3 May 2022–Including topography—the hills, cliffs and valleys of a landscape—in ground motion models shows where shaking might be most amplified during an earthquake, researchers demonstrate with detailed new models of the Puget Sound region.
In their models, Ian Stone of the U.S. Geological Survey and colleagues simulated several magnitude 6.5 to 7.0 ruptures on the Seattle Fault, which would affect the city of Seattle and most of its surrounding metropolitan area. Their study in the Bulletin of the Seismological Society of America examined how the region’s detailed topography, from steep slopes along Puget Sound to the 681-meter tall Squak Mountain, influenced ground shaking from the simulated quakes.
The average ground motions calculated in the simulations were similar with and without considering topography, they concluded. But they confirmed that shaking is typically amplified at topographic high points like hilltops and cliffs, and de-amplified at lows such as channels and valley bottoms.
This trend holds across even small features that are less than 500 meters in width, Stone and colleagues found. Amplification patterns can also overlap on the landscape, leading to more complicated ground motion responses at these features.
The researchers were surprised to see that their models only broadly predicted how topography would affect ground motion, and that “the response on any given feature was quite variable,” Stone said. The topographic amplification also differed depending on variables such as local focusing and scattering effects on seismic waves, along with details about the earthquake rupture itself.
For instance, “we did find that in our models, some of the steep coastal bluffs in the region reliably scattered and trapped seismic energy during earthquake shaking, which caused amplification at some frequencies and de-amplification at others,” Stone explained.
“Part of this variability may be related to the relatively amorphous topography in our model region, the central Puget Sound area of Washington State. There aren’t as many sharp cliffs or well-behaved triangular mountains like you see in some of the fundamental studies on topographic effects,” he said. “But in my opinion, that makes these results a good indication of what we may expect during a real-life earthquake in a typical, non-mountainous region.”
The research team simulated 20 earthquake scenarios along the Seattle Fault, varying the slip distributions and hypocenter locations, in models with realistic surface topography for the region.
Until recently, seismologists have had difficulties modeling fine-scale topographic effects on ground shaking. Stone said lower costs of supercomputing, along with improved seismic wave propagation models and higher resolution mapping of the crust to improve 3-D seismic wave velocity models, have now made it feasible to consider regional topographic effects.
“We can now model large finite fault earthquake ruptures while also considering fine-scale features that affect high-frequency shaking,” Stone said.
The consistent effects of topography suggest that it should be considered to some degree in seismic hazard studies, the researchers concluded.
“Paired with a method that estimates the topographic curvature across a particular region, you could potentially begin to estimate the likelihood and severity of topographic amplification at a given location,” Stone said. “Though more testing in areas with different topographies would need to be done first, likely paired with more real-world measurements of topographic amplification during earthquakes.”