Frozen in Time
Jan 19, 2012
The Icelandic Saga Continues

By Paul Homewood

Just to recap, we have learnt that GISS temperatures for Iceland and Greenland have been artificially adjusted, with the result that current temperatures appear much warmer than when compared with the warm period during the 1940s. Temperature data for Reykjavik from the Iceland Met Office confirmed that this adjustment was wholly artificial and resulted in a net warming of about a half a degree centigrade since 1940 and that the actual mean temperatures in the last decade are about a degree less than GISS show.

I also have data from the Iceland Met Office for two other stations, Stykkisholmur and Akureyri and these show the same pattern of adjustments as the graphs below illustrate.

image
Enlarged.

image
Enlarged.

In both cases the temperatures from 1940 to 1964 have been adjusted downwards, and as with Reykjavik the overall effect is to create about a half a degree of warming.

On further investigation, it appears that the adjustments have actually been carried out by GHCN, whose figures GISS use. The changes seem to have taken place when they issued a revised version, 3.1, of their database in November 2011.  The GHCN website gives access to all their stations and shows both adjusted and unadjusted data. Examination of these records confirms that, out of eight stations in Iceland, seven have had such artificial warming applied, e.g

image
Enlarged.

The top right hand graph (red) is the unadjusted version, middle one is adjusted, and the bottom one graphs the adjustment (blue is minus, red plus).

Evidence is already building up that these adjustments are not limited to only Iceland. Similar adjustments have already been found in Greenland, Ireland and Scotland.

This issue raises several points of concern :-

1) These are palpably not “one-off” adjustments, which might be justified for station location changes or other local reasons. Have they been made as a result of a deliberate decision by GHCN, or are they the result of an error or a faulty piece of software?

2) If the result of error, what does this tell us about the quality control procedures at GHCN and GISS?

3) How many other similar adjustments have been made previously that have not been spotted? Would these have been uncovered without the attention of independent observers?

4) If GHCN believe the adjustments are justified, why have they not published their results and reasons for discussion, before issuing the revision? According to their CHANGELOG “GHCNMv3.1.0 is released with several minor corrections and a reworking of internal arrays for more efficient operations.” No mention of large scale temperature adjustments!

5) What assurance do we have that more changes of this sort wont be made in future?

Page 1 of 1 pages