Skip to content

The European Heatwave Of 2003

September 21, 2017

By Paul Homewood


 Difference in average temperature (2000, 2001, 2002 and 2004) from 2003, covering the date range of 20 July – 20 August

There is little doubt that the European heatwave of 2003 was one of the defining meteorological events of the last hundred years.

Climate scientists have often held it up as an example of what the future has in store, as global warming takes hold. Indeed, every time the sun comes out in summer, we seem to get the mantra that “heatwaves are becoming more common”.

We heard the usual nonsense this summer when southern Europe had a hot spell. It was even given a name, Lucifer. The Telegraph reported:

Unusually high temperatures, in some cases unprecedented, are being recorded across an area spanning much of the Iberian peninsula, southern France, Italy, the Balkans and Hungary.


But was this summer abnormally hot, and have we been seeing 2003 style heatwaves since?

It was France which was most seriously affected back in 2003, so let’s start there, with Limoges in the south west of the country:


The summer of 2003 stands out very clearly. This summer, while warmer than average, was 2.1C cooler than 2003. It was also no hotter than 1976.


In Spain, the 2003 heatwave was less pronounced. As the temperature record at Alicante shows, despite the hype, this summer was not even as hot as 1952.


And at Pisa in Italy, this summer was 2.7C cooler than 2003. As in Spain, summer temperatures in 1950 and 1952 were at a similar level.



All of these places are, of course, heavily urbanised, so much of the recent increase in temperatures may be due to UHI. There are very few genuinely rural stations in these countries with long term records, at least not in the GHCN record anyway.

There is one, however, at Mont Aigoul, in central France. When we look at the record there, we find that it was as hot or hotter than this summer in 1949, 1950 and 1951.



The heatwave in 2003 was clearly an anomalous meteorological event, but there is no evidence that it has any climatic significance, or that it tells us anything about future summer weather. It is similar in this respect to the dustbowl years in the US.

Climate scientists have used it to scaremonger about global warming. It is time for them to drop this pretence, and admit they were wrong.


All temperature data from GISS:

  1. September 21, 2017 2:10 pm

    Reading and sharing an article from “Not a lot of people know that” in France re smart meters….

  2. Jack Broughton permalink
    September 21, 2017 2:14 pm

    How can they possibly admit that they are wrong when everything in weather is consistent with global warming theory? Sadly, our free-press has lost its freedom and only reports what Big Brother requires of it.

    • Adrian permalink
      September 21, 2017 2:55 pm

      I don’t think the brain-free press is doing it deliberately as part of some Machiavellian plot, it is simply that they employ the same arts degree (at best mebbe 2.2 level) ‘left liberal’ semi-sentient drones that infest our political classes. They confuse mantra with thought and ranting with reason. They understand nothing because they have never been taught to actually think. They repeat what those around them say because that is all the wit they have.

      Is it going to turn out well? The history of Man is replete with examples of such behaviours, from religions to mass political movements leading to persecution, warfare and genocide. I can’t think of any of those that finished well so take you guess on this one.

  3. September 21, 2017 2:21 pm

    The european heat wave of 2003 has been officially “attributed” to AGW with “event attribution science”. It’s a flawed device as shown here in a case study of the same kind of attribution of the 2000 floods in england and wales.

  4. jim permalink
    September 21, 2017 2:47 pm

    We spend April to October 40km from Toulouse, SW France. At the end of this August Meteo France were telling us that the summer of 2017 was the hottest summer in France since 1900. We read this as we shivered at the beginning of September!
    It was definitely not as hot as 2016 where we are. There were two hot spells, 10 days in June and 10 days in August. July was a wash-out, and the heating has had to be put on in September ( although its got a little warmer today).
    I understand Provence has had a long hot and dry summer, but Provence is not ‘France’.
    They lie, there is no other way of putting it, they lie.

    • Ben Vorlich permalink
      September 21, 2017 8:08 pm

      I can confirm your weather report, it was the same for Northern Nouvelle Aquitaine.

  5. September 21, 2017 3:09 pm

    The Wiki article you link to is very revealing. The section on France lays the blame in part at the door of the French government for not reacting quickly enough, due to the French habit of effectively closing down government for the month of August. No mention of climate change. Then click on the French version and you’ll see an entirely different (though still critical) article.

    I trawled Wiki on this subject a few years ago, and the articles were entirely different. The Italian version was much longer than now, with Italy apparently vying with France for the record of number of deaths. There was also a version in Czech, I think, which seems to have disappeared.

    IPCC AR4 has a chapter on this, I believe, which manages to cite a French Senate report in support of its thesis that AGW is to blame, though the Senate said nothing of the sort, plus an entirely irrelevant paper by the French medical whistleblower, apparently cited for no other reason than that it’s in English.

  6. Broadlands permalink
    September 21, 2017 3:18 pm

    “There is little doubt that the European heatwave of 2003 was one of the defining meteorological events of the last hundred years.”

    So was 1921…Maybe even worse?

  7. Athelstan permalink
    September 21, 2017 5:28 pm

    Paul, 2003 was a few hot days, 76 was a summer of hot days.

    Besides which the T data sets have been so ‘fixed’ it’s difficult to tell – yeah it was hot but not as hot as they [Exeter UEA, GISS et al] like to make it out and I don’t set any store in a single figure from Gravesend, Heathrow nor Greenwich either.

  8. Athelstan permalink
    September 21, 2017 9:16 pm

    and btw, the top graphic depicts Euro heat and further east it being somewhat cooler.

    If I recall correctly [maybe] then, Jet stream ‘meanders’ was posited as a causal climatic factor during 2003. Britain and western Europe were in a high pressure loop, which was drawing warm air up towards these shores and all the way from the north African continent, whereas in the east over Russia they were suffering a series of depressions drawing in a cold plunge from the Arctic, or so it was told and all due to man made CO2 apparently, whodathunked that – eh?

  9. Ian permalink
    September 21, 2017 10:08 pm

    ‘Scuse my ignorance, but whilst the 2003 results to stand out, all the charts seem to show a gradual increase in temperature. Doesn’t this support the alarmists’ claims, whatever the reasons for the 2003 spike?

    • 1saveenergy permalink
      September 22, 2017 5:45 am

      Which alarmists’ claims ?
      They claim everything is bad…. that’s the nature of being alarmist

  10. CheshireRed permalink
    September 22, 2017 8:57 am

    Notice how this years heatwave was given a suitably scary and foreboding name as Lucifer, with all its devilish connotations. Let’s be honest they were never going to call it Gerald.

  11. Green Sand permalink
    September 22, 2017 10:00 am

    Here we go, as predicted!

    ‘E.On ‘ploy’ to persuade customers to get smart meters with cheaper deals ‘

    “One of the UK’s biggest energy firms will for the first time offer cheaper deals to customers if they agree to have a smart meter installed.

    The move by E.on, which has nearly 5 million customers, the vast majority of which do not yet have a smart meter, is being described as a “ploy” to get more people to accept the new style meters…….

    Well I am sure they will pleased to know their little ‘ploy’ has just cost them one customer.

  12. Geoff Sherrington permalink
    September 23, 2017 2:32 am

    The following show heatwave data for 5 Australian capital cities, home to most Australians. They do NOT support the official line that heatwaves are becoming longer, hotter and more frequent. You guys in UK should make some of these for your cities. Inarguable data soon trumps myth. Geoff

  13. robinedwards36 permalink
    September 23, 2017 2:28 pm

    The assessments of Limoges temperatures in this thread all miss an important point regarding observed temperatures that applies equally well to all NW Europe sites. This is that the linear model that seems to be applied universally to temperature time series over whatever scale interests the observer, cannot be regarded as a reasonable approximation to the real underlying behaviour of the series. In Europe it is simple to demonstrate that in late 1987 (often September) a discontinuity in the series occurs. A sudden change of around 0.7 to 1 deg C occurred in a series that had previously been almost constant for quite a number of years, and which after the discontinuity again remained remarkably stable.
    If you do not believe this simply “cherry pick” by dividing the series into two main segments, pre 1988 and from then onward to the present. Fit linear models to the two separate segments, plot the regression fits, and note the difference in the two means. It all becomes obvious. Try it with any NW Europe temperature data.

    The rationale behind this apparent “cherry pick” is that cumulative sum diagrams derived from this type of data invariably have a grossly obvious discontinuity at around this date. It cannot be an instrument problem. It is too consistent over widely different sites. In my view it is not an arbitrary choice of dates but one dictated by the data themselves, once you know how to examine them properly.
    It all becomes much more obvious if you “de-seasonalise” the data, for instance by subtracting monthly averages from each of the monthly data, to help eliminate the gross month to month differences prior to further analyses.

    Regarding “outliers”, the Limoges data display some really substantial ones, and formal outlier tests (Grubbs’ and Rosner’s) demonstrate just how odd some seasons or individual months’ data are. If the “outliers” are omitted the remaining data conform closely to “normal distributions” as one might expect. I can supply all the numerical details here, but cannot upply graphics (except via email) if any of you are interested.


Comments are closed.

%d bloggers like this: