Shaping policy on earthquake risk estimation and forecasting

Research on earthquake risk estimation and forecasting has directly impacted the policy recommendations of the International Commission on Earthquake Forecasting and, in turn, the policies of governments in numerous countries.

Photo of an earthquake-struck building in L’Aquila, Italy
An earthquake devastated L’Aquila in Italy in 2009

Can science predict when and where an earthquake will occur? Given the enormous impact earthquakes can have, the ability to predict an earthquake (i.e. state that an earthquake of a certain magnitude will occur in a specific region over a certain period) would be of great value and save many lives.

The L’Aquila earthquake in Italy in 2009 killed 309 people, injured 1,500 and displaced 65,000. The breadth of devastation, paired with the lack of short-term warning from government and scientists, spurred questions about the role of science in civilian protection.

Headlines asked, ‘Should scientists stop giving advice?’ and six members of the Italian Grand Risks Commission were convicted of manslaughter as a result of their perceived poor communication of the risks involved, although they were later acquitted on appeal.

The Director of Civil Protection in Italy appointed Professor Ian Main (School of GeoSciences) as the sole UK member of the International Commission on Earthquake Forecasting (ICEF), based on Professor Main’s record of research into multiple aspects of earthquake predictability.

Underpinning research

Professor Main’s research, using both statistical and rock physics approaches, has illuminated the difficulties of earthquake prediction. He found that global earthquake data can identify increases in the probability of earthquakes, whereas deterministic predictive models for individual events, sufficient to justify a general evacuation, are unreliable.

Emphasising the uncertainties involved in estimating recurrence rates, he also found that models for seismic hazard can easily be altered through the inclusion of single extreme events (for example the earthquake off the coast of Sumatra that led to the devastating tsunami on Boxing Day 2004).

Professor Main’s rock physics studies included the development of a model for the quasi-static nucleation of unstable fracturing and slip. The resulting highly nonlinear behaviour within the model provides a rationale for the practical difficulty of accurately forecasting the failure time in brittle failure events - even in controlled laboratory and engineering settings.

While reliable short-term prediction of individual earthquakes remains elusive, long-term probabilistic forecasting provides a useful baseline for developing building design codes, and the shorter term clustering properties provide an opportunity to calculate and communicate periods of heightened risk.

Research impact

Professor Main’s research conclusions directly impacted the policy recommendations of the final Reports and Recommendations of the ICEF.

The report stated: ‘Any information about the future occurrence of earthquakes contains large uncertainties and, therefore, can only be evaluated and provided in terms of probabilities’.

The report emphasised the lack of clear and reliable precursors needed for deterministic prediction, and recommended investment in ‘operational’ forecasting with clear communication of probability and uncertainty.

The ICEF report stimulated policy innovation across the world. The Italian Department of Civil Protection committed €1 billion to a 10-year research project on operational earthquake forecasting and implemented a public education programme to better communicate probability and risk.

Authorities in the US and New Zealand also took their cues from the ICEF report and initiated similar programmes. The report has also influenced policy development in Greece, Japan and Russia.

Another outcome of the L’Aquila earthquake and the legal proceedings against officials was media discourse about the role of scientists in providing advice or commenting on risk. Although the six Italian officials were eventually acquitted, it is clear that the communication of risk by scientists to the public and to governments is an area where all parties are in need of greater mutual understanding.

Relevant publications

Main (1996) ‘Statistical physics, seismogenesis, and seismic hazard’. Rev. Geophys. 34, pp 433-62

Main (1999) ‘Applicability of time-to-failure analysis to accelerated strain before earthquakes and volcanic eruptions’. Geophys. J. Int. 139, F1-6

Huc & Main (2003) ‘Anomalous stress diffusion in earthquake triggering: correlation length, time dependence and directionality’. Geophys. J. Res. 108 (B7), p 2324

Seher & Main (2004) ‘A statistical evaluation of a ‘stress-forecast’ earthquake’. Geophys. J. Int. 157, pp 187-93

Main, McCloskey & Naylor (2008) ‘Effect of the Sumatran mega-earthquake on the global magnitude cut-off and event rate’. Nature Geoscience 1, p 142

Jordan et al. (2011) ‘Operational earthquake forecasting’. Annals of Geophysics 54(4), pp 361-91

Download this case study

Download a printable version of this case study as a PDF.