NOTHING like a mighty earthquake demonstrates the power of nature. People who live through one remember it forever. California's Loma Prieta and Northridge earthquakes were disastrous, but experts warn that more and possibly bigger earthquakes threaten the state. We're powerless to prevent them, but we can prepare. That is just what the University of California (UC) is doing, with the help of the Campus Earthquake Program (CEP), a partnership of Lawrence Livermore National Laboratory and seven UC campuses. The CEP is helping UC prepare for large earthquakes by determining what can be expected at specific sites on various campuses. Livermore's François Heuze, a geotechnical engineer in the Energy and Environment Directorate, initiated and leads the CEP. Heuze explains, "Campus structures were damaged from the moderate Loma Prieta and Northridge earthquakes. In larger tremors, University campuses could suffer loss of life and serious damage." The work has been funded under the Campus-Laboratory Collaboration Program of the UC Office of the President. It has also received funding from the campuses that had sites evaluated as well as from Lawrence Livermore's University Relations Program.

From Research to Reality The Campus Earthquake Program had its genesis as a research project in Livermore's Laboratory Directed Research and Development (LDRD) Program in 1991. In this project, started by the Engineering Directorate's Gerry Goudreau, Laboratory seismologists and geotechnical engineers used site-specific records from small earthquakes to predict strong ground motions at those same sites during large earthquakes. Then structural engineers on the team used these strong-motion estimates to calculate the response of specific structures, such as the Dumbarton Bridge crossing San Francisco Bay (see Energy & Technology Review, September-October 1993, pp. 7-17). With LDRD results in hand, Heuze met with UC officials to determine their interest in conducting similar studies at the campuses. At the time, UC's seismic policy was quite brief. "It basically said that if you had any earthquake concerns, you should call a structural engineer," recalls Heuze. UC campuses at Riverside, San Diego, and Santa Barbara expressed interest in acquiring additional information about specific sites on their campuses. About the same time, the UC Office of the President initiated the Campus-Laboratory Collaboration program to encourage cooperative research between the national laboratories and the campuses. Of 120 proposals submitted to the program in 1995, the CEP was one of five that were funded. Bringing together experts in geology, seismology, geophysics, and geotechnical engineering, the CEP in 1996 began a four-year examination of three specific sites: the Engineering 1 Building at Santa Barbara, the Thornton Hospital at San Diego, and the Rivera Library at Riverside.

Listening to the Underground Given a site's ground motion-that is, how the site's geology responds to earthquakes-a structural engineer can design a building to withstand that motion. However, estimating the range of possible ground motions at a particular site requires a detailed knowledge of the site geology, the regional earthquake faults, and the site's response to seismic waves. A geotechnical engineer draws on a variety of methods to determine what level of ground shaking should be considered in designing a structure to withstand earthquakes. One method, deterministic hazard estimation, focuses on designing a structure that could survive the largest earthquake expected at that site. Before this method is applied, the source (fault) and size (magnitude) of the threat must be determined. Another method, Probabilistic Seismic Hazard Estimation (PSHA), combines a variety of uncertainties-such as the likelihood of an earthquake of a certain magnitude occurring on a given fault-to estimate the probability of the structure experiencing a certain level of ground motion (or greater) during a specified time period. The CEP provides site-specific seismic analyses by characterizing the geology of a given site, monitoring small local earthquakes at depth and on the surface, and using this information to model the strong motions of large earthquakes. The first step is to identify faults that could produce moderate to strong ground motion (defined as earthquakes of magnitude 6 and above on the Richter scale). The next step- characterizing the site-involves working out the details of the geology and the stratigraphy (the succession of geologic layers or strata). This work consists of drilling and sampling boreholes, collecting geophysical soil logs, making geotechnical measurements on soil samples, and pulling it all together to paint a detailed picture of subsurface geology. "This kind of exhaustive site characterization is seldom performed," says Heuze. "It is expensive and goes well beyond common site investigations. However, that is the price we must pay to obtain the credible site-specific knowledge required for predicting strong earthquake effects." On each campus, the researchers placed seismic stations in vertical arrays as deep as 90 meters, far beyond the 30 meters typical of most geophysical examinations. The stations recorded small earthquakes from the local faults as well as regional events. Large earthquake motions for a given site were simulated using these data in combination with rupture scenarios of the faults identified as the main threats. Heuze explains, "For our calculations, we divide the fault surface into many subzones, sum up the contributions from small events in each subpart, and thus obtain the strong motion in rock under the site. We then calculate how that earthquake propagates up to the surface through the different soil layers." Rock's response to earthquakes is linear and fairly straightforward. But soils respond nonlinearly. What can't be predicted is how an earthquake will break on a fault. It could fracture at one end and travel the length in one direction, or start at the opposite end. It could begin anywhere along the fault surface and travel in both directions, splitting the energy. Typically, the CEP analyzes over 100 rupture scenarios for each fault for a given earthquake magnitude. At the UC Santa Barbara site, for example, the team used 240 scenarios. The results of these estimates are then presented in terms of a stochastic distribution of possible motions for the campus. "It is important to understand the difference between our approach, which is stochastic but deterministic, and the PSHA, which is strictly probabilistic," says Heuze. "We are not putting probabilities on our motions. We say that, whether the likelihood is low or high, they can happen, because nobody knows how the fault will rupture."

Rock-Bottom Line Study results will be presented in a series of Lawrence Livermore reports, prepared with the campuses and available to the general public. In the reports, the site-specific ground motions calculated by CEP are compared to results obtained by other methods. Heuze says, "Eventually, the University's decision on which motions to use as the design basis will depend on the combination of all the information it acquires. We are working closely with UC's consultants to combine the deterministic and probabilistic assessments, while also accommodating the regulatory constraints of building codes." The CEP also has a proposal to perform a similar assessment for the site of the future UC Merced. Merced is the ideal study site, Heuze notes, because its ground-motion information would be available to the architects and structural engineers before they design any campus buildings. Heuze adds, "This program is unusual in drawing, for the first time, upon the brain power within the UC system-professors, postdocs, and students at the campuses, scientists and engineers at Lawrence Livermore-to address the ground-motion problem facing the University. I see earthquake exposure to be the single greatest threat to the welfare of the University. Now, through this multidisciplinary effort, we can understand much more clearly the earthquake exposure that each campus faces." —Ann Parker

Key Words: Campus Earthquake Program (CEP), Campus-Laboratory Collaboration Program, ground-motion analysis, Probabilistic Seismic Hazard Analysis (PSHA).

For more information contact François Heuze (925) 423-0363 (heuze@llnl.gov).