A new step towards earthquake forecasting?
The most difficult task facing an Earth scientist lies in advising upon the likelihood of natural hazards occurring within a definite time frame so that the difficult and expensive decision of whether and when to evacuate the populations at risk. Call it wrong in either direction you get an expensive and useless evacuation or a casualty filled disaster, and 7 Italian seismologists were actually sentenced to 7 years inside for a wrong call over the Aquila quake some years ago. No wonder that anything that helps improve accuracy is much sought after, since many lives are potentially at stake in many danger zones around the world such as Turkey and California.
The problem is that its bloody hard. One can measure the stress the rocks are under, estimate the tension, but telling an oncoming major rupture of devastating proportions apart from a bit of inconsequential slipping and sliding along the fault is a really tricky business. There are no clear symptoms as exist for volcanic eruptions (where the accuracy of prediction is reasonably good, except in some impoverished parts of the world), and the methods that worked once don’t seem easily applicable to different geological contexts. Previous attempts have explored such ideas as radon gas leaks, heat maps and animal behaviour. So far the best we can do is estimate x risk of a y magnitude quake sometime during the next z years.
A new lead may be about to improve matters…Iceland is a volcanic and quakey place, sitting on the Mid Atlantic Ridge where new crust is being created as the North American and European plates split apart, and recent research there has shown that changes in groundwater chemistry may help predict a larger rupture months in advance. Scientists sampled groundwater weekly over five years analysing the samples and found big spikes in the months before two major quakes in 2012 and 2013 over 80km away from the sampled area. The changes occurred as crustal stress built up to breaking point allowing aquifers to mix as they start to fracture.
Similar changes occurred before the 1995 Kobe quake that killed 6,500 people, but a single event does not build up sufficient data to justify further expensive study, so the fact that two quakes elsewhere were preceded by groundwater changes provides us with a route for further exploration that may lead to more accurate forecasts. Much further study is needed both in Iceland and around the globe, particularly to see if this technique can be used in different geological environments, but it is the best lead towards improving earthquake forecasting we have so far.
Image credit: Iurie Belegurschi
Original paper, paywall access: http://www.nature.com/ngeo/journal/v7/n10/full/ngeo2250.html
Reblogged from earthstory