On March 28, 2006, scientists unveiled results from the most comprehensive study to date of how hard and how long the ground shook in the 1906 San Francisco earthquake.
Over a 2-year period, scientists from the U.S. Geological Survey (USGS), Stanford University, Lawrence Livermore National Laboratory, URS Corp., and the University of California, Berkeley, simulated ground shaking caused by the 1906 earthquake, using a new three-dimensional geologic model of the San Francisco Bay area. The simulations, which can be viewed on the Earthquake Hazards Program's 1906 Ground Motion Simulations page, have great potential for research, hazard-loss estimation, and public education.
The simulations show how the 1906 earthquake spread from its offshore epicenter, about 2 miles west of the San Francisco Zoo, and grew to cause strong shaking and damage along more than 300 miles of the San Andreas Fault. Understanding the huge geographic extent of the earthquake's impact is important for the San Francisco Bay region's citizens and decisionmakers.
"We want to emphasize that a large earthquake such as 1906 is not just a San Francisco quake but a northern California earthquake," said Brad Aagaard, the USGS geophysicist who led the effort. "Earthquakes greater than magnitude 7 are going to cause intense shaking over a large areaeveryone needs to be prepared."
Multiple studies that use the new simulations to assess the likely impact of such a quake occurring today were among the scientific results presented at the 100th Anniversary Earthquake Conference, held April 18-22, 2006. The conference was the science, engineering, and policy centerpiece of a suite of activities commemorating the centennial anniversary of the 1906 San Francisco earthquake (for a complete list of activities, visit the 1906 Earthquake Centennial Alliance Calandar of Events).
The new simulations use a three-dimensional geologic and seismic-velocity model, released last October by the USGS (see the 3D Geologic and Seismic Velocity Model of the San Francisco Bay Region), as well as a model for how the fault ruptured in 1906.
"For the new fault-rupture model, we re-analyzed century-old surveying and seismographic data," said Seok Goo Song, a geophysics graduate student at Stanford University. "Our model confirms that the rupture was about 300 miles long with a magnitude of 7.8 to 7.9. Our work gives new insights into the strength of strong ground shaking in large earthquakes."
The simulated ground motions were compared with ground motions reported in a famous report on the 1906 earthquake written by Andrew C. Lawson and published by the Carnegie Institution of Washington in 1908 (Publication 87, v. 1, reprinted 1969), which USGS scientists recently mined to create detailed maps of ground-shaking intensities in that earthquake (see Modified Mercalli Intensity Maps for the 1906 San Francisco Earthquake Plotted in ShakeMap Format).
"These simulations give us a much more realistic and detailed picture of the strong shaking levels used for hazard-loss estimation and disaster-preparedness exercises and longer-term planning," said Mary Lou Zoback, a USGS seismologist and co-coordinator of the Earthquake Hazards Program in northern California. "These easily understood simulations should increase public awareness of our hazard."
The scientists have also calculated the ground motions for the 1989 magnitude 6.9 Loma Prieta earthquake, which occurred beneath the Santa Cruz Mountains and caused damage from Watsonville to San Francisco.
"It's not that the ground motions in 1906 were significantly larger than those in 1989it's that the area experiencing intense shaking was much, much greater," Aagaard said. "We don't know whether the next rupture will look like the 1906 earthquake, but we know that many of the same areas hit hard in 1906like San Francisco, Santa Rosa, and the Santa Cruz Mountainswill be hit hard again in the next large earthquake on the San Andreas Fault."
"When the 1906 event reoccurs, we'd expect tremendous damage to structures around the Bay Area," said Rich Eisner, regional director of the Office of Emergency Services. "We need to rebuild now to reduce our vulnerability before an earthquake hits."
"I'm excited about the future potential of these computer models as much as anything," said Rob Graves, a seismologist at URS Corp., an international engineering consulting firm, and project participant. "With earthquakes, we've had to learn from the past to understand or perhaps predict the future. If we can build confidence in these models' ability to simulate earthquakes, we may have a way to essentially predict earthquake effects and present different earthquake scenarios."
The simulations were run on computers at the USGS and Lawrence Livermore National Laboratory (LLNL), as well as the Southern California Earthquake Center's facilities at the University of Southern California. LLNL scientists Shawn Larsen, Anders Petersson, and Arthur Rodgers took advantage of access to the world's 45th-fastest computer in running their simulations of the earthquake.
in this issue:
Scientists Recreate Shaking from 1906 San Francisco Earthquake