Skip to main content

When can we attribute extreme events to climate change?

Dan Farber, professor of law | October 31, 2011

Moscow suffered from a severe heat wave in the summer of 2010, with temperatures reaching 101 degrees and an average temperature 14 degrees higher than normal for July.  What are the odds that the heat wave was due to climate change?

Wildfire sparked by Russian heat wave

Wildfire sparked by Russian heat wave (National Geographic)

RealClimate presents the results of an analysis that was just published in the Proceedings of the National Academy of Science.  The analysis indicates that there was an 80% probability that climate change caused the heat wave.

A previous study showed the opposite, so it’s natural to ask the reasons for the difference.  There seem to be two basic reasons.  First, the earlier study made an erroneous correction for the urban heat island effect, which incorrectly assumed that the effect is equally strong all year around.  The newer study used satellite data to correct that error.  Second, the earlier study assumed that climate change was linear over the past century, which is unlikely since CO2 concentrations have had a non-linear increase.  Not surprisingly, a non-linear trend fits the data better.

You can actually get a sense that something unusual is happening from this simple graph.  Blue dots are record colds and red dots are record highs:

graph of Moscow temperatures

Moscow July temperatures (Centigrade)

There are a couple of things that are suggestive about this graph, besides the overall upward trend.  First, there hasn’t been a new record cold temperature in a century, and the coolest months are well above the previous record.  Second, the last record-breaking high temperature only broke the previous record by a small amount, but the 2010 broke its predecessor by a much larger amount.  These facts suggest — but obviously don’t prove — that there is an underlying shift in temperature over the past century which is affecting the record-breaking events.

The article also makes a more general point about record-breaking events.  If natural variability is large relative to trend, we shouldn’t expect to see a lot of record-breaking events because they’ve already had many opportunities to occur in the record.  But as the trend grows larger, the number of broken records also increases.  So we can expect many more broken records to come as climate change continues to grow relative to background natural variability.

Cross-posted from the environmental law and policy blog Legal Planet.

Comments to “When can we attribute extreme events to climate change?

  1. I find it interesting that this article is written by the Dean of the Law School–a professor of law, not a scientist. Yet, the article attempts to draw a scientific conclusion rather than a legal one. Please stick to your area of expertise.

    • EG, the fact is that scientists need all the help they can get to explain facts to the public.

      The credibility of scientists has hit rock bottom and no one pays attention to what they say anymore because global warming is at or near the bottom of most priority lists.

      We (humanity) desperately need some one to motivate us to demand action from our politicians.

  2. Hi Dan–

    The statement “there was an 80% probability that climate change caused the heat wave” is extremely difficult to make sense of. Either climate change caused (or contributed to) the heat wave, or it did not. There is no “probability” about it.

    The data–temperature measurements–might have a random component from instrumental errors, for instance. But using probability to describe the state of the world is fraught with serious philosophical and statistical problems.

    For instance, it generally makes no sense to say “there’s an 80% chance that the defendant committed the crime.” Either the accused party committed the crime, or not. We might not know whether the defendant is guilty, but that doesn’t turn guilt into something that’s random.

    It is quite different to say “if the state of the world is X, there’s an 80% chance we would observe Y” than to say “we observed Y, so there’s an 80% chance that the state of the world is X.” The former can be sound statistics. The latter is generally nonsense. Probability might describe the data, but generally does not describe the state of the world–quantum physics aside.

    Similar problems arise in making probabilistic forecasts about the world. Statements like “There’s a 90% chance that global temperature will go up by 3 degrees in the next 50 years” don’t parse as scientific claims if you really try to make sense of them. This is a different issue from whether the underlying data are reliable, whether the underlying models are realistic, and so on.

    For a discussion of such difficulties, see Stark, P.B. and D.A. Freedman, 2003. What is the Chance of an Earthquake? in Earthquake Science and Seismic Risk Reduction, F. Mulargia and R.J. Geller, eds., NATO Science Series IV: Earth and Environmental Sciences, v. 32, Kluwer, Dordrecht, The Netherlands, 201–213. (preprint:

    Fiat lux,

  3. I wish the term used were “climate disruption” as more people could associate recent massive rains, destroyed crops, untimely snows and huge fires as the peril we are already experiencing. The warnings about “climate warming” of a degree or two or even “climate change” are not breaking through the general consciousness, but seems to encourage doubt and challenge.

  4. Nicely done. Regrettably, no event will “prove” that global warming is a reality, or as I like to say, God is not going to write it in the sky. We will each have to decide when it is proven to our satisfaction, and how high to set the bar. For me, James Hansen’s Congressional testimony in 1988 was persuasive, and the balance of evidence that has accumulated since that time–in particular the rapid deterioration of Arctic sea ice–has simply added to the case that global warming not only exists, but poses a profound threat to civilization.

    • Tom, the greatest threat of all was stated by the Royal Society’s president Martin Rees who questions whether we are smart enough to use what we’ve learned to save ourselves.

      Scientists have been paid more than enough to produce hydrogen bombs that can destroy humanity, but no one is paying scientists enough to produce hybrid fusion to eliminate CO2 producing power plants around the world to save humanity or it would have been done by now.

      To make this worst case scenario even worse, we have out of control world population growth together with insatiable consumption that is using up and/or contaminating resources to sustain acceptable quality of life for the population we have already.

      And our politicians and intellectuals have no answers whatsoever that will solve these problems before we pass the point of no return.

      Rees is right, we truly are not smart enough, and altruistic enough to save ourselves, or we would have done it by now.

  5. “First, there hasn’t been a new record cold temperature in a century,”

    Lies, media reported them ad nausea in the last 5 years.

    Global Warming = Raises for Filthy Rich proffesors, case solved ! 😉

    • Xinghu–I suggest you read the article more carefully and look at the graph in detail before writing an idiotic and reactionary comment. The statement “First, there hasn’t been a new record cold temperature in a century,” is referring to the temperatures in Moscow, based on the graph, not world-wide temperatures or temperatures in other cities. If you look at the graph the last record cold temperature was in ~1905 and was about 14.5 degrees C. Since then, over a period of more than a century (exactly what the quote says) there has not been another cold temperature lower than 14.5 degrees C (another new record cold temperature) in Moscow but yet there have been three new record high temperatures. I hope now you understand the quote you called a ‘lie’ a little more clearly or should I get a ‘filthy-rich’ professor to explain it to you?

  6. Dan, thank you very much for your continuous efforts to educate us on the latest climate change facts so that we might start demanding actions to protect the long-term quality of life for our latest and future generations.

    Unfortunately, we live in an era when candidates in presidential debates are cheered for attacking human rights, so we can’t expect enough people to begin caring about climate change even “as climate change continues to grow relative to background natural variability.”

    Once again humanity is most at grave risk because of a paramount fact of life that Will and Ariel Durant proved in “Lessons of History”:

    “When the group or a civilization declines, it is through no mystic limitation of a corporate life, but through the failure of its political or intellectual leaders to meet the challenges of change.”

Comments are closed.