Sunday, April 26, 2015

GWPF inquiring into temperature adjustments

Well, this is an interesting one. GWPF has announced an inquiry into temperature adjustment practices. It is being boosted by Booker at the Telegraph ("Top scientists start to examine fiddled global warming figures") which is a bad dent in its credibility. However, it has a reasonably qualified panel, so it may be interesting to see what they develop. They have asked for submissions by June 30, so I might try to come up with something. But the remit sounds like they have been advised by Paul Homewood. For now, I'll just review that:

Are there aspects of surface temperature measurement procedures that potentially impair data quality or introduce bias and need to be critically re-examined?

Are they talking about current practice? Or that of years past? Are they going to recommend how it should have been done back then?

How widespread is the practice of adjusting original temperature records? What fraction of modern temperature data, as presented by CRU/GISS/NOAA/BEST, are actual original measurements, and what fraction are subject to adjustments?

This is framed Steven Goddard-style. How widespread is the practice of doing arithmetic? People use records to calculate spatial averages etc, and may well wish to adjust for that purpose. As they should. Is detrending, say, an adjustment? Or calculating an anomaly?

 CRU/GISS/NOAA/BEST do present data. NOAA explicitly offers unadjusted and adjusted data. GHCN Daily is largely an unadjusted copy of what the national Mets have, and GHCN Monthly unadjusted is a straight average of that (for long-record stations). But the primary business of CRU/GISS/NOAA/BEST is the calculation of spatial indices. That is basically a spatial integration, and requires manipulation. They have to calculate estimates of what actually happened in sub-regions, and they publish that.

Are warming and cooling adjustments equally prevalent?

Goddard again. No. Why should they be? Adjustments are for the repair of bias. If the identified bias is down or up, the repair will rightly have the opposite effect. TOBS, in the US, has a clearly cooling bias. The reasons are well established in terms of observing practices of the past, and so correction is clearly warming.

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

This is straight Homewood. There may be some. The inference is that there was a cooling bias that masked all of the warming. But it isn't common. If it helps them, the broad regional breakdown is here.

Are the adjustment procedures clearly documented, objective, reproducible and scientifically defensible? How much statistical uncertainty is introduced with each step in homogeneity adjustments and smoothing?

Well, it will be interesting if naysayers finally bring themselves to read the papers of Menne and Williams etc. But agenda may intrude. Statistical uncertainty of what, I wonder? Anyway, I'll probably send in a submission. I've been working out how to convert the html of blog posts to pdf. We'll see.


Update. I noted something amusing. Booker's articles (based on Homewood) have clearly been influential in setting up the inquiry, and at GWPF, Peiser has prominently featured Booker's announcement in the Tele. But he modified the headline to
"Top Scientists Start To Examine Adjusted Global Warming Figures"
As I noted above, Booker's headline is actually:
"Top Scientists Start To Examine Fiddled Global Warming Figures"

I guess GWPF likes to sound a little more objective.

Update again. Here at Homewood's blog is Booker celebrating their achievement in staging the inquiry
"It was entirely prompted by the two articles I wrote in the Sunday Telegraph on 24n January and 7 February, which as I made clear at the time were directly inspired by your own spectacular work on South America and the Arctic."


  1. I wonder if they will look at the "original" satellite measurements too, before they are calibrated processed and adjusted, as described by RSS

    "Calculating TB from raw radiometer counts is a complex, multi-step process in which a number of effects must be accurately characterized and adjustments made to account for them. These effects include radiometer non-linearity, imperfections in the calibration targets, emission from the primary antenna, and antenna pattern adjustments. "

    1. They say explicitly that satellite measurements are outside their scope.

      They don't mention SST, which has generally been adjusted downward (cooling).

    2. Aye, there is a huge correction in SST (due to AFAIK the previous use of canvas buckets) that the deniers do not want to talk about. They won't because the correction goes the wrong way, something that their leader Steven Goddard will never admit.

    3. Any idea who's going to fund all this? Presumably not the Koch brothers since they've alredy coughed up to fund BEST, which was basically an attempt to pick holes in the analysis of temperature data , and which is itself, according to GWPF, in the line of fire.

  2. I left the following comment on Homewood's site - it didn't make it through moderation. Ironic that the name of the blog is "Not A Lot of People Know That" When it comes to accuracy in describing the global temperature record Homewood wants to keep it that way - at least for his readers :)

    "The effect of all the adjustments is to *** REDUCE *** the increase in global temperature - not raise it.

    Big DUH! that deniers always fail to understand.

    Let me repeat, with no adjustments the increase in global temperatures would be even LARGER.

    The comparison is readily available from numerous sites, here's just one:

    Left at

    Was in moderation for a couple hours - then disappeared completely.

    1. Damn. I so didn't realize this.

      I need to spend a little more time dabbling with the science guys and less wading into the denial lair.

  3. re: "However, it has a reasonably qualified panel, so it may be interesting to see what they develop."

    Also reasonably counter-consensus in their purported views on climate change.

    1. I would have characterized it as, "a panel composed of people who have had demonstrated competence in some scientific endeavor at some point in their careers, but who have also more recently demonstrated ignorance and bias with regards to at least some issues involving anthropogenic climate change science".

      I think some people didn't get the answer they wanted out of the BEST effort, so they decided to try again. Pielke at least can be counted on to cite approximately 3 of his papers in every 5 sentences he writes to support his contentions that surface measurements are flawed and uncertain and that climate is influenced by more than just greenhouse gases. (to give him some credit, neither contention is entirely wrong, just not as important as he claims).


  4. I thought Berkeley Earth had solved these issues once and for all...

    I dont believe in crap like fraudulent adjustment practices making the Earth warmer. On the contrary, I suspect that GHCN adjustments have cooled the temperature trend of the recent century, and that adjustments in "key areas" is one possible cause of the alleged global warming slowdown.

    I looked at adjustments in GHCN stations in the Arctic area (>70 degrees N), which had been active after January 2000. I found that 11 of them had been adjusted down, and no one up, during this period.
    There are few stations in the Arctic area, which makes them relatively important in global temperature series that use infill techniques over empty areas, like GISTEMP and Cowtan & Way kriging.

    Nick, if I have got it right, Your TEMPLSmesh-index have a large degree of infill, and uses unadjusted GHCN data.
    Actually TempLS mesh show the largest trend (1.34 C/century) of all global indices in the recent century (Jan 2000-Mar 2015). It's more than C&W kriging (1.09 C) or C&W hybrid (1.19 C). According to the tool in your February 5 blog post, the use of unadjusted data in TempLSmesh increase this trend with about 0.3 degrees.

    If I've got everything right, this is really interesting, and maybe stuff for a scientific paper... I can see the headlines "GHCN adjustments have slowed the global warming, no real hiatus, etc" :-)

    1. Olof,
      That's interesting about TempLS mesh being the highest. But you're right that adjustments lower the trend for periods going back to about 1970. So it's the result of that combined with coverage about the same as C&W.

    2. At the moment all I can say is: read the next GHCN article.

    3. Victor, is it this we are waiting for?
      "The next version of GHCN-Monthly is expected to be released in 2015. Version 4 will contain more than 30,000 stations with monthly mean temperature data and improved quality control and bias correction processes."
      Just wondering if they have recognised an Arctic cooling bias in v3 and fixed it...

  5. Nick, looks like I did not have had to write my own post. Our responses to the naive and loaded questions are quite similar, although mine got a bit too long.