Author Topic: Darwin Zero: Homogenizing Temp Data  (Read 1639 times)

0 Members and 1 Guest are viewing this topic.

Offline SSG Snuggle Bunny

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 23048
  • Reputation: +2232/-269
  • Voted Rookie-of-the-Year, 3 years running
Darwin Zero: Homogenizing Temp Data
« on: December 14, 2009, 07:57:41 PM »
Far be it for anyone to claim we don't run an honest shop around here. We can provide platforms for both sides of the debate (unlike climate change advocates).

First things first, the raw data from Darwin Zero:

Quote
People keep saying “Yes, the Climategate scientists behaved badly. But that doesn’t mean the data is bad. That doesn’t mean the earth is not warming.”

...

There are three main global temperature datasets. One is at the CRU, Climate Research Unit of the University of East Anglia, where we’ve been trying to get access to the raw numbers. One is at NOAA/GHCN, the Global Historical Climate Network. The final one is at NASA/GISS, the Goddard Institute for Space Studies. The three groups take raw data, and they “homogenize” it to remove things like when a station was moved to a warmer location and there’s a 2C jump in the temperature. The three global temperature records are usually called CRU, GISS, and GHCN. Both GISS and CRU, however, get almost all of their raw data from GHCN. All three produce very similar global historical temperature records from the raw data.

...

YIKES! Before getting homogenized, temperatures in Darwin were falling at 0.7 Celcius per century … but after the homogenization, they were warming at 1.2 Celcius per century. And the adjustment that they made was over two degrees per century … when those guys “adjust”, they don’t mess around. And the adjustment is an odd shape, with the adjustment first going stepwise, then climbing roughly to stop at 2.4C.

...

Yikes again, double yikes! What on earth justifies that adjustment? How can they do that? We have five different records covering Darwin from 1941 on. They all agree almost exactly. Why adjust them at all? They’ve just added a huge artificial totally imaginary trend to the last half of the raw data! Now it looks like the IPCC diagram in Figure 1, all right … but a six degree per century trend? And in the shape of a regular stepped pyramid climbing to heaven? What’s up with that?

Those, dear friends, are the clumsy fingerprints of someone messing with the data Egyptian style … they are indisputable evidence that the “homogenized” data has been changed to fit someone’s preconceptions about whether the earth is warming.

One thing is clear from this. People who say that “Climategate was only about scientists behaving badly, but the data is OK” are wrong. At least one part of the data is bad, too. The Smoking Gun for that statement is at Darwin Zero.

http://wattsupwiththat.com/2009/12/08/the-smoking-gun-at-darwin-zero/

Read the whole article. Willis Eschenbach details the raw data, the trends and the process by which the data is "homogenized."

However, The Economist was unsatisfied with Eschebach's findings.

Quote
So is it reasonable, if the GHCN is using complex statistical tools to adjust the temperature readings at Darwin based on surrounding stations, that they might come up with the figures they came up with? Sure. No. Yes. I have no idea. And neither does Mr Eschenbach. Because in order to judge that, you would have to have a graduate-level understanding of statistical modeling. For example, one paper describing the BOM's homogenisation approach includes this discussion:

Trewin (2001) comments that the median reference series used by Torok (1996) can have biases introduced when converting the median interannual differences back into an absolute reference series by the accumulation of rounding errors in the interannual differences (also well documented in Peterson and Easterling (1994)). This can introduce spurious trends in the reference series. Trewin (2001) uses a distance weighted mean of highly correlated reference station anomalies to create the reference series and so avoids the problems associated with the conversion of interannual differences. The weights, Wi, are given by Eqn 1 where di is the inter-station separation measured in degrees, and ri is the interstation correlation and N is the number of stations.

Wi =    {ri ( 6 – dI) 2,    r > 0.6, dI < 6}  i=1,...,N    ...1

   {0 ,         r < 0.6, diI> 6}


I don't understand that formula. I don't have the math for it. The paper goes on to reject the Trewin formula for reasons which, again, I don't have the math to understand. This is academic-level statistics. You can't render judgment on it by plugging disparate data series into Excel and eyeballing the trend lines, as Mr Eschenbach does. Here, for example, is a recent example of the kind of stuff climate scientists have to be able to do with statistics in order to get accurate results:

Two discontinuities were detected in the air-temperature time series at the meteorological station of the National Observatory of Athens. The first discontinuity reflects the instrumental change, which took place in June 1995 and the second discontinuity (and most pronounced) the application of a correction factor to the temperature values (in January 1997), after a calibration of the new thermometers. As a result, a cooling bias was observed after June 1995 and a warming bias after January 1997. The magnitude of bias exhibited a seasonal variability being more pronounced and reaching up to 0.67°C during the warm period of the year.

Judging by his post, Mr Eschenbach doesn't have the expertise to assess issues like these any more than I do. Mr Eschenbach is not a scientist; he's an amateur. His first effort in climate scepticism apparently came in 2002 while working as the construction manager for a beach resort in Fiji, when he published a non-peer-reviewed article claiming to have found that sea levels in Tuvalu were not actually rising, and that claims that they were stemmed from attempts by locals to blame subsidence problems on the developed world, and cash in on it. He's been beating this drum for years; he does not approach this issue from a position of neutral scepticism, he approaches it from a position of certainty that AGW is a hoax.

Look back, for instance, at the way Mr Eschenbach starts off his discussion of the Darwin data. He makes it sound as if he's just happened to stumble across this one site whilst perusing a debate over climate change in northern Australia. But as his link to that conversation from 2000 makes clear, Mr Eschenbach is already aware that climate change denialists have been trumpeting the apparent anomalies at Darwin for nine years. They do so because of that errant data at Darwin from before 1941, which makes it look as though there was a cooling trend there. The fact that climate-change researchers have to do a particularly strong correction on the data at Darwin, because they moved their dang instruments from the downtown post office to the airport, makes Darwin a perfect place to look for support if you want to claim that climate-change scientists are cooking the data.
http://www.economist.com/blogs/democracyinamerica/2009/12/trust_scientists

Eschebach responds...and he ain't happy:

Quote
On Dec 11th, the Economist published an unsigned article attacking both me and my work. This open letter is my reply.

TO: The Person Unwilling to Sign Their Economist Article

Dear Sir or Madam;

Recently, you wrote a scathing article about me in the Economist discussing my post called  The Smoking Gun At Darwin Zero. Some of it was deserved, but most was undeserved and false. The URL for your unprinicpled attack is http://www.economist.com/blogs/democracyinamerica/2009/12/trust_scientists … trust_scientists?

Trust_scientists?? Have you read the CRU emails?

...

So, to sum up your first arguments, changes in the Stevenson Screens and other local conditions cannot be the explanation for any of the GHCN adjustments because 1) the GHCN doesn’t use local conditions to make adjustments and 2) the timing of the screen change is wrong. In addition, there was no “dramatic change in 1941″.

Next, you point out two actual mistakes I did make.

First, in my proofreading I did not catch that I that I had written “the 1941 adjustment” when I meant the 1930 adjustment. That should have been obvious to me, because there is no 1941 GHCN adjustment. My bad.

Second, I had said that the Darwin temperature data couldn’t have been adjusted by using the GHCN method. This method requires five neighboring stations to which Darwin can be compared. Why couldn’t the GHCN method be used? I said it was because in the earlier time periods like the 1930s, there were no such stations covering that time period within 500 km of Darwin. I was wrong, it fact there is one such station.

Neither of these errors of mine affect my point, which is that there are not enough neighboring stations to adjust Darwin using the main GHCN method. The GHCN folks mention this possibility, saying:


Quote
Also, not all stations could be adjusted.

Remote stations for which we could not produce an
adequate reference series (the correlation between
first-difference station time series and its reference
time series must be 0.80 or greater) were not adjusted.
The homogeneity-adjusted version of GHCN includes
only those stations that were deemed homogeneous
and those stations we could reliably adjust to make
them homogeneous.

SOURCE:http://www.ncdc.noaa.gov/oa/climate/ghcn-monthly/images/ghcn_temp_overview.pdf

Unfortunately, they adjusted Darwin anyway. Consider the GHCN adjustment in 1920. To find five stations around Darwin covering 1920, you have to go out 1,250 km. Nor is there any guarantee that those stations will be suitable. You need to have five stations with an 80% correlation with the Darwin record … I wish you the best of luck finding those five stations.

So while my statement about stations nearer than Daly Waters was wrong as you point out (there is one nearer station that covers the 1930 adjustment), my point was correct – there are not enough neighboring stations to adjust Darwin using the GHCN method. The first GHCN adjustment to Darwin was a single year adjustment in 1901. To get five “neighboring” stations for that adjustment, you have to go out 1728 km. You fail to deal with that issue at all. Instead, you say:

“So is it reasonable, if the GHCN is using complex statistical tools to adjust the temperature readings at Darwin based on surrounding stations, that they might come up with the figures they came up with? Sure. No. Yes. I have no idea. And neither does Mr Eschenbach. Because in order to judge that, you would have to have a graduate-level understanding of statistical modeling. … I don’t understand that formula. I don’t have the math for it.”

“Surrounding stations”? We’re talking about stations a thousand km away and more, not surrounding stations.

And while I am sorry to hear of the lacunae in your math education, please don’t make the foolish assumption that others are similarly limited. I have no problem with the GHCN math. If you truly have no idea on the question as you say … then why are you excoriating my ideas on the question?

http://wattsupwiththat.com/2009/12/12/sticky-for-smoking-gun-at-darwin-zero/

They felt that all the down in Australia.
« Last Edit: December 14, 2009, 08:03:03 PM by Mr Snuggle Bunny »
According to the Bible, "know" means "yes."

Offline Oceander

  • Hero Member
  • *****
  • Posts: 1196
  • Reputation: +1/-0
Re: Darwin Zero: Homogenizing Temp Data
« Reply #1 on: December 15, 2009, 08:23:16 PM »
Quote
Judging by his post, Mr Eschenbach doesn't have the expertise to assess issues like these any more than I do. Mr Eschenbach is not a scientist; he's an amateur.

Funny, that.  I wonder if anyone's told most of the really great scientists down through history that they really aren't "scientists" because they were amateurs (albeit wealthy amateurs, which is why they could fool around with things like radio waves and electricity and telescopes on their own dime and not have to spend all day working to earn their keep).  Speaking of which, I wonder if Mr. Anonymous Economist Author quite grasped the irony of his accusing Mr. Eschebach of being an "amateur" because, after all, the original meaning of the word is someone who is a "lover, devoted friend, devotee, enthusiastic pursuer of an objective," and Mr. Eschebach is certainly an enthusiastic pursuer of the truth regarding climate change and climategate.  That sense of "amateur" would apply to all of the great scientists down through history.

The Word History for "amateur" from the American Heritage Dictionary is:
Quote
Word History: When Mrs. T.W. Atkinson remarked in her 1863 Recollections of the Tartar Steppes and their Inhabitants, "I am no amateur of these melons," she used amateur in a sense unfamiliar to us. That sense, "a lover, an admirer," is, however, clearly descended from the senses of the word's ultimate Latin source, amtor, "lover, devoted friend, devotee, enthusiastic pursuer of an objective," and from its Latin-derived French source, amateur, with a similar range of meanings. First recorded in English in 1784 with the sense in which Mrs. Atkinson used it, amateur is found in 1786 with a meaning more familiar to us, "a person who engages in an art, for example, as a pastime rather than as a profession," a sense that had already developed in French. Given the limitations of doing something as an amateur, it is not surprising that the word is soon after recorded in the disparaging sense we still use to refer to someone who lacks professional skill or ease in performance.