Skip to content

Who Needs Thermometers When You Have Algorithms?

May 11, 2012
tags:

By Paul Homewood

 

You will no doubt recall reading how GHCN decided to change the original Reykjavik temperature record from this

https://notalotofpeopleknowthat.files.wordpress.com/2012/02/image29.png

to this

https://notalotofpeopleknowthat.files.wordpress.com/2012/02/image30.png

thus doing away with the warm period in the 1940’s and creating a false warming trend. Well it seems our friends at GHCN must have decided their adjustments were wrong and so have now produced yet another set of adjustments, so their latest version is :-

rek

The period up to 1920 is now as warm as the warm 1940’s, which any Icelander with a bit of knowledge about their history will tell you is a load of old b****x. A quick look at the GHCN adjustment graphs below will show what they have done. (Sorry for the fuzzy quality).

 

 

ftp://ftp.ncdc.noaa.gov/pub/data/ghcn/v3/products/stnplots/6/62004030000.gif

Is it a coincidence that the latest incarnation looks distinctly like a hockey stick?

Not just Reykjavik either. In March GHCN changed Alice Springs from

https://notalotofpeopleknowthat.files.wordpress.com/2012/03/image23.png

to

https://notalotofpeopleknowthat.files.wordpress.com/2012/03/image24.png

Well now they have totally reversed this and made most of the past record a degree warmer comparatively, thus

station

It is surely time that GHCN admitted that the software modifications made by Google Summer Student, Daniel Rothenberg, and introduced in Nov 2011, have been a disastrous mistake.

 

For anyone who has missed the story so far, there is a good summary here and here.

 

FOOTNOTE

It is worth looking at the latest temperatures for Reykjavik on the GISS database, particularly the “after homogeneity adj” figure (which is after allowing for UHI effect).

 

  Original Unadjusted After GHCN Adjustment After GISS Homogeneity Adj
1910 3.63 4.22 3.33
1920 3.66 4.76 4.36
1930 4.64 3.92 3.92
1940 5.08 4.38 4.58
2011 5.58 5.58 5.58

Degrees Centigrade

To make proper allowance for the UHI effect, past temperature need to be increased (or current ones reduced) in cities such as Reykjavik. In 1940 this adjustment is only 0.2C and in 1930 there is no adjustment all to the GHCN data. (This of course ignores the fact that the GHCN temperatures have already been arbitrarily reduced in these years). Is 0.2C really an adequate allowance for UHI in a city such as Reykjavik?

5 Comments
  1. May 11, 2012 7:04 pm

    Reblogged this on Climate Ponderings.

  2. June 3, 2012 2:29 pm

    Paul:
    I have tried to get old GISS graphics from other stations (Carrasco, in Uruguay, and various others in Spain) but it only appears the most recent.
    I tried at WayBack Machine, but I can’t get nothing (“Wayback Machine doesn’t have that page archived” message)
    How could I fix that?

  3. June 3, 2012 7:28 pm

    Paul:

    I have just posted a commentary in a meteorological Spanish forum with a comparison between v2 and v3 graphics of Carrasco Airport station (it also has a link to this article):

    http://foro.tiempo.com/cambio-climatico-debates-varios-t135847.0.html;msg2966875#msg2966875

    It is, of course, in Spanish.
    Thanks again.

Comments are closed.

%d bloggers like this: