Skip to content

Why Do GHCN Adjust Temperatures?

March 3, 2012
tags: ,

By Paul Homewood




There has been much discussion recently about how GHCN have adjusted temperatures at many stations around the Arctic Circle. There have been certain misconceptions about why this should have happened, so I want to explain the logic behind what both GHCN and GISS do.


Let’s start with GHCN.

I have seen it claimed that GHCN adjustments are designed to “average out” regional temperatures – so, say, it is OK to alter Reykjavik’s temperature record if it shows a different trend to other stations in Iceland. I have also seen it claimed that the sole objective is to arrive at a “global average temperature”, and that, therefore, individual stations are irrelevant. Neither of these claims are true.

In their technical report GHCN make it absolutely clear why they perform “homogeneity adjustments”.


Surface weather stations are frequently subject to minor relocations throughout their history of operation. Observing stations may also undergo changes in instrumentation as measurement technology evolves. Furthermore, observing practices may vary through time, and the land use/land cover in the vicinity of an observing site can be altered by either natural or man-made causes. Any such modifications to the circumstances behind temperature measurements have the potential to alter a thermometer’s microclimate exposure characteristics or otherwise change the bias of measurements relative to those taken under previous circumstances. The manifestation of such changes is often an abrupt shift in the mean level of temperature readings that is unrelated to true climate variations and trends. Ultimately, these artifacts (also known as inhomogeneities) confound attempts to quantify climate variability and change because the magnitude of the artifact can be as large as or larger than the true background climate signal. The process of removing the impact of non-climatic changes in climate series is called homogenization, an essential but sometimes overlooked component of climate analysis.

To isolate these “abrupt shifts”, they use an algorithm. And it was changes to this algorithm in July 2011 by a Google Summer Student, Daniel Rothenberg, that suddenly produced this swathe of anomalous adjustments in Greenland, Iceland and Siberia. The Icelandic Met have confirmed that there have been no station  moves or other non-climatic factors, which would have created the need for the adjustments in Iceland, and of course the algorithms in use previously in GHCN V2 and V3 did not spot anything unusual in the temperature data. So it appears that the new algorithm is not working correctly.

It is also perhaps worth pointing out that as the algorithm only looks for “abrupt shifts”, there is no UHI adjustment in GHCN, as the UHI effect would occur much too slowly.

Let’s move onto GISS then. GISS start with the GHCN adjusted data, but then carry out their own “homogeneity adjustment”. This is designed to take account of the UHI effect. Usually the older temperatures are increased at urban and peri-urban sites, so that at Reykjavik, for example, the 1940 temperature is adjusted upwards by 0.6C. The effect can be seen below.



                            Before UHI Adj                                                                After UHI Adj


Whether 0.6C is enough is another question!!

  1. Coldish permalink
    March 3, 2012 11:03 pm

    This is interesting work, Paul. Keep it up.

  2. Anything is possible permalink
    March 4, 2012 7:25 pm

    If a station lies close to the average position of the Arctic front,

    then the abrupt shifts in mean temperature levels being detected by the GCHN algorithm may actually be natural in origin, with temperatures rising as the front shifts to the north, or falling as it shifts to the south. Nothing to do with anthropogenic factors, whatsoever.

    So where might one expect to find such stations?

    Northern Canada?
    Northern Russia?

    Ringing any bells yet? (:-

  3. John Kannarr permalink
    March 5, 2012 9:37 pm

    I thought, based on other posts at this site, that many of the GISS adjustments were to make early 20th century temperatures lower, not higher.

    • March 5, 2012 10:08 pm

      There are valid reasons for adjusting either way, John. The GISS adj for UHI is obviously a good example, although one might query the amount.

      However GISS use as their starting point, temperatures already adjusted by GHCN, and it is these adjustments in the latest version 3.1 that seem to have no justification.

    • Shawn permalink
      December 20, 2016 12:15 pm

      You are correct John, the two graphs they have superimposed over each other have had their temperatures (on the left) shifted to make it look like it’s actually the later temperatures that were increased. The truth is that it was earlier temperatures (in the 40s) which were decreased.

      I can’t say whether that was on purpose or not, but I hope they correct it to help people understand the issue.This looks extremely sketchy as is.

  4. John Cunningham permalink
    March 9, 2012 5:04 am

    from what I have read, both GHCN and GISS have pretty widely reduced station temps for the period 1900-1940 by a full 2 degrees C. what I have never seen is how that 2 C was arrived at.

  5. March 9, 2012 6:30 am


    The unadjusted (correct) graph is to the right here above. It was really as warm in Reykjavik around 1940 as today.

    The temperature drop around1965-1980 on the graph is real. I remember well the sea ice (very unusual here) and the severe economical impact of this “micro ice age”. (Mainly because of fishing stocks that fell). I was 20 in 1965 🙂

    The adjusted (wrong) graph is to the left.

    There should be no need for UHI adjustments for stations where the wind is constantly blowing around them. The weather is almost never calm in Reykjavik 🙂


    The correct graph for Reykjavik on the Icelandic Met Office web is the second one from the top: (This one:

    Best regards


Comments are closed.

%d bloggers like this: