The "Not Evil, Just Romm", 2% solution

ERRATA: I made a mistake regarding the 2% figure, I misheard what was being presented during my visit with the BEST team at Berkeley. As many of you may know I’m about 80% hearing impaired and the presentation made to me was entirely verbal with some printed graphs. Based on the confidentiality I agreed to, I did not get to come back with any of those graphs, notes, or data so I had to rely on what I heard. I simply misheard and thought the 2% were the Japan station analysis graphs that they showed me.

I was in touch with Dr. Richard Muller on 3/28/2011 who graciously pointed out my misinterpretation. I regret the error, and thus issue this correction about the 2% figure being truly a random sample, and not just stations in the Japan test presentation shown to me.

I am told of another correction, and that is that Dr. Caldeira was shown a paper they are working on related paper on oceans, and that contained the preliminary 2% graph from the surface analysis, but was not the full surface analysis paper, which hasn’t yet been written.

According to Dr. Muller, that 2% test run does not contain all the bias corrections they plan to apply for station moves, discontinuous records, UHI and other station effects. I look forward to seeing the data when those are applied to the full dataset.

This episode where Mr. Romm gets an email from Dr. Caldeira and creates a “finding” illustrates the danger in rushing to judgment on snippets of preliminary results, and as  BEST says: “The Berkeley Earth team feels very strongly that no conclusions can yet be drawn from this preliminary analysis.”

I still believe that BEST represents a very good effort, and that all parties on both sides of the debate should look at it carefully when it is finally released, and avail themselves to the data and code that is promised to allow for replication.

– Anthony Watts

Last week Willis told you about how Joe Romm at Climate Progress botched a blog post so bad, Joe had not only to fix his own post by removing false claims about population trends, so did the paper’s authors. Then, rather than simply admit a mistake and move on, he spun it into some sort of twirling victory dance, bizarrely claiming that because Willis put up a chart of CO2 rates, “he” got us to admit that CO2 rates were increasing because Willis chose it as a reference. Heh, well if that floats your boat, you go Joe. WUWT has quite a history in discussing CO2 with graphs, rates, and guest essays, no news there.

This week, it’s the old pea and thimble trick combined with desperation and some silly claim of “exclusive”, like some cheap MSM news labeling graphic where they’ve caught some sex poodle on tape. After earlier writing a piece condemning the Berkeley Earth Surface Temperature Project (BEST) he’s now put up his interpretation of an email from scientist Ken Caldiera who said:

I have seen a copy of the Berkeley group’s draft paper, which of course would be expected to be revised before submission.

Their preliminary results sit right within the results of NOAA, NASA, and HadCRU, confirming that prior analyses were correct in every way that matters. Their results confirm the reality of global warming and support in all essential respects the historical temperature analyses of the NOAA, NASA, and HadCRU.

Their analysis supports the view that there is no fire behind the smokescreen put up by climate science deniers.

The only problem is this: there’s no “draft paper” yet, there’s nothing that is “submission ready”, not even close. It hasn’t been written. In fact, BEST hasn’t even done a full global analysis yet. How do I know this? It’s simple; I’ve visited the team and asked them directly, something Romm has not done.

Caldeira was simply looking at the same set of data (some preliminary charts and graphs that Richard Muller carries around with him in a file folder), that BEST has been showing to several people, including me. The only difference is that people like myself, Steven Mosher, and his friend Zeke Hausfather who visited BEST with him, haven’t run off the rails to make early and unsubstantiated claims about it “confirming” anything yet. And now, Romm’s adding to his original blogpost, is backpedaling, while at the same time picking a fight with Steven Mosher for notifying him on the issue in comments. It is sad, comical, and oh-so-typical of the sort of thing we’ve come to expect at Climate Progress. Romm simply got excited and jumped the shark. He’s not doing himself any favors with this sort of thing.

Here’s the Initial Findings statement from BEST, written by lead scientist Robert Rhode (of globalwarmingart.com) which pretty much mirrors what Caldeira is saying:

The Berkeley Earth Surface Temperature project has not yet done the analysis of the full data set with the corrections to produce a global surface temperature trend. We are first analyzing a small subset of data (2%) to check our programs and statistical methods and make sure that they are functioning effectively. We are correcting our programs and methods while still “blind” to the results so that there is less chance of inadvertently introducing a bias.

A preliminary analysis of 2% of the Berkeley Earth dataset shows a global temperature trend that goes up and down with global cycles, and does so broadly in sync with the temperature records from other groups such as NOAA, NASA, and Hadley CRU. However, the preliminary analysis includes only a very small subset (2%) of randomly chosen data, and does not include any method for correcting for biases such as the urban heat island effect, the time of observation bias, etc.

The Berkeley Earth team feels very strongly that no conclusions can yet be drawn from this preliminary analysis.

That 2% subset they refer to is some weather stations in Japan. They chose Japan because it made for a compact insular test case for the code, combining rural, urban, and airport stations under one organization’s output to keep it simple. Like Ken Caldeira, I’ve seen that preliminary 2% output. I’ve also seen a lot of other things, some things Caldieira hasn’t seen that the BEST team has shared with me. So has Zeke and Mosher, but neither they nor I are screaming “exclusive” and jumping to conclusions like Romm is doing over Caldeira’s general statement on that 2% sample run to test the code.

What’s even funnier is that whenever we mention USHCN trends for USA stations, AGW proponents are quick to point out that the USA has only about 6% of the land surface area of the Earth (USA: 9,629,091 km², Earth: 148,940,000 km² source), but they are now willing to go with the weather station data from 377,930 square kilometers of Japan’s land area which is 0.25% of the Earth’s surface area,  as enough for “confirmation” of a global trend.

In response to this latest yapping from Romm, BEST has also now updated their FAQs page here, and says this:

NEW – What do your results show?

The Berkeley Earth Surface Temperature project has not yet done the analysis of the full data set with the corrections to produce a global surface temperature trend. We are first analyzing a small subset of data (2%) to check our programs and statistical methods and make sure that they are functioning effectively. We are correcting our programs and methods while still “blind” to the results so that there is less chance of inadvertently introducing a bias.

A preliminary analysis of 2% of the Berkeley Earth dataset shows a global temperature trend that goes up and down with global cycles, and does so broadly in sync with the temperature records from other groups such as NOAA, NASA, and Hadley CRU. However, the preliminary analysis includes only a very small subset (2%) of randomly chosen data, and does not include any method for correcting for biases such as the urban heat island effect, the time of observation bias, etc. The Berkeley Earth team feels very strongly that no conclusions can yet be drawn from this preliminary analysis.

Compare Romm’s “exclusive” to Zeke’s writeup over at Lucia’s last week which preceded Romm’s. Zeke’s essay has firsthand accounts, is a lot easier to read, and doesn’t need gratuitous exclamation points. For those who don’t know him, Zeke Hausfather is very much in the warming camp, but he’s also a reasonable person. Zeke wrote about a technique that I agreed to keep in confidence until they had a paper accepted for publication or chose to announce it on their own, but it may have been just a slip or communications misunderstanding:

Their major innovation, in addition to those that overlap the work of other bloggers, is to treat inhomogenities as the start of separate records. The least squares method of record combination has the major benefit of allowing relatively short records to be combined together without introducing biases. This means that instead of trying to artificially correct inhomogenities detected by comparing individual stations to their neighbors, they can simply treat these as break points, where subsequent measurements from the same site are treated as a separate record and are optimally fit to the larger series using the LSM approach.

The issue hasn’t been the slight warming over the past century, we’ve always conceded that there is some. The issue has always been magnitude, uncertainties, and cause. With the BEST project, we’ll get closer to the ground truth of magnitude and uncertainties, but it will say nothing about the cause, except perhaps to help define the contributions of UHI and station siting.

I’ll repeat what I said earlier about BEST:

And, I’m prepared to accept whatever result they produce, even if it proves my premise wrong. I’m taking this bold step [providing my surfacestations data to them] because the method has promise. So let’s not pay attention to the little yippers who want to tear it down before they even see the results. I haven’t seen the global result, nobody has, not even the home team…

…

My gut feeling? The possibility that we may get the elusive “grand unified temperature” for the planet is higher than ever before. Let’s give it a chance.

More science, less barking.