The catastrophic tsunami that killed more than 2000 people in Palu, Indonesia, in September 2018 occurred as a culmination of unusual seismic events, report two papers published in the journal Nature Geoscience.
The event started with a 7.5 magnitude earthquake that shook the island of Sulawesi, starting around 80 kilometres from Palu, the provincial capital. An unusual feature of the earthquake was its supershear speed.
Earthquakes are caused when rocks either side of a tectonic fault – a crack in the Earth’s crust – move suddenly in opposite directions, like fingers snapping, explains Jean-Paul Ampuero, lead scientist from the Université Côte d’Azur in France and senior author of the first paper. The shift starts in one part of the fault, then tears along it like a zipper opening.
The rupture generates two types of waves – S-waves, or shear waves, which travel at about 3.5 kilometres per second, and P-waves, which travel faster, at around five kilometres per second. Most earthquakes tear open at slower speeds than the shear wave velocity.
Ruptures that travel faster than S-waves are known as “supershear” earthquakes. According to Ampuero, these rare phenomena create intense shaking, comparable to the sonic boom of a supersonic airplane.
Both papers established that the Palu earthquake travelled at a speed of 4.1 kilometres per second along the Palu-Koro fault, its impact magnified by a devastating succession of events comprising soil liquefaction, landslides and the tsunami that hit the capital and nearby settlements.{%recommended 4268%}
The earthquake occurred on a strike-slip fault, where two blocks slide horizontally past each other.
From a “classical” perspective, Anne Socquet, professor at the Université Grenoble Alpes in France and senior author of the second paper, says strike-slip faults are not expected to cause a major tsunami.
“This earthquake showed that this is clearly a crude underestimation of the tsunami hazard,” she adds.
Previously observed supershear earthquakes travel along faults that are remarkably straight, like “geological superhighways that present little obstacle to speeding earthquakes,” Ampuero says.
What was surprising, he adds, is that optical and radar images recorded by satellite showed the speed of the Indonesian earthquake was maintained along the length of the 150 kilometres rupture despite at least two sharp bends in the fault.
Socquet’s team analysed the satellite imagery to work out how much slip occurred on the fault plane both at the surface and at between 15 and 20 kilometres depth.
Earlier studies have found less slip on the surface. “One explanation is that immature faults feature more structural complexities and damage around the fault, which may accommodate some of this missing slip,” she says.
Contrary to those studies, Socquet’s team found maximum slip near the surface of the Palu fault. This suggests “limited deformation”, she explains, proposing that is due to structural smoothing that occurred as the fault matured.
“For earthquakes to rupture at supershear velocities, the fault must be free from structural or geometric complexities, which will slow the rupture down.”
She adds: “Our results are of great significance, as this earthquake represents one of the only instances where both the shallow and deeper slip on the fault plane can be very well-resolved for a mature continental plate-boundary strike-slip fault.”
The studies add new insights that can be added to earthquake hazard predictions.
“Earthquake speed affects the shaking intensity, but is not accounted for in conventional hazard assessments,” explains Ampuero, “because we don’t know exactly why sometimes (rarely) earthquakes run very fast.
“If we could identify the factors that determine the speed of earthquakes, we could estimate better the shaking intensity. Our work suggests one factor worth exploring: a possible relation between the age of the fault and rupture speed. It also gives a counter-example to the idea that fast earthquakes happen only on straight faults.”
