Seismic hazard models attempt to predict the long term risk of earthquakes by looking at, among other things, the slowly changing stress that builds on faults.
However, there is new evidence to suggest that megathrusts and other very large earthquakes fundamentally alter the nature of the hazard.
It seems that these large events actually slow down seismic activity at the earthquake’s epicentre, while turbo-charging it in surrounding areas.
A new study by Temblor (a catastrophe risk modelling company with a focus on earthquakes) and Tohoku University in Japan suggests that, following a megathrust earthquake, aftershocks on the rupture surface quickly shut down. Meanwhile, large aftershocks light up the surrounding area, putting coastal population centres at risk.
The researchers looked at earthquakes since 1960 with a magnitude of nine or above. They found that after each event, seismicity rates dropped well below the rate observed in the decades before the megathrust.
Immediately after each of these events, a surrounding “corona” of seismicity was activated, and continues to be active to this day. The mechanism by which this happens, whereby stress is transferred from the rupture to surrounding faults, is known as the Coulomb stress transfer.
This has obvious implications for reinsurance companies that deal with natural catastrophes. If risk models fail to take into account the way the picture is altered by seismic events, they can’t hope to accurately gauge the level of risk they provide cover against.
Most current hazard models assume that the risk of further seismic activity drops after a megathrust. The new research suggests that there is a net hazard increase, lasting about half a century, spread over an area up to ten times larger than the initial core.
In support of this interpretation, the researchers use the 2011 Tohoku earthquake as an example of the phenomenon. In the decade prior to the eruption there were four or five shocks of magnitude 6.7 and greater. In the decade since there have been 22 such events. They say their research exhibits a forecast skill that could be used to improve hazard assessments.
