The Northridge earthquake in 1994 killed 72 people and damaged elevated roadways, bridges, and residential and commercial structures. New research suggests that the background seismic fields created by ocean waves hitting the earth’s crust can help seismologists predict the nature and impact of future earthquakes in the Los Angeles region. FEMA News Photo/Robert A. Eplett
New research at Stanford University has capitalized upon the ambient seismic field—created by ocean waves as they strike the earth’s crust—to forecast potential earthquakes in the Los Angeles basin.
February 11, 2014— The city of Los Angeles is all too familiar with earthquakes. Although minor earthquakes are a near daily occurrence in Southern California, the last major earthquake in the region, in 1994, was devastating. The Northridge earthquake measured a magnitude 6.7, killing 60 people and injuring 7,000, collapsing infrastructure, and damaging more than 40,000 buildings in the Los Angeles area. Preparing the city for the next major earthquake before it arrives is thus of critical importance to the city.
So research conducted by Stanford University that has resulted in a new method of forecasting the potential ground motion in the region before a major seismic event is making news. The method capitalizes on the ambient seismic field created by ocean waves as they hit the earth’s crust, and was used to validate, for the first time, the ground motion simulations conducted by the Southern California Earthquake Center in 2006 and the U.S. Geological Survey (USGS) in 2008. The “virtual earthquakes” created by the Stanford study confirm the computer models’ projections that greater shaking would be experienced in the sedimentary basin of Los Angeles than in other nearby areas.
The approach was conceptualized by Greg Beroza, Ph.D., a professor of earth sciences in Stanford University’s department of geophysics and the deputy director of the Southern California Earthquake Center (SCEC), and implemented by Marine Denolle, Ph.D., currently a postdoctoral scholar at Scripps Institution of Oceanography at the University of California, San Diego, who conceptualized and implemented the field work as part of her doctoral work at Stanford. Both Beroza and Denolle wrote in response to written questions posed by Civil Engineering online.
Denolle’s work was also overseen by Eric Dunham, Ph.D., an assistant professor of geophysics at Stanford, and Germán A. Prieto, Ph.D., an assistant professor of geophysics at the Massachusetts Institute of Technology.
The ocean waves the team studied create small pressure pulses that are constantly picked up by seismometers. By measuring the time it takes these pulses to move between sensors, the underlying soil conditions and their response to ground motion can be mapped. Denolle’s examination of the Los Angeles basin represents the first time that these pulses have been used to forecast potential ground motion during earthquakes. She placed temporary sensors along the San Andreas Fault at a depth of 1 to 2 ft and coupled them with existing sensors to track the movement of the pulses as they moved from the fault through the basin. Denolle then processed the signals and “developed corrections for depth, sense of slip in the earthquake, and finiteness of the fault,” Beroza explained.
The research produced an effective new approach for measuring, understanding, and anticipating the effects of geology on seismic waves: “Rather than waiting passively for an earthquake to occur that will test our predictions, we can move actively to deploy instruments in areas of concern to learn what we want to know before an earthquake happens,” Beroza said.
Robert Graves, Ph.D., the Southern California coordinator for the USGS’s Earthquake Hazards Program, sees the potential of the method to work alongside computer simulations. In written responses to questions posed by Civil Engineering online, Graves said, “The benefit of this new approach is that it maps out the potential ground shaking levels for future earthquakes using information based solely on existing ground motion observations.” Computer simulations, on the other hand, can be “incomplete or uncertain in various regions, and thus the predictions of ground shaking based on the simulations also will have uncertainties,” Graves said.
Using the ambient seismic field to map potential ground motion “potentially reduces the level of uncertainty with the ground motion predictions,” Graves said. However, both computer simulation and virtual earthquake analysis, as was done by Denolle, are both needed. “I view the virtual earthquake approach and the simulation approach as being complementary approaches to the ground motion prediction problem,” Graves said.
This proof-of-concept study is just the first step, and was limited to long-period seismic waves because the temporary sensors were placed for short periods of time just below the ground’s surface, Denolle pointed out. “Only tall buildings, and large infrastructures (bridges, for instance), are sensitive to those frequencies,” she said.
Still, the practical implications of Denolle’s findings are immediate. Los Angeles Mayor Eric Garcetti recently announced that the city was partnering with the USGS to develop earthquake resilience strategies for the city. As part of this partnership, Lucile Jones, Ph.D., a seismologist with the USGS, will be working as the science advisor to the mayor for seismic safety. She will spend the coming 12 months applying the lessons learned from the region’s annual earthquake preparedness drills, called ShakeOuts, to develop recommendations for reducing seismic risks in the Los Angeles region, focusing particularly on four major areas: communications systems, water infrastructure, nonductile reinforced concrete buildings, and buildings with “soft” first stories.
California’s “crust is young and hot and broken up with faults, so it’s a poor transmitter of energy,” says Jones. As a result, ground motion “dies off really quite quickly, but then when you have these local soil conditions it can amplify the shaking,” she says. As waves travel through the ground, Jones explains, certain ground conditions—such as the chain of sedimentary basins that connect the Los Angeles basin to the San Andreas Fault—can act as an “energy channel,” focusing the energy in a certain direction. Before this study, the theory was that the band of thick sediments in front of the San Gabriel Mountains would act as a wave guide, directing seismic waves to the Los Angeles basin, where they would bounce back and forth within the basin for a longer period of time than what is experienced at the epicenter. This new study has confirmed that the ground motion behaves in precisely that manner.
The research conducted at Stanford “really changes the answer of what the maximum credible earthquake shaking will be [within the basin],” Jones says. “Especially in the long-period motion, which will impact the big buildings.” The study confirms that previous computer modeling of Los Angeles’s ground motion, which was criticized as showing too much motion, “is the appropriate one to work from,” she notes. “So it’s greats to have, especially at the beginning of the project.”
Structural engineers are tasked with designing the strongest buildings possible while being conscious of their client’s budgets, Jones points out, so knowing the potential ground conditions during a maximum credible earthquake is crucial, determining the seismic strength required of a building.
Which is why the existence of a “wave guide” in the Los Angeles basin is of such paramount importance, Jones says. “If we said, ‘Okay, it’s only going to be the smallest shaking that you’d expect from an ordinary location,’ and that’s how you build the building, and then we get these extra strong motions, [because the] wave guide causes specific amplification, you’re talking about the potential collapse of high-rises,” Jones says. “Especially because the biggest issue is these really long-period ground motions that are strongly affecting the tallest buildings.”
Measuring shorter wave frequencies, which affect smaller buildings, using the ambient seismic field is also possible, but the sensors would need to be deployed for longer than 5 months, Denolle said. Additionally, high-quality instruments will need to be deployed in vault-like environments and attached to concrete bases to ensure their ability to measure small ground motions. “We would need borehole instruments: [such as] when people drill through the ground, maybe 100 meters down, to deploy similar instruments, much like the Japan High Sensitivity Seismograph Network,” Denolle said.
The usefulness of ambient noise mapping and virtual earthquake forecasting is not limited to sedimentary basins. “It is worth mapping anywhere people are affected, or anywhere the knowledge we would learn from that area would be useful to our global understanding,” Denolle said. “Sedimentary basins exacerbate the strength of the shaking, but being close to a fault stays a first order effect. Anywhere with rockier areas would also work.”
Denolle is currently working with colleagues in Tokyo to similarly map potential earthquake ground motions in Japan. “The Alpine Fault in New Zealand could be the next target as [the city of] Christchurch also sits on a sedimentary basin and could feel the large Alpine Fault future earthquake,” Denolle said. “[The] possibilities are endless.”
The findings from the study were first published in the January 14, 2014 issue of Science.