Is betting on the future climate a good way to make money? That clearly depends upon the bet, and how much it is. Recently there was a bet proposed in The Durango Herald, A chilly response: Durangoan willing to bet $5,000 Earth won’t be hotter in 10 years. At Stoat, William Connolley has a few bets going on sea ice; I wonder if he’d be interested in this one.

Should he, or you, take the bet? (I have no money, so I’m right out.)

Let’s have a looksee at the data. He wants to use ‘authoritative “HadCRUT3? data set’. Fine, here’s the HadCRUT3 global data set. The first, and easiest thing we can do with the data is to just plot it versus time. The results are not exciting.

This looks like every other temperature plot I’ve ever seen. The temperature is getting warmer. That’s the hockey stick thingy at the right side of the plot. He wants to bet on temperatures 10 years from now. Okay, let’s look at the 9-year linear trends. The folks at RealClimate did a similar thing recently.

The solid black line shows the 9-year temperature trend (C per decade) as a function of the middle year. The heavy, dashed blue line represents no trend. The light, dashed black line represents the trend of the 9-year temperature trend. A couple things are obvious. Since 1980, the 9-year trend has always been positive. Most years in this record have had a positive trend. In fact, only 55 of the 148 years have a trend that is less than zero.

Gerald Baumann, who has a doctorate in engineering, raised a couple of caveats for anyone considering taking up Cohen’s challenge.

“Global temperatures from 1920 to 2007 have increased relentlessly and indisputably,” Baumann said. “But the randomness of year-to-year temperature change actually results in some years cooler than their 10-year predecessor, even in strong trends. If Cohen wanted a serious bet, he would have offered comparing the next 10-year average to the prior 10 years.”

While that’s an interesting notion, we actually have the data. So we can look at the data. Let’s plot the the temperature in one year and subtract the temperature from 10 years ago.

It is a little noisier, but the plot looks very similar to the one directly above; as it should. Again, the plot shows that the temperatures have been rising. And since 1980, there has been only one year that has been lower in temperature than 10 years prior. For this set, there are only 46 years out of the possible 148 where the temperature has been lower 10 years after the initial temperature.

If you thought that the current observed temperature trends are not caused by humans, you should still take this bet. Of course, you could always take that $5000 and invest it for 10 years. I have no idea how much money you’d make (see above regarding me being broke). Given that for each of the two “trend” analyses done above about 2/3 of the years since the beginning of the record have positive “trends”, and if I thought that there was no anthropogenic signal in the data, and if I thought that the temperature signal was “white“, then I would set the odds at 2:1 against the proposer of the bet.

Of course the temperature record isn’t “white”, it’s “red”. The temperature this year is related to the temperature last year. This should favor anyone taking the bet. And a vast majority of scientists think that most of the warming is caused by humans, which increases your odds even more. Therefore, I would realistically expect the odds to be about 15:1. This is solely due to the natural fluctuations called weather.

Should you take this bet? I don’t think so. It’s just a publicity stunt by an oil company. He’s retired, but formerly the manager of strategic planning for Exxon Mobil. Of course there’s no connection.

(No Ratings Yet)

 Loading ...

Related Posts:

Trackback URI |