It's Really Not That Difficult

After my last post I thought it'd be wise to take something of a break from blogs. As you can imagine though, that's been fairly difficult. The internet has made it so news organizations and blogs have become so interconnected you practically can't avoid one without seeing the other, and I want to keep up on the news of some things. As a result, I've largely accepted I'll just have to watch and try to resist saying anything about how stupid what people are saying is. And yes, it is all quite stupid.

Today I'd like to talk about one of the more remarkable examples, largely because it involves an update on a story I posted about here a while back. As people who've followed this blog know, I've discussed the Berkeley Earth Station Temperature project, obnoxiously titled BEST (seriously, who is arrogant enough to title their work "BEST"?), a number of times. There work has received some attention in the global warming debate, and I want to make it clear I don't focus on it because I think there is some great fraud going on in which they manufacture global warming to advance some political cause. I just think their work is crap.

Oh, and I don't like that the group compared me to a Holocaust denier. Yeah, that actually happened. It's not important in the grand scheme of things, but I do think it's rather remarkable the sort of response one might get from BEST representatives when pointing out issues with their work can be something like:

But you realize that the brandons of the world would remain skeptical. You know there is uncertainty over exactly how many died in the holocaust.

That was a remark from one Steven Mosher, comparing me to a Holocaust denier in a discussion I wasn't even involved in because he dislikes the fact I criticize his group's work. It's about as pathetic of behavior as you can find on a moral level.

But today I'd like to discuss behavior that's even worse. Nor on a social level, mind you. It isn't as socially disgraceful as smearing people like that simply for having the audacity to question your work. It is as morally reprehensible though. You see, the BEST group has actively misled the public for about a year now by making false statements which nobody in the media appears to have made any effort to verify.

To understand what I'm talking about, the head of the BEST team, Richard Muller, gave an interview for the Carbon Brief website last April which led to an article by on Roz Pidcock titled:

Prof Richard Muller: Not adjusting global temperature records would be “poor science”

This article gives Muller the opportunity to respond to questions/concerns raised by a number of people about the surface temperature record used to monitor global warming, but it is really nothing but an empty puff piece for Muller to say whatever he wants, regardless of whether or not what he says is true. While there are a number of problems with the piece, the one which I have focused on deals with the hot topic issue of adjustments made to the data. There has been all sorts of talk in certain crowds about how these adjustments exaggerate, or even manufacture, global warming. In response to these concerns, Muller said:

“Furthermore, because of the interest, we re-analyzed all the data with ZERO adjustments, just to see what we would get. These results have been made available online. What we found was that the conclusions we had previously drawn were unchanged. The data are available here

This was a hugely important claim for BEST as it had been making the rounds promoting this idea, with them having posted pieces on many sites about the topic (e.g. here). Naturally, when you go around telling a bunch of people what your results are, some people might want to look at those results for themselves. That's why Muller provided a link so they could.

The problem is the link Muller provided didn't go to a page with those results. In fact, those results weren't available anywhere. BEST had never published them for anyone to look at. Nobody anywhere could see them, unless maybe they spoke to BEST directly and got a copy. I know this for a fact because that's exactly what I did. When I saw this article claiming the results had been posted online, I sent an e-mail to BEST asking about what Muller had said:

Heya. A few days ago, an article
was published at The Carbon Brief which quoted Richard Muller
extensively, primarily to address the five questions the GWPF poses
for it's recent "temperature review." When answering the second
question, Muller said:

"Furthermore, because of the interest, we re-analyzed all the data
with ZERO adjustments, just to see what we would get. These results
have been made available online. What we found was that the
conclusions we had previously drawn were unchanged. The data are
available here"

I don't see the data for the results he refers to on that page. I
don't even see a graphics for those results. Was there supposed to be
a link to the results on that page?

I've looked at all the other data pages on the BEST website, and I
couldn't find any links to what BEST would get if it did its analysis
without adjusting the data. Muller says that has been published. Am
I missing something, or is he just mistaken?

Simple enough, right? I didn't accuse anybody of any wrongdoing. I didn't use any harsh language, insult anyone or suggest there was some nefarious plan going on. And I got a response, which was equally friendly. It pointed out the existence of things I already knew about, suggesting they might have been what Muller had in mind, but as I pointed out in a follow-up e-mail, none of them remotely matched what he had said. There was no explanation offered for that, but within a couple days, I was sent a copy of an unadjusted global temperature series created by BEST. I responded:

Cool, thanks. Do you happen to have any info on if/when the gridded results for the same comparison will be public?

I got a prompt answer:

I'm waiting to hear back from Robert our meeting today was cancelled. He and zeke may still be out of country

But it turns out now, seven months later, those results still haven't been made publicly available. In fact, the unadjusted global temperature series I was e-mailed hasn't been made publicly available either. It was e-mailed to me, but it hasn't been published online for people to examine themselves.

Normally, you'd think a journalist would be troubled at finding out there was an error in an article they had written. Based on that expectation, I contacted Roz Pidcock, the author of the Carbon Brief piece on Twitter to point out none of these results had ever been published. This led to her tweeting:

And updating the article to have this note:

[ Link not currently working, BEST tech team are aware of the issue and we will update when we have more information]

As I've said before, it takes a huge stretch of the imagination to read, “Link not currently working” to mean, “The data was never publicly available like Muller claimed, and the link he provided isn’t currently working because there was never any real link he could have provided.”

That's literally what I've said before. That's taken straight from a post I wrote a month after BEST e-mailed me the results it hadn't published for the public to examine. From then to a few days ago, the Carbon Brief piece continued to portray the reason these results aren't available as simply being a broken link. That is, a link was supposedly broken for eight months, and that's why nobody could see these results.

Obviously, that's all BS. BEST creates a temperature record for the world so people can see what temperatures have been like in different areas over the years. It then compresses that information down into a single, global record for simplicity so we can talk about one value for how much the planet has warmed. It publishes both of these on its website.

It then, as Muller put it, did the same thing without adjusting the data "just to see what we would get." Based on what it supposedly found when it did this, it went around giving media interviews and whatnot. During these interviews, the head of BEST claimed these results had been posted online so people could examine them for themselves. That was completely and utterly false.

When BEST was contacted to inform them of the error, it didn't try to get the error corrected. Instead, it apparently fed the Roz Pidcock misinformation. Pidcock in turn, apparently didn't attempt to verify what she had been told because if she had ever looked, she'd have found out these unadjusted results had never been posted online for the public to examine. And in fact, they still haven't been posted online for the public to examine. That means people can't check for themselves if Richard Muller is right in exchanges like this from his interview:

4. Are there any regions of the world where modifications appear to account for most or all of the apparent warming of recent decades?

“For large regions, I think not.

But fortunately, I am not the average person. Because I e-mailed BEST directly about this, multiple times, I was sent the unadjusted global temperature series. Because I then wrote multiple blog posts and comments, as well as more e-mails, I was, months later, e-mailed a private link (which I cannot share) to the unadjusted gridded results. This meant I could create maps showing rates of warming with and without adjustments. For instance, here is a map showing temperature trends since 1960 in the unadjusted data set:


Here is the same map, shown with adjustments:


There are significant differences. To highlight the one most relevant to Muller's interview, look at this map I made showing the difference between trends in the unadjusted and adjusted data sets:


While Muller said he doesn't think the warming of any "large regions" was largely or wholly accounted for by the affect of adjustments on his results, the reality is much of the United States wouldn't show warming to any meaningful degree in the 1960-2013 period if not for those adjustments. South America wouldn't show much warming without those adjustments either. Other time periods might show other areas with similar results.

Plus we can already see the adjustments BEST introduce significant systematic changes in other areas. People may be focused on the potential for adjustments to introduce spurious warming, but them introducing spurious cooling would be just as wrong. It's important people be able to look at why these adjustments happened. To do that though, they need to be able to see what the adjustments were. People can't hope to tell if BEST's adjustments do a good job if they can't look at what was adjusted and try to figure out why it was adjusted.

That's why BEST was supposed to publish these results. That's why BEST told everybody it had published these results. BEST knew if it told everybody it wasn't going to show them what its results were without adjustments, many people would distrust those results. BEST knew more people would trust its claims if those people believed both the unadjusted and adjusted results were publicly available. So it told everyone both were even though they weren't.

Now, I don't know how the initial error happened. I imagine Richard Muller genuinely believed these results were all published when he said they were. I don't think he intentionally lied. I just think it's a bad sign when the head of a project doesn't even know what results have been published.

Well that, and I think none of this is very difficult to understand. So imagine my surprise when I spoke to people who completely failed to understand it. A few days ago a BEST team member, Zeke Hausfather, tweeted about a new Carbon Brief article Roz Pidcock had written:

When I saw this, I checked the previous article and found the note blaming a broken link was still there. I thought blaming a broken link for results not being available for almost a full year was rather remarkable so I tweeted:

This, unfortunately, led to confusion as can be seen here:

You see, I had called the results gotten when one redoes the BEST calculations without adjustments "data." I did this because they are routinely called data, with that being what they were referred to as in the e-mails I exchanged with BEST. Pidcock presumably found that the raw data was available on the page linked to in the article and thought that was the "data" I was referring to.

This is a bit weird, of course, as the page she linked to hadn't been changed to add that data or anything since the time she added the note saying the link didn't work. Since Pidcock had recognized there was a problem eight months ago, I had assumed she would be able to recognize the same problem when she looked at the page now. That was apparently a mistake on my end. Regardless, what's more interesting is how Hausfather responded:

I find this fascinating. As Hausfather say, the raw data is available online. Nobody disputes that. It's just obvious people are, by far and large, not going to redo millions and millions of calculations to take that raw data and, hopefully, come up with the same results BEST came up with while not even being able to know if they succeeded since they wouldn't have anything to compare to.

So what's at issue is the results. Muller said BEST did the same calculations for the data without adjustments as it had done with adjustments and posted the results online. BEST routinely publishes charts showing temperatures for the globe, area by area. It even has an entire section of its website titled "Results by Location" where one can look up results by specific area.

In comparison, Hausfather said BEST publishes a single temperature series showing what the global average is without adjustments. I would say that doesn't remotely match what Muller said. Hausfather would call me pedantic. I think most people would call me reasonable. After all, all I want is for people to be able to look at maps like those I created above to see how BEST's adjustments affect its results. Muller claimed BEST had done the analysis to allow them to, and he claimed BEST had published the results of that work. I would like to think I am being reasonable when I say BEST should actually publish those results rather than just claim it has.

I'd like to think that's being reasonable, not pedantic. But at risk of being further labeled pedantic, I should point out one other tiny, little detail: Hausfather is full of it. See that part where he claims the "global land unadjusted time series" is posted online? Not true. He just made that up. I pointed this out to him:

And asked about it, multiple times, in the several days since. Strangely, he's chosen to ignore me. Hausfather's apparent strategy is to label pedantic pedantic for pointing out BEST hasn't published results it claims to have published, falsely claim BEST has published other results then just shut up when it's pointed out he's wrong. I can't say I get that.

Another thing I don't get is why Pidcock tweeted this to me:

To inform me the note in her piece has been changed to say:

[CB note: The raw station data is available in the ‘source’ and ‘intermediate’ sections on this page. BEST does not publish the raw gridded data.]

Leaving aside that we just had confusion over the difference between "data" and "results," so it seems strange to call anything "raw gridded data," there is no amount of semantic parsing which will change the fact BEST hasn't published any unadjusted results on its website. All it has published at the page linked to in Pidcock's piece is the raw data. Her update does nothing to address that Muller was completely and utterly wrong to say BEST has published unadjusted results. This note does nothing to address the issue at hand.

The worst part of all this is none of this is difficult. There are a lot of technical issues I don't expect people to follow in the discussions about how one creates temperature series for the planet. There is, however, no reason for people to have to disagree about what effect adjustments have on temperatures. There's no reason for people to be unable to figure out what data is and is not available. These are simple issues anyone should be able to figure out. All we need is to be able to point to each data sets and show what its results are.

The only reason there's any discussion over any of this is people keep managing to fail at incredibly simple things like... publish your results. Really, it's that simple. You do a bunch of calculations for an analysis because you think it might be interesting, and afterward, before you go around telling everyone what your results are, you toss those results up on a web page so people can download them. That's it.

As for people like Richard Muller and Zeke Hausfather... I don't know. I have no explanation for why these people keep saying things that aren't true. It's not a topic which interests me. The only reason I ever started looking at BEST's work is I find some of the technical details interesting. I'd much rather be discussing those. I just don't see how anyone can be expected to discuss the technical details of work when the people responsible for that work keep going around saying things that aren't true.

As for Roz Pidcock, this is an example of what laziness does for journalism. We all like to think of journalists spending months researching a story, taking weeks tracking down every lead they can find, but the reality is that usually doesn't happen. Puff pieces like Pidcock's are far more common. I mean, if we're being honest, Pidcock's article was really just a puff piece for BEST so you almost shouldn't blame her for anything in it, because really, she might as well have not written anything.

Oh, and as a book end, the BEST team member who compared me to a Holocaust denier did chime in on this topic. I want to point this out because I want to show whatever problems there are going on at BEST, it seems to run from top to bottom.


That's right. Despite Muller clearly and explicitly stating results were published online, BEST member Steven Mosher called me a liar because I "know exactly he meant by raw data being available." You can see more from Mosher in the discussion I had of his comment here, if you care for some reason.

Personally, I wouldn't. I just also wouldn't recommend criticizing BEST. Who knows what they might say about you?


  1. 'Other time periods might show other areas with similar results.'

    Do you mean that they didn't send you the pre-1960 data, or that you haven't gotten around to analyzing it yet?

    Great post, although I think you could've been a bit more clear on the difference between raw data and gridded unadjusted results, perhaps explaining it at the beginning; I don't think the issue is at all obvious to the average reader (just see how BEST gets away with conflating them).

  2. Heya. Thanks for the feedback. I'm sure I could have explained things more clearly. To be honest, I was spending so much effort withholding snark/avoiding abusive rhetoric, I didn't put as much effort into the contents of this post as I'd like.

    For your question, they didn't actually send me the data file itself, but the link they sent was to a file (so it showed me a private location I could find the file) which has gridded results for the globe as far back as their results go. I spent a fair amount of time looking at it about half a year ago, and the page where I originally posted these maps has similar maps for some other periods. The spatial patterns definitely change over time.

    I just didn't want to refer to specifics because my laptop died a couple months back, and it had all the results and analyses I ran on it. Referring to specific results would require I download the rather large file again and redo the comparisons to refresh my memory on the details. I'm sure I'll do it at some point in the mext few months (or recover the files from my dead laptop's harddrive), but for a post like this one which isn't technical, I didn't want to go bogged down.

    Additionally, there's always the problem of data versions. It may be we see particular patterns if we make this comparison for a time period now, but BEST ran that analysis in 2013. If it redid the calculations with the updated dataset, including additional stations covering past periods, the patterns we see now could change. A different version of the BEST data set might show different signals.

    Given things like that, I just didn't want to dwell on or focus too much on specific examples.

  3. Regarding 3000 quads, Fuller has a link to something similar from Pielke Jr
    "Let’s try to put these numbers into perspective with respect to total energy consumption in 2035. In terms of quads, the low, medium and high scenarios imply a total 2035 consumption of 940, 1,500 and 2,310 quads, or an increase over the EIA 2035 projection of 170, 1,000 and 1,810 quads. "

  4. For those who might not know what MikeN is talking about, he's referring to a previous post of mine about claims made by Thomas Fuller, particularly in a book he wrote, in which he uses a completely insane methodology to claim humans will use six times as much energy in 2075 as they did in 2010. This was particularly noteworthy because the analysis he used to come up with those numbers originally led to him saying:

    straight line extension of consumption trends gets you to about 2,100 quads in 2035, and about 3,000 in 2075.

    But when writing his book, Fuller quoted much of the text he had previously written verbatim yet changed that line to say:

    straight line extension of consumption trends gets you to about 1,000 quads in 2035, and about 3,000 in 2075.

    Even though the methodology hadn't changed at all, and consequently, the results it gave couldn't have possibly changed. The change of 2,100 to 1,000 however, brought those results more in line with the results of a different analysis he published in his book, meaning he simply rewrote his previous results to pretend they matched his new ones even though the two were wildly different.

    The primary problem of the methodology Fuller is that he assumed the entire world would become as developed as the United States by 2035. This, of course, is ludicrous. That couldn't possibly happen. If by some sort of magic all of the lesser developed portions of the world suddenly had lifestyles equivalent to the United States, that would cause a massive jump in energy consumption, but simply assuming such would happen then claiming it as evidence there will be a massive rise in energy consumption is insane.

    I'm not sure what the quote MikeN offers is supposed to contribute. Roger Pielke Jr. makes it clear in his piece he isn't projecting future consumption rates. For instance, he explains:

    Rather than starting from today and asking how much energy the world might consume in 2035, let’s turn the question upside down. Let’s postulate different levels of energy access, efficiency, and equity for 2035, and ask what it would imply in terms of required energy supply, applying an approach that policy wonks call “backcasting.”

    The point of his piece is not to claim we will reach energy consumption rates like those MikeN quoted. Instead, what he does is describe the amount of energy consumption that would be necessary to bring lesser developed portions of the world more into line with countries like the United States. That's why the conclusion of his piece is:

    This article makes an important point. To raise the quality of life in the developing world we will need dramatically more energy available. Virtually every quality of life indicator: life spans, rates of disease, rates of education, life choices for women, access to technology, improved transportation systems, etc, all directly correlate to energy consumption.

    As life improved in the developed world in the 20th century energy consumption increased. India, China and the rest of the developing world want the same improvements for themselves. Who are we to deny them?

    Instead of crafting policy around doing more with less (which may be OK for the gluttonous and wasteful USA) we should be crafting global policy around producing vastly more energy than we have today.

    Saying we should produce far more energy than we do now so that all countries can consume energy at rates similar to the United States obviously does nothing to support the idea in 20 years, all countries will have reached the level of the Untied States.

  5. I didn't notice the change in prediction times, 2075 vs 2035 which would put 3000 a definite possibility. I suspect Fuller used much the same method despite what he wrote. He has written along the lines of 'A adopting the lifestyles of B' many times.

  6. MikeN, I have no idea what you're talking about. The difference between Thomas Fuller's 2075 project and 2035 projection (when you account for him surreptitiously altering it) is the difference of saying humans will consume six time as much energy in 2075 as they did in 2010 and saying they will consume four times as much energy in 2035 as they did in 2010. The idea humans would quadruple their energy production between 2010 and 2035 is beyond ludicrous. There isn't any course of action which could possibly make that happen that has ever been proposed by anybody.

    As for what Fuller wrote, I don't know why you think anything "despite what he wrote." Fuller was quite clear in what he was doing. The only thing that's not clear is why you think what he did does anything to suggest Fuller's numbers are remotely plausible. All Fuller did is say how much energy humans would need to produce to meet the demand Fuller refers to. He never said a word to suggest humans actually would produce that much energy. In fact, he strongly indicated humans won't produce that much energy, saying that's a bad thing which people should work on changing.

    If anything, Fuller's piece you link to just shows how absurd Fuller's projected values are by showing there are no plans to come anywhere close to meeting those projections.

Leave a Reply

Your email address will not be published. Required fields are marked *