Home > Deconstructing Watts > Watts: SMHI apologizes to Dr. Jones

Watts: SMHI apologizes to Dr. Jones

2010 March 9

Of course, that’s not the way that Anthony Watt’s reports it. But that’s tone of the letter he links to.

With reference to the current debate regarding, amongst other things, access to climate data we have found that our letter to you dated 21 December 2009 unfortunately have rendered bad publicity both to SMHI and to the climate research community. We understand now that our response to your request forwarded by UK MetOffice 30 November 2009 may have been misinterpreted, maybe due to the fact that the formulations may have been a bit harsh.

SMHI: Letter to Jones (Mar 4) (WUWT)

Sweden (SMHI) is apologizing to Jones for their earlier denial on Dec 21 which has caused embarrassment to both parties.

Did Jones misinterpret what SMHI said on Dec 21? Here is what they said.

Given the information that the version of the data from the SMHI stations that you hold are likely to differ from the data we hold, SMHI do not want the data to be released on your web site.

SMHI: Letter to Jones (Dec 21) (WUWT)

This is the timeline as I understand it, posted with the understanding that I have not followed this closely. I’d be happy to see comments regarding errors, corrections, or missing pieces.

CRU received data from many national weather services, including Sweden.
CRU homogenized that data
(filled in missing data, removed blanks, adjusted for UHI, etc …)
CRU was asked to release its data (free the data!)
CRU said “we can’t, we have service agreements that say we can’t”
Nov 30: CRU then asked Sweden for permission to release the homogenized Swedish data
Dec 21: Sweden said “no, you can’t, you changed it”
Mar 1: Jones testifies that Sweden won’t let him release the homogenized data
Mar 4: Sweden is embarrassed and apologizes to Jones.
Mar 4: Sweden agrees to let Jones release homogenized data, asks for a link back to Swedish web site for data.
Today: On the Swedish weather data web site, you find this license.

3.2 The Licensee owns no right to use the data or products provided under this agreement for commercial purposes and not for development or production of meteorological, hydrological and oceanographic value added-value services. The licensee does not own nor authorized to redistribute, sell, assign or otherwise transfer data products or documentation without further processing to third parties unless the parties have received written permission from SMHI.

Emphasis mine. Translation by Google.
http://data.smhi.se/met/climate/time_series/html/essential20.html

Advertisements
  1. Carrick
    2010 March 10 at 12:44 am

    I think it’s equally obvious that the “retraction” itself is politically motivated, that the original denial was standard operating procedure for SMHI. As I read the situation, they wanted to retain control of the data to prevent people from using out-of-date versions from other sites. When Watts made a mountain out of a molehill, they were forced to move against their best interests to protect Jones from a ridiculous PR campaign.

    That said, just in terms of the science, I really think allowing Jones to publicize a bastardized version of the data to be a mistake. Nor should Jones publish his modified data sets.

    Instead, he should do like GISTEMP does and publish the algorithms and code for modifying the raw data into his format, and give instructions on how to download the raw data from his various sources.

  2. 2010 March 10 at 12:54 am

    I see your point. Jones is not a ‘data-monger’ and shouldn’t be forced into that role.

    On the other hand, just telling people to go out and roll their own is going to work better when your dataset is GHCN+Antartica+USHCN+Hohenpiessenburg.

    It’s going to be sort of unworkable if your data set is WMO + 20 individual National Weather Services + a University Department in China + a couple of other insider channels.

  3. Derek
    2010 March 11 at 12:06 pm

    I’m not sure I read the Mar 4 letter as an apology. I read it as a clarification:

    – We objected to you releasing our data on your website because you changed it. Refer people to OUR website. (Dec 21 version)
    – We have no problems with you releasing data on your own website as long as you acknowledge that it may be BASED on our data but you’ve changed it so it is no longer SMHI product.
    – We would like you to provide a link back to our own website so readers can get the original SMHI product.

    I just don’t see that as an apology or vindication for Jones. I see it as a way that lets him provide his modifications without presenting those as if they were the original data or blessed by SMHI.

    Jones SHOULD have to be a data monger any time he changes the data. Other researchers should have the ability to validate and verify his results which means he needs to provide that data if they can’t get it from original source (e.g., when he processes it).

    I generally prefer the approach that keeps weights and biases separate from the raw data itself and then explains how to use the weights and biases to develop the product.

  4. 2010 March 12 at 12:04 am

    Jones SHOULD have to be a data monger any time he changes the data. Other researchers should have the ability to validate and verify his results which means he needs to provide that data if they can’t get it from original source (e.g., when he processes it).

    It’s a rare case that you have to have his data to validate and verify his results.
    Having his data and his code allows you to audit his code, but it’s not a much of scientific verification.
    That’s better accomplished with different methods applied to similar data sources (GISTEMP with GHCN v2.mean) or, better yet, independent data sources (satellites for the last 30 years).

    The “problem” is that CRUTEM is not just “Jones.” It is not just research. It is part of a policy making machine deciding how to allocate resources and direct economies. CRUTEM and GISTEMP are not just scientific products of interest to a few thousand scientists. If those temp records are going to be waved around to influence political policy, taxpayers have a right to demand full access to the process that created them, and politicians have a responsibility to ensure that these products have it ‘right.’

    That’s NOT normal science – but attempting to redirect entire economies based on some scientific findings is not “business as usual” either. Hansen calls for action now. Part of the action is, rightly, an audit of the research leading to that call – an audit ‘outside’ the scientific process and open to review. Because the kind of money being talked about, and the kind of regulatory community being suggested, is big stuff. Jones and Hansen are suffering precisely because they have been successful in moving the topic out of the scientific community into the political. The rules of the game have changed.

  5. Derek
    2010 March 12 at 8:40 am

    Ron Broberg :
    It’s a rare case that you have to have his data to validate and verify his results.
    Having his data and his code allows you to audit his code, but it’s not a much of scientific verification.
    That’s better accomplished with different methods applied to similar data sources (GISTEMP with GHCN v2.mean) or, better yet, independent data sources (satellites for the last 30 years).

    Having his data and code allows you to at least try to replicate the results he claims. Isn’t replication part of the scientific process — isn’t that where Pons and Fleischmann got hung up?

    Jones and Hansen are suffering precisely because they have been successful in moving the topic out of the scientific community into the political. The rules of the game have changed.

    I agree with you completely there. In fact, my observation was that Hansen was unsuccessful in the scientific community 20 years ago so moved his game to the political and social community where he has been wildly successful. People sometimes forget that scientists are human too and subject to the same emotional sways and group dynamics as anyone else. It’s helpful therefore to get back to the science by seeing where someone got their data, how they manipulated it and filtered it to see how they can make extraordinary claims.

    I’m not a climatologist but I do have a background in physics and it’s a wonder to me that other climatologists didn’t ask Jones and Mann for the data supporting claims that the historically well-established Medieval Warming Period didn’t exist, that the 1930s really weren’t as warm as our grandparents (or crop records) thought they were.

    If someone were to come to me and claim they could something revolutionary, I’d want to see their data, their processes and have enough information to confirm it independently. In fact, that’s what I did for many years and usually the stories would fall apart — their claim required a priori knowledge that wasn’t generally available in the Real World or processing which wasn’t feasible to meet mission timelines. Once in a while you find a real game changer — but I want to see the data, the processing, etc. to confirm it!

  6. 2010 March 12 at 12:57 pm

    Isn’t replication part of the scientific process — isn’t that where Pons and Fleischmann got hung up?

    P&F failed because no one could independently verify their claims (at least as far as my limited knowledge of it goes). It wasn’t because some blogger requested access to their data and code and found flaws in the method. P&F failed because when other scientists tried it for themselves, they could not replicate the results.

    GISTEMP is an a real sense, a replication of CRUTEM. So is the NOAA/NCDC gridded anomaly with adjustments. So is Zeke’s homegrown method. So is my home-grown (see CRUTEM tab). All of these used GHCN v2.mean as a base data set. Dr Spencer went out and got an independent data set for the US. Same basic results. Satellites have been doing this for 30 years. Completely different data set and different methodology. Same basic result.

    Here is my handy-dandy verification breakdown:

    1) Same data, same method
    Independent group can verify no mistakes in implementation

    2) Same data, independent methods
    Independent group bolsters or undermines original methodology

    3) Independent data, independent methods
    Independent group bolsters or undermines original findings

    I suppose we can add a another
    This is what Spencer’s use of a different weather data set looks like to me

    2b) Independent data, similar methods
    Independent group bolsters or undermines original data set

    It’s a wonder to me that other climatologists didn’t ask Jones and Mann for the data supporting claims …

    I think that’s a misunderstanding of the way science works. You don’t ask Jones or Mann for their data. You go out and collect your own tree rings if you want to verify their findings. Or you go out and slice up clams looking for another angle on the question of past climate. You add to the accumulated knowledge base, to the work that came before. If mistakes were made, they are revealed by the accumulation of new data that casts into doubt previous conclusions.

  1. No trackbacks yet.
Comments are closed.