Skip to content

No Greenhouse Warming In Arctic – Mark Serreze 1993

October 22, 2014

By Paul Homewood





Back in 1993, Arctic alarmist, Mark Serreze, was joint author of a letter to Nature regarding research they had done on the evidence for greenhouse warming over the Arctic. 


ATMOSPHERIC general circulation models predict enhanced greenhouse warming at high latitudes owing to positive feedbacks between air temperature, ice extent and surface albedo.

Previous analyses of Arctic temperature trends have been restricted to land-based measurements on the periphery of the Arctic Ocean. Here we present temperatures measured in the lower troposphere over the Arctic Ocean during the period 1950–90. We have analysed more than 27,000 temperature profiles, measured by radiosonde at Russian drifting ice stations and by dropsonde from US ‘Ptarmigan’ weather reconnaissance aircraft, for trends as a function of season and altitude. Most of the trends are not statistically significant. In particular, we do not observe the large surface warming trends predicted by models; indeed, we detect significant surface cooling trends over the western Arctic Ocean during winter and autumn. This discrepancy suggests that present climate models do not adequately incorporate the physical processes that affect the polar regions.



Before anybody jumps on me, yes this is only up to 1990, but it nevertheless shows there is no evidence at all for polar amplification, at a time when CO2 emissions were rising rapidly.

It is also interesting that they realised they could not trust the small number of land based measurements.


But what is most interesting is that, according to GISS’s hand selected handful of land based stations, much of the Arctic did show moderate warming between 1950 and 1990, particularly over the western Arctic.




More tellingly, if we use GISS annual data, we get the increasing trend below.





From all of this we can gather that:


1) GISS’s temperatures for the Arctic are worthless.

2) Grant funding can trump the best of research.






The map presented originally was incorrect, and has been replaced with the above.

  1. David permalink
    October 23, 2014 1:32 pm

    The abstract refers to “temperatures measured in the lower troposphere over the Arctic Ocean”; so comparison surface station data isn’t valid. Also, the GISS chart only shows September data. GISS annual (D-N) Arctic data shows practically no change at all between 1950-1990:

    UAH NoPol/Ocean shows -0.18C/dec cooling in the 145 months from Dec 1978-Dec 1990. Sticking to 145 months, Arctic LT (ocean) warming didn’t really begin until late 1991. Between Nov 1991-Nov 2003, LT above the Arctic ocean warmed at a rate of +1.35C/dec.

    Interestingly the running 145 mth trend is now currently negative: back to -0.18C/dec again!

    • October 23, 2014 6:15 pm

      Thanks David. Good spot on the map. I have replaced it, and also added a graph which shows the trend better.

      Serreze does not seem to think surface and air temperatures cannot be compared. For instance he says

      we detect significant surface cooling trends over the western Arctic Ocean during winter and autumn.

      Interesting that you found yesterday that UAH temps match closely with surface ones! Now you say they cannot be compared.

      The bottom line though is that

      a) Serreze felt he could not do a proper analysis based on a handful of surface sites
      b) There was no evidence of GHG warming. This would suggest that natural cycles have had at least some effect in both the 1950-90 period and in the subsequent warming.

      • David permalink
        October 24, 2014 11:11 am


        According to GISS, the rate of warming in the Arctic between 1950 and 1990, as shown on your chart, is +0.05C/dec. In the text you refer to this as “moderate warming”.

        The rate of global warming in GISS since 1998 is +0.07C/dec. Do I take it then that you now accept there has been at least ‘moderate’ global warming since 1998?

      • October 24, 2014 11:37 am


        When satellites and surface sets disagree, we are told they are measuring different things. When they agree, we are told that this confirms the accuracy of surface sets.

        Too funny!

      • David permalink
        October 24, 2014 11:44 am

        But I am referring to surface data in both cases Paul. In fact, to GISS in both cases.

        You appear to be saying that when GISS measures a rate of +0.05C/dec in the Arctic it represents “moderate warming”; but when it measures a rate of +0.07C/dec globally this doesn’t represent any sort of warming.

        That’s like saying a rate of 5 mph is ‘moderately fast’ but a rate of 7 mph isn’t fast at all.

      • October 24, 2014 12:20 pm

        I’ll amend it to “moderately insignificant”!!

        But you have not responded to me previous point about satellite v surface temps. Or the role of natural cycles in all of this. Or the fact that Serreze and co felt they could not rely on a handful of surface stations.

        As an aside, there appears on the GISS map to be up to 1C of warming in the western Arctic, where Serreze found significant surface cooling trends during winter and autumn.

    • David permalink
      October 24, 2014 12:58 pm


      Moderately insignificant. Like it. You’d accept that 0.07 is moderately less insignificant than 0.05!?

      Re satellites; the point I intended to make was that Serreze et al. were specifically measuring ‘the lower troposphere over the Arctic Ocean’. As you pointed out, the GISS data measures surface temperatures from land-based stations; though these may be extrapolated over the oceans (I’m not sure).

      The best modern equivalent we have for the data Serreze et al. were analysing is undoubtedly UAH NoPol/Ocean. This agrees with Serreze et al. re flat temperatures up to 1990. As I mentioned above, the faster rates of warming observed in the Arctic didn’t really start until the early 1990s.

      I’m not suggesting that natural conditions, such as AMO, haven’t played a roll in this.

  2. Derek permalink
    October 23, 2014 7:17 pm

    I was interested in the temperature scale on the graph – degrees C x 100 , giving a rise of 0.2 degrees C over the 40 years. I am sure that I have read some reports that give a much higher rise than that. If this is correct then what has caused the recent melting? I’m no believer in CAGW, but I always thought that the Arctic was one place where warming had taken place.

    • October 23, 2014 9:51 pm

      Don’t forget this is 1950-90.

      But the major cause of ice melt has been the incursion of warmer waters from the south, combined with wind blowing the ice out through the Fram Strait to melt in warmer waters.

    • David permalink
      October 24, 2014 11:34 am


      The total rise in Arctic surface air temperatures between 1950 and 2013, according to the NASA data Paul posted above, is 1.8C.

      UAH data only runs since December 1978, but this shows a total warming of the lower troposphere above the Arctic (land and ocean) of 1.6C to date. Between 1979 and 2013 NASA shows 1.9C Arctic warming.

      This confirms that the Arctic actually cooled slightly (about -0.3C in total) between 1950 and 1978.

      • October 24, 2014 12:26 pm

        Interesting that, on GISS, all of the 1.9C increase took place between 1979 and 1981!

        The anomalies were +1.28 and +1.26C for 1981 and 2013 respectively. As we all know, 1979 was an exceptionally cold year in the Arctic.

      • October 24, 2014 12:31 pm

        Also, on GISS, there was an increase of .59C between 1950 and 1990, the period where Serreze found none.

      • David permalink
        October 24, 2014 1:11 pm

        The total rise in GISS between 1950 and 1990 was 0.20C Paul, not 0.59C. 0.59 is what you get if you deduct the 1950 anomaly from the 1990 one. All that tells us is the difference between those two years; not the full change over time.

        Full change over time is found from the ‘linest’ function (Excel); highlight the range in brackets and multiply by the total number of data points (in this case 41). The decadal rate is just 0.05, which as you’ve just said is not significant.

      • October 24, 2014 9:41 pm

        One of the serious criticisms of the GISS record is the way it has cooled the past back to the 1940’s, particularly in the Arctic.

        The raw data shows that temperatures in the 1940’s were not much different to now there.

        As Serreze (no doubt the junior partner) and co spent ages analysing radiosonde data from then, it would be interesting if he resurrected it and compared it to UAH/RSS data since 1979.

        I wonder how much real warming we would see?

  3. October 25, 2014 6:55 pm

    Isn’t it relevant that the current increase in average Arctic temperature is due to the winter months (Jan-Mar) when the sun isn’t out? That suggests to me that this increase has nothing to do with AGW. If there were an AGW-related increase, wouldn’t it be greatest during the summer months, when the sun is out – in contrast to what DMI have reported over the last few years?

Comments are closed.

%d bloggers like this: