Global Warming and Probability

Posted by Eric Bastholm on October 19, 2011
Oct 192011
Cyclone Yasi, global warming and probability
Cyclone Yasi approaching Queensland, Australia on 2 February 2011 (NASA; MODIS)

Whenever I read an article in the news about some extreme weather event like the occurrence of tropical cyclone Yasi in North Queensland in Australia the discussion inevitably turns to the question – was this storm a result of global warming? It’s the same for any type of unusual event; floods, storms, extreme cold, snow at the wrong time of year, extreme heat, bush fires, etc, etc.

The truth is that a lot of the complex modelling that the weather boffins do does point to there being an increase in these types of events over time. Not straight away. It is not right to blame the occurrence of any one event on a single cause. Especially, as in this case when the event is dependent on so many variables.

Humans are adept at pattern matching and connecting disparate information to form new facts, but sometimes we don’t do it well and probability theory is one of those areas that generally confuses us. Also, when something bad happens we like to blame something for it.

To say that cyclone Yasi was caused by global warming is like saying that a biased die is the cause of someone throwing three 6s in a row. It might be, but it is more likely that this is one of those times that a unlikely event has actually occurred just because the laws of probability allow it to. You have to look at events over a longer time to make any statements on their cause.

So, if you throw the die 1000 times and 6s come up way more than 1/6th of the time, then you might have something to say about the fairness of the die. The same for the weather. In time we will have enough history to say for sure if any increase in global temperatures has had an appreciable effect on weather events. But NOT YET. And even then, it still only makes sense to talk of trends in the weather. You can never use the data to explain a single event. The mathematics does not allow it.

So if we can’t blame cyclone Yasi on global warming should we be worried about it at all? The problem we have is that if we wait until we have the data it will be too late to do anything about it. The weather may have changed catastrophically and we won’t be able to change it back. This is why scientists try hard to get the best data they can so we can analyse the trends and extrapolate the effects of those trends. Given that the predicted effects of global warming are so bad, even in the most conservative scenario, and we know that lower temperatures than those predictions are harmless, it would seem prudent to err on the side of caution and try to keep our effect on the environment to a minimum. At least that gives us a state we are familiar with, rather than a new state which may be difficult to live with.

In any case, the real point of this article was to point out some bad reporting that you see in the press regarding probability and statements of cause and effect. It’s a subtle point. Yasi is possibly one tiny part of a long term observation in the trend of unusual weather events. Global warming did not cause Yasi. The science says that it is possible global warming increased the probability that that event would occur. A very different statement.

© 2011 phinetune Suffusion theme by Sayontan Sinha