New ground motion simulations inform seismic hazard estimates for Cascadia megathrust earthquakes

New 3D simulations of ground motion for a hypothetical set of 36 full-margin Cascadia megathrust earthquakes can help inform seismic hazard estimates for the region, according to a study published in the Bulletin of the Seismological Society of America.

The simulations expand on previous ground motion scenarios studied for possible megathrust events in the region, by including a variety of earthquake magnitudes and newly mapped offshore faults using recent high-resolution seismic data, among other new details.

The study was funded as part of the Cascadia Coastlines and Peoples Hazards Hub (CoPes Hub), a U.S. National Science Foundation hub that aims to improve the adaptive capacity and resilience of Pacific Northwest coastal communities to natural hazard risks.

“The Hub brought together researchers across many disciplines, so we were able to incorporate a lot of expertise into these simulations,” said Audrey Dunham of the U.S. Geological Survey and lead author of the BSSA paper.

Paleoseismic data suggest that an earthquake along the full margin of the Cascadia subduction zone may occur roughly every 500 years, with a 10-15% probability that one of these massive events may occur within the next 50 years.

Dunham and her colleagues modeled ground motion for 36 possible full-margin events. For the first time, they included a variety of earthquake magnitudes in their simulations, finding that on average, larger magnitudes generate stronger shaking, as expected.

However, different magnitudes did not appear to increase basin amplification—the increase in the strength of shaking due to low-velocity sediments—for the Seattle Basin, the researchers found.

“We wanted to know if we should expect a magnitude 8.7 to amplify the Seattle Basin in the same way as a magnitude 9.1,” Dunham said. “The finding that basin amplification is magnitude independent for these full-margin megathrust earthquakes is important for engineering applications.”

Their simulations suggest that the amplification factors currently used in the National Seismic Hazard Model for the Seattle Basin, which guide building codes, should continue to be taken into account and do not need to be adjusted based on earthquake magnitude, she noted.

There are also sedimentary basins around the Portland area that could amplify ground shaking, but less is known about their characteristics, Dunham said. “Getting better constraints on the basins in this region would help to improve our estimates of ground shaking and seismic hazard there.”

The researchers quantified how the location of an earthquake’s hypocenter affects regional inland shaking. Offshore hypocenters (located in the shallow, seaward boundary of the subduction zone typically called the updip) generated on average 45% higher shaking at periods longer than 6 seconds compared to inland hypocenters (the deeper, landward boundary of the subduction zone called the downdip), due to the direction of the predicted ruptures, they found.

Estimates of long period shaking from the simulated earthquakes were larger than estimates derived from current empirical ground motion models or GMMs, Dunham noted.

“This could have to do with the distribution of large earthquakes used to constrain the GMMs, that they primarily have more downdip hypocenters, and our dataset is composed of an even distribution of updip and downdip hypocenters,” she said.

Dunham and her colleagues also included data on regional offshore splay faulting, where secondary faults branch off a primary fault, that had been mapped recently using new high-resolution seismic reflection data collected by the CASIE21 seismic survey.

“These types of faults have never been included in estimates or simulations of ground shaking from megathrust earthquakes, so we were really interested in understanding if they impacted results in the Pacific Northwest,” Dunham explained. “It turns out that because of how far offshore these were, they didn’t cause any changes to inland ground motions when we did or did not include them in a simulation.”

Splay faulting can impact tsunami hazard, however. “The reason these faults are important is that they can change how the seafloor deforms—moves up or down—during an earthquake, compared to a rupture just on the megathrust, which can change tsunami generation in deep water and therefore potentially how that tsunami will inundate along the coast,” Dunham said.

The researchers are now working with tsunami modelers at the University of Washington and the National Ocean and Atmospheric Administration (NOAA) to include their new findings “to understand the differences in hazard between scenarios with and without these splay faults,” Dunham added.