Risky things

By Marcus Wilson 21/09/2009

I love the headline at the end of last week ‘Wellington quake risk halves’. As if you can wake up one morning and find that the chances of an earthquake happening today are suddenly half of what they were yesterday just because someone says so.  What next – someone decreeing that summer will last 12 months of a year and so magically it does?

Of course, what is really meant (and to be fair the second paragraph of the article on stuff.co.nz gets it right) -our best estimate of the risk ‘of the big one’ is now half what was previously thought.

To be honest, I don’t know a lot about predicting earthquakes, but what I do suspect is that a major problem the scientists have is that there isn’t a great deal of data to go on. Really big earthquakes don’t happen all that often in any given place. Hence the difficulty in knowing what the risk really is.

There is a whole branch of physics devoted to things that occur randomly. Some random processes are rather easier to work with that others. For example, radioactive decay. Now, if I have an atom of (say) Caesium 137, I don’t know when it will decay, but I can be very specific about its chances. I can say that the chance of it still being a Cs 137 atom in 30.1 years time is exactly half. Why? Because people have measured the rate at which decay events happen to Cs 137 atoms. There are a lot of atoms around (contrast the case with earthquakes) which makes it easy to construct very accurate predictions. It doesn’t help me predicting the exact day that my atom will decay on, but what I can say is that if I had a million atoms, and waited 30.1 years (the half-life of Cs 137) I would have pretty well half a million left.

Radioactive decay is an easy example because the underlying process is what we describe as ‘Poisson’ – which leads to a very simple mathematical description. Other random processes are not always so nice. A good deal of research in this area is driven by insurance companies, and for good reason – they are the ones that need to properly assess and quantify the risk of something, because they have to put money on it. The Earthquake Commission will be very interested in the Wellington fault-line study, because it may well affect how much insurance levy they feel the need to collect from us NZ residents.

As I said, I’m not an earthquake expert, but I can hazard a guess that a big earthquake in Wellington isn’t going to be great for the economy, whether you live in Wellington or Kaitaia.