By Mike (originally posted 22nd May 2012)

I picked up a copy of the New Scientist a couple of days ago to shorten the bus ride to Sydney, and one particular article caught my eye, an interesting piece of recent work by some Japanese geophysicists relating to earthquake prediction. The recent post by Evan about the tragic Italian quake reminds us of a sadly acceptable truth; that even in Europe with its high concentration of science and scientists, we cannot predict earthquakes. We can identify regions in which they are likely (like near tectonic plate boundaries or large fault lines) and we can assess the likelihood of a particular sized event occurring in a given time period (say, I predict one magnitude 7 event will occur in region x over 100 years). However, knowing exactly when and where a quake will occur is still about as accurate as my dart throwing ability after 4 pints of beer, or my pool-playing ability before the same amount (i.e. not very accurate).

However, this new work suggests that the tide (which can be viewed in this case as a fluctuating mass of water on top of the earth’s crust) causes small earthquakes in sections of the ground that are ready to rupture. The frequency of these little tremors increases before a big earthquake, so theoretically by modelling and then monitoring the small quakes, the big one could be predicted.

Unfortunately, all research in this field has been done retrospectively, after the big earthquakes themselves. Lots of data is needed to amass enough information to begin to predict quakes, and unfortunately this means lots of earthquakes will have to happen first before the data set reaches its critical mass of usefulness.

Whilst I hope that this method eventually yields useful fruit and can start to predict quakes, for now we will have to keep relying on barking dogs, snakes emerging from hibernation and odd toad behaviour. Science is fun.