Mathematics, Media Criticism, Political Commentary, Science And Society, Technology and Society

Climate Change And Extreme Weather

11 Comments

People like Seth Borenstein were excited to link to this paper yesterday.

“This new study helps get the actual probability or odds of human influence,” said University of Arizona climate scientist Jonathan Overpeck, who wasn’t part of the research. “This is key: If you don’t like hot temperature extremes that we’re getting, you now know how you can reduce the odds of such events by reducing greenhouse gas emissions.”

Lead author Erich Fischer, a climate scientist at ETH Zurich, a Swiss university, and colleague Reto Knutti examined just the hottest of hot days, the hottest one-tenth of one percent. Using 25 different computer models. Fischer and Knutti simulated a world without human-caused greenhouse gas emissions and found those hot days happened once every three years.

Then they calculated how many times they happen with the current level of heat-trapping gases and the number increases to four days. So three of the four are human caused, the team said.

This is crap science, because it’s based on crap models, that have been failing.

Interestingly, even Kevin Trenberth agrees with me:

“The paper is interesting and has some results that may be reasonably OK,” he said. “However, the paper is based almost entirely on models with little or no validation or relations to the real world. None of the models do precipitation realistically, and some are quite bad.”

You don’t say. Garbage in, garbage out.

11 thoughts on “Climate Change And Extreme Weather”

  1. The “hottest one-tenth of one percent” translates mathematically to “the hottest 1 out of 1,000”. There are 365 days in a year. The hottest 1 day out of 1,000 days would therefore occur 1.095 times every three years.

    Their “research” indicates that such an event, absent AGW, “happened once every three years”.

    Really? You don’t say! An event that happens once every 1,000 days would only happen once every 3 years? That’s some powerful statistical modeling you’re using!

    People may also be surprised to know that, absent AGW, the hottest 1% of days would only occur 36 days every decade.

    I’d reveal more “startling” statistics such as these, but I want to save them for my upcoming paper. Plus, now I have a grant proposal to write.

    1. Also, it’s complete bullhockey to claim that the hottest 0.1% of days happen any more than once per thousand days, absent or present AGW. It is statistically impossible to have 4 “hottest” days. It’s either the hottest or it isn’t, unless you’re recording at such imprecision that you have large groups of days that all share the same temperature, in which case you’re not really proving anything, other than that you have low precision in your recording instruments (or in your “massaged” data set).

  2. At 58, I’m old enough to remember when some people actually said things like “If it comes from a computer, it must be right.” I’m also old enough to remember when “Made in Japan” meant “junk.” We all know that neither of those statements are accurate.

    If your computer model doesn’t agree with real-world observations, it isn’t the real-world that’s wrong.

    2. “If it comes from a computer, it must be right.”
      Those of us in the computer field have long had another phrase, “Garbage in, garbage out.”

  3. Hard to see the point of all this. Obviously, if a system is on the average a little bit warmer, the far right-hand edge of its Bell Curve will be a little bit to the right. You will get more of the very rarest days in the right-hand-tail. The farther out you go on the tail, the more it will dominate the same curve shifted a little bit to the left; that’s how Gaussians work.
    The same would be true on the left-hand-side: There will be considerable fewer of the very very coldest days. Which may be even more important.

    1. Or it could be that the mean hasn’t changed but the standard deviation is larger, in which case you’d get both more of the hottest and coldest days.

  4. If you don’t like an ultra hot day every 3 years, then you know what to do: stop emitting greenhouse gas, such as CO2 emission from breathing.

  5. Most people live in cities, and cities cause urban heat island effect and this can cause hot summer day to be few degrees warmer than it would be otherwise.
    global temperature measurement try to remove this warming effect caused
    by UHI effect [as it’s not part of global temperatures]. But in terms of effecting most people the UHI effect causes higher air temperature in the daytime and mostly higher temperature at night.
    So obviously humans have caused this local warming effect and unlike the slight rise in global temperature this can be much larger amount of warming.
    And can solved by people moving out of cities.
    Or cities spending billions of dollars to lessen the effects- how they could do this [and whether tax payers want to do it] would vary.

    1. And can solved by people moving out of cities.

      YMMV, but I solved it by running my AC longer and staying indoors.

Comments are closed.