Highly Allochthonous

Is a pet toad the new must-have earthquake detector? This paper, which is getting a lot of media attention today, claims that in the 5 days before the L'Aquila earthquake in April last year, toads which should have been massing to engage in their annual mating instead left their ponds and ran for the hills. The authors speculate that this exodus was linked to very low frequency electromagnetic perturbations in the atmosphere which happened around the same time.

I'm not going to go into detail on this, because Michael Reilly has produced an excellent write-up already, even interviewing Susan Hough, who has written a book on the science of earthquake prediction that is high on my to-read list. Dave Petley also has some interesting thoughts. Given my previous thoughts on claimed earthquake precursors, it should come as no surprise that I share their scepticism about all of this. One data point does not a robust detection system make.

Not a seismologist.

Of course, it's no surprise that this story has legs: earthquake detection is extremely topical in the wake of two large earthquakes in the last three months, and being able to give it the "serendipitous discovery" spin (study toad breeding, publish on earthquakes!) just adds to the appeal. However, this study suffers from the usual problems with claimed earthquake precursors: is this 'precursor' replicated in other places, before other earthquakes? More importantly, is it only associated with earthquakes? Is there a consistent correlation between the appearance of the precursor, and the size and/or the timing of the earthquake that follows? These are the questions that need to be answered before you can move from anecdata to something that is actually useful. They are also questions that have yet to be definitively answered for any claimed percursor, be it animal, electrical, or gaseous. It's not that the various types of signals that have been observed before large earthquakes have no connection to what's going on beneath our feet - it's entirely plausible that in some cases they do - but I suspect that without some new theoretical insights into what that connection might be, we're never going to be able to interpret with confidence what these signals mean before it's too late to matter.