What do you mean, the Gulf Stream doesn’t keep Europe warmer than North America? How even scientists are afflicted by urban myths

A post by Chris RowanIn science, you discover that you’re wrong at least as often as you’re proven right – and the things that you end up being wrong about can be quite surprising. Prior to last week, if asked I would have confidently confirmed that the reason the UK does not have a polar bear problem, despite being located at the same latitude as Hudson Bay, is the heat supplied by warm water transported into the northwest Atlantic from the Gulf of Mexico by the Gulf Stream. But then, courtesy of Kevin Anchukaitis, known on Twitter as thirstygecko (and someone you really should follow if you’re interested in things paleoclimatic), I found out that this is not the case at all.

. @ 'The Gulf Stream-European climate myth' http://t.co/kfOWaDYd :)
Kevin Anchukaitis

If you follow the link, Lamont ocean and climate physicist Richard Seager explains how even in models of Earth’s climate where ocean heat transport had been turned off, the UK and western Europe have milder winters than the eastern US at the same latitude. Both sides of the Atlantic got colder as the static oceans made heat linger in the tropics, but the large trans-Atlantic temperature difference remained constant.

The Gulf Stream is part of the global ocean conveyor, but it doesn't convey much extra warmth to Europe. Source: Marika Holland.

The oceans do still play an important role in keeping Europe’s winters mild, but it is nothing to do with the Gulf Stream. The winds that blow northeast onto Europe from the Atlantic carry with them air that is relatively warm even in the winter; the large heat capacity of water means that the sea cools off more slowly in the winter, and this also moderates the temperature of the air above the sea surface. This contrasts with the eastern US, where the prevailing southeast winds are cold, having lost their heat to the thousands of kilometres of land surface they’ve already passed over.

So the oceans are still involved, but it is the atmosphere that is the true key to explaining Europe’s mild winters; and in a pleasing geological twist, it is apparently the presence of the Rocky Mountains that causes the large scale waviness of atmospheric circulation that magnifies the temperature contrasts on either side of The Pond (Richard Seager explains this in more detail in this more comprehensive piece at the American Scientist).

How the Rockies cause the westerlies to wiggle, magnifying the trans-Atlantic temperature difference. Source: The American Scientist

According to Seager, the notion that the Gulf Stream was warming Europe can be traced back to a book first published in 1855, and is “the climatological equivalent of an urban legend”. It is certainly persistent enough, although given that this particular “fact” has been promulgated not by a friend’s sibling’s cousin’s friend over a pint on a Friday evening, but by scientists and educators in newspapers, televison programmes and lectures, it’s arguably even more pernicious. But how did this happen? Isn’t science meant to be self-correcting?

Here’s how it can happen. In the introduction to your average paper, you’ll often see sentences along the lines of:

The link between [phenomenon] and [process] has long been known (Bloggs, 1996).

The implication is that everyone knows and accepts this, so it’s not worth wasting time going through the evidence in painstaking detail; but if you’re interested, you can look up the given reference for the gory details. Most of the time, this is exactly what you get when you track the given reference down; but sometimes, you find that it is nothing more than the oldest reference to this fact that the original paper’s author was willing or able to look up, and all it says is:

There is strong evidence that [phenomenon] and [process] are linked (Obscuro, 1982).

If you persist further, you may find yourself going through the process of looking up a reference, only to be directed to an even earlier one, several more times before you finally reach the canonical document, the one that contains actual data and discussion. And this is what you find:

Based on [invalid data] collected using [method known to be inaccurate] and assuming [long disproven assertion], we conclude [process] causes [phenomenon].

This is how a scientific urban myth is born: by the time you reach a citation 3 times removed from the supporting observations, a conclusion becomes something ‘everyone knows’ despite very few people ever being exposed to the evidence it was based on. “I’m telling you, this paper told that paper that this other paper has compelling evidence for this! Compelling! Well no, I haven’t actually read it myself…”

I don’t believe this is a hugely common phenomenon. But science nowadays is such a vast body of knowledge that there are bound to be a few zombie ideas traipsing around in it, managing to survive because no-one has really properly examined them for a while. It is only when a scientist is inspired to chase one of these ideas back to its origin that they are brought into the light.

Categories: academic life, climate science
Tags: , ,

Comments (17)

  1. Torbjörn Larsson, OM says:

    And of course by bloggers et al doing the public service to broadcast the message.


    [/goes off to at least check _this_ source one step back. =D]

  2. BDoyle says:

    “This contrasts with the eastern US, where the prevailing southeast winds are cold,”

    I will admit that this got me confused for a minute. Isn’t it sort of odd that if you are out standing in your field, a southeast wind is obviously one that blows FROM the southeast, but if you are looking at a map, a southeast wind is intuitively one that blows TOWARD the southeast?

    • Chris Rowan says:

      As I understand it, you use the ‘-erly’ suffix when you’re talking about winds blowing from a particular direction: westerlies blow from the west, to the east. It can get confusing, though, I agree.

  3. Hi Chris,

    Very nice write-up and thanks as well for the mention. For anyone that wants to get a bit more into the weeds on oceanic heat transport and read some counterpoints to Seager et al, Rhines and coauthors have a book chapter here (Google Books link): http://goo.gl/46J0S called ‘Is oceanic heat transport significant in the climate system?’


  4. Kim says:

    That Scientific American article might be really good reading for a process of science discussion in an intro Earth systems or oceanography class. I might bring the article to a workshop I’m going to next week, and think about how to set up class to make it work.

  5. Gavin says:

    I love the whole ‘debunking of urban myths’ meme, but I think you are in the middle of promulgating another one. The implication that the Gulf Stream (or more correctly the meridional overturning circulation whose surface expression adds to the wind-driven component of the Gulf Stream – MOC/GS for short) does not impact European climate is a fallacy. The differences though are all in the precise nature of the questions being asked.

    If someone asks, does the Gulf Stream (or MOC/GS) make Europe’s winters mild, the answer is yes. This is an absolute statement with no reference to conditions anywhere else. In Seager’s paper and in many other (Rind et al; Vellinga et al, Stouffer et al etc.) climate in Europe without the MOC reduced or shut off, gives a European cooling of around 3 to 6 deg C. Estimates of cooling associated with proxy estimates of MOC slowdowns are comparable.

    If however someone asks whether the MOC/GS makes Europe milder with respect to Labrador or Newfoundland, then Seager’s argument is correct. MOC slowdowns affect both sides of the Atlantic pretty much equally.

    But to automatically assume that all questions posed like the first are actually questions posed like the second is an error. Often people are indeed trying to discern what might happen if there is a MOC slowdown (‘day after tomorrow’ scenarios etc.), and so the answer to the first question is most relevant, not the answer to the second.

    This discussion might be better served by acknowledging that different questions have different answers and helping people work out what question they really want answered is the way to go, rather than going down the urban myth/debunking path which can get tangled up in some unnecessary strawmen.

    • Chris Rowan says:

      Thanks for dropping by, Gavin.

      I did try to make it clear in the actual post that we’re talking about a temperature gradient across the Atlantic, rather than an absolute temperature – and the ocean circulation does affect that. The title, though, does potentially mislead, and I’ve modified it in an attempt to rectify that. Growing up in the UK, the Gulf Stream was always discussed in the sense of it making Europe warmer than we’d expect, so I perhaps didn’t think as carefully as I should have about the possible confusion.

      • Gavin says:

        But Europe is warmer than you would expect because of the Gulf Stream (MOC etc… – caveats always included so as not to induce the wrath of the Wunschians).

        That of course begs the issue of what one bases those expectations on. If they are based on the North Pacific (where there is no MOC to speak of), then the UK is warmer than expected. (Compare the climate of Prince Rupert, BC with Manchester for instance). It is not at all clear to me that every time someone says ‘warmer than expected’ or the ‘Gulf Stream warms Europe’ that they are strictly referring to the relative temperature difference across the Atlantic. Indeed, I doubt very much that this is the case.

        Rather, the plain interpretation of whether Europe would or would not be warmer with the Gulf Stream/MOC based on absolute temperatures appears the most obvious, and in that sense there is no problem with the public understanding.

        That isn’t to say that there aren’t some terrible leaps made in popular coverage of this – there are many (here for instance). But this has much more to do with over-simplification of the system rather than basic error in conception or an urban myth.

      • Hello Chris, Gavin,

        I guess I don’t personally see any promulgating of myths in the post above. And I also disagree that a statement like ‘the Gulf Stream is responsible for Europe’s mild climate’ isn’t (sorry for the double negative) more often than not referring to the cross-Atlantic temperature contrast — based on that observation that one seldom seems to hear said that ‘the Gulf Stream is responsible for Newfoundland’s mild climate’ :)

        With respect to AMOC shutdown, indeed the Seager presentation at the link in my original tweet notes, as Gavin indicates, the basin-wide temperature decrease that accompanies a hosing-induced shutdown.


      • There is a figure illustrating the temperature effect from AMOC shutdown here:


        This isn’t precisely the one I was looking for, but it illustrates the temperature decreases in response to AMOC shutdown (I think this is from Stouffer)


      • Gavin says:

        Kevin, the absence of a (correct) observation in the public discussion is not a particularly strong basis upon which to claim that a related but oft cited observation is wrong. This may be in no small part related to the observation that prominent media outlets based in Newfoundland are in rather short supply – relative to their concentration in Europe of course. ;-)

        It is one thing to find examples of the specific error targeted by Seager et al (that the difference with respect to Labrador is related to the GS) and correct them. It is another thing to take examples of a common truism that makes no reference to Labrador at all, and insist that they must all be interpreted in the least favorable way.

        Does the Gulf Stream keep Europe warmer? The answer is yes.

        (Some relevant figures: fig 4 in Lewis et al (2011), Fig 2 in Wiersma and Renssen (2005), LeGrande et al, 2006; Vellinga and Wood (2002) etc.)

      • Gavin says:

        Just for a little context, these arguments have been ongoing for 100 years at least:
        James Croll, 1885

        Page 41 describes what Croll considers “Europe to be warmer than” (due to the Gulf Stream). He specifically uses the temperatures with respect to total longitudinal mean at that same latitude. And this is very clearly warmer because of the GS as well as being affected by the position on the eastern part of the ocean basin.

    • Hi Gavin,

      Thanks for the Lewis reference — that was the figure I was thinking of (I think).


  6. jbwarriner says:

    ‘Science’ by press conference, by media ‘science writers,’ by (dare I say it) bloggers and blog commentators, headlines, etc. is a pernicious virus. Just today, in addition to your well-done description of the problem, along came two separate instances. We got headlines that Neanderthals were the first artists and we got an origin for the 1258AD eruption — that was neither specified nor described. The climate change fight is replete with examples of the problem from both sides. I guess the problem is an intersection of minimal general science education, near-zero practice in critical thinking, and an explosion of interconnected communication. Lots of places to lay the blame but generalities like ignorance, politics, money, and ego won’t fix it.

  7. Lab Lemming says:

    As a non-climate earth scientist, when I hear “gulf stream” I think North Atlantic Gyre, not AMOC.
    Are people possibly not even arguing about the same thing here?

  8. Chris Nedin says:

    Has anyone calculated what the european winter would be if the north atlantic was the temperature it is around Newfoundland, instead of what is currently is? Since without the gulf stream, the north atlantic would be the same temerature as it is around Newfoundland (north polar currents flowing south). If the gulf stream has no effect, then the temperature would not change, but, if the north atlantic is kept warmer because of the gulf stream then it does affect the european winter. If warmer winters are due to warmer air and warmen air is due to warmer water and warmer water is due to the gulf stream, then the gulf stream is having an impact.

    Links (2)
  1. Pingback: What’s up? The Friday links (35) | paleoseismicity.org

  2. Pingback: Europe Is Warmer Than Canada Because of the Gulf Stream, Right? Not So Fast | Smart News