‘Synthetic’ quakes to help predict big ones
Kiwi scientists have created a millionyear catalogue of “synthetic” earthquakes that could help answer just the kind of big questions now swirling around Friday’s shaky morning.
It may also help tell us what kicked off the 2016 Kaikoura earthquake — one of the most complex ever recorded — and how certain elements of New Zealand’s landscape can worsen shaking effects. In the past decade alone, our country has experienced more than 226,000 quakes large enough to be felt — including half a dozen that topped a magnitude of 7.0.
All were detected by GeoNet’s impressively dense network of seismometers, providing a constant feed of activity in real-time. Yet, despite some fascinating advances around the world over recent times, there’s still no way to predict precisely when or how quakes will happen in the future. While GNS Science does give model-based aftershock forecasts — the probability of an event the size of Friday’s 8.1 tsunami-making Raoul Island jolt over the next 30 days is about 15 per cent, and less than 1 per cent for one 8.5 or higher — New Zealand’s next “big one” will strike as a shock. Looking at how past quakes might set up others to come has proven a particularly difficult puzzle to researchers, who only have a narrow number of events, over a short geological timeframe, to calculate seismic hazard from. “The full cycle of New Zealand’s earthquake activity spans thousands of years,” explained Dr Bill Fry, a GNS scientist and a co-leader of a major new programme.
“Modern observations and data only cover the last few decades, and this combined with the incompleteness of geological information means we can’t capture the entire range of possible natural earthquakes, especially the largest, most devastating events.”
Fortunately, there’s much that earthquake physics alone can tell us about future risk.
We know that earthquakes kick off when stresses acting on a fault became larger than the strength of the fault itself. It was likely these bending stresses within the Pacific tectonic plate, as it subducted beneath the Australian plate, that triggered Friday’s first two 7.3 and 7.4 quakes. It was by simulating these bending forces that Fry and his Canterbury University colleague Professor Andy Nicol were calculating “synthetic” quakes, and exploring how these might redistribute stress release into nearby faults. “Our new models benefit from the existing statistical and geological information but go further by highlighting a more complete range of possible natural earthquakes.”
The team have already used their new catalogue to test the effectiveness of the recently-deployed New Zealand DART (Deepocean Assessment and Reporting of Tsunami) buoy array. The buoys, which proved crucial in monitoring Friday’s tsunami scare, are part of the tsunami early warning system for big earthquakes that strike off the country’s eastern and northern coasts. “We’ve also started work looking at how major earthquakes might change our rivers and groundwater systems.”
The scientists are also using the catalogue to better understand what happens after large quakes. “The big prize is to use the catalogue to underpin future earthquake and tsunami hazard models, but there are less obvious opportunities too,” he said.
“For example, we plan to use it to better understand the way New Zealand’s landscape intensifies earthquake shaking, causing landslides and damage to buildings. Another key application is to improve our understanding of the way one earthquake triggers another.”
A case in point was the 7.8 Kaikoura quake, which set off more than 20 faults in a seismic cascade that spread across the country. Fourteen of those faults ruptured violently enough to displace land by more than a metre. One of the most dramatic examples was along the Kekerengu Fault in Marlborough, where the land offset was as much as 12m, and in some places created walls of raised-up earth.
The study — supported by a $4.2m grant through Resilience to Nature’s Challenges — was a world-first in many respects, and what was learned here could open up insights elsewhere. “Every time we get a big earthquake, we learn something new,” Fry said. “Our ultimate goal is to ensure this new knowledge helps us be much better prepared for the next big, complex earthquake.”
Fry and Nicol are presenting the model in a webinar tomorrow.
Every time we get a big earthquake, we learn something new.
Dr Bill Fry, GNS scientist