when it was put there.
Now that would be ironic.
If Muller was to be targeted for a smear campaign directed against his work.
Then the circle will be complete.
Should be easy. There’s a lot of data files and I am sure a few stations have been labeled wrong.
Looks like there are a few urban pixels just at the outskirts of the .1 deg boundary, so they would classify this as urban and not very rural..
.08333 versus .1C. took a bunch of compute time. arrg
An interesting set of papers by BEST which reconfirms the other land surface temperature time series. I would like to comment on the UHI paper as I am surprised with the methodology they used.
Roy Spencer did some work on UHI early last year. He used NOAA’s International Surface Hourly (ISH) weather data from around the world during 2000 plus 1 km gridded global population density data.
http://www.drroyspencer.com/2010/03/the-global-average-urban-heat-island-effect-in-2000-estimated-from-station-temperatures-and-population-density-data/
His abstract:
“Global hourly surface temperature observations and 1 km resolution population density data for the year 2000 are used together to quantify the average urban heat island (UHI) effect. While the rate of warming with population increase is the greatest at the lowest population densities, some warming continues with population increases even for densely populated cities. Statistics like those presented here could be used to correct the surface temperature record for spurious warming caused by the UHI effect, providing better estimates of temperature trends”.
See the graphs of “station warm bias” versus “pop density”.
Other interesting analyses follow; scroll forward.
I suspect all BEST have done is pick out a low population density subset of stations which have a higher rate of warming over the time period than the total population and viola, these stations report a higher warming rate than the total population. I’m not surprised they got the results they did.
I suspect all BEST have done is pick out a low population density subset of stations which have a higher rate of warming over the time period than the total population and viola, these stations report a higher warming rate than the total population. I’m not surprised they got the results they did.
Geoff – Roy Spencer often describes his data in ambiguous fashion, particularly when the data are unpublished. In this case, he refers to greater “warming” in low population density areas than higher density areas, but from the description, he is not actually referring to “warming” (temperature increase with time), but to a one-yer averaged temperature as a function of population density – 1000 more people in a very sparsely populated region is associated with a greater temperature difference than an extra thousand in a more populous area. I’m not sure this would be particularly relevant to the BEST analysis unless it turns out first that the correlation reflects causation (population growth causes the temperature to rise) – something that is not implausible – and second that there has been a far greater percentage population increase in the BEST “very rural” areas than in more populous areas – a possibility that would need to be very consistent to result in the magnitude of the effect in the Wickham paper, and which seems somewhat doubtful.
Others can check the Spencer blog to see if they think he was saying something else.
Fred:
Since you link your page I went and listened to three of your songs on YouTube. They were folk and eco-preachy. Got anything more jazz or blues oriented and less eco?
On my web page, listen to Midnight Fugue and Waiting for the Midnight Blues.
Okay, that’s more like it. I liked them both. But I didn’t like the soliloquy in Midnight Fugue. Lisa and Maria can both sing. Lisa has that special touch.
Fred to me there is nothing ambiguous in Roy Spencer’s post; maybe it is a little difficult to understand at first. He makes the point: –
“Some readers were left confused since my posting was necessarily greatly simplified; the level of detail for a journal submission would increase by about a factor of ten”.
Hopefully, as suggested in a following post, a paper will be published: –
“John Christy has agreed to co-author a paper on this new technique, since he has some experience publishing in this area of research (UHI & land use change effects on thermometer data) than me. We have not yet decided what journal to submit to”.
http://www.drroyspencer.com/2010/03/global-urban-heat-island-effect-study-an-update/
http://www.drroyspencer.com/2010/03/urban-heat-island-a-us-versus-them-update/
I will read the BEST paper in detail when I get some time however these posts of Spencer’s and others, are a big red flag to me as far as the methodology used in the BEST UHI paper, I’m afraid. However I’m willing to be persuaded.
Geoff – My point was that Spencer referred to “warming rate”, but he never looked at how temperature changed with time, only at the correlation between temperature and population density. Since his language suggested that he was evaluating how a location warmed over time, I thought it was ambiguous.
The updated Spencer posts you have since linked to seem to show the same thing. Different years are looked at but not temperature change with time, only temperature correlation with population density.
Fred, I’m sorry, but this is a complete red herring. When one refers to rate of something why do you automatically think time has to be in the denominator??? Clearly he refers to warming per population density increase in the abstract & the analyses he performs.
Geoff – If you don’t look at temperature change with time, you can’t draw conclusions as to whether one location will warm faster or slower than another. Without additional information, it tells us nothing about differential warming rates at the heart of the UHI effect concept..
Were his posts ambiguous? Other readers can visit and say whether they immediately understood he was not talking about “warming” in the usual sense of growing warmer with time. It took me some scrutiny to figure that out. In any case, his material can’t be applied to the BEST data as indicative of, to quote your comment above, a “higher rate of warming over the time period”.
Geoff quoting Roy Spencer: “Global hourly surface temperature observations and 1 km resolution population density data for the year 2000 are used together to quantify the average urban heat island (UHI) effect. While the rate of warming with population increase is the greatest at the lowest population densities, some warming continues with population increases even for densely populated cities. ”
Bingo. This is what I have been trying to explain for about a dozen posts now. I didn’t know about Roy’s paper, but he basically confirmed the little thought experiment that I posted way above yesterday. Thanks Geoff.
Tilo – As I stated above, reading Spencer’s blog indicates that he did not examine the “rate of warming” but only the correlation between temperature and population density. There was no attempt to look at temperature change with time, which is at the heart of the BEST findings.
Fred’
He did examine the “rate of warming”, Fred:
“RESULTS & DISCUSSION
The following graph shows the average rate of warming with population density increase (vertical axis), as a function of the average populations of the station pairs. Each data point represents a population bin average for the intersection of a higher population station with its lower-population station mate.”
Wouldn’t the “change with time” depend on the rate of change in population over time, Fred? Maybe Roy will get around to that, as soon as he gets some funding. Or you could do a study.
See my comments to Tilo and Geoff.
I have seen your comments. Fred. I will make the question easier for you: If the population of Palo Alto, CA, (or you pick it) were doubled overnight, with all the associated additions in infrastructure and life-style accoutrements, would the temperature likely be warmer in that locale tend to be warmer, or cooler? That eliminates the time BS.
I already addressed that issue. It might be warmer, although not necessarily. As I suggested, that was not the weak point in the Spencer argument.
OK, Fred. Stop dodging. If the same thing occurred in a thousand cities, would you expect most of them to get warmer, or not? Assume that they were not moved to a cooler climate, Fred.
How about a city in a fast-growing region like Phoenix – where along with increasing population, more roads and cars, what once was desert now has lawns, trees, swimming pools, golf courses, parks?
Would the rate of change in temperatures be the same sign and/or degree as an increase in Palo Alto’s population increase? In comparing those two cities, can you predict changes in temperature merely as a function of population increase?
I don’t have any more patience for your foolishness, Josh-ua. But I am going to help Fred. All else being equal, every one of those thousand cities would get warmer. And that is just what has happened in thousands of cities across the world, over the last century. That is where the term UHI comes from, Fred. And for BEST to suddenly discover that the UHI effect is negative, is ludicrous.
How about a city in a fast-growing region like Phoenix – where along with increasing population, more roads and cars, what once was desert now has lawns, trees, swimming pools, golf courses, parks?
Would the rate of change in temperatures be the same sign and/or degree as an increase in Palo Alto’s population increase? In comparing those two cities, can you predict changes in temperature merely as a function of population increase?
Phoenix is well know for it’s UHI effect. The increase is mostly night time temperatures:
“a phenomenon that makes the Phoenix nighttime low temperatures 10 to 15 degrees Fahrenheit warmer than temperatures in rural areas. ”
http://schoolofsustainability.asu.edu/news/greentalk/urban-heat-island-affects-phoenix-all-year-round
I would guess Palo Alto’s UHI would largely be from all build in that area- not just the town itself, I would guess it’s much less than Phoenix and different- a more of balance of daytime and nite time warming from UHI.
gbaikie,
Josh-ua was thinking about all those air conditoners cooling off the Phoenix area, that weren’t there before they built a big freaking Urban Heat Island. They call it that, for a reason.
gbaikie –
Thanks for the interesting link.
Having camped often in cold desert nights, it occurred to me that the lack of nighttime cooling would add to the UHI of a place like Phoenix.
The link is interesting – not the least because it describes how different the effect is in central urban areas than it is in more residential urban areas like those that surround vast amounts of landscape outside of central Phoenix.
It also fully supports the notion that it would be very difficult to extrapolate changes in UHI as a function of population increase – too many uncontrolled variables there.
You are a moron, Josh-ua. Let’s pretend that suburban swimming pools and landscaping negate the effects of big freaking Urban Heat Islands. Never mind the houses, shopping centers, office buildings, parking lots, streets and all the other parts of the infrastructure that supports the lawns and swimming pools. You are really a trip.
Check out all that greenery in Phoenix:
http://maps.google.com/maps?q=google+maps+phoenix&oe=utf-8&rls=org.mozilla:en-US:official&client=firefox-a&um=1&ie=UTF-8&hq=&hnear=0x872b12ed50a179cb:0x8c69c7f8354a1bac,Phoenix,+AZ&gl=us&ei=b1aiTqqlL6ebiAKyyfyJAQ&sa=X&oi=geocode_result&ct=title&resnum=1&ved=0CBsQ8gEwAA
You obviously havent looked at modis data.
If you look at grid cells where modis500 says there are no built areas
guess what?
the population there is non existent.
“The ‘built environment’ includes all non-vegetative, human-constructed elements, such as buildings, roads, runways, etc. (i.e. a mix of human-made surfaces and materials), and ‘dominated’ implies coverage greater than or equal to 50 percent of a given landscape unit (here, the pixel). Pixels that are predominantly vegetated (e.g. a park) are not considered urban, even though in terms of land use, they may function as urban space.”
They disagree with you Mosher.
http://www.sage.wisc.edu/people/schneider/research/data_readme.html
Bruce I’ve said exactly the same thing elsewhere. parks and golf courses will show as un built. Its one of the tests I ran on Modis
“We defined a site as “very rural” if the MOD500 map showed no urban regions within one tenth of a degree in latitude or longitude of the site.”
7 miles? Very rural if an Urban site is 7 miles away?
Ha ha ha ha ha!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
No wonder they could find 16,000+
Droll.
bruce. you failed to read my comments as usual
http://judithcurry.com/2011/10/20/berkeley-surface-temperatures-released/#comment-125123
Bruce.
.1 degrees is 11.12 km at the equator
As I pointed out this changes with Latitude.
A UHI “bubble” is a function of many things, further there are many kinds of bubbles. Not every urban location has them. They are not uniform.
They are not present every day or every season.
So here is a question. If you have two sensors.
A. located 10km away from any human structure
B. located in the midst of human structures
Which will be warmer?
Do you expect a difference?
How large?
Are you willing to change your mind in the face of a test?
Steve you say: “located 10km away from any human structure”
The MODIS definition of urban is 50% of a pixel or more and a minimum of 1 square kilometer in size.
So you could have an interstate highway running 5 feet away from the MMTS sensor as long as the rest of the pixel is not urban this would be classified as “very rural”.
As Anthony Watt has shown over and over again, you need real metadata – not inferred metadata.
Steve McIntyre’s latest post strongly implies BEST got UHI wrong.
“I suspect all BEST have done is pick out a low population density subset of stations which have a higher rate of warming over the time period than the total population and viola, these stations report a higher warming rate than the total population. I’m not surprised they got the results they did.”
Since you have not looked at Modis500 data and I have I can tell you that you are flatly wrong.
Modis500 data identifies “built” pixels. That is 500 meter squares that have human built structures. Next BEST looked .1degrees around each pixel and determined that there were no “built” areas within that radius. Thats 11km at the equator and less as you go toward the poles.
You can also do a correlation between Modis pixels that are not “built” areas and population density. In almost all cases there isnt any population in these areas. Where there is some population it is rural and not urban. you can also cross check with land use data. land use in these areas is non urban:
croplands, forest, sparse vegetation, grasslands. I know, I checked using land use datasets for the entire globe. you can also cross check using impervious surface data. Again, Modis ‘non built’ pixels are also 0% impervious surface. You can do yet another cross check and look at nightlights. Again, you’ll find these locations are un lit.
While there are some exceptional cases, in general you don’t know what you are talking about.
I look forward to seeing which sites are which and checking them.
But 6.5 – 7 miles is not outside an urban heat bubble and does not preclude A/C units from being 5 feet away from a station.
Bruce,
MMMMMM. maybe more sewage plants and such???
a 7 meter per second wind is strong enough to eliminate UHI.
6,5-7 miles is plenty of distance for cities with no tall buildings.
Anthony proved that microsite doesnt matter
next.
I thought 7m/sec was the minimum wind speed necessary to eliminate just the UHI within the MMTS enclosure itself.
And again, using the MODIS500 definitions, to be urban it has to be more than 1 square kilometer and have 50% or more of a pixel filled with structures.
That leaves many,many scenarios that could contaminate a site.
Real metadata is better than inferred metadata.
Just a reminder. If a pixel is 49% covered by pavement, and the rest by grass Modis500 does not classify it as built/urban.
And it has to be a certain size.
“Finally, we also define a minimum mapping unit: urban areas are contiguous patches of built-up land greater than 1 km 2”
Again … no wonder they could find so many non-urban stations.
That’s why you go to other satellite products and check. You are welcome
to go get Modis and look at the earth 500 meters at a time. Its a painful slow process, trust me. However, you can do things like look up the latitude and longitude of golf courses and see that Modis gets them right.
You can look for small patches of grassland and see that it gets them right.
If a pixel if 49% concrete or asphalt, then a cross check of Modis and ISA picks that out usually, and if its lit a cross check with Nighlights gets that as well. You can also cross check with land use data from other platforms. There are about 8 different datasets.
If you like you can do the work yourself
The package is called “metadata” its written in R. I usually answer users questions. In your case, I’ll make an execption
What you are suggesting is the lazy man’s method of metadata inference. It is not real metadata.
As Steve McIntyre says in his latest post:
““CRS urban bad” had a corresponding trend of 0.42 deg C/decade.”
And 27% of sites are Urban.
I’ll quote his rural findings when I’m sure about the possible typo.
Once more, Mosher, you are out to lunch. Areas on a Modis map are designated as urban if they more than 50% built and built contigously that way for more than 1km square. Whether Modis can recognize 0% built area is completely irrelevant, because everything from 0% to 50% is classified as rural. So the areas that BEST uses as rural can be 50% built. Or they can be 90% built if they are smaller than 1km square. Or they can have multiple areas that are 90% built, but smaller than 1km square, seperated by areas that are only 45% built. All are classified as rural.
Here, have a good cry…. I mean read.
http://www.sage.wisc.edu/people/schneider/research/schneider_erl2009.pdf
And the “very rural” designation are only used by BEST. They have nothing to do with build. They only indicate distance from an area classified as urban.
Tilo, You really do need to write to the PI download the data and do some work. While it is possible for a pixel to be “half built” and count as rural, when you actually go to sites that classify as rural you will see that this possibility you worry about is very rare. here is what you should do.
1. get modis
2. get ISA
3. get Grump
4. get Nightlights
5. get the landcover data
6. get the global pop data.
Then you locate all the sites in these datasets.
then classify
Then build KML
Then talk to me when yo u’ve viewed 1000s
Did BEST do any of that? Their paper only says MODIS500.
And they came up with a designation MODIS500 never mentions – “very rural”.
How did they find 16,000+ “very rural” stations?
I don’t know if this has been commented on, yet, but have you guys seen BEST’s video of the globe with anomalies running up and down the side?
Go here, http://www.berkeleyearth.org/movies.php
Look at the coverage they are claiming and the year they are claiming the coverage. BEST simply made up temps. There is absolutely no way they can claim 100% coverage for Africa and South America in 1891. They also show full Antarctica coverage in 1956. That’s crap! They don’t have full Antarctic coverage now!
You can go here, http://suyts.wordpress.com/2011/10/21/is-that-the-best-they-can-do/ for a couple of screen shots.
I noticed the same about Greenland in 1850s. I just put it to being more of an artists representation than anything else. The red hot colouring is also overly dramatic.
suyts:
BEST uses a form of statistical extrapolation and interpolation called kriging to cover areas where they have no thermometers. It’s what GISS does, except that GISS does a more standard type of extrapolation and interpolation.
But it seems to me that kriging probably suffers the same problems as the standard extrapolation. For example, in Greenland and Antarctica you have a lot of shore stations, but almost no inland stations. But when the ice near a shore station melts, they are suddenly exposed to a local warming source from the open water. But when you get away from the shore that warming source doesn’t exist. That mean that the extrapolation, or kriging, from a place with a special warming source to one without one is invalid. But, nevertheless, it’s what they do. That is how GISS gets it’s Arctic coverage. And of course it is invalid – and always invalid on the warm side.
Not sure if it changes the global temp though. I mean it would be consistently warmer, it not as if they are dropping stations and replacing them with kriging (is it?). And if kriging gives you a warmer temp, then when you add a real station temp should go down. Over time the number of stations is going up, normally. However, in the 70s or 80s it looks like there was a sudden drop in the station numbers.
Kermit:
It makes a difference because global temperature comes from gridded cells of the globe. The idea is to have every area equally represented in the average. In other words, you don’t want a place with 10 thermometers to have ten times the influence as an equal area with only one. So, if you are using shore stations warmed by nearby water to produce gridded cells inland to ice covered areas where there are no thermometers, then those cell values are invalid, and too warm. Over the years the ice has retreated from those shore stations earlier. So that local effect is spread to all the extrapolated areas earlier. The result is that you get a divergence between GISS, and probably BEST on the one hand, and HadCrut – which does not extrapolate, on the other.
And yes, if you start getting real thermometers in those areas, the effect will go away.
More explanation and examples can be found here:
http://reallyrealclimate.blogspot.com/2010/01/giss-temperature-record-divergence.html
Yes, I think I understood your point. I just didn’t explain mine well. Stations A, B and C are on the coast of Greenland, D is interpolated on top of the ice sheet. This interpolation is shows D is warm but in fact its on top of an ice sheet so its cold. My first point is that this would alway be warmer so until they use real data it will not make a difference since it is artificially warmer from the beginning. Okay, maybe if it gets hotter over time then there would be some magnification. My other point is that once they the use a real station, then that real temp would have dropped the global average, so . I also think they would be able to visit the D station to validate the kriging. Greenland would be a current example but many frontiers in the past would have gone through this.
Tilo,
I agree with what you are stating EXCEPT GISS has had at least one embarassment when their Arctic station came in cooler than expected and it affected the global average the wrong way!! 8>)
Tilo, thanks, I understand interpolating, but in 1890s? The gridded areas must be continent size.
Kriging is valid in some applications and ventures such as mining. In this particular instance, its tantamount to just making things up.
Fred: “I’m not sure this would be particularly relevant to the BEST analysis unless it turns out first that the correlation reflects causation”
LOL. Isn’t that what the structure of the BEST UHI test is assuming?
Fred: “and second that there has been a far greater percentage population increase in the BEST “very rural” areas than in more populous areas ”
More LOL. Yes, you would think that the BEST designations of “very rural” and “not very rural” would actually have something to do with how population dense an area is. Unfortuantely the designation has nothing to do with that. The designation is only reflective of how far a thermometer is from and area that is designated on the map as urban. In reality, therefore, you could have many “very rural” areas that have greater build and population density than “not very rural” areas.
Come on Fred, let’s face the facts here; the test was designed to quantize how human construction effects temperature, and it doesn’t come close to doing that.
I believe you’ve misinterpreted the BEST data as well as the points I made above. Both are available for others to review. What I perceive, Tilo, is a fierce determination on your part not to accept the finding that there is no strong UHI effect in the land data. You have correctly raised questions about possible misclassification, but as I pointed out earlier, it is unlikely that misclassification would have been so severe as to not merely overturn the BEST finding of a “reverse UHI effect”, but further convert it into a strong positive UHI effect. However, if you want to go through all the data and test out a series of reclassifications, you should do that and report how much change would be needed to reach the conclusions you favor. Otherwise, the discussions are likely to grow repetitive.
BEST knew their result disagreed with the previous one and it shows a counterintuitive negative result. They should have been the ones to reconcile or offer an explanation. What they said we obviously think our result is right and the old one is wrong but they don’t have any explanation why.
I don’t think they said that, but if you find that claim, could you quote the passage? They did suggest possible explanations, but didn’t claim they were necessarily correct..
I think it was on the Keenan post over at BishopHill. I’ll try and find the exact passage and link.
In the Keenan post, Muller said that they have more confidence in their analysis than in the analyses that contradicted theirs in some aspects. Presumably, the reasons lie in what they consider to be improved methodology – improvements that were (prior to the “inconvenient” results) endorsed by “skeptics.”
Ironic, that.
Fred: “Tilo, is a fierce determination on your part not to accept the finding that there is no strong UHI effect in the land data.”
There was no such finding Fred. The test as designed by BEST has zero ability to determine UHI effect.
Fred: “but as I pointed out earlier, it is unlikely that misclassification would have been so severe as to not merely overturn the BEST finding of a “reverse UHI effect”, ”
First of all the classification of what is Urban and what is Rural is arbitray.
But it’s not the classification that counts. What counts is the amount of build that is happening around thermometers and how close it is to the thermometers. BEST’s test doesn’t deal with that at all. Obviously there was more build in the urban areas at one time. But as time moved on, the possibility that more build happened in areas classified as rural than in areas classified as urban became very real. Spencers results confirm that. BESTS results confirm that, and the thought experiment that I did yesterday on this thread confirms that. Hope that doesn’t ruin a UHI song that may have been brewing in your soul.
Well, I think you’re illustrating my point, but I’ll leave it to others to judge. I don’t think we have the final word on UHI effects, but the evidence does seem to be gathering that they are unlikely to be strongly positive and may even be negative, although the latter is still subject to much further analysis. I’ll leave it at that.
Sorry, Fred, but Tilo Reber’s point is important, and you are either missing it entirely or trying to talk past it. Somehow, the “identified rural” spots in the BEST data showed slightly greater average warming than the sample as a whole. Until that is investigated, and we have lots of evidence about what distinguishes “warming” stations from “non-warming” stations, the “UHI monster” remains alive and well — and large.
Consider Steve McIntyre’s concrete plant above. Until all such cases are thoroughly studied, or else a large random sample of stations that display “warming” vs. “non-warming” is studied, it is not possible to say how much of the 20th century observed warming is due to UHI.
Unless I have missed it, no one has suggested that a large random sample be studied. Anthony Watts has tried to investigate them all, which has yielded a partial sample with a large convenience bias. Various posters on the web have selected samples that they like — I mean actual site visits, on the ground, in situ, “up close and personal”, etc. The indirect approaches, population density and nighttime lights, are too indirect.
MattStat – I have no problem with investigating. Above, I stated why I thought misclassification is possible but very unlikely to convert a slight “reverse UHI efffet” into a strong positive UHI effect. Previous evidence also indicates any UHI effect would be small. Completely unrelated to these measurements, the ocean/land temperature relationships are of the magnitude expected for the observed land warming rate, and so a large UHI spurious warming would be difficult to reconcile with the ocean data, although a small UHI effect would still be compatible. In that regard, I agree with Dr. Curry that the forthcoming ocean data analysis will be informative, including light it throws on this issue.
Does the difference reach statistical significance?
Joshua,
the Berkeley report is useless for determining UHI as all they did was compare stations similar to others previous work. Comparing stations that may not even be in the same climate area will not show much. Without having stations baselined where the UHI has been determined by setting up a station outside the UHI dome and actually MEASURING, the relative measurements between stations only tells you the difference in trends between stations. This trend may be UHI in part, wholly, or a negative number reducing the actual warming trend. The trends, due to microclimates, can be different so again this is not a definitive method of determining UHI.
http://www.informath.org/apprise/a5700/b1104.pdf
is a copy of the Berkeley UHI paper. Mr. Keenan has the other three papers hosted at his site also and the links in his post at WUWT.
They actually show larger warming trends for Rural sites than Urban but not at the magnitude of Dr. Spencer’s work. While I agree with Fred that some cities could actually have a negative UHI due to changes around the observation site, I think onsite validation would show that there are few like that. Basically they have added very little to the UHI question other than confirming Dr. Spencer’s work showing that rural is generally higher than urban and the GISS practice of averaging rural and urban not quite as insane as thought.
One other issue I have been wondering about is whether the flat temps in the last 10 years are at least partially due to the UHI of some cities’ maxing or the moves to airports limiting it!!
Oops. Steve McIntyre’s example was a nickel smelting plant.
Tilo –
Do you have any evidence to support a conclusion in that regard? I can think of a number of cities, say Shanghai, were the buildup was exponentially greater than the development in vast regions of rural China.(which have seen significant drops in population, relatively speaking)..
Joshua: “Do you have any evidence to support a conclusion in that regard?”
There are Spencer’s results. And the negative BEST UAH results suggest the same thing. But then it’s really up to BEST to show that urban build was greater than rural build isn’t it. In fact, for what they are trying to measure with their method they need to show that there was a large urban build and zero rural build. But as any fool can see, the names “urban” and “rural” don’t mean “build” and “no build”. But the design of their UHI test implies that urban means build and rural means no build. So their test doesn’t demonstrate a thing about the magnitude of UAH.
“I can think of a number of cities, say Shanghai, were the buildup was exponentially greater than the development in vast regions of rural China”
First of all, your cherry picked thinking is irrelevant to the argument. Second of all, Shanghai has been urban for a very long time. A thermometer in Shanghai is likely to see old buildings destroyed and new buildings put in place, but it’s unlikely to change from being in an area of no build to an area of build. And Shanghi may also grow at the edges, but that growth would be at a distance from the thermometer. The vast regions of rural China are also irrelevant. The thermometers are not on the farms and in the wilderness. They are mostly in areas where build can and does happen.
They admit that “the influence in recent decades of oceanic temperature cycles has been unappreciated and may explain most, if not all, of the global warming that has taken place” stating the possibility that the “human component of global warming may be somewhat overestimated.”
You heard it here first back in May 2008:
http://climaterealists.com/index.php?id=1302&linkbox=true&position=9
“The Real Link Between Solar Energy Ocean Cycles and Global Temperature”.
Stephen – Please state where you extracted that quotation from in the BEST preprints. – which paper and what page. I’d like to read the entire context to see whether it conveys the same impression as the section you excerpted.
Stephen Wilde
You are seconded by non other than Michael E. Mann himself:
Using a 1400 year climate model calculation, we are able to simulate the observed pattern and amplitude of the AMO. The results imply the AMO is a genuine quasi-periodic cycle of internal climate variability persisting for many centuries, and is related to variability in the oceanic thermohaline circulation (THC).
….
This effect needs to be taken into account in producing more realistic predictions of future climate change.
http://bit.ly/nfQr92
I relied on the gwp report but on checking I see it was set out slightly differently on page 12 from here:
http://www.berkeleyearth.org/Resources/Berkeley_Earth_Decadal_Variations
I can’t seem to quote the right bit here because of a formatting problem but you can see for yourself.
In the larger context, the statement is of the “anything is possible” type:
“If the long-term AMO changes have been driven by greenhouse gases then the AMO region may serve as a positive feedback that amplifies the effect of greenhouse gas forcing over land. On the other hand some of the long-term change in the AMO could be driven by natural variability,e.g. fluctuations in thermohaline flow. In that case the human component of global warming may be somewhat overestimated.”
It’s interesting that there were some clauses left out of Stephen’s quote.
I’ll highlight them a few of them:
Interesting how some context didn’t make it across into the quotations in certain realms of the “skeptical” blogosphere. Precisely the same truncation has been repeated numerous times over at WUWT.
Even elipses were not used in some cases, as in Stephen’s “quotation” above.
Still, they do seem to be giving approximate equivalence between a natural and anthropogenic scenario and do not try to choose between the two.
The possibility of GHG changes from humans driving the AMO changes on the timescale observed is very unlikely given the huge thermal capacity of the oceans as compared to the atmosphere.
Which brings it all back to my 2008 article.
Introducing the THC as a potential factor opens a can of worms too because the THC takes 1000 years or so which means that a part of current SSTs could be attributable to MWP warming.
Overall Berkeley gives a lot of support to sceptical positions.
Fred,
Yes, if a negative UHI effect is possible, then “anything is possible”.
Dr. Curry,
The BEST press release states:
“Four scientific papers setting out these conditions have been submitted for peer review and will form part of the literature for the next IPCC report on Climate Change.”
How is it that BEST can know and state that these papers, as yet peer reviewed, will be part of the IPCC report?
Seems to imply that the peer review and acceptance into the IPCC report are already decided.
Will you renounce this statement and join Anthony Watts in calling for an open peer review process to ensure it is not pal review?
Will you recommend that Doug Keenan be selected as a reviewer?
This statement by BEST demonstrates their intentions are not scientifically, but policy-oriented. is this your motivation as well?
KSD,
“How is it that BEST can know and state that these papers, as yet peer reviewed, will be part of the IPCC report?”
With such a high powered team including a Nobel prize winner, I think it is probably fairly safe to say that the report will both be peer reviewed and make it into the next IPCC report.
Will Judith nominate Doug Keenan as a reviewer? Not a chance. She won’t nominate Anthony Watts, Andrew Montford, Christopher Walter Monckton, 3rd Viscount Monckton of Brenchley, or Snr James Inhofe either. At least IMO !
It looks like you’re scuppered, mate!
And I’m sure, tt, that you wonder why so many doubt the credibility of climate “science”.
If you ever ask yourself that question, just send yourself a link to your reply above. That is why the credibility of climate “scientists” is now on a par with that of Bernie Madoff.
There’s an interesting element to these discussions that may have escaped adequate attention. The Wickham et al paper looked at 16,132 “very rural” stations and found their warming rate slightly greater than the rate for the entire population of 39,028 stations. This finding has been criticized on the grounds that some of those 16,132 stations may have been misclassified, and should have been put in the category of “not very rural” or even urbanized stations. The question arises – how much misclassification could convert the Wickham “reverse UHI effect” into a strong positive UHI effect?
An interesting thought experiment would be to consider half of the 16,132 – i.e. 8,066 stations – to have really been urban stations, a 50% error rate. What would that have done to the Wickham findings?
Well, for those stations to have created the “reverse UHI effect”, they would have had to warm not at the same rate as the remaining urbanized stations within the 39,028 total, but significantly faster, if the remaining truly rural 8,066 were actually warming slower via a real UHI effect. If the misclassified group had been warming at the same rate as the towns and cities, their inclusion in the 16,132 could not have created a spurious “reverse UHI effect”.
Is it possible the study inadvertently put 8,066 out of 16,132 stations in the wrong group, and these 8,066 somehow happened to be exerting a “hyper-UHI effect”?
It’s possible. Maybe stations that fool Modis into thinking they are still very rural warm faster than anywhere else, and there are a very large number of them? Is it likely?
I’m not absolutely convinced there is a true “reverse UHI effect”, but the ultimate explanation for that finding is unlikely in my view to support the existence of a strong positive effect.
Okay Fred, I’m having to tell you the same things all over again and I’m getting tired of it. Try to remember this time.
A. Classification is irrelevant. Build is relevant.
B. Classification of “very rural” as opposed to “not very rural” have nothing to do with build or population. They are strictly given on the basis of proximity to a designated urban area on a map.
Fred: “Maybe stations that fool Modis into thinking they are still very rural warm faster than anywhere else”
Modis doesn’t have a designation of “very rural”. When build density and area get beyond a certain point, then that area gets an urban designation. Everything else is rural. A city of 10,000 would be rural. If that city went from 10,000 to 20,000 it would still be rural. Areas between cities that have extensive build, thermometers, and rapid development are still rural for Modis.
The Modis requirement for urban is that the land is covered by greater than 50% build, plus, that such areas must be contiguous and larger than 1km square. I live in the Denver burbs. My area would be considered as rural because the yards cover more area than the houses and streets. But it’s city from here all the way to the center of Denver. And there are the occasional undeveloped strips just large enough to break the “contiguous” requirement.
http://www.sage.wisc.edu/people/schneider/research/schneider_erl2009.pdf
Fred’s logic is that they couldn’t be that far wrong, so they must be on to something. It’s like the GCM’s. Most of them are so far off that they make the ones that are not quite as far off look good in comparison.
I believe that Tilo, Don, Berenyi (see below) and possibly one or two otjhers have missed the point, which means I probably didn’t explain it well.
The Wickham paper suggests a “reverse UHI effect” – i.e., rural areas warm faster than more built-up areas. This requires an explanation, and ultimately a proper explanation may show the conclusion to be wrong. However, we have to ask, “can the reverse UHI effect be explained by postulating that locations classified as very rural were in fact contaminated with some (or even a large number) of locations that, for whatever reason, were not behaving “rurally” but were instead exhibiting the behavior of built up areas?”
Well, the simple answer seems to be, no – that can’t explain it If there is indeed an actual tendency of truly rural stations to warm more slowly than those that are built up (or have become more built up over time), and some of the latter are erroneously mixed in with the rural data, the result would be to dilute the evidence that rural areas warm more slowly, but not to reverse it into making them appear to warm faster. The only way to make them appear to warm faster is to contaminate their data with locations exhibiting a “hyper-UHI effect” – i.e., stations that warm faster than ordinary urban areas that warm because of their tall buildings, asphalt, industry, and so on. Given the causes understood to underlie the warmer temperatures of urbanized areas, or even modestly developed areas, it is perhaps not impossible for this “hyper-UHI” contamination to have occurred, but it is certainly implausible (or at least I have not seen any mechanism postulated for such a coincidence of “hyper-UHI” with the potential to be mistakenly considered “very rural”.)
One other aspect of this should be considered. If, in fact, there is some undetected tendency of areas generally thought of as rural to include very fast warming examples, this would be consistent with the notion that we are underestimating rather than overestimating the rate of land warming, because it is the apparently rural areas that are underrepresented in the station data, but which constitute the much larger fraction of global land surface..
In developing these points, I haven’t concluded that the reverse UHI effect is real – in fact, I’m skeptical. What I have concluded is that it hasn’t yet been explained, and that the profferred explanations don’t work. There are now sufficient data, including these latest, to indicate that any real UHI effect is likely to be minimal, but it is still likely in my view to operate in the forward direction.
Also a cooling effect would be bad for proving AGW directly. This is already hard to prove, some would even say impossible. It would mean that the AGW footprint is worse for rural than urban but urban is where there are more humans. Surely urban areas give off more CO2 and other greenhouse gases than rural. I know co2 is a well mixed gas but still cooling means a disagreement with AGW theory. Perhaps your right there is some contamination but that is still a big problem. This should be more a problematic finding for the AGW believers than the ‘skeptics’.
Kermit – I’m not sure I understand your point. The data show that both urban and rural areas have been warming, and the question is not whether either is cooling but rather which warms faster. There is also no question that urban areas have higher temperatures than rural areas, but whether they warm faster or slower is also a separate question from whether they are warmer.
“We observe the opposite of an urban heating effect over the period 1950 to 2010, with a slope of -0.19 ± 0.19 °C/100yr”
-0.19 degrees is a relative cooling, no?
I’ve also read they have different results for different regions. Certainly a cold northern urban area could be more susceptible to a UHI effect than a hot southern one but I haven’t heard they offered any such explanation.
The effect is faster than average warming in the very rural group, compared with the overall average for all stations, whose warming trend is not very different from that reported previously by others. Within each group, individual stations may have warmed or cooled depending on location, but that is already well known.
Given the causes understood to underlie the warmer temperatures of urbanized areas, or even modestly developed areas, it is perhaps not impossible for this “hyper-UHI” contamination to have occurred, but it is certainly implausible (or at least I have not seen any mechanism postulated for such a coincidence of “hyper-UHI” with the potential to be mistakenly considered “very rural”.)
I repeat what I wrote above. This question can’t be answered without inspection of the sites. Since they can not all be inspected (Watts has tried), there has to be a large random sample. This, by the way, is the common argument for sampling: a full census is impossible. And a random sample (or stratified random sample, with strata clearly defined in advance) is necessary to avoid (most likely) conscious and unconscious biases.
Looks like Wickham, confirmed by BEST, have overturned the old urban legend about UHI:
http://en.wikipedia.org/wiki/Urban_heat_island
This is very quaint., in light of what we know now. They actually thought UHI melted arctic ice packs:
“Recent qualitative speculations indicate that urban thermal plumes may contribute to variation in wind patterns that may influence the melting of arctic ice packs and thereby the cycle of ocean current.[4]”
Thanks for helping us understand this, Fred.
The BEST UHI test has nothing to say about UHI. Maybe you should read some of the discussion.
Tilo,
Maybe you should read some more of my posts, before you jump to any dumb conclusions.
Got it.
Thank you.
Dr. Curry, the interesting part is that you used the NCDC data, processed it similarly and got a similar result??? If you HADN’T gotten a similar result you would really have a problem unless you actually fixed the adjustments!
Your study shows the UHI of the rural stations is higher than the Urban. How did you determine the amplitude of the Urban UHI??? Without this piece of information how do you know how much to adjust the records? Without this information and adjustment your new spiffed up record is just as pointless as the current one. I believe a certain Mr. Keenan found the same issue and has a post over at WUWT covering that and other issues he has with the reports. Probably already mentioned upthread, but, I don’t have time to read the whole dang thing!!
Would someone point me to the answer to the UHI question please??
Another issue is that preliminary testing with raw data was only done on the US stations. The US has a generally lower rate of warming than the rest of the Northern Hemisphere. It also probably has a better set of stations than the ROW. Could you explain how doing testing on only the US subset of data is meaningful for the whole globe with the certainty that there is wide variance in quality and issues from country to country??
Again, if this has been addressed please point me.
Thank you.
PS: Here is an interesting old thread from Lucia’s Blackboard. The primary post is OK , but, the really interesting part is the discussion between MikeC and Steven Mosher. Mosher was a real animal on this one. Nick Stokes barely needed to comment. Were MikeC’s issues addressed in your adjustments? In fact, were any of the adjustments for homogenization or other verified and improved??
http://rankexploits.com/musings/2010/a-cooling-bias-due-to-mmts/
OK, I might as well get this question in also. Does your processing address this gentleman’s concerns?
http://tallbloke.wordpress.com/2011/10/22/signal-processing-and-normal-meteorological-maths/
kuhnkat:
This is not Judith’s report. I believe she served as advisor for some of the issues. Don’t ask her to defend the results. The UHI issue has been well discussed above. You will have to break down and read some of it if you want the answer.
“Would someone point me to the answer to the UHI question please??”
Check my posts and Geoff’s last post.
Don’t ask her to defend the results.
Look. Judith can’t have it both ways. Either she puts her name to the report or she doesn’t. She either defends the results or she doesn’t.
Here is what she said:
“The fact that my name appears as second author on some of these papers is attributable to my last name starting with the letter “C”. The group has taken a “team” approach to authorship on this set of papers. My contribution to these papers has been in the writing stage and suggesting analyses. I have not had “hands on” the data, one of the reasons being that I do not have funding to do any analysis.”
I’ll just previously repeat what I said previously. It doesn’t matter whether they’re the first or the last in the list. All potential authors have a choice. If they don’t agree with the contents they can opt out completely.
Nope, defense by who did the work.
======
I’ll just previously repeat what I said previously. It doesn’t matter whether they’re the first or the last in the list. All potential authors have a choice. If they don’t agree with the contents they can opt out completely.
This has been debated forever. It is not possible for each author to independently check all of the work of all the other authors. It would be nice, but it isn’t possible. The usual standard is that all of the intellectual content of the paper can be explained by at least one of the authors. That standard at least is achievable, although there may be some dispute about what exactly is intellectual content. Who, for example, verifies that the microprocessor chips adequately adhere to the IEEE floating point standard, or that the compiler has never produced errors on the code used in the particular (novel) application? Examples of every sort of possible error have been published amidst much discussion, and no achievable standard of “joint and several” responsibility has been found.
Urbanization results in global cooling trend comes from the mystics of Gaia, not science.
I really think this classification and misclassification argument is dancing on the head of a pin. That we have a supposedly high quality data set of temperature stations and nobody actualy knows their environment is astonishing.
Unfortunately Watts’ survey is not, I believe, sampled systematically. Until each station is surveyed systemically, I really do not think that one can draw conclusions. The temperature bias should be measured.
Scientific method:
Raise a hypothesis
Design experiment
Perform experiment to get as high quality data as possible that will allow hypothesis to be tested.
Analyse results.
I am sorry but I find the whole idea that massive number crunching can enable one to draw conclusions, when nobody knows what the data means, absurd.
“design experiment” ??
You mean like doubling the amount of CO2 in the atmosphere and observing what happens?
Don,
You can’t assume there is something wrong with the data just because it produces a surprising result. I can understand Fred thinking there probably isn’t a negative UHI, but there could be. It could be something counter-intuitive. Possibly. Or, it could be there is a an error somewhere but it can’t be large. Otherwise, the good agreement with previous surface temperature records wouldn’t be as it is.
Unless the whole thing is a complete fabrication that is! But if that were the case why introduce a negative UHI ?
tempt,
I am sure you are just as open to counter-intuitive results arrived at by the skeptics. Get real. Everybody knows about the UHI effect. It’s hotter in the city than it is in the country. Period. The data is screwed up, or the analysis is faulty. Either way, it makes all of the BEST results suspect.
Don,
The question on the UHI is whether it is correctly accounted for in the temperature record. Not whether it exists. Yes, it probably does exist
http://en.wikipedia.org/wiki/Urban_heat_island
But it would still be possible for urban sites to warm less than rural sites over a period of time.
I think you need to wait and see exactly how the Berkeley group tackled the problem and not dismiss the study out of hand just because you don’t like their conclusions.
tempt,
You know that the UHi effect exists, what are you talking probably. And you know the magnitude of the effect is obvious and large. Certainly much larger than the supposed increase in average world temperature since 1950. It is not possible that rural sites warmed faster than urban sites, as the world’s population grew from 2.5 billion in 1950, to 7 billion today, and at the same time the population migrated from rural areas to cities. Where is the science?
I used to live near chicago and weather reports contained the terms warmer near the lake and cooler near the lake with near seasonal correlation of 100%.
In winter Chicago is warmer thean the rural areas and in summer is cooler.
So much for your period
bobby d,
You are a character. Remove the lake from the scenario, or remove the big freaking UHI of Chicago from the scenario, and rethink your dumb comment.
Remove the lake from the scenario, or remove the big freaking UHI of Chicago from the scenario, and rethink your dumb comment.
Yes, remove every data point that he tells you to. Then submit it to a journal with each of those points annotated with the note “removed at Don Montford’s request”.
A large proportion of the population lives within 50 miles of the coast
http://oceanservice.noaa.gov/facts/population.html
Temp,
other than differencing stations this study did NOTHING to determine UHI. In other words, the UHI for the Urban stations could be .5c/c and the UHI for the rural .7c/c OR the urban could be -.7c/c and the rural -.5c/c. THEY WOULD NOT KNOW!!! It is not likely to be lthat large, but, until someone goes out and actually runs a high quality study on various types of sites NO ONE KNOWS.
Bob Droege,
unfortunately your story shows us little. The UHI could STILL be affecting those gross temps positively or negatively by .5c/c and people wouldn’t notice for decades, if then.
Increasing vegetation cover in Urban areas is claimed to reduce the UHI effect. I’m not sure if Australian cities are typical but, looking at old photos, it does seem quite apparent that there is much more tree cover now, than there was in previous decades, in many urban areas.
It wouldn’t mean that the UHI effect would disappear altogether, but it could well be less now, than it used to be, in many cities. Isn’t that what a negative UHI result would really mean?
PS It is also worth just taking a look at the proportion of land area which can be considered ‘urban’. Its not very high, I’d say less than 1%. Anyone have an exact figure?
If so, then neglecting the UHI effect altogether, or even doubling it, isn’t going to make much difference to the final answer.
tempterrain, I’m not sure what makes you think that is “what a negative UHI result would really mean.” Look at figure five of the UHI paper. It clearly shows rural sites are warmer on average than sites in general. How do you think a lessening of UHI would cause this?
Temp,
“It is also worth just taking a look at the proportion of land area which can be considered ‘urban’. Its not very high, I’d say less than 1%. Anyone have an exact figure?”
Not off hand, but, I believe that we can all agee that even if we inlude grass huts the area of human habitation (excluding agriculture areas) is small compared to land area of the earth. Now, we only measure temps in or adjacent to that small area and mostly affected by that small area. Maybe an issue since we have virtually NO pristine temperature records as a baseline to determine the extent of warming of any type much less the amount that should be assigned to UHI, GHG’s, or any other reason??
Personally I think this is part of the reason that we see the land temps rising faster than the ocean temps. We aren’t measuring the surface temp, we are measuring the inhabited area contribution. We would see a similar problem if we measured CO2 in, or next to, inhabited areas!!!
One additional thing Temp,
this could partially account for the current flat temps if a number of areas have stabilized their UHI or actually lost some due to more efficiency and trees/foliage/green areas added to inhabited areas…
Brandon,
You write ” Look at figure five of the UHI paper. It clearly shows rural sites are warmer on average than sites in general.”
You mean figure 5 of this paper?
http://berkeleyearth.org/Resources/Berkeley_Earth_UHI
If so, no it doesn’t. It shows that if you remove the ‘rural sites’ from ‘all sites’ the temperature anomalies , relative to a 1950 -1980 average , are very slightly cooler than with all sites.
The paper then goes on to say
We observe the opposite of an urban heating effect over the period 1950 to 2010, with a slope of -0.19 ± 0.19 °C/100yr. This is not statistically consistent with prior estimates, but it does verify that the effect is very small, and almost insignificant on the scale of the observed warming (1.9 ± 0.1 °C/100yr since 1950 in the land average from figure 5A).
tempterrain, your response here is rather asinine. You didn’t respond to what I asked you. Instead, you only disagreed with the premise of my question. This would be okay, except you didn’t indicate how your disagreement actually changed my question. Your disagreement was:
Yes, I referred to the stations being “warmer” rather than saying their anomalies were warmer. Does this change anything? No. Everybody knew we were discussing anomalies, not absolute temperatures. As for your comment about the baseline, changing the baseline wouldn’t change the results, so there’s no benefit in adding that clarification. Ultimately, both of these points clarify what I said, but not in a way which affects anything I said.
On the other hand, you do directly dispute one thing I said. I said Figure 5 shows rural temperatures were warmer. You say it shows what happens if you remove “rural sites” from the rest. In actuality, it does both. I described 5a; you described 5b. Despite the fact both graphs were in Figure 5, you claim what I describe didn’t exist (nevermind the one is just a transformation of the other).
You raised points of clarification which had no relevance, and you gave a deceptive portrayal of the figure being discussed. You did this in what seems to be an attempt to avoid answering a very simple question. As I said, asinine.
“Yes, I referred to the stations being “warmer” rather than saying their anomalies were warmer. Does this change anything?”
Yes it does. But we’ll leave that aside.
What does “their anomalies were warmer” mean? It means that rural stations have warmed faster than urban stations since they were last normalised to a 1950-1980 baseline.
Does that mean that the UHI effect is negative? No, you’d have to know what the UHI effect was in 1950-1980 to be able to say that. All we can say is, that within the confidence levels specified in the paper, ( and that’s not the same thing as ‘for sure’ ) the UHI level now is less than it was then.
Yes, it is a surprising result. IMO it may well turn out to be a statistical quirk. What isn’t a statistical quirk is that the UHI effect is very small.
Why? The BEST team clearly report a statistical difference between their findings on UHI and previous results. And yet their overall agreement with previous results is good. It wouldn’t be, if the UHI effect was anywhere near as large as some claim.
tempterrain, I see you’ve dropped a number of points. I suspect it’s for the better, so I’ll leave them be as well. Unfortunately, you’ve still said nothing to justify your initial claim of what a negative UHI “would really mean.”
Even more unfortunate, what you claim is “[a]ll we can say” is not all we can say by any stretch of the imagination. We can compare current UHI results not just to the baseline used, but also to other periods in the graph. When we do so, we see there is no statistically significant warming in any period. This means we have no period during which we can say UHI caused warming. Without such a period, there couldn’t be a decrease in UHI. A negative trend would have to show urban cooling/rural heating, not just a diminished UHI.
Of course, the father back in time one goes, the larger the uncertainty bars become. Eventually those bars do get wide enough that they could hide warming from UHI. This means we could possibly fit your idea to the data. It would require a fairly large number of assumptions (with no physical basis), and it would be impossible to prove right, but it wouldn’t be completely impossible.
Ultimately, what you said a negative UHI “would really mean” is something we cannot completely rule out, but have no reason to believe.
It may be possible to measure an absolute UHI effect in one of those new cities being built in China if reasonable temperature data were available from the site before any building construction took place. Although, even there the vegetation may have been changed from what is completely natural due to agricultural development.
In nearly all other sites, which are suspected of being influenced by an UHI effect, all that can be measured is a relative change from an older urban setting to a newer one. And of course, it is quite possible, if a little surprising, that the older setting may be more heavily influenced than the newer one. That’s what a negative trend means.
temp,
the question is still how you are magically coming up with the attribution of the trend. Where is the work that shows GHG’s cause x amount of the trend in Stockon, Ca as opposed to UHI and Solar.
The simple fact you and the rest refuse to admit is that there is NO attribution possible with the data available and if y’all wave your arms any harder about it you will probably achieve flight.
No I do not. You design an experiment to test whether the thermometers are measuring the representative temperature of its surroundings. You actually measure if there is a bias. Once you have done this, you can more confident of the data. Once you have data and you understand what it means, you analyse it meaningfully.
When the data does not fit the consensus crowd’s world view, there is something wrong with the data. We know that there is a Urban Heat Island effect. If the data says that it is actually a Rural Heat Island effect, there is something wrong with the data, or somebody’s interpretation of the data.
There is certainly an UHI, but it’s less certain that the actual temperature time series follow it, because there are so many different changes that affect them.
If there is certainly an UHI, then it is certain that as the world’s population has increased from 2.5 billion in 1950, to 7 billion today, the UHI effect has not been negligible, or negative, for chrissakes.
The basic argument is really that the difficulty of carrying out quality control of all sites is so high its not actual do able , although I not seen anyone provide research to prove this . While to date its still not know how many stations currently meet the standards required of this types of installation.
So a ‘statistical’ technique is be employed to smooth the data and so allow for all the errors, which ever admits exist . Of course at the root of this is the assumption made as adjustments for errors ‘ The classic problem is the more assumption you make the more errors are likely to find in the end results when its compared to reality .
Its worth bearing in mind that sites at airports were designed to meet the needs of air movements in and out of the airport , they were not designed to be used for a wider area let although a much wider area as they become used. So although they may do a good job for the air movements its far from clear they do the same job for anything else .
The Watts survey included 80% of the stations the last time I looked at it. What do you consider an acceptable “sample?” 100%?
The problem is not the size but the bias. First of all, mostly the most accessible sites have been visited; secondly, very few of the station histories (necessary for assessing biases in trend estimates) have been recorded.
Mat,
it is the magnitude also. With a station with a .8c/C trend, how much of that trend is GHG caused, how much sun direct and indirect, how much UHI…The magnitude tells us whether it is important or not and whether it significantly replaces some of the assumed GHG contribution.
I am sorry but I find the whole idea that massive number crunching can enable one to draw conclusions, when nobody knows what the data means, absurd.
I agree with that.
There seems to be a dichotomy between people who believe that the sites need to be studied, and people who believe that the problem can be solved without studying the sites.
Think for a moment that we don’t want to solve all the problems immediately and can monitor the climate metrics over time. One of the rules of measurement metrics is that many of the concerns over accuracy, precision, and bias are less important than sticking to a consistent approach over time.
If the BEST results end up crashing and burning, we will eventually find this out over the course of time.
In the meantime, if someone wants to invest in a completely new infrastructure or wants to meticulously check and update all the current stations, nothing is stopping them. Of course, should this ever happen and we get more accurate and precise measurements as a result, the next complaint will be that this is useless as the historical results remain uncalibrated.
Consider comparing global temperatures with that of reported oil production numbers. This is not a well-kept secret, but the metrics used for calculating oil production change over time. Together with the conventional crude oil count, the official numbers are now starting to include the “all liquids” category, which can include natural gas liquids, coal to liquid, biofuels, etc.
http://www.crudeoilpeak.com/wp-content/gallery/biofuels/iea_biofuels_adjustments_comparison_eia.jpg
This may be accurate accounting but it misses the extra energy needed for processing in comparison to conventional crude. The gap is growing over time and the uncorrected numbers are being used as propaganda by the elites to suggest that natural resources availability is stable. And this is really no different to what we are dealing with with respect to arguments over surface station temperatures. Who is trying to hide what? In oil production we are seeing the real “hide the decline” agenda, the global temperature “hide the decline” is tinker toy FUD in comparison to the real issues we face.
Everything new will always be debated, but that’s hardly an argument that less knowledge is better than more.
I expect most of the debate to be settled over the next 2 decades, but it would still be informative to investigate why some locations are showing declining temperature trends and others are showing increasing temperature trends.
I would like to see a probability density function for the rate of change, R=dT/dt. This will show the dispersion of the results. If there is a 2:1 ratio of positive to negative trends, the tails would tell us something. There is not a natural reflecting boundary at R=0, so I don’t see why the rates can’t go negative.
Ricahrd Saumarez writes:
‘That we have a supposedly high quality data set of temperature stations and nobody actualy knows their environment is astonishing’
It is also a devastating indctement of the organisation of the field of climatology. In 30-odd years of existence they haven’t even managed to construct a robust network for the collection of their basic data.
In UK we have a term that rhymes with ‘Bankers’ to describe such people.
Okay, it’s my bed time. But before I go, I want everyone to understand what the rural classification really means. So here is the paper that will tell you. If you don’t care about the whole thing, start reading at page 2 where it says:
2. Defining Urban Extent
http://www.sage.wisc.edu/people/schneider/research/schneider_erl2009.pdf
I also want to say that I plotted the BEST monthly data. The thing that jumps out at me about it is that the monthly variations appear to be about three times as large as those of HadCrut3, RSS, and UAH.
Good night.
I just watched the debate between Lindzen and Dessler and I was struck by the discussion of the “hot spot.” It is strange that Dessler claims that we can’t trust temperature measurements form baloons and satellites and instead must calculate temperature from wind speed. This strikes me as bizarre given the fact that wind speed is much noisier than temperature. Andy Dessler, you are straining credulity here. I know from everyday experience that wind speed is MUCH noisier than temperature measurements. This is a sign of desparation and lack of honesty.
I’m somewhat confused about this. I understood that the basic dimensions were mass,length,time,charge and temperature. It is proposed to measure temperature with a variable with dimensions of length and time? Am I missing something here?
Well, it’s traditionally measured by the length of a column of mercury.
and depends on coefficient of expansion (length and temperature)
Yes, temperatures are measured by proxies, and they all have non-temperature units. Can you cite any measurement scheme that works in uhits of Celsius?
UHI is clearly an attractive issue to argue upon. It’s potential effect on warming was taken up early (in the 1980’s) by the main stream scientists. John Arnfield lists in his review on urban climate research (Int. J. Climatol. 23: 1–26 (2003)) many such studies, but the review is not on the influence to the temperature trends. The early studies made various tests to conclude that the effect doesn’t influence the warming estimates too strongly. CRU chose to leave it uncorrected, but states a small additional uncertainty based on that, other analyses have corrections based on apparent discontinuities in the data series, which are likely to include also automatically UHI related corrections.
The Spencer study discussed above shows a correlation between population density and temperature based on comparison or pairs of nearby stations after removal of the influence of altitude and making sure that nearby large water bodies are not present. His results indicate that there is a correlation at all population densities from less than 10 persons/sq.km to 7000 persons/sq.km. The dependence of the observed temperature is strongest at low population densities, when given relative to the absolute change in population density.
Superficially this might explain the negative relationship of the BEST analysis as it might tell that the UHI-bias would be largest in rural areas, if there is some increase in population there. The Spencer analysis is, however, of little value there, because he is not controlling at all other contributing factors than altitude and population density. It’s, however, likely that there are other differences between two rural (also urban) points with clearly differing population densities than those influenced by a change in population. The causal relationship is certainly very often the opposite: some location has properties that keep the population density low and these same features may also affect the local temperature. It’s easy to imagine that this opposite causality may be behind Spencer’s results. This is only speculation, but so is also the assumption that the population density would be the dominant cause for the result.
When population is increasing and new construction work is done, that leads often to some relocation of the weather station. These relocations have not been recorded well in all cases. Various operations of maintenance and modernization have also been made to the weather stations and they may lead to unrecorded transitions. It must be common that the newly located or renovated weather station has a lesser UHI effect than the earlier station. This may contribute to the negative trend observed by BEST.
There histories of all weather stations have not been recorded well enough to allow for removal of all these problems. Therefore various corrective measures have been applied in the statistical analysis. One straightforward solution is to cut every time series at points, where it has an unlikely discontinuity. Both resulting peaces are then used as separate datasets. This makes sense, but cannot correct for creeping effects. The method may also introduce some bias, when it’s used to remove evidence on a real change in temperature. This approach is used also by BEST, but they are not the first ones to do that. (As far as I understand correctly, the data series used for the analysis of UHI effect are broken only at points, where the changes are recorded in metadata.) The BEST work is one more analysis. As all earlier analyses it’s forced into making some assumptions, which are not exactly true, but reasonable enough for being used.
With all the caveats the evidence supports still the conclusion that UHI has not had essential influence on the calculated warming trends. Many people have tried many different methods, but none has found a significant effect on the level of trends in average temperatures. UHI is real, but it is unlikely to contribute much to these trends.
I think a larger point is being missed here. The UHI effect is partly due to the replacement of vegetation with asphalt, concrete, and steel; but it is also due to energy use. The heat released by burning petroleum products is concentrated in cities, but is actually ubiquitous in modern societies. Has the extra heat input been factored into the global temperature increase? Does anyone having familiarity with the literature know the answer? Is energy use a significant factor in global temperature? I’m thinking probably not, but not sure.
The role of energy production and use (excluding that through CO2) has been discussed many times on this site. It’s far too small to be significant on global scale, but it’s certainly visible in the UHI effect in heated cities.
Plus the heat released is transient and it does not accumulate like the CO2 itself does.
I would think the heat never has time to fully dissipate, either, especially in a larger city.
Webby,
another pointlees point until it is proven that the radiative properties of the GHG’s actually assist in warming as opposed to warming and cooling.
Hey Kumkwat, over 50% of the population of the USA lives within 50 miles of the coast. Keep on talking about the heat island effect and keep wasting your time. On second thought this is the perfect stagnant backwater for you guys to play around in.
Webby,
I lived across the street from Ocean Beach in San Francisco for 20 years. Please oh please explain to me how being next to the ocean, or being on a man made island ON the ocean, is going to make physics stop working?? (snicker)
Thanks for the useless anecdote from your diary, kumkwat. I am sure your important data point will make a difference in the BEST statistics.
Granted that, I am only making a hypothetical inference. Follow the inference chain:
The key statement is that warming is less over the ocean than land.
Coastal areas are modulated by the ocean temperature, therefore highly populated urban areas may show less of an increase and it may explain the data trend. That is the way that statistics work — you have to look at the preponderance of the data, and how the factor weighting can influence the results.
Your (snicker) single data point don’t impress me much.
Webby,
You really won’t allow yourself to think on this one will you.
“Coastal areas are modulated by the ocean temperature, therefore highly populated urban areas may show less of an increase and it may explain the data trend.”
Your statement is that coastal areas are modulated by the oceans. I don’t think anyone has an issue with this. You seem to be conflating lower absolute temp with a TREND. We have no way of attributing how much of what we are measuring to any specific drivers. That is, has the relative effect of the ocean cooling changed or stayed about constant? The only way the TREND could be affected by being next to the ocean is if the cooling effect CHANGED in a systematic way. Have you seen any studies that the oceanic cooling effect has increased over time to cause a lower TREND in marine affected stations?
This still does not prevent there being substantial UHI RELATIVE to the base climate that would exist without the city!!!!!!!! Where I was in in SF, and the area I am currently in Southern California both are subject to the cooling effect of Marine layers. They both are WARMER and have(possibly had) higher TREND than if there was no city here!! The UHI effect MAY be smaller due to the more efficient cooling due to the oceans, it isn’t non-existent!!! Because appropriate studies have never been done we simply do not KNOW how much or the sign even.
If you wish to say that comparing coastal and inland areas is dumb fine, we can agree on that. It doesn’t change the fact that there is probably significant UHI positive trend that needs to be deducted from the records of many inhabited areas. It is conceivable that there is UHI negative that needs to be added to some stations!! It doesn’t change the fact this is so whether the area is coastal or inland.
Which brings us back to the point I keep harping on. You and no one else, including the Berkely Bozos, have definitive studies to which they can refer as to how much UHI is in most locations on earth. Without having any idea what part of the trend we think we are measuring is attributable to UHI, how much of that trend is GHG, solar, cloud, aerosol is also not possible to know at this time. Basically they have been making stuff up since this Climate Scam began and they still are. As has been broached elsewhere, the UHI could be positive or negative or flat for each of the points in the study. Without extensive studies we simply do not KNOW!
You should go back and read Mr. Keenans’ and Mr. Briggs’ critiques. One of the major issues with how they have done this study, outside of GUESSING about attribution, is that the certainty has been increased by incorrect methods. Of course, doing it right would have blown the error bars up making it look like they have no idea what they are talking about when they claim there is a problem. For a guy who claims to know a lot about maths you are certainly talking like you are lacking in this area. Then again, the EXPERTS recruited for the Berkeley Bozos project appear to be just as ignorant, unless they did this knowingly or were ignored by the head activist.
I hope you noted Muller told us in that old interview that:
1) it is OK to distort and mislead because this is so important.
2) that he has been an environmentalist who BELIEVED in Goreball Warming since the 80’s. He claims he left the Sierra Club because they wouldn’t back Nuclear Power to reduce CO2.
Where was the 30 years of data to drive that belief? He didn’t even have 5 years of statistically significant warming at that point!! The GHG’s radiative transfer effect has been computed and speculated on for over 100 years. Why has no one taken it seriously or worried about it till recently?? Because there was no evidence of it being a problem in the real world. There is STILL no evidence of it being a problem in the real, highly variable world. Without activists like Phil Jones, James Hansen, Michael Mann, and now Muller making stuff up it would STILL be a non-issue!!!
Pekka,
How do you reconcile findings of large UHI in some rural cities with other studies claiming UHI is small and can be ignored?? This would seem to indicate that we simply do NOT have good studies to settle the question in any reasonable manner. ASSumptions that it is small are as bad as ASSumptions that it is large or non-existent. We should find out before making pronouncements.
With the recent measurement of the large temperature differential of UHI domes in some metropolitan areas by NASA we KNOW that their temperature records had to have increased trend due to this even if we do not have details of how much over what periods. Yet, you seem to ignore this also with your continued assumption that it is not a significant factor.
It is interesting to see what different sides of the discussion concentrate on and what we discount.
How do you reconcile findings of large UHI in some rural cities with other studies claiming UHI is small and can be ignored??
One very obvious explanation is that cities are extremely varied; some may have a large heat producing industry; some not; some have new well insulated houses, others old leaky houses that take twice as much energy to heat and half of it goes to warming the outside air by a degree or 2 or 3.
Cities are bound to be a source of more than their share of “noise” or variation unrelated to the global tendency one is trying to measure, so if a result you don’t like can be demonstrated by leaving out the cities, these protests sound like grasping at straws. Like trying to stop a drug study from being released by saying “wait! I looked at some of the data and it sort of looks like green eyed people didn’t do as well as the rest!”.
The issue here is not the size of UHI as a phenomenon. The issue here is much more specific: Are temperature measurements affected more by UHI than before? The answer to this more specific question depends on how the weather stations are located in their immediate neighborhood. It may depend also on their construction. The practices in these areas are changing, but all changes are not listed in the metadata and the effects of the changes may be small enough to avoid recognition in the time series.
Relocation of weather stations must have been common in urban areas, as the weather stations are not important enough to be taken into account in building codes.
I’m not making any claims on, what has really happened, I wanted only to note that the overall increase in UHI is not necessarily visible in the temperature data. It’s a bit surprising, but not impossible.
Somebody mentioned better insulation, and somebody else mentioned factories being offshored (or just going out of business). Personally I’ve always suspected that the immediate local effect of the UHI might often be negative: I remember decades ago seeing a picture of a thunderstorm over Paris, putatively caused by the effect of urban heat on conditionally unstable air. If the negative effects have come to be more important since the ’50’s, that might be able to do it. Possible reasons for that would include changes to the aerosol regime, larger urban areas being more likely to set off a thunderstorm, increased dumping of heat into the air above ground (20-200 meters), and many others.
And what about the effect of regulatory changes on smokestack height? That could result in more of the heat staying too high to affect thermometers.
Pekka,
While you are certainly correct that the amount of change in UHI over time is important, without knowing how much of the temperature observation is UHI means you CANNOT KNOW this. All the information we have is the TOTAL amount of change in temerature with no way of knowing WHAT has caused it. With Dr. Spencers preliminary study and now this one even the view that generally, large cities should have a higher rate of UHI increase, has even been shown to be in question.
The statistical games being played with the temperature record CANNOT attribute the change. It can only tell us the total changeand the total differences between areas (if you accept it can even do that within reason) that must be separated between several possible sources.
It must be a conceit that Climate Scientists and others think we can intuit the amount to assign to the sources of these changes without detailed, rigorous attribution studies at a number of differing sites.
Hal,
I would point out that UHI appears to be more from the materials such as blacktop and concrete absorbing energy during the day and releasing it at night than from waste heat. Waste heat would probably increase the max temps more than the min, but, we see the min increasing more than the max in most areas where we would expect to see UHI!! It is an influence, but, based on the poor data we have at this point would appear to be a secondary part of UHI. Who knows, with rigorous studies I might be shown to be wrong!!
Pekka, I think on the whole that your conclusions are not supported by your detailed points. Consider this: It must be common that the newly located or renovated weather station has a lesser UHI effect than the earlier station. This may contribute to the negative trend observed by BEST.
There is no relevant evidence on that point, none at all. You could be right, but there is no reason to believe that you are right. There is no reason to believe that you are wrong either. The lack of evidence should not be used to support a conclusion. There are lots of examples to show that you could be right or wrong, but not a reasonably unbiased survey.
“The urban heat island effect is locally large and real, but does not contribute significantly to the average land temperature rise. That’s because the urban regions of the Earth amount to less than 1% of the land area.”
The Urban Heat Island effect paper is a nice study, and based upon the reporters’ questions, seems to be the primary paper of interest. Is this paper the final word on the UHI issue? Probably not. I will be interested in the reactions to this paper, since UHI is a contentious issue.
Judith,
you should look harder than that. It is a well established fact that for high enough population densities each doubling of this density contributes 0.25-0.27 °C to local warming. That is, the relation is logarithmic.
The big (and unanswered) scientific question is what level of population density counts as high enough in this respect. That is, what is the population density below which this logarithmic dependence breaks down?
If it is significantly lower than the lowermost population density measured in the most densely populated 1% of land surface, your proposition simply does not make sense. But that question was not even asked.
We do know, that population of the entire Earth has doubled almost twice during the last century. Therefore, if logarithmic dependence of UHI on local population density extends to areas categorized as rural, effect of UHI alone could amount to a rather significant portion of the entire warming observed.
Or it may be the case that in an area with low local population density, doubling has no effect on local temperature, but if it is so, it should be still proven. What is more, the accurate form of the population density – temperature anomaly function should be determined and quantified. That’s the way science is done, not by hand-waving.
Additional note: It is quite easy to demonstrate, that population distribution is fractal-like over land. It is also reasonable to assume location of temperature sensors (weather stations) is not randomly distributed relative to this fractal. It should be that way, because if a weather station is too far from any human habitation or place of activity, maintenance & operational costs can rise to prohibitively high levels. Therefore, even if percentage of land area with “high enough local population density” is small, it is still conceivable that almost all thermometers happen to be located in this tiny (fractal-like) segment.
The only way to get around this problem is to consider how local population density has evolved in the vicinity of a large sample of weather stations. That is, it is not a very rural / whatever categorization what you need, but one with categories like declining / stationary / increasing population density, along with absolute population densities for those places. It would make possible to specify the local population density – temperature anomaly function mentioned above.
If you disclosed the list of weather stations categorized as “very rural” in your study (as you should, in the spirit of BEST), one could use county level census data to look for such a relation in case of “very rural” stations located in the U.S. of A.
Barrow, Alaska has a population well below 5000, so it is rural by any standard. Still, a sizable UHI effect is observed around the township.
INTERNATIONAL JOURNAL OF CLIMATOLOGY. 23: 1889–1905 (2003)
DOI: 10.1002/joc.971
THE URBAN HEAT ISLAND IN WINTER AT BARROW, ALASKA
KENNETH M. HINKEL, FREDERICK E. NELSON, ANNA E. KLENE and JULIANNE H. BELL
Taking this one step further, one needs to distinguish urban heat island (UHI) effects from local heat island (LHI) effects. The various kinds of heat contamination that Anthony Watts people photographed have nothing to do with large scale urban growth per se. They are extremely local, typically involving heat sources a few meters away. There is no way to isolate or exclude them for analysis using population based filters. The difference between UHI and LHI is a fundamental confusion in the debate.
Yes, I agree. The LHI is an error whereas UHI is an effect that can be modelled.
A more general question, is the UHI seasonal? I could imagine that effects in high summer such as hot tarmac could raise summer temperatures, but does human activity heat cities as much in winter, or is increased energy consumption during winter a factor?
Would you expect DT/dt over a monthly scale to be higher in cities during spring an autumn?
At one point EPA was considering making UHI a regulatory program, so I imagine this seasonal issue has been studied. It is also related to ozone formation and regulatory ozone level exceedences. I think Atlanta GA did a lot of work a decade or so ago. That’s all I know.
LHI may or may not affect temperature trend. If it does, the bias could go either way.
Cooling is hard to create. Very few of the problems the Watts folks identified would create cooling. There is also a study suggesting that a paint change alone created a significant warming bias. Irrigation is an exception but airports are seldom irrigated.
Re LHI, cooling can be created by moving a thermometer from a hot spot to one that’s not or by changing what’s causing the spot to be hot.
These comments come back again and again to the necessity to investigate actual sites and document what has actually been changing in them. I have read almost every post on this thread, as well as the posts on other blogs, and I see no way around this. In my reading, Anthony Watts had the fundamental insight, but the large scale statistical analyses (like BEST) avoid carrying out the procedures that could possibly settle the issue. Much as I respect the BEST work, it has highlighted, not solved, the problem of estimating the UHI effects; that is, as everyone generally recognizes, a common outcome of good scientific and good statistical work..
The next study that has to be funded is systematic field work.
From 1998 to 2008 I referred to the Pierce College weather station website for local conditions. The weather station is located atop a low hill in Woodland Hills, CA, at the west end of the San Fernando Valley, where temperatures range from the high teens to a record 119F. During this time period an empty field at the western base of the hill was developed into high-density housing (condos). On hot days, with air conditioning working hard and wind gusts varying wildly in direction, I could watch in real time as temperature rose when wind from the west carried heated air from the development and fell when the direction changed toward north or south.
The record high temp occurred in 2006, after the development was built and occupied.
This weather station has archived daily temperature records online from 1949, precipitation records from 1950.
That is something I’ve wondered for awhile –
wouldn’t one need a different set of adjustments for virtually every site experiencing either a LHI or UHI?
In other words, two similarly poorly sited stations might vary differently from a correct reading if one was in a more cloudy environment one day than the other, or the AC unit that was nearby wasn’t run as much over a weekend or a day it was non-functional, etc.
JohnWho
Several UHI/LHI differences:
1.) UHI can be attributed to a real, known trend of urbanization; LHI doesn’t match any specific single identifiable trend that might not have been uniformly present in one way or another since temperature recording at a site began.
2.) UHI has a specific heating mechanism or set of mechanisms proposed (and too a specific cooling mechanism or set of mechanisms) that can be considered as a single effect; LHI is just bad measurement practice, and can as easily be cooling as heating (simply casting a shadow for example can cool). No mechanism, no hypothesis. No hypothesis, it can only be treated as observational error.
3.) UHI was painstakingly considered in BEST using techniques specific to the UHI hypotheses, established, and narrowly defined; LHI was clumped in with other sorts of measurement error using techniques appropriate to observational error, treated as ordinary error, and did not rise to such high levels as to be detected by BEST’s very exhaustive methods, so far as I can tell (I’m glad to be corrected, if I’ve missed it in their reports).
So.. LHI looks like it happens, as with other sorts of error in measurement, but at such a small rate and of such a small type as to be of no real consequence.
The data’s already bad enough that added LHI/LCI doesn’t make it appreciably worse; even with all this, the dataset is now fit for much, much better use than the datasets before it, and tends to confirm most of the major AGW hypotheses capable of being based on the surface temperature record.
Thank you for the response. I understand mostly what you’ve said, but this still doesn’t remove the “common sense” situation that I see. First, each station that is not properly sited may have its own unique set of requirements to get a reasonably close to accurate set of readings and these adjustments may also vary from day to day. It would seem that the only proper thing to do would be to identify these incorrectly sited stations and either correct them and then use only the data from the corrected position, or not use the data they collect since it is “tainted” in a manner that for all practical purposes can not be corrected. After removing these from the data base, then consider the UHI influenced sites. My question regarding these would be, “if there is more than one site under a specific UHI “umbrella”, shouldn’t each have its own specific adjustment criteria”? Maybe they do, as you describe/suggest. If they do not, then this adds just a little more uncertainty to the mix. Then there are the LHI influenced stations – I’m assuming there would be some left. Again, each would need its own specific adjustment criteria as well.
I know, call me a skeptic here, but I do not believe we have enough data on every station to perform the necessary adjustment(s) needed – do we have cloud cover for each day/hour/minute(?) of every reading, was it raining at the time of data collection or just prior to the data collection, for example. Wonder what the analysis of only the properly sited, non-UHI/LHI stations shows? I would think we would all trust the trends from that over any other one. Maybe not the “global temp” (whatever that is), since these sites aren’t even close to covering the entire globe, but the trends might be the best we can get from surface measurements.
Again, I’m not trying to be argumentative here, just trying to understand the rationale behind adjusting the temperature data and then “assuming” it is now accurate. Why not use only the temperature data that does not require an “adjustment” and accept that it is the best we have?
JohnWho
These are important and significant questions you ask, of such a type as ought be considered in every set of observations.
Some of the solutions you propose, however, would themselves tend to skew the data, by replacing actual observations with arbitrary (even if well-meaning) decisions of the observer.
Indeed, to some extent, all observations deal with this issue.
The methodology BEST describes covers many of these points in the technical language of experts in observation, and addresses them with other solutions than you suggest, that achieve many of the results you seek (and many more, besides, that BEST have considered in their approach).
For instance, finding ways to add more observations to the dataset tends to smooth out the problems of local error. As with anything this complicated, it comes at some trade-offs, but on the whole it is a preferred way to help reduce local errors.
Another way LHI (and LCI) effects are dealt with are by the multiple layers and types of goodness-of-fit and other error-checking done by BEST, which would tend to spot irregularities for each site. (Though of course this isn’t perfect, it would catch most issues large enough to bias the whole dataset.) Again, there’s a cost for this, a trade-off in what the dataset looks like and can be used for, but BEST appears to have been scrupulous in evenhanded application of error-checking, and in being transparent about what they’ve done.
Some of the issues you mention are covered by general principles, such as Occam’s Razor (don’t make more assumptions, for instance assumptions of errors that can’t be proven and don’t show up in the error checks, than are minimally necessary) and Ceteris Paribus (treat all other things as equal in absence of good reason to believe otherwise).
The datasets produced by BEST are now purpose-built and fit (pending review) for climatology.
LHI/LCI, while real, fit into the +/- 0.042 C since the 1950s (95% confidence for statistical and spatial uncertainties) portion of the 0.911 C figure, along with all other observational error. They amount to part of the less than 5% of the overall, one would expect 19 times in 20, statistically.
Clouds, site situation, shadows, asphalt and all, the math’s been done to ensure it all comes out about even overall.
There are many other considerations, too, to consider, from aliasing to why was 10-year smoothing selected and does it much affect the fitness for purpose, to what improvements ought be made to future data collection.
The LHI/LCI one is.. one of the smaller considerations overall, I’d think.
Excellent post.
“We do know, that population of the entire Earth has doubled almost twice during the last century. Therefore, if logarithmic dependence of UHI on local population density extends to areas categorized as rural, effect of UHI alone could amount to a rather significant portion of the entire warming observed.”
Where is the science?
“If you disclosed the list of weather stations categorized as “very rural” in your study (as you should, in the spirit of BEST)”
I want to mention again that the “very rural” designation is not related to population density in the BEST study. It is only related to distance from areas that are designated as urban. This means that “very rural” areas could actually be more built than “not very rural” areas.
Understand that it is alleged that due to increased green house gases in the atmosphere, heat is trapped that cannot escape from earth. So if an increase in green house gases is to blame for the warming, it should be minimum temperatures (that occur during the night) that must show the increase (of modern warming). In that case, the observed trend should be that minimum temperatures should be rising faster than maxima and mean temperatures. That is what would prove a causal link.
What I have discovered so far from my (silly?) carefully chosen sample of 15 weather stations is that the overall increase of maxima, means and minima was 0.036, 0.012 and 0.004 degrees C respectively per annum over the past 35 years. So the ratio is 9:3:1. Assuming that my sample is representative of all those stations listed, I have to conclude that it was the maximum temps (that occur during the day) that pushed up the average temps. and the minima. So either the sun shone more brightly or there were less clouds. Or, even, perhaps the air just simply became cleaner (less dust? Are there records on that?).
I also noted that the warming on the NH is totally different to that of the SH. There is virtually no warming in the SH as seen by the means and minima whereas in the NH, the ratio of the increase in maxima, means and minima is about 1:1:1, amazingly.
Again, if it were an increase in CO2 or GHG’s that is doing the warming, you would expect to see exactly the same results for NH and SH because these gases should be distributed evenly in the whole of the NH and SH hemisphere. So, even here, we again must conclude that it never was the increase in CO2 that is doing it. The only logical explanation I can think of is the difference in the rate by which the earth is greening. In South America we still had massive de-forestation over this period whereas Australia and Southern Africa have large deserts. Obviously, the NH has most of the landmass and here everyone seems to be planting trees and gardens. A recent investigation by the Helsinki university found that 45 countries were more green then previously out of a sample of 70.
Paradoxically, the increase in greenery is partly due to human intervention, partly due to more heat coming available (increase in maxima!) and partly due to the extra CO2 that we put in the air which appears to be acting as a fertilizer/ accelerator for growth.
For my data, see:
http://www.letterdash.com/HenryP/henrys-pool-table-on-global-warming
(make a copy for yourself of the tables)
Now, if we could have the 3 plots for the Best figures? That would help.
Henry P,
UHI shows a very similar pattern of higher night time warming trend.
Yes, it is silly. No one takes undercounted cherry-picked statistical studies seriously.
As opposed to those wonderful studies showing minimal UHI by Phil Jones where there is no data to back them up!!!
I am saying all of your results (and berkeleys & & &) even it is from 1000s of stations without showing me what the rise in maxima and minima is, are all pretty useless because it does not tell me why the temperature rose.So why don’t the experts concentrate on giving us the rise in minima and maxima as well?
Now that you have done it for 15, do it for 150, then 1500. Be careful as you increase the sample that you don’t bias the sample by oversampling where you expect your ideas to be confirmed. If you are going to compare regions of one type (reasonably well defined) to regions of a different type, sample randomly within regions.
BEST have reported that 1/3 of stations show cooling. By cherry-picking some warming and some cooling stations, you can support almost any hypothesis.
I had hoped that the BEST people would explore the statistical strangeness of the grid-based area averaging method, but that seems not to be. Here is a simple analog to illustrate the issue.
Suppose one is a pollster tasked with polling an issue. You send two takers out to two towns, to poll individuals. One is successful and gets 1000 responses. The other has problems and only gets 10 responses. In normal statistics one would first combine these responses then take the average. But in our case one first averages each set of responses then averages these results. Thus the 10 count as heavily as the 1000. The result is not an average at all, as that term is normally used.
Let’s now make it a bit worse. Suppose we sent out two other takers as well, to two other towns. One got only one result and the other got none. First we will count the one result as equal to the 10 and the 1000. Second, we create a fictitious result for the town that produced no results, based on the results from some of the other towns, say the 10 and the one. Then we average these four so-called averages to get our overall average result.
I can’t imagine anyone who understood them accepting these poll results as accurate, but this is how global “average” temperatures are calculated.
David,
they also WEIGHT the temps to attempt to deal with the issue you state. Of course, the method of determining the weighting becomes an issue. In areas with few stations, a station with a high trend then has much more affect on the overall trend than many stations with small trend in another area. We see this issue most dramatically in the Arctic where GISS uses one station to “spread the wealth!!” Another station in the Argentina or Bolivia area close to sea level gets to speak for a lot of mountainous regions also!! Basically, gridding without reasoable coverage in the grid most likely distorts the record even with weighting as there is no survey to see if the stations are representative of the grid.
Now that I’ve read the study fully, a very interesting possibility is suggested in the second paragraph before Acknowledgements on page 12.
The numbers suggest the possibility that human climate forcings — from CO2 emissions, methane, ozone, black carbon, deforestation, land use changes, and reduction of sulfate — might have been responsible for about 0.25 degrees warming since 1975.
Back of the envelope can be simplistic, the AMO might not drive temperatures as the authors imply it MIGHT, and the AMO and the land temp records might instead both be driven by climate forcings (as the authors also say MIIGHT occur). But given the links between the AMO and PDO and temperature records going back over a century, it seems plausible to me that these oscillations could drive land based temperatures, with human influences added on top.
The paragraph states that the AMO has risen sharply since 1975, by 0.55 degrees. The authors previously showed that the AMO is the oscillation most tightly correlated with temperature increases, and that the PDO leads the AMO by about 2.5 years. Land based temperatures, the authors state, have risen by 0.8 degrees in the same time frame. IF the AMO is a driver of land based temperatures, as the authors suggest it MIGHT be, THEN subtracting 0.55 from 0.8 suggests that anthropogenic influences have caused temperatures to increase by 0.25 degrees in 35 years. A current rate below 1.0 degrees per century.
These words come before the words, “In that case, the human component of global warming may be somewhat overestimated.”
If my memory holds, there have been a couple of papers by IPCC authors suggesting that the PDO is now in a declining mode, which will hold temperature increases down for a while. The Berkeley work suggests that these earlier papers MIGHT be correct.
If these hypotheticals turn out to be correct, Richard Lindzen will feel vindicated. But it is the third or fourth inning, not the ninth. I’m not placing my bets yet.
The link to the study, for those who haven’t gone to the study yet:
http://www.berkeleyearth.org/Resources/Berkeley_Earth_Decadal_Variations
John,
the AMO is still strong. We should wait to see what happens when it starts dropping.
According to Don Monfort | October 22, 2011 at 2:31 am
“We know that there is a Urban Heat Island effect. If the data says that it is actually a Rural Heat Island effect, there is something wrong with the data, or somebody’s interpretation of the data.”
Interesting, so maybe we don’t need no steenking data. Aristotle didn’t when he wrote his laws of physics.
But have you really thought clearly about what the UHI effect is? I would think it obvious that urban areas are warmer than rural areas, but wouldn’t the UHI effect say that the average temperature in urban areas is RISING FASTER than in rural areas? Have you heard that building insulation has improved in the last couple of decades? And isn’t much of the UHI affect due to LEAKAGE of heat that was intended to warm the insides of buildings? And there is less of that, hence energy is used more efficiently to head buildings? And haven’t factories been leaving urban areas in the U.S. in droves?
If you want to know why the right-coalition is putting the “We’ll primary you” gun to the heads of Republican politicians, and very effectively, consider this: http://therealtruthproject.blogspot.com/2010/07/corrolary-to-big-lie-theory.html
Thank you Hal. I never thought of the insulation, and the LEAKAGE thing. The heat don’t LEAK out no mo’. Please forgive me for not commenting on your link, whatever it is.
Hal,
waste heat definitely is part of the UHI in cities. Of course, even in temperate climates where there is no, or little, heating or cooling you can still have serious UHI as it is a function of the absorption of energy by the anthro materials being so much higher than most natural materials. This is why the min trend rising is a much stronger indicator of both AGW and UHI. The absorbed energy is radiated at night keeping the min higher. If it was primarily waste heat we would see little min increase in temperate areas as the least energy would be used and lost then.
John says
The numbers suggest the possibility that human climate forcings — from CO2 emissions, methane, ozone, black carbon, deforestation, land use changes, and reduction of sulfate — might have been responsible for about 0.25 degrees warming since 1975.
Henry@John
It would seem to me
http://judithcurry.com/2011/10/20/berkeley-surface-temperatures-released/#comment-125821
that the 0.25 is for the NH only, and then forget about all the GHG’s.
It does not do anything.
It is the increase in vegetation that traps that heat.
Looking again at the BEST UHI results it looks like it is produced by subtracting a non urban anomaly curve from an urban anomaly curve. It seems to me that the problem with using anomaly curves is that the build that already exists in an urban area at the time that the thermometer is put there cannot be reflected as a part of the anomaly. In other words, the UHI contribution that exists at the time of thermometer instillation is not recognized as being part of the urban anomaly. Only the change in urbanization of an area that is already partially urbanized is accounted for when using anomaly. And it is easy to believe that the population change of places that don’t meet the high density standards of being urban could actually be more UHI influential than the changes in urban areas for some later period of time.
It seems to me that if you want to get your hands around the UHI problem that there is a better way to do it. Why not construct a global land temperature curve that consists of 10% of the thermometers in the least densly build areas. Construct another that consists of the 10% of the thermometers in the most densly built areas. Then compare their absolute temperatures rather than their anomalys. That way the UHI effect that was added to the urban areas before the thermometers were installed would not be lost.
Tilo,
It would be interesting to compare the absolute changes in temps between North and South Korea, since about 1950. As soon as Big Climate gives me the funding, I will work on it.
http://bigthink.com/ideas/21270
David Wojick | October 22, 2011 at 7:45 am | Reply
Thus the 10 count as heavily as the 1000. The result is not an average at all, as that term is normally used.
Temperature has ‘conservation’ in the sense that it is likely to be nearly the same in neighboring areas. Assume that the areas nevertheless have somewhat different temps, say 20C and 16C as measured by three thermometers in each, making the average a reasonable 18C. Now, imagine that somebody installs 3 thousand thermometers in the one of the areas with 20C. The average of those is likely to be close to 20C. It is clear that one should not compute the average of both areas as (3*16+3000*20)/3003=19.996, but as (3*16/3+3000*20/3000)/2 = 18. Each area must be averaged first and then the total computed as the simple average of the two individual ones without regards to the number of stations involved.
I was just reading the decadal variations report, and it looks like the BEST global land data that was used there has a negative trend for the last dozen or so years. That chart comes from 2000 GHCN temperature records that had not been used before by the other surface temp records. Last night I plotted the monthly BEST data that comes from the 39,000 records that they use. I trended it, and the trend is very positive for the last dozen years. I’d like to get a hold of the data from the 2000 records and compare the trends. Based on an eyeball of the 2000 records, it looks like the trend difference between one data set and the other is huge – at least over the last 12 years. I would think that such a difference would deserve comment by BEST. Has anyone seen a link to where I can get the data for the 2000 station chart?
Tilo: ” Last night I plotted the monthly BEST data that comes from the 39,000 records that they use. I trended it,”
How did you do this? What exactly is this data?
Joel:
Go to here:
http://www.berkeleyearth.org/analysis.php
Go to the bottom of the page and click “ananlysis chart data”
That will download the data for you. Unzip it. Open the file named Full_Database_Database_Average_Complete in Microsoft Word. Use Word to cut the column of data that you want. Past it into Excel. Have Excel create a chart for you. Make it pretty. Have Excel add a trend line for you. What the data is is explained on the page and in the file header. Good luck.
Dear lord, I must be blind. I was looking for that exact file earlier, and I swear I checked that page at least half a dozen times. I have no idea how I missed that link. Thanks for pointing it out so clearly!
For what it’s worth, you can copy the data directly from the file into Excel. I found it easier to do that, then pick/delete whichever columns in Excel.
Now if I could only find that 2000 record reconstruction.
Ah, well. I thought you meant the data.txt file. It is headed as Tavg – detrended, but it doesn’t make any sense to me. But thanks anyway.
It isn’t just the last 12 years. The plot from the 2,000 subsample is dramatically unlike the rest in several multiyear epochs.
TT
Thanks for your reply. I have read all MIchael Manns papers not just the one on the MWP. In many ways Dr Manns papers are very good and exhibited some novel detective work, however he got several important details wrong which is precisely why I have just spent two days ay the MET office archives (see other thread for context)
tonyb
Tony, I’ve been a “context addict” for virtual eons! And this sounds fascinating … but could you possibly provide either a link or some hint as to which other thread contains this context?!
Thanks :-)
[Hilary … going back to lurking ‘n learning]
hro001
It was in the changing minds thread. The context is that I am writing an article looking at the respective output of Hubert Mams and Michael Mann as regards when the warming started. My research shows it to be 1607.
tonyb
As early as 1975, Wally Broecker wrote an article in Science that describes the current situation with breathtaking foresight:
“If man-made dust is unimportant as a major cause of climatic change, then a strong case can be made that the present cooling trend will, within a decade or so, give way to a pronounced warming induced by carbon dioxide. By analogy with similar events in the past, the natural climatic cooling which, since 1940, has more than compensated for the carbon dioxide effect, will soon bottom out. Once this happens, the exponential rise in the atmospheric carbon dioxide content will tend to become a significant factor and by early in the next century will have driven the mean planetary temperature beyond the limits experienced during the last 1000 years.”
http://www.jstor.org/stable/1740491
Given Keenan’s, and now Brigg’s criticisms (http://wmbriggs.com/blog/?p=4530), would seem that peer-review approval should be anything but straightforward…
I haven’t looked at Brigg’s complaints in any detail, but like Muller, I found Keenan’s critique confused at best. Yes, using an incorrect noise model will result in artificially narrow confidence intervals, but AR(1) is not a bad approximation of the noise in the temperature record although ARMA(1,1) (as shown by Tamino) is a better model.
Briggs and Keenan complained about analyzing data after it has been smoothed. Briggs has written about this several times over the years and so I have come to view the practice as a rookie mistake. Muller attempts to defend it but I don’t see the value of the practice unless one is trying to hide something. You can always go back and smooth a prior step for presentation purposes if you want.
Even if this complaint is correct (and it is not always correct, depends on what you are looking at), this only effects the decadal variations paper, which, to my mind at least, is the weakest of the the 4. It does not effect the methods paper or the overall reconstruction.
“which, to my mind at least, is the weakest of the the 4.”
No, the UHI paper is the weakest of the four. It fails completely to do what it set out to do.
Rattus Norvegicus, your comment misrepresents my critique. The BEST team did not use an AR(1) model. Rather, they wholly ignored autocorrelation in time, which is a serious error. And they did that in their methods paper, as well as in the decadal variations paper (as noted by Matt Briggs). Their papers would not pass in an introductory course in time series.
Doug Keenan – Maybe you can help me to understand your point about autocorrelation. I ask out of genuine curiosity – it’s not disguised snark.
Imagine we have a temperature value at time t (with variance), and another value at t+1. Can we use these two temperatures as separate indicators of a trend? The question relates to their independence. My surmise would be the temperature at t+1 is correlated with the temperature at t because whatever is warming the planet at t is likely to be doing something similar at t+1. In that sense, they are not independent of the physical phenomena underlying temperature. However, it is not apparent to me that the recorded number value for the temperature at t affects the number value at t+1. If the first one happens to be off because the measuring devices failed to record the actual temperature accurately, would that affect the number recorded at t+1? If there is a systematic error affecting both, would that affect the trend (as opposed to a detrended average)? I have a sense that when autocorrelation describes a genuine physical process, multiple data points can legitimately be interpreted as multiple points, and that the need to correct for non-independence is not an automatic consequence but depends on what we are trying to ascertain. In the case of trends, there might be a danger in overcorrecting for autocorrelation. Your comments would be welcome. Also comments from other statistical experts.
The purpose of these filtering models is to draw the desired data signal out of the signal+noise. The model of the process at time t+1 representing a Markovian update from time t is indeed part of the physical mechanism and is applied to help filter out the noise that is largely uncorrelated with itself from t to t+1. The moving average is there because you actually don’t trust the current value at time t to give you the update at time t+1, because that could be noisy as well. I come academically from the DSP world where the MA is the FIR filter and the AR+MA is the IIR filter.
I think more interesting of course is to use these techniques to cross-correlate the temperature with the CO2 measurements.This is closer to the transfer function application of the model. When you apply an ARMA model to this interacting system you can see how symmetric the modeling parameters can become. Are the CO2 the MA parameters with the Temperature the AR? Or are the CO2 the AR and Temperature the MA? It’s tricky because they are coupled by positive feedback.
Douglass –
As someone who can’t evaluate the technical arguments about disagreement in methodological approach – I am left to focus on related aspects of the debate.
How can you explain why a team of highly trained scientists, some with extensive experience and expertise in statistical analysis, would produce a paper that “would not pass in an introductory course in time series?”
I have no idea – factually – whether you are correct in your assertion. Such a debate would likely boil down to yet another diametric disagreement about basic issues between experts (a phenomenon that never ceases to amaze me).
So – to judge whether your statement is plausible, it would help to see an explanation from you how such a counter-intuitive result could have taken place.
BTW – I still have that question after reading your email exchange with Muller. Are you ruling out that the details of their methods section might explain how they would produce a result that you feel is below introductory-level science?
Joshua,
why are you requiring Mr. Keenan to speculate on why some people didn’t use the correct procedures. It is enough if he can prove that they did not use a more accepted method that is better. Why they did what they did is for them to explain if it is shown to be non-optimal.
See Tamino’s. He agrees with Muller, you are confused.
I think the correct Tamino link is here
Geez, how did that get in my paste buffer instead of the right one? I am mortified.
I always respected Richard Muller because he was willing to criticize Michael Mann and his Hockey Stick. I just read the following quote which bothers me and makes me wonder about Muller’s own ability to be fair with the data:
“If [Al Gore] reaches more people and convinces the world that global warming is real, even if he does it through exaggeration and distortion — which he does, but he’s very effective at it — then let him fly any plane he wants.” See http://www.grist.org/article/lets-get-physical
Commenter ZT quoted this on ClimateAudit. Does this trouble anyone else?
Well, there you have it.
You liked him when he affirmed your prejudices, and now that he’s challenged them you will go to whatever lengths to rationalize disliking him.
Robert,
His criticism of Mann did not affirm my prejudice. Mann was wrong, demonstrably wrong. Muller did the only decent thing he could do and criticized him for it. In some way it is comforting to see someone criticize another in their own tribe when their actions were terribly, terribly wrong – not just in a scientific sense but in an ethical sense. But in this quote Muller is praising Gore for the same kind of exaggeration and distortion Mann was guilty of. Why is it okay for Gore to do it but not Mann?
This is troublesome to me. Is it not to you? Or would you rather Muller praised both Gore and Mann?
Did you notice the date on that interview?
Well, it was Oct. 2008. But, FWIW (which may or may not be very little!) what I found interesting about this interview was:
Perhaps someone should send Muller a copy of Donna’s book! But I digress …The interview continues:
And here we are three years later and – as his oh-so-timely WSJ op-ed demonstrated – Muller appears to be playing the role of “effective advocate” … for the same side!
I wonder if Muller realizes that, not unlike like Hansen, he’s “doing the politics” and is “no longer a neutral party”!
AK, 2008 was two years after Muller had first criticized Michael Mann. Do you think that is significant? Or did you think it was before his criticism of Mann?
hr001, very good point about Muller becoming an advocate. His statements are not those of a careful scientist but of an advocate. Steve McIntyre seems to be on the best track for getting Muller to issue a Corrigendum or clarification.
Funny then how no one was able to demonstrate how Mann was so very wrong, demonstrably wrong.
Instead multiple independent groups of researchers found him to be right, demonstrably right.
Where are you hiding the proof that Mann was wrong all along? Can we look forward to your peer-reviewed research being unembargoed soon?
Or is this yet another case of sour grapes by a “skeptic” who lost a few teeth to Mann’s hockey stick?
Robert,
go back and read where Mr. Muller thinks it is OK to misle and distort the Science to SAVE THE WORLD!! Mann and those other researchers have done the same. Mann’s work and those associated have used a number of progressively less effective TRICKS and manipulations of the data including inappropriate proxies to SELL their myth. The latest studies are almost back to a realistic record due to all the TRICKS being caught and publicized primarily by Mr. McIntyre. If you wish to continue in your delusional world there is nothing the rest of us can do about it, but, you are wrong.
Try comparing Mann’s latest efforts to his first Iconic distortion and see for yourself how much his work has “evolved” under scrutiny which he went to great lengths to prevent!!
Ron,
thank you for the heads up. Apparently Mr. Muller comes under the same leftard heading as others who have stated it is OK to misle, distort, and LIE to SAVE THE WORLD (even when you have no evidence of it needing saving)!! This puts paid to any idea that he is an honest, objective broker. That was already starting to come across from his MEDIA WHORE activity and dismissive treatment of Mr. Keenan’s concerns, not to mention the shoddy treatment of Mr. Watts.
Extremely interesting that in the article he recounts how he broke with the Sierra Club over not supporting Nuclear Energy forcing us into more CO2 production.This was IN THE EARLY 80’s. This man BELIEVED in Goreball Warming BEFORE there was a shred of temperature data to support the theory, just like James Hansen and others!!!
Judith,
You’ve recently asked the question “Does the Aliasing Beast Feed the Uncertainty Monster?” with the implication that it did perhaps overfeed it, I would say.
So maybe the BEST group have now put your pet on new regime of a reduced diet and daily vigorous exercise?
And it could well be quite a lot fitter, leaner and smaller already?
Is it right that the essence of trhe Berkely findings is that the rural and urban sites have been warming at a sinilar rate and so it is assumed that UHI effects are not a significant factor for determining the direction and rate of the temperature trend ?
Well, if so, how about the proposition that the incremental rate of nearby development is on average the same for both rural and urban sites ?
Wouldn’t that produce just such an outcome ?
During a period of development isn’t it just as likely that it will occur near a rural site as near an urban site ?
Yes. I will say it again: rural/urban is a false dichotomy regarding UHI growth.
Stephen:
Correct Stephen. The areas classified as truely urban have very little space for further development. Those areas can grow around the edges, but that is likely to be far from an urban thermometer. But in the rural areas, and this would include small cities and suburbs, it’s more likely that there can be new build very close to the thermometers. And in highly urban areas they often have to tear something down to build something new.
Henry@Stephen
Hi Stephen!
forget about GHG emissions and UHI effects, etc.
I discovered that the SH does not have (much) warming at all
http://judithcurry.com/2011/10/20/berkeley-surface-temperatures-released/#comment-125821
that is apparently also confirmed by some antarctic measurements
http://wattsupwiththat.com/2011/10/20/pielke-sr-no-surprise-about-best/#comment-774013
I think there is a relationship between deforestation (which causes cooling) and the “greening” of the planet which appears to be more prevalent on the NH and that is what causes some heat entrapment, in addition to natural warming.
Cheers! (I raise my glass full of red wine!)
Henry
Hi Henry, keep plugging away with your stuff.
I think the reason that SH shows little warming over the period concerned would primarily be a result of oceanic thermal inertia.
I’m sure the amount of vegatation is relevant but it would be hard to put a number on it in order to rank the effect amongst all the other variables.
I prefer to focus on the cloudiness/albedo/jetstream aspect which appears to me to solve a whole lot of observational features.
The UHI aspect is a side issue as far as I am concerned. I am satisfied that there was some warming from LIA to date but I think the more recent numbers might be a little too high due to largely incompetent ‘adjusting’ for the undoubted UHI effect.I seem to have got some agreement here that the split between rural and urban is a false distinction that does not help in ascertaining the presence or scale of the UHI effect.
A small change in a rural area will have a larger effect than a large change in an urban area so the two types of site may well show a similar rate of temperature change on average but that does not mean there is no significant UHI effect. The Berkely conclusion to that effect is flawed and deeply unscientific.
I’m also pretty sure that my sun/ocean combination is the real big hitter for climate change as the Berkeley group is hinting with its comments about the AMO and oceanic effects generally.
Judith –
What do you think about being lumped into a group called “media whores” by Willis?
Is that yet another example that shows the “asymmetry” in the impact of tribalism on the opposing sides of the climate debate?
Perhaps yet another example on the list of “intemperate” remarks on blogs that you dismiss as being irrelevant? That list seems to be getting longer by the day.
Judith –
More mere intemperance. Muller = Tito
http://calderup.wordpress.com/2011/10/22/hoodwinked-by-berkeley-earth/
Joshua says:
What do you think about being lumped into a group called “media whores” by Willis?
Henry@Joshua
which Willis are we talking about here, and where is this comment that contains this remark?
Rg\rds. Henry
Check out WUWT.
Willis goes on another of his error-filled, over-reaching narcissistic rants – personally attacking climate scientist to prove that he’s above their tribalism. It’s hilarious.
He was also corrected and admitted his mistake(s). When will Climate Scientists reach that level??
Henry@Stephen
Look, if I were Berkeley, I would limit my sample of stations to say, 50 , and they must be carefully chosen,
1) they must be well established for at least 50 years (airports are good, there is no urban build up and the CO2 emissions is supposedly to make things “worse” but we know that that does nothing to temps anyway)
2) they must represent the earth 70/30 water ,
70% near coast or in or on islands in the oceans
30% on land, in land
3) you have to balance NH and SH similar to my tables here,
http://www.letterdash.com/HenryP/henrys-pool-table-on-global-warming
that is what I can think of off-hand.
reason why I found a connection with vegetation are the results that came from an Argentine station, cooling rate quite dramatically@ 0.066 degrees C per annum since1974. Can only be due to de-forestation in that area.
Obviously the opposite is happening in the NH since the green movement saw everyone planting trees and gardens.
Note that the main driver of the increase in warmth is natural, between 0.03 and 0.04 degrees C /per annum (past 35 years) on maxima.
My point is that we must be looking at how much of that heat is trapped and why.
One over the square root of 50 is 14%.
Do you have the slightest idea what the implication of this number is?
no.
please ignore him then.
I am saying all of your results (and berkeleys & & &) even it is from 1000s of stations without showing me what the rise in maxima and minima is, are all pretty useless because it does not tell me why the temperature rose.So why don’t the experts concentrate on giving us the rise in minima and maxima as well?
http://www.letterdash.com/HenryP/more-carbon-dioxide-is-ok-ok
There is a common narrative in the breathless, orchestrated press accounts of the BEST results: Climate skeptics do not believe that the World has been warming. The BEST results confirm what everybody else (the smart people) already knew; the World has been warming. The Global Warming debate that was over, is over once again. The science that was settled, is settled again. The team has a new convert and hero, Dr. Richard Muller.
But this media blitz is hyperventilating BS propaganda. Most climate skeptics know that the World has warmed in recent history. And most of those who agree that there has been warming believe that some of that warming is the result of radiative forcing from anthropogenic CO2 accumulation in the atmosphere.
Richard Muller, the orchestra leader/former skeptic, sets up the straw man:
http://online.wsj.com/article/SB10001424052970204422404576594872796327348.html#printMode
“The Case Against Global-Warming Skepticism
There were good reasons for doubt, until now.”
Dr. Muller goes on to lay out the straw man version of the significant objections of the skeptics, which were the basis for their erroneous disbelief of the team’s story on AGW. Bad surface stations, and UHI effect. Well, he promptly lays those issues to rest with his as yet non-peer-reviewed study results. The problem is, those aren’t really the most significant objections of the skeptics, and he really did not lay them to rest. See Anthony Watts on the surface stations issue:
http://wattsupwiththat.com/2011/10/21/best-what-i-agree-with-and-what-i-disagree-with-plus-a-call-for-additional-transparency-to-preven-pal-review/#more-49721
See William Briggs reaction:
http://wmbriggs.com/blog/
Briggs sums it up succinctly:
“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.”
Don –
It is to laugh!
and
Does Anthony Watts imagine we have goldfish memory, or what?
6 months ago, he, along with many other “sceptics” were making strong assertions about the temperature record which have now been shown to have been wrong.
Now he’s pretending he never held that position.
Is this respectable behaviour?
Well said Don!
I can never understand why all these so called experts come with data that do not show us the increase in maxima and minima that would tell us why the temps. are rising.
http://www.letterdash.com/HenryP/henrys-pool-table-on-global-warming
“I think everyone in the room agrees that the climate has warmed over the last 50 years, for whatever reason: we saw plots of land atmospheric temperature, marine atmospheric temperature, sea surface temperature, and (from Prof Svensmark) ocean heat content, all with a rising trend.”
******************************************************************************************
Incorrect. Not “everyone” agrees the earth has been warming for 50 years, indeed this has been a bone of contention among the so-called skeptics for years. A very cursory review of the skeptics blogs would easily reveal a plethora of comments and posts yucking it up about “global cooling”. And these skeptics have made a cottage industry out of claiming that the existing temperature reconstructions were fraudulent, and were constructed by scientists who manipulated data to exaggerate, perhaps even fabricate a warming trend.
In short, there’s undeniably a whole lotta goal-post moving going on.
A. Wattts himself, while supporting the BEST project, said he didn’t know if the BEST temperature reconstruction would show cooling, or warming. Those are his own words….
***************************************************************************************************
A. Watts, March 2011, on the BEST project: “I have no certainty nor expectations in the results. Like them, I HAVE NO IDEA WHETHER IT WILL SHOW MORE WARMING, ABOUT THE SAME, NO CHANGE, OR COOLING in the land surface temperature record they are analyzing”…..” And, I’m prepared to accept whatever result they produce, even if it proves my premise wrong. I’m taking this bold step because the method has promise.”
http://wattsupwiththat.com/2011/03/06/briggs-on-berkeleys-best-plus-my-thoughts-from-my-visit-there/
and….
Joseph D’Aleo and Anthony Watts, 2010:
“Instrumental temperature data for the pre-satellite era (1850-1980) have been so widely, systematically, and uni-directionally tampered with that IT CANNOT BE CREDIBLY ASSERTED THERE HAS BEEN ANY SIGNIFICANT “GLOBAL WARMING” IN THE 20TH CENTURY.”
From: “Surface Temperature Records: Policy-Driven Deception?”, June 2010. SPPI paper.
“Most climate skeptics know that the World has warmed in recent history.”
Even a moron should be sensible enough not to argue with that meticulously constructed factual statement. I challenge the hapless moron who would be so foolish as to make the argument, to prove that most climate skeptics do not know that there has been some period of warming in recent history. Show me one skeptic who believes that the temperature always stays the same, or declines.
Of course some morons will mistake the lack of confidence in the quality of the temperature record, and the ability to reliably use the record to prove “global warming”, to mean that some prominent skeptics deny global warming.
Skeptics want to see some proof that warming is unusual, man-made, and dangerous. What is so freaking hard to understand about that?
Does anybody here really believe that rural areas, far from those big Urban Heat Islands, warmed more than those big freaking Urban Heat Islands warmed as they grew ever bigger, from 1950-2010? If so, please explain how that happened.
Rural area – creating a large paved area where there was previously grass would make quite a difference if close to the sensor.
Urban area – replacing one large building with another larger building won’t make much difference.
Development around both rural and urban sites occurs in parallel once the urban location has become established. The largest difference is created in the early stages of urban development. Once past that the two types of site would develop pretty much in parallel but with occasions when differing types of development could make either site warm or cool faster or slower than the other.
The development of the motor vehicle made it far easier to develop more rural sites over the period 1950 to 2010. Indeed due to preferences for living and working in more rural communities one could envisage that many rural sites caught up with the urban conurbations to some extent.
Proper research into the complexities of the UHI effect has never been carried out.
I think I explained it more than once above.
Don,
you have to step back and consider whether you are postulating a non-ending warming to destruction for UHI like was postulated for GHG’s. Even with open blast furnaces there will be a rational limit to how much a city will warm. Insulation is better, heaters, A/C’s, and engines are more efficient, no more room for more blacktop, the higher it builds the less insolation reaches the ground, etc.
A large city may actually have an internal 10C UHI. It may have a growth rate of that UHI of only .1c/C now. That is, its fast increase in UHI may be over. A small rural area may have a 1c UHI with a growth rate of 10c/C. The city is a lot hotter than the rural area but the rural area is warming a heck of a lot faster.
What these studies are showing is that there is no simple correspondence between size and UHI or temp. There are too many factors to be able to draw useful conclusions about UHI from straight statistics when the information needed to separate out the contributing factors are absent..
et tu steven,
Look, I have heard that story. Muller says about the study:
“We then conducted a temperature analysis based solely on “very rural” locations, distant from urban ones.”
If those sites are classified as “very rural” today, how much increase in population density and infrastructure could they have experienced, since 1950? I will help you; not much. How many buildings higher than three stories do you think you will find in those areas now? Go to Google Earth maps and have a look. Zoom in. Then have a look at the 1,000 largest Urban Heat Islands and go to wiki to find out how much those things have grown in population and pop density, since 1950.
And on the car thing, “very rural” is not suburban. I am pretty sure they mean that they took locations as far away from Urban Heat Islands, as they could get.
“Very rural” simply means any place that is marked on the Modis map as being more than10 kilometers from a place that is marked urban. Modis has no such standared. BEST made it up. In theory, very rural can be more heavily built than rural. Modis simply has urban and rural. Urban is any place where the surface is more than 50% built and where that built is unbroken for more than 1 square kilometer. “Very rural” areas could be as much as 49.9% built and urban areas could be as little as 50.1% built.
Tilo,
OMG! It’s worse than I thought! If that is what they are passing off as “very rural locations, distant from urban ones”, then they are either deliberate deceivers, or they speak some cryptic language meant to be intelligible only to themselves. You would think that if it is the latter case, they would have the common decency to translate to plain English, when they are writing articles in the Wall Street Journal and putting out press releases.
Judith, is this the best that can be done to determine if there is any significant UHI effect? Isn’t there data from areas that are truly rural? I have flown over a lot of very sizable open spaces. Are there thermometers there? And if not, how do we know what has happened in those big sparsely populated areas, during the period under study? Isn’t this important enough to get right? Jump in here any time. Your name is on this stuff.
Don: Here is the description from Modis.
http://www.sage.wisc.edu/people/schneider/research/schneider_erl2009.pdf
They have rural and urban only. You can go to the BEST web site to the UHI paper and see that BEST extended the classification to “very rural” and “not very rural”. The only difference being that the “very rural” must be 10 kilometers from any place on the map that is classified as urban.
Thanks Tilo,
I looked at the paper. I see what you are saying. Not exactly as advertised to the unwashed masses. Most people think very rural means raccoons and squirrels outnumber the people. I wonder if Judith has any issues with the way this crap is being portrayed in the popular press.
Tilo, if you read the paper even moderately closely, you would find that they have a good reason for their choice of .1 degree. It is because many of the station positions are only reported to that precision, so they are not pulling this out of their nether regions, a very rural station is a station which is classified as “rural” and is at least .1 degree away from an area classified as “urban”.
You are basically arguing that the very rural stations are all built up and that they have a UHI effect equivalent to Tokyo. BEST basically came up with an answer of 0 to the question, which is consistent with previous research on this topic, some of it going back more than 20 years. UHI has basically no effect on the trend, although it does have a large effect on absolute temperatures. I think their approach is basically good (it is fairly obvious) although it may miss some subtleties, but I doubt that going the next step is likely to change the result from “basically no effect” to “OMG, the effect is huge!”.
They do seem bothered, as they should be, by the fact that they came up with a negative sign for the effect and they don’t really have a decent explanation of why that is. Some of it may be leakage from the siting bias (urban stations may be more likely to be poorly sited). I don’t know, they don’t know, and you don’t know.
“a very rural station is a station which is classified as “rural” and is at least .1 degree away from an area classified as “urban”.”
Yes, you are telling me what I already know.
“Tilo, if you read the paper even moderately closely, you would find that they have a good reason for their choice of .1 degree. It is because many of the station positions are only reported to that precision, so they are not pulling this out of their nether regions,”
Yes, I know that also. Apparently you are complaining about how closely I read the paper without having read what I said before. Their motivation for picking the distance is completely irrelevant. What is relevant is whether or not that tells you anything about how built up an area is that is classified as “very rural”. And it obviously tells you nothing.
“You are basically arguing that the very rural stations are all built up and that they have a UHI effect equivalent to Tokyo.”
No, I’m not, because it’s not the total build that shows up in the anomaly, but only the change during the time when the thermometer was present.
“UHI has basically no effect on the trend, although it does have a large effect on absolute temperatures.”
Absolutely wrong. The change in UHI is relected as a change in trend. So the question is not how big is the UHI effect, when looking at trend, but what is the increase in UHI effect. And it’s very easy to imagine that the change could be larger in what is classified as rural areas than in what is classified as urban areas. In fact, both the BEST results and the paper on the subject by Roy Spencer verify that. You can’t increase the absolute temperature with UHI without increasing the trend. Where did you ever get such a dumb idea?
Rattus,
If it’s I don’t know, you don’t know, they don’t know, what do we need them for? They are stating conclusions for things that they don’t know. The reason they got a negative number is because the data is not so good, or their method was inadequate to the task.
They got a number which is indistinguishable from 0. It was -0.19 deg/Century +- 0.19 deg/Century. As I stated, this is consistent with previous research.
Let me add one more thing, since you will probably miss it. When you install a thermometer in an urban area that is already built, you will not see, in the anomally, the amount of UHI that it had already acquired prior to that thermometer being introduced. The anomally record will only show the anomally since the thermometer was introduced. So you can’t assume than an anomaly temperature for a city will reflect all of the UHI influence that the temperature of that city is experiencing.
Don Montford and Tilo Reber
The preliminary UHI report of BEST has certainly not yet laid the UHI question to rest.
It concluded slight (if statistically insignificant) COOLING from urbanization, which is clearly in error, based on the many studies from all over the globe on the impact of urbanization on local temperature (and hence on globally averaged temperatures).
But the question is still open: has this spurious warming signal had a significant impact on the globally averaged temperature record or not?
The preliminary BEST report is unable to answer this question.
If one simply compares the post-1979 satellite and surface station records (UAH/GISS) one sees that the difference in global warming trend was 0.03C per decade (UAH = 0.14C, GISS = 0.17C).
The troposphere should theoretically warm more rapidly according to GH theory and there are serious questions regarding the SST record prior to around 2000, so a conclusive comparison cannot be made, but it would appear to me that an upper limit for the UHI would be around 0.03C per decade (or 0.3C over the entire 20th century).
Several studies also point into that general direction.
Basing its conclusion on “calm night / windy night” comparisons rather than direct comparisons of several urban and rural stations, IPCC states that it is less than one-fifth of that amount over land (and zero over the ocean)..
In all likelihood, it lies somewhere between these two limits.
But one thing is CERTAIN. It is NOT NEGATIVE as the preliminary BEST report states.
Max
manacker
It concluded slight (if statistically insignificant) COOLING from urbanization, which is clearly in error, based on the many studies from all over the globe on the impact of urbanization on local temperature (and hence on globally averaged temperatures).
Oh really?
So.. you trust your own broad reading of “the many studies from all over the globe” over the actual data?
I’ve seen plenty of studies from all over the globe discussing cooling effects of urban aerosols in the wind shadow of cities.
I’m unconvinced by your reasoning, as either you’ve seen such studies too, and are cherry picking, or you’ve never seen such studies and are not very well-informed on your one and only piece of evidence.. which is your lack of reading about something.
By such logic, the Universe and each of its parts only come into being when you read about them.
Judith…I can’t believe your name is on this piece of rubbish. You lot seem to have concluded there has been warming since the middle of the Little Ice Age. Over to Homer Simpson. “DOH!!!”
What have you accomplished Judith? Have you found the missing link that connects the warming to anthropogenic CO2? No, Have you refuted 30 years of satellite data, that scans 95% of the atmosphere while sampling billions and billions of oxygen molecules? No. Have you refuted what John Christy of UAH has claimed, that most of the warming is in the northern part of the Northern Hemisphere, and not global in scope? No. Has your statistical analysis identified whether the warming is local or global? No. Have you verified that all the data you used is accurate? No. Have you explained the lack of a warming trend the past decade, which is plain on satellite data sets? No.
What exactly have you done? You have re-iterated what everyone knows, that the world has been recovering from a period of cooling.
Please tell me how you can give so much credence to surface stations that take two a day temperatures and average them, when 72% percent of the surface, covered by oceans, is barely monitored. Please tell me why you see fit to ignore the modern technology employed by satellite telemetry, which renders your two a day surface readings obsolete.
I really thought you had an open mind.
Gordon, did you read what Judith’s involvement in this project is? I have yet to see her post a word of defense for it.
I have been looking at BEST’s methods and results, and I’m pretty sure they got it wrong. Not wrong in the sense that there is no warming, but wrong in the sense that the answer is much closer to HadCrut3. But it might be difficult to demonstrate. The clue to their error is in what you mentiond above, that the majority of the warming is in the northern part of the northern hemisphere. And their methods amplify the difference in a way that it should not. But they are using Matlab code, and I’m not anxious to shell out two grand for a copy of Matlab. GNU has a free Matlab like option called Octave, but it sounds like there are enough differences that conversion for that much code would be a pain. Know of anyone that has a copy of Matlab they don’t need?
By the way, do you have a link for that Christy paper?
Tilo,
“Gordon, did you read what Judith’s involvement in this project is? I have yet to see her post a word of defense for it.’
Yes, you are right. But hopefully it’s because she needs to wait for the full release of the papers before being able to comment further.
I would suggest that it would be beyond belief that she’d try to walk both sides of the street on this one. That is, pick up the academic points associated with having her name on a scientific paper, but which she wasn’t prepared to defend.
Tilo…you have to dig for it with Christy. Here are three links that cover most of what I have said about him.
http://meteo.lcd.lu/globalwarming/Christie_and_Spencer/25years_highlite.pdf
note the temperature contour map on page 9 which covers 1978 – 1993. I have seen the same maps up till 2010, and the temperature distribution is similar. All of the maps, including a 25 year version are available on his site, in a hit or miss situation. The server goes wonky at times.
http://science.house.gov/sites/republicans.science.house.gov/files/documents/hearings/ChristyJR_written_110331_all.pdf
http://discovermagazine.com/2001/feb/featgospel
In the last link, the 0.2C warming from 1979 – post 1998 is mentioned. He also laments how eco-alarmists have tried to distort his satellite data sets through innuendo.
What’s your problem Gordon Robertson? If the paper just “re-iterated what everyone knows” then why are you so offended by it?
Maybe it is because they have only re-iterated what everyone knows, but the media blitz they have unleashed can be summed up by paraphrasing the common headline, “Knuckle-dragging Climate Skeptics Stuffed, Real Good!”
“Knuckle-dragging Climate Skeptics Stuffed, Real Good!”
Yes. I must admit that sounds good to me, too!
I am offended because the lead author has suggested skeptics are trying to infer there has been no warming. He seems to have confused a specific reference over 30 years by John Christy of UAH, who has claimed his satellite data showed no warming from 1979 till after the 1998 El Nino extreme.
I don’t know of any skeptic who has argued there has been no warming since the pre-Industrial period. Most skeptics seem to agree there has been warming, they are questioning the connection to anthropogenic warming.
-Gordon Robertson: “Judith…I can’t believe your name is on this piece of rubbish.”
-Tilo Reber: “Gordon, did you read what Judith’s involvement in this project is? I have yet to see her post a word of defense for it.”
*************************************************************************************
Tilo, Perhaps because you didn’t even read her entire post?
Judith Curry: “I am honored to have been invited to participate in this study, which I think was conducted very well.”
What’s with all the attacking of her? Did anyone really believe the claims of many skeptics that NASA, NOAA and HADCrut temperature reconstructions were based on the fraudulent work of scientists who faked and manipulated data to fabricate temperature reconstructions? Was there some anticipation that BEST would blow the lid off this alleged conspiracy?
“Did anyone really believe the claims of many skeptics that NASA, NOAA and HADCrut temperature reconstructions were based on the fraudulent work of scientists who faked and manipulated data to fabricate temperature reconstructions? ”
Yes, many did believe that and probably still do. Big mistake. They should have taken a lesson from OJ Simpson’s lawyer, Johnnie Cochran. Yes, he did an excellent job of arguing his client’s case and getting the acquittal, but he wouldn’t have allowed himself to be influenced by his own beliefs!
I did come across something that looked a bit suspicious, to me,
when I checked some data that came from the Gibraltar, UK
and compared it with data from neighbouring Spain and Morocco.
http://letterdash.com/HenryP/what-hanky-panky-is-going-on-in-the-uk
peterstone,
Can you furnish any other of Judith’s comments on the study, or the characterization of the findings in the media blitz? Is she honored by the way they are spinning this? Maybe you can get her to talk.
“Maybe you can get her to talk.
Maybe. We’ll see. She seems to want to use this blog to raise doubt and increase uncertainty and when she does “talk” she doesn’t actually say much. IMO.
tempterrain
You do our host injustice when you spout off things like:
It appears to me that she simply wants to point out where there IS uncertainty, which has been understated or simply ignored by the IPCC “mainstream consensus” presentation of things..
This may end up raising doubt about some of these more uncertain IPCC claims but, hey, that’s what science is all about.
It’s called “rational (or scientific) skepticism” in the scientific sense (check it out on Wiki).
Max.
Really? You just endorsed a statement – in direct contrast to Judith’s opinion – that BEST’s results are “absurd.”
Are statements (with absolute certainty) that BEST’s results are “absurd” uncertainty?
I guess in a post-normal science world, statements made in absolute certainty are expressions of “uncertainty?”
It boggles the mind.
Joshua
BEST has already stated the uncertainty of their preliminary UHI result.
I have simply said that it is absurd to believe that urbanization has resulted in cooling, as the preliminary result would indicate (although stating that the cooling is not statistically significant)..
If you think that urbanization results in local cooling, and hence a globally averaged cooling, please state your reasons for this opinion.
I would personally call such an opinion “absurd”.
Max
manacker –
I’m having trouble reconciling two statements of yours:
Here you say “preliminary results would indicate”:
Yet earlier you endorsed the following statement:
So – which is it? Is it that the BEST report “would indicate” (meaning that their analysis of the data suggest) that UHI would result in cooling? Or is it that the BEST report “stated” that UHI has a statistically insignificant cooling effect?
Scientific analysis requires that you sometimes analyze the data in ways that return counter-intuitive results. The appropriate response to that is to further analyze the data – to find if the first analysis was flawed, or contrary-wise if your intuition was flawed.
The language used there is important.
But either way, when you say that an analysis or a conclusion are “absurd” without proving the “absurdity” scientifically and simply stating that it is so because they contradict your intuition, it is, IMO, a textbook case of being overly certain and confusing opinion with fact.
The results of a data analysis are neither “absurd” or not “absurd.”
They might be the product of a valid process or not. Intuition can clue is in to when analyses return invalid results. But when you call a data analysis “absurd” without proving the methodological flaws, you aren’t following what I consider to be valid scientific principles. And further, even if you consider a statement of conclusion based on what you suspect but haven’t proven to be flawed analysis – if the conclusion follows from the analysis, by definition it may be wrong, but it is not absurd.
Max and Joshua,
I agree that the BEST results on UHI are surprising. However, only a fool would call them absurd until the the papers are released in full and have been thoroughly understood. Everyone can then take a good look at how they’ve been arrived at, and what they mean in terms of the accuracy of the surface temperature record. Is the UHI effect significant in terms of the overall record or does it have just a small effect which can safely be neglected?
What isn’t surprising is that the BEST results agree so well with previous studies.
Also not surprising, but is nevertheless worrying, is Judith’s unwillingness to say anything in reply to the many legitimate questions, about papers which bear her name, which have been been raised on their treatment of the UHI effect. It may be she needs to wait for a time before she is able to comment further. If this is the case, why cannot she say so?
Tilo,
I went back up in the thread to see if Judith had been responding to any questions-got tired of looking-but I noticed a post of yours I had missed. Very perceptive and I believe correct:
“No, you are confusing classifications with reality. Certainly it is true that the more built up an area is, the more UHI effect you would expect there. But all we are talking about with UHI is the effect on instruments that results from the addition of man made structures around the instrument. So, indeed, if you have an instrument on a farm and if you put a concrete parking lot next to it, the instrument is going to start yielding higher temperatures. And that is what we care about – not the fact that the name has “urban” in it. On the other hand, if you have a thermometer in the middle of a city and it’s already built up, you won’t see much change if you put up another building or parking lot a mile away. Of course total city build up will effect the temperature over time because the larger the warmed area is the harder for the heat to dissipate. But when you start to build in a rural area, immediately around the instrument, the effect can increase even faster than it can in an already built up urban area. And that rural area won’t change its classification status for a very long time and until the building gets really dense.
I’m only pulling this out of my backside, but I would guess that out of 39,000 thermometers in their study they would be lucky to have a thousand with no UHI effect. These things are almost never put in national wilderness areas. They are near roads, close to where someone is able to read them and maintain them. And such areas are naturally subject to population growth.
But really, do you need to know any more than that they came up with a conclusion of “none, or negative, UHI effect” to know that they are doing it wrong. Come on – a little common sense please.”
It seems to me that anywhere you have human habitation (a mud hut with a family of five and a cook-fire) you have some human warming effect, on the local environment. The population of the earth grew from 2.5B in 1950, to 7B today. The thermometers are mostly located where the people are. To believe that there has been no significant human heat island effect, as population increased around those thermometers, is not sensible.
There are 105 people per square mile of land on the Earth. That’s probably about 10 to 25 fire/high heat sources each night per square mile on average. For an urban area that must be 10 to 100 or more times more than that. It doesn’t seem likely there could not be any UHI. Look at how just a few smudge pots can keep frost away from oranges or airports.
Kermit,
The effect of UHI on the temperature of the earth is very small. But the effect of UHI on measurement of the temperature of the earth is significant because such a high percentage of the thermometers are where there is a lot of build. You don’t get many thermometers out in the middle of the national parks, on the farms, or in the desert and in the wild.
If it were possible to put a thermometer on every acre of land there would be variation in temps for differing reasons. The argument against uhi is that it all evens out in the end due to the large sample size. Your argument is that measurements are not diverse enough for that to be true.
Don Montford
You conclude:
Let me add to that:
Max
Ok –
So a number of Judith’s “denizens” call BEST’s report “absurd,” (it’s all over this thread) and Judith says the following about BEST’s report:
What are the criteria on which Judith can characterize the expectations of her “denizens” WRT to “expectations for understanding and characterizing uncertainty” when so many of her “denizens” express absolute certainty in their opinions in direct contradiction to her own conclusions?
Any study that does not show the development of maxima and minima TOGETHER with the averages is pretty useless.
http://www.letterdash.com/HenryP/henrys-pool-table-on-global-warming
Note that you did not reply to me here,
http://judithcurry.com/2011/10/20/berkeley-surface-temperatures-released/#comment-126505
http://judithcurry.com/2011/10/20/berkeley-surface-temperatures-released/#comment-126933
Josh-ua,
You are a little putz. We are not Judith’s trained seals. She allows us to think, and disagree. This ain’t realclimate. Maybe she has some explanation for the finding that UHI don’t mean squat. Maybe she will actually help us out here and explain how they got that absurd result. Maybe the study was conducted very well, but the data ain’t good. Try using your head, Josh-ua. And try being something other than a nuisance.
Stephen,
I owe you an apology. Upthread I was intemperate and disrespectful in replying to a post of yours, that I subsequently learned I had misinterpreted. Sorry for that.
This particular discussion has got me more agitated than usual:) And I am disappointed to find that some, who I considered to be intellectually honest, are letting us down on this one. Fred and Pekka are dodging and weaving, and Judith remains mute.
Don,
“Fred and Pekka are dodging and weaving, and Judith remains mute.”
Makes you wonder if this is a trial balloon in the Communication area to see how much they can get past the blogging and Science community donut!!
Joshua
You really should work on your reading skills.
I have not said that the BEST report is “absurd” (as you claim), simply that a conclusion in one small preliminary sub-report on UHI which postulated that urbanization had caused cooling was “absurd”.
Show me, Joshua, where urbanization can cause a spurious cooling signal, if you can.
Obviously you cannot, because such a hypothesis would be “absurd”.
Max
Max – The BEST study did not purport to show urban cooling. Rather, it reported that a subset of stations classified as “very rural” warmed faster than the overall average. The urban stations, or more precisely those not classified as very rural, also warmed on average rather than cooling. The cause of the disparity is not known, but if the disparity is real, it doesn’t necessarily signify a cooling influence on urban locations that reduced their warming rate. That is possible (and possible mechanism have been listed above), but it is also possible that there are features of the “very rural” subset that caused those locations to warm particularly fast.
It is true that to claim overall urban cooling would have been hard to justify, but the actual conclusions, which don’t make that claim, remain as a finding that must be further evaluated empirically, and can’t be dismissed as absurd.
Max –
You endorsed this comment by reproducing it:
Did the BEST report “state” what is reported in the comment you reproduced and (presumably) endorsed?
As I understand them, BEST’s results seem to indicate that the UHI effect in urban environments is less than that of rural environments. That’s what their analysis shows. Various conclusions can be drawn from that. Some would be absurd, and some would not be absurd. The absurd conclusions would be those that aren’t supported by the data.
If the data suggest conclusions that you find counter-intuitive, you have two reactions available that would be scientifically valid. The first is to prove flaws in their methodology. The second would be to reevaluate the basis of your intuition.
And please keep in mind, Max.
When a statistical analysis returns results that are not statistically significant, they are not considered valid evidence to support a conclusions (except that no statistical significant was found).
Fred,
The actual conclusion is that “very rural” areas warmed more than the average of other areas that includes all more densely populated (more built up) areas. That is not credible. It ain’t physical, Fred.
The most likely explanation is that UHI affects the temperature time series so little that some random variation was enough to produce a negative result on statistically non-significant level.
There are certainly also other possibilities, but this seems most likely.
Muller said in the Wall Street Journal:
http://online.wsj.com/article/SB10001424052970204422404576594872796327348.html#printMode
“To study urban-heating bias in temperature records, we used satellite determinations that subdivided the world into urban and rural areas. We then conducted a temperature analysis based solely on “very rural” locations, distant from urban ones. The result showed a temperature increase similar to that found by other groups. 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.”
If they had truly selected only “very rural” areas that are distant from urban ones, they would have not gotten that result. The analysis is only done on land stations, so it is deceptive to make the point that urban regions cover only 0.5% of the globe. He also does not mention that the thermometers are located where the people are, disproportionately to where they are not. Is this really science?
Pekka,
I don’t care what the explanation/excuse is. The conclusion that the UHI effect is negligible, or negative, based on that study is bullcrappy. It looks like deception to me. Why are they trumpeting this crap?
The sites they used for comparison are not “very rural” and they are not far from urban areas, and they know that. A real study would have used sites that were very sparsely peopled, or empty. But those areas don’t got thermometers, so they just make it up as they go along. I repeat the question: Is that science?
Joshua: “You endorsed this comment by reproducing it:”
I’m happy to say that the BEST results on UHI are absurd, even downright stupid. That doesn’t mean that they didn’t get the results that they got. It means that the test, as they designed it, had no ability to isolate and quantify the UHI effect. The reasons are multiply given in the posts I wrote above, and I don’t see your post addressing those reasons at all.
Pekka: “The most likely explanation is that UHI affects the temperature time series so little that some random variation was enough to produce a negative result on statistically non-significant level.”
No, Pekka, that isn’t even close to being the most likely explanation. The most likely explanations are given in the discussion above; which you either didn’t read or didn’t understand.
.. or didn’t agree with.
Hi Pekka.
The thing I love about blogs is that even the best (including me) can make daft comments :)
Of course UHI has a negligible effect on global temperatures. However it makes a significant contribution to the temperatures recorded where the thermometers are placed.
Haven’t you ever driven from country to town and watched the vehicle sensor rise by up to 5C ?
For accessibility and cost reasons thermometers tend to be situated nearer the town than the country so the temperature bias reflects that fact.
Even if a thermometer were placed in a pristine site you could bet that pretty soon the mere presence of the thermometer itself and the traffic it induced would cause more of other activities to be attracted to the area of the recording site.
Wherever one man goes others will follow.
Pekka – As I read the paper, I perceived it to claim that any reverse UHI effect would have very little importance compared with the overall global warming trend. The slope difference had a 95% confidence range between 0 and -0.38 C/100 years, and that still requires explaining before dismissing it as spurious. However, from a climate perspective, the actual value within this range has little significance.
When i said, what I consider the most likely explanation, I didn’t mean that the issue should not be looked at more carefully.
It’s possible to check the present situation at many weather stations and it’s possible to collect additional data on recent history. Such information would help in judging, what kind of changes have been possible or likely in the past. That cannot be done on all weather stations, but collecting such data might help in understanding the data. That way the reliability of the temperature time series may be improved further, although it may impossible to base a better overall estimate on such considerations.
Fred,
You are missing the point. They are advertising this as putting to rest the issue of UHI effect on the accounting for global temperature change, over a period of time when the population of the earth increased from 2.5B to nearly 7B, and most of that increase was in urban areas, and suburban areas close to cities. Actually, they are claiming their studies have eliminated all good reasons for Global-Warming skepticism, without making any legitimate attempt to define Global-Warming skepticism.
“The Case Against Global-Warming Skepticism
There were good reasons for doubt, until now.
By RICHARD A. MULLER ”
Don’t you think they are overselling their results a tad? What about global warming skeptics’ doubts about the level of knowledge of clouds, natural variability, water vapor feedback, etc.?
Do you believe that their study on UHI effects is good science, Fred? Do you believe that the sites they selected for comparison are actually “very rural”, and distant from urban areas? When they say the satellite separated the urban areas form the rural areas, and they selected only the “very rural” areas, out in the freaking boondocks, isn’t that misleading the public?
Stephen,
The point where the thermometer is placed has an influence. A few weather stations have been found, which are located in very bad places and obviously subject to increasing UHI effect, but that leaves it totally open, whether that is common enough to be significant for the global average. The statistical analysis of BEST does not support the existence of such an effect, neither has any earlier statistical analysis supported the existence of significant effect of that type in the averages.
Some people have argued that the result does not make sense at all or is even absurd. That leads one to wonder, whether the problem is in the data or in their ability to see the truth.
The list of possible explanations for the observed effect is long. I stated that for me the most likely basic reason is that the whole effect is small on average. That allows any weak effect including just random variability to determine even the sign of the outcome of the analysis.
Pekka,
I think you are missing the point.
ALL sites are affected by UHI as soon as ANY change to the site characteristics occurs. Sites are usually placed in a location where access is easy and cheap which implies SOME nearby development already and with time that gets worse.. Whether urban or rural development the consequent UHI effects occur on average in parallel so it is not significant that there is no difference in the rate of change between so called rural and so called urban sites. One must focus on INCREMENTAL and not ABSOLUTE UHI effects and they are more likely to be larger at less developed sites simply because replacing one building with another in an urban centre will make little difference but construction near a rural site will make a big difference.
It is not for nothing that the site requirements of the UK Met Office back in the 60s were so rigorous that I never considered myself capable of meeting the necessary standards. Those standards have now been abused at virtually EVERY site as Anthony Watts’ work testifies.
The surface temperature record is affected to a significant degree and the BEST findings fail to deal with the issue for the reasons I have stated.
Stephen,
I haven’t seen any analysis that tells on a serious error due to UHI in the temperature time series. I have seen a lot of claims and speculation, but that’s just claims and speculation.
Can you give a reference to some analysis that indicates a significant effect.
Isn’t it enough to know that the climate professionals take it seriously enough to make adjustments to try and compensate for it?
Perhaps we should be asking them ?
My beef is that they don’t know enough about changes in individual site characteristics for those adjustments to be accurate and as a consequence I suspect that the adjustments are inadequate resulting in a net warm bias on average.
Tilo –
Yes – at least you’re being consistent.
I’ve read some of your posts on that topic – but don’t know enough about the details to determine their merit. Sure, if your analysis is valid, then mistakes in classification could invalidate their conclusion that their findings invalidate a differential in the effects of UHI in a way to significantly distort the rate of change in identified temperature trends. If they compare “good stations” to “bad stations” or “urban stations” to “rural stations,” and their classification structure is inconsistent, then your argument seems logical to me.
I assume their methodology to be valid in that regard. If your argument is correct (and the arguments of others that I have no reason to trust any less than I trust your arguments), then your point seems like a rather basic one of methodology. My assumption is that such a basic mistake in methodology will be identified in the peer review process. I do make a certain estimate of the probability that such an experienced and recognized group of experts in highly technical statistical analysis would make errors of the basic type you mention – but it certainly is within the realm of possibility.
Your argument is distinguished from that of others. You are saying that their results are “absurd” because they don’t follow from a valid methodology and you identify what you see to be flaws in the methodology. That is different from not identifying the flaws in the methodology and: (1) attributing conclusions to the authors that they never made or, (2) saying that the conclusions that they do make – which follow from their analysis of the data – are “absurd” because they seem to some people to be counter-intuitive.
Tilo –
Actually, i have another question for you:
Did they state state that UHI results in (statistically insignificant) cooling?
First, I don’t see how any reasonable statistician would say that a finding that doesn’t reach statistical significance would warrant a conclusion as you describe.
Second, there is a difference between stating that UHI results in cooling in some areas as opposed to others, and saying that it doesn’t have a statistically significant impact on analyzing trends in such a way as to give a false impression of degree of warming. As I recall, they acknowledged quite forwardly that UHI has a warming effect – but that it doesn’t have that effect in a way that warming at urban, or “bad” stations relative to rural, or “good” stations significantly distorts an analysis of trends.
Some quick clarifications:
No. First, they said the trend in their “very rural” sites was higher than the trend in the total dataset. While it was discovered during their test for UHI, they didn’t attribute it to anything in particular.
Second, it was not statistically insignificant. The trend since 1950 wasn’t statistically distinguishable from 0, but that is a single trend. Had they picked a different starting point, it would have given a statistically significant result. This is made clear by their paper which says:
Joshua
Since you apparently missed the second half of my sentence when you read (and later quoted) it, let me repeat it for you again:
Got it this time? (Please read several times, if you did not grasp it at first.)
I also asked you to explain to be how urbanization can cause a spurious cooling signal (which you have, so far, been unable to do).
Max
Two issues, first UHI is the absolute change of temperature above the theoretical background. If there are no further changes and the excess energy does not increase the UHI will remain stable and contribute no more to the trend even if it is 15c higher than the baseline!!! For UHI to affect the temperature anomaly it must be dropping due to energy efficiency, more trees, green areas… or increasing due to more waste energy, blacktop and concrete…
If the city has increased area vegetation, removed blacktop, and increased the efficiency of energy use along with better insulation I can conceive of a NEGATIVE UHI TREND until the changes stop or are balanced. In other words, a thermometer in Central Park may have NO trend from UHI while having substantial UHI from the surrounding city. Also, a number of large and not so large cities may have topped out on a positive TREND in UHI at specific stations due to lack of changes in the environment this decade contributing to the flat temps!!
Second, since we are talking about the TREND, a small area will have more bang for the buck as it starts to grow as the change from natural or fields to concrete and blacktop is much larger than most other changes. A large urban area will not have as easy of a time causing a large change in its UHI without changing concrete for blacktop and painting roofs black!! The small area may have a UHI measured at 1c compared to the large cities 10c but have a TREND double the large city!
It should be noted that this would indicate that the amount of UHI TREND could be quite variable even among areas of the same size. My old Denier meme that Rural means low UHI trend and Big City means high UHI trend is not tenable and must bow to reality, The warmer meme that UHI is stable or it all averages out also needs to be tested against hard evidence gained by high quality studies in varied environments.
Finally the differencing used by the Berkeley Report tells us nothing about UHI. It tells us there is a difference in the composite trend between the average of the selected rural stations and all stations. It is impossible for it to attribute this change to UHI as it is only an ASSUMPTION that only UHI is variable between locations. As water vapor is the primary contributor for GHG’s it is certainly possible that GHG contribution would be different between areas. As the Texas drought shows the humidity certainly changes if only due to irrigation in some areas. The cloudiness definitely varies and would vary the solar contribution. Nope, the Berkley study has only shown us different stations have different trends and his subset warmed faster. Not much else can be taken away from the study.
I went back up in the thread to see if Judith had been responding to any questions-got tired of looking
No need to get tired.
Just press CNTL -F
Type in “curryja”. You’ll see two very short comments. That’s all.
The location of the center of gravity of a passenger aircraft also gives no information about its dynamics, but a sequence of such locations can tell you its destination and time of arrival. It is possible that a sequence of mean earth surface temperatures might tell us what to expect globally in the next few decades.
This analogy is not even wrong. I hope no casual reader of this thread believes such arguments because they only show a deep misunderstanding of physics.
The size of a plane is so small that the force fields that act on it (gravity,lift, propulsion and drag) may be considered constant and as the plane is a rigid body, considering that the integral of these forces applies at a point like object (center of gravity or whatever other point) allows to determine the complete dynamics!
Under some heavy assumptions this allows to make estimations about the trajectory and the time of arrival.
Of course these estimations could be totally wrong if these assumptions have not been met.
Just 2 examples :
1) the plane flies in a place with a very strong gravity gradient. The estimation of arrival by observing only the center of gravity are wrong because the plane will be torn in pieces by the tide forces. The right prediction should have been : the plane will be destroyed
2)By considering only the center of gravity, all rotations are lost. So if the plane rotates with variable speed (f.ex around its horizontal axis), drag and lift vary in an incomprehensible way for the observer. Not knowing what happens he gives up on predicting anything about the movement of the point.
Now there is nothing such with a temperature field. There is no equivalent of the gravity center theorem which allowed to apply the integral of a force field like gravity in a single point of a rigid body.
The spatial average of temperature is relevant to nothing. The temperature field itself is a result of interaction of all atmosphere&ocean fields (velocity, pressure, cloudiness, density, albedo etc). The coupling is local, there is no relation between “averages” of the fields for the non stationary non linear cases we are dealing with here.
To use the plane example. Spatially averaging temperatures and hoping to predict them is like wanting to predict the trajectory of the center of gravity of a plane which is simultanously rotating around the 3 axis.
As I have been flying myself, for those of you here who did too, you know that this is a VERY bad case of stall where anything can happen :)
Joshua makes certain allegations regarding “Willis” and when I ask for details he says:
“Check out WUWT”.
Good heavens. Joshua, you are not serious. You don’t even deserve you name.
I picture Joshua standing at the base of a large brick wall, blowing as hard as he can on his little horn (the same plaintive note over and over and over again), completely oblivious to the lack of impact he is having, certain that if he blows just one more time, the whole thing will come crumbling down.
Hmmm… maybe it’s more like one of those whistles you get in a box of Cracker Jacks….
I left some links to John Christy article for Tilo, in his thread. Here is a link to the UAH site which has temperature contour maps:
http://nsstc.uah.edu/climate/
Note that the middle column lists the monthly maps. If you compare September 2011, which is winter in the southern hemisphere, to January 2011, which is summer, you can get a visual of the problems with the study in question in this thread.
When you do a statistical analysis, you must be careful to explain what your numbers mean. The temperature contour maps show immediately that global warming is not what it implies.
If one studies the accompanying graphs from column 3, it becomes apparent that a linear trend line is see-sawing about 1998. Prior to 1998, average temps have been influenced by natural variables, which produce cooling, and post 1998 (including ’98) reveals more natural variables (El Ninos) which are temporarily boosting the average.
Satellite data shows a flattening of the warming trend post 2001, which is more apparent here:
http://www.drroyspencer.com/latest-global-temperatures/
The overlaid polynomial shows the extent of the see-sawing. There were several spikes from 1998 onward that were caused by El Ninos, particularly 2010, but they were followed by dips in the average. The study in question here must acknowledge such natural phenomenon, and cannot claim anthropogenic warming as the cause. The question arises as to how much of the warming in the past 200 years has been caused by natural variations, particularly in the ocean.
I think Stephen Wilde may have more to sat about that, if he is the same Stephen Wilde I am thinking about.
Remember that satellites scan 95% of the atmosphere using advanced technology that samples billions and billions of temperature dependent microwave emissions from oxygen molecules. Surface data, on the other hand, upon which this study is dependent, relies on the average of a maximum and minimum daily temperature garnered from surface stations which are incredibly sparse when compared to satellite coverage.
Yes Gordon, that’s me.
I’ve said lots about the interplay between sun and oceans and think I have constructed a scenario which complies with basic physics, fits all available observations and is capable of explaining all or nearly all of the observed temperature changes in the troposphere (which is not the same as system energy content).
Satellites record the overall temperature of the globe which varies very little.
Surface thermometers reflect the flow of energy past them as the rate of energy flow from surface to space varies and that changes quite a lot from region to region without necvessarily indicating any change in total system energy content.
Stephen…your explanations re ocean warming make a lot of sense to me. Perhaps you could include a link to your theories for people who are wondering what I am on about.
It’s apparent that the temperatures on the west coast of Canada and the US, as well as the west coast of Europe, are moderated by ocean currents. On the other hand, the Canadian prairies produce a temperature differential of up to 50 C from the west coast, at near equal latitudes, in winter. That differential is due to freezing Arctic air being pushed south.
I am convinced that what Lindzen claims is correct: that the greenhouse theory, and it’s extension, the anthropogenic greenhouse warming theory, featuring radiative effects, is far too simplistic, and that convective currents are more important. I think you are saying essentially the same thing. Although surface stations are enclosed to protect them from direct solar radiation, they are placed in the paths of convective currents that are modifying temperatures, not CO2.
Thanks Gordon, there is overlap between me and lots of others but I think my scenario is more complete and just need more observations to confirm or rebut it.
For anyone interested see here:
http://www.irishweatheronline.com/category/news/environment/climate-news/wilde-weather
Convection does put a cap on surface temperatures, but convection stops at night. That’s when greenhouse gases can be effective by narrowing the “radiation clear window” and reducing the cooling rate and hence the diurnal temperature range. This would result in some warming. What would cause the BEST warming remains a matter of interest.
Solar induced convection reduces at night but does not stop for as long as the surface remains warmer than the air above.
Water induced convection hardly reduces at night at all because of the thermal inertia of water.
Thanks Gordon.
I would like to see more discussion of the _pre_-1850 BEST estimates. Isn’t this important? Dr. Curry said, “there are no particularly surprising results here”, but that graph sure gives a different impression from the graphs of the original data sets. (I see that Steve McIntyre has already mentioned this.)
Since Judith is apparently not going to enlighten us any further, on BEST’s methods and conclusions, or on her opinion about the way the studies have been used to make monkeys out of “Climate Skeptics” in the media blitz unleashed by the Best Team, I will move on to something else.
I will offer this material on a non-funded and maybe non-professional look at ‘rural’ station data. This seems to be to be less mechanistic, and more imaginative than the analysis done in much of the BEST study; I would be interested to know what those who are stats literate think of this; that leaves you out, Josh-ua.
forgot the link:
http://joannenova.com.au/2011/10/messages-from-the-global-raw-rural-data-warnings-gotchas-and-tree-ring-divergence-explained/
Sorry folks, more to the story. The RUTI analysis compared wtih BEST:
http://hidethedecline.eu/pages/posts/ruti-global-land-temperatures-1880-2010-part-1-244.php
Don:
I like the paper. Some interesting ideas there. It kind of ties in with some of the things that I was posting about a couple of years ago here:
http://reallyrealclimate.blogspot.com/2010/01/giss-temperature-record-divergence.html
The errors that are produced by GISS by extrapolating across the arctic ice from stations with shore ice melt can also apply to extrapolating inward from warmed shore stations to inner land areas that have no such warming. This error also applies to the kriging that BEST does. And it is part of the reason why the BEST model is too hot, especially in the last 50 years.
Okay folks, I can’t believe that we are still talking about UHI. I didn’t think that it was that hard. But many people are not getting it. I’m going to write it up so simply that a 10 year old should be able to get it. But right now I have to take my 9 year old daughter to piano lessons so it will have to wait a couple of hours.
First, I want to answer Joshua’s question regarding classification:
Joshua: If they compare “good stations” to “bad stations” or “urban stations” to “rural stations,” and their classification structure is inconsistent, then your argument seems logical to me.
Joshua, I am not claiming that there are any classification errors according to the definitions that they use. I am claiming that the definitions they use make it impossible to determine the UHI effect. Now I’m going to give the highly simplified explanation that I promised above. Examples can help make things clear.
Let’s say the year is 1950 and we are going to put a thermometer in a growing city. But the city is already there and already has a very high built density. So, let’s say that the city already has 1C of UHI effect. Over the next 60 years the city continues to grow, mostly around the perimeter. The UHI effect goes up, and by 2010 there is 1.5C of UHI effect. The thermometer was only there since 1950, so the thermometer will only see the delta UHI change from 1950 to 2010 as an anomaly. So, by that thermometer, the delta UHI effect for that period is .5C.
Now, in the same year, 1950, we put another thermometer into a medium size town. Let’s say that it has a UHI effect of .1C at the time we put the thermometer there. The town grows over the next 60 years, there is a lot of building that happens close to the thermometer, and by 2010 it has .6C of UHI effect. Again, the thermometer will not register that first .1C as anomaly. But it will register the next .5C as anomaly.
So, in 2010, what we end up with is that the urban thermometer has 1.5C total of UHI effect, and the rural thermometer has .6C of total UHI effect. But, the delta UHI for both thermometers since they were installed is .5C. It is that .5C that both of them will show as anomaly.
Now BEST comes along and decides that they will measure UHI by subtracting rural anomaly from urban anomaly. Let’s also say that there has been .3C of real warming over those 60 years. So the rural thermometer shows .8C of warming anomaly and the urban thermometer shows .8C of warming anomaly. BEST subtracts rural from urban and gets zero. Their conclusion is, “either there is no UHI or it doesn’t effect the trend”. But, as we have just seen, .5C of the .8C in the trend of both the urban station and the rural station were UHI.
With their results, BEST has failed to discover the pre thermometer urban UHI effect, the post thermometer urban UHI effect, the pre thermometer rural UHI effect, and the post thermometer rural UHI effect. They have also failed to discover the UHI addition to the trend in either place. In other words, their test is a total fail. Even if they did their math perfectly, ran their programs perfectly, and did their classification perfectly, their answer is still completely wrong. Why? Because the design of the test never made it possible to quantify UHI. Now, many of you may object to my scenario. But I’m going to post this, get a cup of java, and speak to those objections next. (continued)
Could it be looked at from the other side? Shouldn’t someone be able to more easily select the least urban sites and show that they are colder(relatively speaking)? How many sites would one need to satisfy the whether it was on to something worth pursuing? 100-200(my wild guess)?
There are two ways to do it Kermit. You can compare absolute temperature of the urban area with the absolute temperature of completely unbuilt areas near that urban area; or you can compare the trend of cities for which you have records that extends over the entire build history of the cities with the trend history of completely unbuilt areas with equally long records.
Tilo,
I agree completely, but it is worse than that. They have made it look to john q public, like they have compared urban data, with data from far out in the boondocks, where there would be little to no human affect on temperature. If they had compared urban sites to truly very rural sites, then the UHI effect, would no doubt in my mind, have shown up. I wonder if anybody cares about this except us.
(continued from above)
Some of you may wonder if it is reasonable to expect a small town to grow at a rate that pushes up the delta UHI as fast as a city. This is where the definitions of rural and urban come in. Modis defines an urban area as an area that is greater than 50% built, and there must be greater than 1 square kilometer, contiguous, of such an area. So, for example, if you have two .75 square kilometer areas that are 60% built, separated by one square kilometer of 40% built, it’s all rural. So the urban standard is high enough that an area must be strongly urban to qualify. The rural standard is anything that is not urban. And that allows for a whole lot of built. 10 square kilometers of 49% built is all classified as rural.
BEST then goes on and further refines the rural standard as “very rural” and “not very rural”. Unfortunately, they make no new build requirements for “very rural”. The only new requirement is that such an area be at least 10 kilometers from an area classified as urban. But a “very rural” place could still have 49% build.
This means that you can have towns, small cities, and even some suburbs that are classified as rural. In such areas there is still plenty of room to build and build close to the thermometer. In the urban areas, there is little room to build. So either structures are torn down in the city to make room for new structures, or structures are put up at the edge of the city, expanding it. The new structures being put up at the edge of the city are far from the thermometer and while they still effect it, the further away they are, the less effect they have.
In the rural area there is still space to grow close to the thermometer. So, in the rural area you can actually have more UHI effect with less change in the amount of build. So, if a rural area goes from 10% built to 30% built it will still be rural and it can have the same UHI effect on the thermometer as the city where most of the new building is around the edges. The urban area may go from 75% built to 85% built around the thermometer, and it may have it’s suburbs growing, but the total effect will be close to that to the rural build.
All of this is essentially confirmed by Roy Spencer’s paper and by BEST’s own test results.
Tilo – I don’t want to prolong these discussions unnecessarily on a thread that’s now receding into the background, but it might trouble you to realize that the type of analysis you’ve provided is exactly what mainstream climate science as well as the BEST group would endorse as a reason to discount the possibility of a significant UHI-based distortion of land warming trends. For the distortion to occur, the urbanized locations will have to add what you call “anomalies” faster than the rural areas rather than at the same rate (or perhaps very slightly slower according to the Wickham et al paper). Unless they warm faster, there will be no distortion. If they warm more slowly (yet to be confirmed), the distortion will be in the other direction – to underestimate actual land warming rates.
Scientific prediction: There’s going to a thread on Urban Heat Dragons, or perhaps a series thereof, just like Sky Dragons.
Just wait and see.
That’s up to Dr. Curry. Perhaps she’ll be guided by her success in conclusively putting to rest the Sky Dragon philosophy.
It was a prediction, not a suggestion (much less a request). But in addition to whether or not the sky dragon threads ever shut down debates about the basic GH effect, there’s also the fact that they’ve pretty much removed them as spam from other threads, haven’t they?
One way to think of this in terms of what might be happening on a general rather than exclusively urban scale is to recognize that “land use changes” have been judged to play a significant, albeit minor, role in the temperature trends of the past century. If both urban and rural areas are becoming warmer beyond the effects of CO2 or other climate forcing agents because fields are being replaced by asphalt and buildings, that is a real effect, and anthropogenic in a different manner from the effect of ghgs.
But it’s local and an artifact WRT global average temperature.
I don’t think it’s an artifact if it records an actual increase in ambient temperature due to the change in land use. We’re not talking about inappropriately putting a thermometer next to a water heater but rather observing extra heat from land that now absorbs more sunlight, emits less IR at a given temperature, or exhibits less heat loss from evapotranspiration.
Maybe, if it’s actually happening all over; that is the stations are a representative sampling of the whole surface. That should probably be entered as a question to be investigated. It’s certainly possible that human changes to the surface are affecting the albedo on a global scale. In fact, IMO something along those lines may well have happened a few thousand years ago in the Sahara, when overpopulations of early domestic sheep/goats destroyed the forest cover.
Sorry Fred, you missed it. It’s not what you are calling the “distortion” between rural and urban that counts. It’s that both have their trend distorted by unrecognized UHI.
See my response to AK above.
Tilo
“here the definitions of rural and urban come in. Modis defines an urban area as an area that is greater than 50% built, and there must be greater than 1 square kilometer, contiguous, of such an area. So, for example, if you have two .75 square kilometer areas that are 60% built, separated by one square kilometer of 40% built, it’s all rural. So the urban standard is high enough that an area must be strongly urban to qualify. The rural standard is anything that is not urban. And that allows for a whole lot of built. 10 square kilometers of 49% built is all classified as rural.”
As I said:
When you look at 11km around the site as they did you are looking at a large number of pixels: roughly 400 pixels. The likelihood that the station will fall on a 500 meter pixel that is 49% built is diminishingly small.
Let me take some examples.
Lets suppose I take 18,000 random locations. For each of those random locations I look around the location for 11km.
Lets suppose I find no built pixels. I’ll call that area “unbuilt”
Now lets suppose I find 1 pixel that is “built” that 1 out of 400
Let me call that area “Built”
Your concern is that somehow, simply because modis has a 50% rule on a given pixel, that somehow you will get many Built pixels in the same area.
I can test that.
I selected 18,000 locations around the globe.
For every location, I looked 11Km ( .1degree at the equator) around the location. Ok?
I created two piles:
Pile A: NO built pixels ( but Tilo worries about that 49% pixel)
PileB: 1 or More built pixel in 11km
can we determine is these areas are really free of urbanization:
lets pick a different variable. A different variable that has nothing to do with the modis sensor. A variable like nightlights.
Will the lights in Pile A be Higher or lower than the lights in Pile B?
Well if you are worried about the 49% pixels in pile A.. what do you say?
Here are the facts:
of the sites I picked about 7500 had no built pixels within 11km
What about the average nightlights? well lets recall that according to Imhoff a nightlights value of 0 to 30 is associated
with rural areas. Whats the average nightlights value in that area?
about 12…. In fact 93% of those areas have lights less than the rural cut off.
What about Pile B. Pile B remember had at LEAST one built pixel
the mean “lights”? recall that Imhoff said that a value over 80 designated Urban areas.. Whats the mean of pile B?
132. Yup. more urban in pile B.
Lets take a different measure. Lets take population density.
population density matters because when you have a dense population you have to build big buildings.
Pile A: mean population density 13 people per sq km
Pile B: mean population density? 503 people per sqkm
Lets take another measure! lets use a different measure this time
lets use percent of the area that is grassland
PileA 28%
PileB 19%
lets switch again and think about how your 49% percent pixel could occur? you are looking for a structure that will be less than 250 meters wide.. for this one I’m going to look at the exact location. the EXACT location shows UNBUILT.. what could be there that is less than 250 meters wide..
A runway!
so out of my 7500 unbuilt pixels? guess what? I find 234 airports, checking you see that these are largely dirt and grass runways ( I have a database of 40K airports)
Pile B? over 700 airports.
Is modis perfect in picking out rural? no. However, its accuracy can be cross checked with other data sources and in the end you end up with two piles of stations. One pile, by all measures has less of the things that cause UHI. Less impervious surface, less electricfication, less population, less built surface, more natural landscape, the other pile has more of the stuff that causes UHI.
steven,
Are you convinced that urban, suburban, small town, and whatever else you want to call populated/ built-up areas are not over-represented in the temperature record? Do you believe that the human heat effect in the temperature record is negligible? Is that proven by BEST, with their “very rural, distant from urban” compared to everything else method?
Don:
Mosher is running tests that have no relationship to the issue we are discussing here. He doesn’t understand the definitions of rural and urban. He doesn’t understand that most of the world is unpopulated. He thinks that one built pixel in an 11 km area constitutes urban and zero built pixels in an area that was randomly selected from the entire globe constitutes rural places where there are thermometers. Simply put, he’s in space cadet land and he’s not listening. So don’t waste your time.
Tilo,
Steven hasn’t discussed the issue of presence of thermometers. That is what I am asking him to address. But it does seem that he has been able to differentiate less populated/less built up areas, from the opposite.
“of the sites I picked about 7500 had no built pixels within 11km”
That would be his group A. , which he says has a pop density of 13 per sq km vs. 503 for group B. OK, what about the thermometers?
PS: about the thermometers. What is the pop per sq km of the BEST “very rural” areas, that presumably do have thermometers?
Don:
Mosher’s built pixels do not correspond to urban areas, because one built pixel in an 11 km area is not urban. So most of his pile B could easily be classified as rural. And his pile A is taken randomly from around the world. And since most of the world is both unoccupied and without thermometers, it doesn’t represent the average place where you would find a thermometer.
Tilo,
I had to look up what population density is generally considered to make an are “urban” to get a better perspective:
http://en.wikipedia.org/wiki/Urban_area
In the US, it is 1000 per sq km. Some of Stevens group B should be urban, but a lot not. Anyway, I don’t think he wants to talk about this any more. He used to be one of my heroes.
I found this interesting from Watts this morning. There are ways to figure out this UHI thing, that the Big Climate crowd are avoiding:
http://wattsupwiththat.com/2011/10/24/what-the-best-data-actually-says/#more-49905
Mosher: “When you look at 11km around the site as they did you are looking at a large number of pixels: roughly 400 pixels. The likelihood that the station will fall on a 500 meter pixel that is 49% built is diminishingly small.
Let me take some examples.
Lets suppose I take 18,000 random locations. For each of those random locations I look around the location for 11km.
Lets suppose I find no built pixels. I’ll call that area “unbuilt”
Now lets suppose I find 1 pixel that is “built” that 1 out of 400
Let me call that area “Built”
Your concern is that somehow, simply because modis has a 50% rule on a given pixel, that somehow you will get many Built pixels in the same area.”
Mosher, you are truly looking dumber to me every minute. Who cares about your 18,000 random locations? Most of the world is unpopulated. When you look at 18,000 random locations around the world you are automatically going to be in much, much more rural space, on average, for those 18,000 than the average space where you have a thermometer. I don’t give a rats behind about 49% built pixels. I’m not talking about 1 pixel out of 400 being built 49%. I am talking about the definition of urban as given by the modis link that I already provided you. Urban requires over 50% build. And it must be that way for a square kilometer, not a friking pixel. Everything that is less than 50% built is rural. I used 49% as an example of the maximum that you could have. In fact, according to the definition, an area could be 80% built as long as it is less than one square kilometer it is rural. But it doesn’t matter if an area is 49% built or 39% built or 29% built or 19% built. As long as it is built. But your random locations that are not built do not represent the average area that is both classified as rural and that has a thermometer. Your locations, on average, have much less built than the areas that are both rural and have a thermometer. So your test is a dumb test that shows nothing about the issue we are trying to discuss. And your test is dumb for another reason. The requirement for your pile B is one or more built pixels. But the requirement for the urban designation is not 1 built pixel, or 2 built pixels, or 3 built pixels. There must be at least 4 of them contigous. This means that you could have a hundred built pixels in your 11 square mile area and they would still be classified as rural. And all of the population findings and light findings that you came up with for your pile B would apply to those rural areas.
A word to the wise is sufficient.
Sometimes, people simply do not realize that a plausibility argument is no substitute for empirical evidence. Until the “cooling” and “warming”stations are studied in detail, there will be know way to assess how much UHI effect there is in the overall (warming) mean trend. You are debating with people who write “If this is true, then the UHI effect most likely is small,” sometimes without bothering to articulate exactly what “this” they are assuming, and always unwilling to address outright the lack of directly relevant evidence.
I will not comment on the four papers. According to Elizabeth Muller in the summary of their work it should cool the debate over global warming by addressing many of the valid concerns of skeptics in a clear and rigorous way. They actually did address the urban heat island issue but to me that was always just a peripheral issue. I did have concerns about published global temperatures but they involved serious discrepancies between satellite and ground-based data, not siting issues. But reading on I find that their most “Herculean” efforts concerned merging available data sets to produce that composite 1.6 billion record monster data collection they put up on the Berkeley web site. As far as I could see creating it fixed none of the real problems that afflict published temperature data which they either did not know about or chose to ignore. To remind them of what still needs to be done I sent an email to Elizabeth Muller that I am including below:
“Dear colleagues:
I have to scold you for not doing what you set out to do, which is to correct errors in the existing temperature record. You just combined records from existing global temperature curves and surprise – your curve looks just like theirs. There are issues associated with global temperature study that range from the nature of temperature variations to falsification of major segments of the warming curve. You either did not bother to read my book “What Warming?“ with comprehension or simply did not know about it for lack of due diligence in your background research, if any. Complaints about such things as the urban heat island effect or placement of thermometers that you make a big deal of are non-issues with me. I am satisfied that the UHI effect can be allowed for with the satellite measurement of night lights that Hansen et al. came up with last year. But there are four major issues that need to be resolved which you totally missed. First is the existence of temperature breakpoints in nature, then falsification of temperature records in the eighties and nineties, the existence of a step warming that started in 1998 and regional temperature variations related to Arctic warming. Additionally, the nature and occurrence of ENSO oscillations and the alleged existence of volcanic cooling episodes must be analyzed. The latter two cannot be separately understood. Looking at global temperature history we notice that since the 1880s there have been several distinct temperature regimes that are separated by short breakpoints. Thus, there was cooling from 1880 to 1910, followed by warming from 1910 to the beginning of the Second World War. These are two distinct temperature regimes that must not be subsumed into a single computer-created temperature curve. Global temperature after the war showed no warming for fifty years while carbon dioxide kept relentlessly increasing. That is another distinct temperature regime that must not be included with the rest of the century. The last part of this temperature regime, the eighties and nineties, is shown as a so-called “late twentieth century warming” on all ground-based temperature curves. It shows up as a wonderful red triangle on NOAA’s version of global temperatures but is wholly imaginary and faked. Average temperature was flat throughout this period which includes five El Nino peaks. How this fakery was accomplished is shown in my book. Within the last 31 years we had just one short spurt of global warming, initiated by the super El Nino of 1998. It raised global temperature by a third of a degree in four years and then stopped. It was a step warming and it was oceanic, not greenhouse in nature. It was responsible for the very warm first decade of our century that included several record years. Presently we are again experiencing ENSO oscillations that restarted in 2008 after the super El Nino temporarily suppressed them. These oscillations are present in all accurate temperature curves. As to Arctic temperatures, these are increasing faster than IPCC has predicted. IPCC’s problem is that they are using models set up for greenhouse warming to predict non-greenhouse warming. There is no doubt that Arctic warming is not greenhouse warming. It started suddenly at the beginning of the twentieth century, paused from 1940 to 1970, then resumed, and is still going strong. Its cause
is warm water carried by Atlantic currents into the Arctic Ocean that started with a rearrangement of the North Atlantic current system at the turn of the twentieth century. Proof of this will appear in a paper in press scheduled for release in December. Except for a short period at the beginning of our century that I call the twenty-first century high the entire temperature curve we know of is covered with ENSO oscillations. They are a normal part of our climate and attempts to eliminate them by various schemes like a trailing average simply destroy information. Enso has existed since the equatorial Pacific current system has existed, which is to say since the Isthmus of Panama rose from the sea. It is important for understanding such climate features such as the alleged volcanic cooling associated with recent volcanism. Twentieth century warming happened in two parts, the first from 1910 to 1940 and the second from 1998 till 2002. In between there was no warming despite a relentlessly rising carbon dioxide curve. If I were you I would consider working on these issues and forgetting entirely about ocean temperatures as your next project.
Sincerely yours,
Arno Arrak “
Arno,
“I am satisfied that the UHI effect can be allowed for with the satellite measurement of night lights that Hansen et al. came up with last year. ”
What about before last year? How do we know that UHI, was not a significant factor effecting temperature data from 1950 to 2010, when the world’s population increased from 2.5B to nearly 7B, with accompanying infrastructure?
they have used it for the last decade for the US and extended it last year. Prior to that they used population for the ROW.
Mr. Rattus N.,
Thank you. Please point me to an explanation on how the data has been adjusted for the UHI effect. ROW?
ROW is the rest of the world. The papers can be found on Hansen’s publications page at GISS. If you know Python, you can read the code at http://www.clearclimatecode.com which should reveal everything.
Rattus,
Doesn’t GISS average stations alledgedly Rural with Urban stations? If both stations have significant UHI what good is the adjustment?
Don Monfort Oct 24 9:18 pm. Your point about previous years is well taken because I don’t know the history. I have generally regarded it as a marginal issue that was also very difficult to resolve experimentally. Partly this was because I saw other much more serious issues with temperature curves that were being ignored and did not want to get bogged down in an unrewarding project.
Arno,
Thank you. It may be a marginal issue, but I would like to see some convincing evidence of that. It seems to me that built-up areas are over-represented in the temperature record, and we know that UHI is a real and substantial phenomenon. Seems to me they should be able to find enough stations in largely unpopulated areas to make a valid comparison. They could use a station near the middle of the Rocky Mountains to get some really rural representation, and employ the Hansen method to extrapolate 1200km, in all directions.
Amazing the amount of dispersion in the Temperature Change Rates
Fig. 2 in http://www.berkeleyearth.org/Resources/Berkeley_Earth_Station_Quality
It will be interesting to watch the histogram change over time.
The individual plots also have this obscure annotation which says “cut outliers”. If they did this too bad, because with the attention paid to fat-tail statistics you have to have a good reason to remove outliers from a data set.
For comparison, this is what the Temperature Change Rates looks like for Vostock. This is on 1500 year boundaries with the fast rises and falls mapping the interglacial transition periods. This qualitatively differs from the Berkeley chart because this is a single location but it gives one an idea of the time scales we are dealing with.
http://img840.imageshack.us/img840/8651/vostockicetempchange.gif
Got the sign of that Vostock data wrong, the ice core data always goes backwards in time, doh.
http://img51.imageshack.us/img51/8651/vostockicetempchange.gif
Note the temperature rises are stronger than the drops in general. The peak is slightly negative because over the course of several hundred thousand years, the average rate has to approach zero.
Rattus,
I once encountered a reticulated python, in the jungles of Vietnam. Tasted OK, under the circumstances. I can’t read that kind of code. I will Google the info. Thanks
Can we just get something straight about the claimed negative UHI? It is all relative. It doesn’t mean that a UHI doesn’t exist or is absolutely negative.
For instance if you look at figure 1 in this graph
http://berkeleyearth.org/Resources/Berkeley_Earth_UHI
It shows an increase of temperatures in Tokyo and Globally compared to, relative to, what they were in 1900 -1920.
It doesn’t mean that Temperatures in Tokyo are now 2 degrees higher than the global land average.
Neither does it mean that the absolute value of the UHI in Tokyo is 3 degrees or even 2 degrees. To know the absolute value, now, we would have to know what it was absolutely in 1900-1920 and also know the amount of warming since that time which can be attributed to causes other than an UHI effect.
This is what I don’t understand about many of the criticisms of BEST that I’ve been reading.
BEST does not, as has been alleged often, conclude that UHI has a cooling effect. In fact, BEST says that UHI has a significant warming effect.
Unless critics remain clear about that, it is impossible for me to know where to start in understanding their criticisms.
I will help you, Josh-ua. They compared what they claimed are areas that are “very rural, and distant from urban areas”, to all other areas. They found that it has warmed slightly more in their allegedly “very rural” areas, than it has in cities, towns, suburbs, etc., where anyone with any sense would expect that the presence of denser human habitation would cause it warm more than in the freaking boondocks. They are saying that the UHI effect is slightly negative/negligible. That it does not skew the temperature record, that the “climate skeptics” are dopes and they should not bring up UHI again. That is bullcrap. Do you follow me, Josh-ua?
http://scholar.google.com/scholar?q=urban+heat+island&hl=en&as_sdt=0&as_vis=1&oi=scholart
http://wattsupwiththat.com/2011/10/24/what-the-best-data-actually-says/#more-49905
Joshua,
exactly how does this differencing stations tell us how much of the trend is caused by UHI, solar, cloud, GHGs…
The Berkeley report simply is showing a difference in trends with absolutely no effort put into verifying how much of the trends are UHI or something else. It is a pointless exercise that tells us NOTHING about UHI.
The basic ASSumption is that the stations would have approximately the same trends without UHI. It would also ASSume that either very rural (a very poor definition) and Urban (another poor definition) would have differing trends. As there has been no definitive work to confirm these ASSumptions and definitions they are unsupportable as a basis for the study.
Dr. Curry,
Is it standard practice for a co-author of a paper to not know if the analysis is correct, or to sign without having read the analysis and found it correct and without being able to comment on it?
Yes it’s common. I have been the principle statistician on papers where the other co-authors could not independently read, let alone assess, my statistical analyses. And I have not known whether their works (bioassays, for example) were either appropriate or done correctly.
I wrote up above that this has been debated forever, and that there is no way that the authors can be “jointly and severally” responsible for what is in the paper. The best that can be attained (and that imperfectly) is that everything that is “intellectual content” can be vouchsafed by at least one author.
Why then the multiple signatures? Isn’t it misleading ?
Why? to honor all authors and to make sure that there is at least one author responsible for each technical claim.
Is it misleading? no.
Consider again what happens when a medical doctor, synthetic chemist, analytical chemist and statistician collaborate on a paper on the effects of a substance on memory in patients with Alzheimer’s disease. The statistician has no way to independently judge the medical diagnoses of the patients or that the analytical chemist has the correct method for measuring the plasma concentration of the medication. In like fashion, the others are not able to verify independently the correctness of the hierarchical nonlinear modeling used by the statistician, or the correctness of the code that was written to use it in the particular study. In some well-known cases, journal editors and tort lawyers have tried to hold every author guilty of every mistake, but it is totally unworkable.
well, two questions:
1. Why not mention each co-author’s contribution and field of expertise ?
2. Is it common practice to have “honorary” co-authors, that made no substantial contribution to the paper?
Jacob,
1. Sometimes it is done. It is irrelevant to your original question.
2. No, it is not common, and this also is irrelevant to your original question.
Jacob,
Maybe I’m rather old fashioned but I’ve never come across anything like this before. I’ve always worked on the basis that if I sign off on something then it’s my responsibility if it turns out to be wrong. Having said that, I’ve not worked in academia for a long time and I do know that there is now so much pressure on academics to produce papers and I do suspect that this may be no longer the case as far as they are concerned.
So when did you ever test for yourself and ensure that the processor chips actually conformed to the IEEE floating point standard, or that the compiler actually compiled correctly? Maybe you did, but most people in teams have had to depend on the technical proficiency of a team member whose technical proficiency they could not actually appraise.
Let me ask it this way:
Papers are supposed to be peer reviewed, the external reviewers are supposed to check the papers and spot obvious mistakes. Can we expect the co-authors to do at least that much ?
I’m not sure this is quite the same thing. Obviously, it is possible for any team member team to let down the rest of the team. It is, of course, fair enough for very straight talking and robust discussion to take place in private. However, the principle of collective responsibility should apply in public. Little hints, and long periods of silence, which suggest otherwise is also letting down the team IMO.
Dr. Curry,
What was the role of Dr. Saul Perlmuter? The same as yours ?
Jacob,
Dr. Curry used to be one of heroes too, but she is MIA on this one. Maybe she is currently constrained on what she can comment on. If so, she should inform us.
I am extremely busy right now, I am barely able to keep up with moderating the blog let alone engaging in substantive dialogue.
Dr. Curry,
OK, I will risk being impertinent and suggest that two of the posts that followed this one are rather lame, and have detracted from a more important issue. But maybe that is just me. Please revisit this topic, when you have time.
By the measure of number of comments they are not so abnormal.
Kermit,
They can be both normal, and lame. Do you think those issues/discussion are more interesting, or more consequential, than the BEST research and the accompanying orchestrated media blitz that portrays the results of BEST, as the death knell for climate skeptics?
Kermit,
do you believe that Dr. Curry does NOT have a day job???
I only have time for “lame” posts at the moment. The BEST topic will hang around for a long time, and I will be putting more posts up as further (outside) analyses are made
I like you, Judy. Please excuse my badgering. I won’t do it any more, to you.
The BEST project declared they’ll be transparent and open. Since they mentioned Dr Perlmuter in their press release, it’s legitimate to ask and interesting to know what his contribution was.
This is amusing:
http://notalotofpeopleknowthat.wordpress.com/2011/10/22/kansas-temperature-trend-updatemuller-confirms-there-is-a-problem/
Look at figure 2. The black dots, according to the text above, are the “very rural’ sites (overlay makes them look more prominent). S. America and Africa are just about empty of “very rural” sites, but you would not be able to distinguish “very rural” sites from any city, town, or burg, in the US. The black dots are obviously in the most populated parts of Australia. Look at Great Britain. To call these “very rural” sites is ludicrous.
Here is another good one:
http://notalotofpeopleknowthat.wordpress.com/2011/10/22/kansas-temperature-trend-updatemuller-confirms-there-is-a-problem/
What’s wrong with this picture?
Dr. Curry, it seems to me, just eyeballing graphs of the PDO and the AMO, that the PDO fits better, e.g., doesn’t lag temperature changes. Graphs of PDO:
http://www.atmos.washington.edu/~mantua/REPORTS/PDO/PDO_cs.htm (fig. 3)
AMO:
http://en.wikipedia.org/wiki/Atlantic_multidecadal_oscillation
While the PDO declines around 1945, and again increases about 1975, the AMO switches modes in the mid 1930s and again in the mid 1960s.
There may be different ways to calculate each, but if one of these is driving temperatures, it looks more like the PDO. Unless somehow there is a lag in how the AMO causes temperature changes, which cannot be discounted.
I’ll ask again (no hurry with the answer).
Is it common practice to have co-authors that don’t make any substantial contribution to a paper ? Is it common to have “honorary” co-authors ?
None of the coauthors are honorary. I have seen email traffic from all of the coauthors discussing the paper, sending in edits, etc. I don’t personally know who contributed exactly what in any of these papers, other than the lead author for each paper.
At what point of urban build-up does the UHI start to become significant, and what is the basis for selecting the arbitrary 50% “urban”, what ever that may mean?
If, say, the UHI was already at 90% of some approximate maximal value at sites that were 20% urban (or vice versa) then what would be the value of looking at sites that were 49% versus 51% “urban”? Does the study assume a linear relationship between 0% and 100% “urbanisation”?
My, unsubstantiated, assessment would be that the first building near the thermometer has more effect than the second building, (though I welcome correction). So I would expect changes in urbanisation to have their greatest effect at thermometers witnessing the greatest changes, i.e. rural ones in the process of becoming urban (but not urban by the definitions as I understand them defined in this study). As Peter Berenyi seems to ask, what land based thermometers are not being affected by global population growth and urbanisation?
Yet the study appears not to distinguish between Steve McIntyre’s nickel-smelter and The-Little-Baby-Jesus lying in the manger next to the sheep and the cattle.
There’s a difference between the effect of buildings near a thermometer and the Urban Heat Island effect. The UHI is a large scale phenomenon in which the air in the entire mixing layer (or a large part of it) is warmed. That goes up 1-1.5 Km (to the top of the Ekman spiral), which means a lateral dimension somewhat larger (3-5 Km?). There are a number of papers describing the UHI, I don’t have time to do a search, but there’s always Google Scholar.
The UHI effect is a proposed artifact of station siting in which stations that were outside any UHI in the past become inside it due to growing urban areas. The technique used in the BEST paper would also have detected any artifact due to increasing temperature of same-sized UHI’s.
AK,
I am not so certain that UHI is properly accounted for by this or other studies. Urbanization is subtle and not simply a switch that is turned on from off.
@hunter…
You’re entitled to your opinion. Personally I doubt a new city’s going to spring up where there wasn’t one before, or that anything large enough to produce a UHI will fail to show up on the MODIS.
Now I’m more confused. Is not a heat island effect caused by buildings (primarily their heat capacitive effects on received isolation)? I’m sure that the effects are felt laterally (esp. down wind) as well as vertically, but are we still not discussing the sensible effects on a thermometer? Specifically land based thermometers?
And how does this study distinguish between sites (and presumably specific thermometers in some, but not all cases) that were facing a field in 1900 and a home 1950 and a Macdonalds in 2000. A part of my point was that any given heat island will not remain the same over time, and that the most sensitive (rural) sites will still be called rural. If the definitions are not appropriate then the study will not be looking at data which is the most significant.
A heat island has to be big due to short-scale mixing. Yes, there’s a separate effect from buildings nearby, but AFAIK it’s not the same thing as UHI.
AK,
please explain to us more ignorant types how the study attributed the difference in trends to UHI instead of other natural factors?
It didn’t, and didn’t have to. The study found no significant difference between the trend for “very-rural” and the total trend. Since UHI was included in the total trend, and not the one for “very-rural”, they determined that UHI, or anything else that acts like it, didn’t cause such an artifact.
It did leave the mystery of the not-quite-significant difference between the trends, opposite in sign to any proposed UHI. As described in the report, IMO this will lead very quickly to at least a few projects to pin down the cause.
AK,
since UHI and GHG’s are not the only items affecting trends the study shows us very little. The only thing it can say about UHI is that it is not greater than the observed trends and is probably less, unless there has been actual cooling for these sites without the UHI.
I will repeat, the study does not tell us anything useful about UHI.
AK,
“It did leave the mystery of the not-quite-significant difference between the trends, opposite in sign to any proposed UHI.”
Dr. Spencer’s study with different data showed a much higher rate of UHI for very small areas than the large areas. As this study wasn’t published it is far from definitive, but, it shows you statement is not correct. A number of us now accept that UHI is very dependent on the actual changes contributing to the environment of the thermometers. The size of the site is not necessarily the most important factor, only that there are changes that increase the heat retention and radiation of the surface or increase waste heat detected by the thermometer..
@kuhnkat…
UHI does not affect the trend, and nobody (who understands the subject) is suggesting that it does.
UHI was thought to be producing an artifact, a false trend in the averages that didn’t represent reality. The study demonstrated that if it is, it’s so small it doesn’t matter.
AK,
I agree, UHI does not affect the trend. I am being sloppy in my terminology.
Take a current large city that has a measured temperature of 12c above the surrounding area as measured by MSU. Do you disagree that at some time in the past there were changes that were made to the natural area where the city is now located that culminated in the current temperature differential? Do you think these changes all happened at once with no change over time or would you agree that there had to be a TREND in the temperature of the cities’ site to reach the current temperature differential from the surrounding area??
@kuhnkat…
Yes, there was a trend, a local trend. It simply wasn’t enough to matter to the global trend.
AK,
I don’t know if you are being obstinate or just naive.
There is a measured TREND which is positive and we can see it in all the surface series. It is only ASSUMPTION that this trend is NOT partially from UHI.
What data do you have that can attribute the TREND that is measured to Solar, Cloud, GHG…? You have the IPCC reports that tell us that ANY warming will create a hotspot with measured trend the lowest at the surface and the highest at the upper trop. We see no hot spot and the inverse with the upper trop trend the lowest and the surface the highest.
Again, what EVIDENCE based on FACTUAL observations can attribute the TREND that is MEASURED to other than UHI. There are NO attribution studies of significance.
You and no one else has the observations that can tell us which driver caused what part of the trend in any of the stations measured whether it was positive or negative. From IGNORANCE the Climate Clique assigns blame and screeches for action.
That is the reality.
Michael,
you know the gas from dirty diapers can be quite effective!! 8>)
Also, I know that the authors specifically exclude whether their results shed any light on anthropogenic warming due to CO2 emissions, but it raises a further question in my mind, which is:
Do urban areas themselves not have significantly higher CO2 levels, and how does this feed into model treatment of surface temperatures with respect to CO2 ?
(If it feeds in to the models at all.)
The curve for UHI effect at ANY and EVERY site starts with a rapid rise and then levels off.
The relative trend in different locations depends on where they are along the curve and the slope of the curve differs in every place.
There are no pristine sites. EVERY site is affected just by being a site. What happens next depends on the individual site in relation to EVERYTHING around it.
Even placing or removing a row of trees or a hedge makes a difference, or the type of material for access and the placement of all types of structures, however small, nearby.
Just look at the MET Office site requirements from the 1950s. I haven’t seen it for decades so readers will have to find it for themselves.
Often a site once placed will act as an attractor for further development nearby or it is initally placed in a location that is targeted for development already .
There is no meaningful distinction between urban, rural, near rural, very rural or any othe words you care to use.
Every single one is affected by UHI to some extent.
True Stephen. One interesting study showed that UHI is related to where you build. For example, if you build in a heavily vegetated area the UHI will be very high. The UHI in a place like Las Vegas, however, if very low. This shows that when you get rid of the vegatation and replace it with anything, you get a UHI effect. So put a thermometer almost anywhere and put a parking lot near it and you’ve got UHI. McIntyre just posted on UHI and he quoted a paper that had a West Virginia town of only 7,000 having a UHI of 5C.
What is so crazy about all of this is that we are expected to believe that there are .6F of positive adjustments that have to be made to the record:
http://cdiac.ornl.gov/epubs/ndp/ushcn/ts.ushcn_anom25_diffs_urb-raw_pg.gif
But, there is no such thing as UHI that would give us a negative adjustment. It’s just daffy!
Hey there! I simply wish to give you a big thumbs up for your great information you’ve got right here on this post. I’ll be returning
to your blog for more soon.
Fantastic goods from you, man. I’ve understand your stuff previous to and
you are just too wonderful. I really like what you have
acquired here, really like what you’re stating and the way
in which you say it. You make it enjoyable and you still care for to keep it wise.
I can not wait to read far more from you.
This is actually a great website.