Abstract

From Zhang Heng’s earliest seismoscope, to Dubliner Robert Mallet’s experiments on Killiney Beach and observations of the 1857 Neapolitan earthquake, scientists have tried to understand earthquakes and use this understanding to reduce their effect. Still, earthquakes have killed almost a million people already this century and the goal of earthquake prediction is now probably farther away than ever.

In this lecture, I will use butterflies and sand piles to demonstrate the fundamental unpredictability of seismicity and show that the tiniest stresses – the force of a gentle handshake – can trigger the biggest and most destructive earthquakes, leaving standard deterministic prediction techniques powerless to make progress.

This is not the end of the story. On the one hand, populations of earthquakes have been found to obey very robust statistical laws permitting forecasts of the probability of an event of a particular size happening in a particular period in a given region. Not prediction, but earthquake-weather forecasting that can inform our assessment of seismic hazard and help us manage the response to destructive earthquakes that is always hampered by aftershocks.  Recent research  has also produced new insights to the mechanics of earthquakes and how they interact with each other and  with the Earth’s near surface, opening up new avenues for the identifying times and places of particular danger.

With 60% of the urban space that will exist in 2030 yet to be built, rapid urbanization, normally seen only as a threat, actually presents a time-limited global opportunity to plan disaster risk out of urban development.  I will conclude with an optimistic look forward.