It has been almost ten years since the Loma Prieta earthquake rumbled through the San Francisco Bay Area. Given the human prediliction for anniversaries, this seemed like a good opportunity to look back and see how things have changed for the Berkeley Seismological Laboratory.
Other sources of Loma Prieta anniversary materials:
At the time of the 1989 Loma Prieta earthquake, the Berkeley Seismological Laboratory (BSL, then known as the Seismographic Station) operated a relatively small network of seismometers in northern and central California. Most of the instruments represented 1960s-level technology and were limited in their ability to record the full range of ground motion. In particular, the network lacked the ability to continue operating when local power was disrupted and to record data locally when telecommunications failed. As a consequence, only the most distant stations operated by the BSL recorded the Loma Prieta earthquake with fidelity. Communication links and power failed at several sites, resulting in loss of data. The BSL provided a preliminary location and magnitude of Loma Prieta within 30-45 minutes, but could provide little additional information to emergency response operators and the press.
The Loma Prieta earthquake provided both challenges and opportunities for the Seismological Laboratory. Since 1991, primarily with funding provided by the University of California at Berkeley, the BSL has been able to upgrade and expand its monitoring efforts in northern California. Additional support has been provided by the National Earthquake Hazards Reduction Program through the USGS, the National Science Foundation, and the California Office of Emergency Services. The efforts of the BSL can be categorized in terms of enhanced instrumentation, improved monitoring and analysis, and expanded archives for long-term research. Although there has been much progress in the last 10 years, there is still much to do.
The 1960s-era seismic instrumentation (which formed the backbone of the BSL's facilities in 1989) has been replaced with modern broadband seismometers co-located with strong-motion accelerometers. This combination of instrumentation means that the BSL has the ability to record the full range of ground motions (from small and large earthquakes) spanning the entire range of frequencies. Data from the seismic sensors are digitized on site and are recorded locally on disk and transmitted over dedicated communication links to UC Berkeley. Each site has batteries to provide back-up power for 2-3 days. These ~20 sites form a sparse array in northern California known as the Berkeley Digital Seismic Network (BDSN).
Complementary to the BDSN, the BSL is collaborating with other agencies to closely monitor the Hayward Fault. Given its location in the urban East Bay corridor, traditional approaches to seismic monitoring with surface sites cannot combat the noisy operating environment. The Hayward Fault Network (HFN) consists of special sensors installed in boreholes at depths ~100 meters in order to record the small to moderate earthquakes which occur regularly on this fault. The analysis and study of these events will enhance our understanding of the behavior of this fault. Similar studies have led to new insights on the San Andreas fault near Parkfield, CA. Today, 11 borehole sites (7 operated by the BSL and 4 by the USGS) are operating. It is hoped that this network will be expanded up to 25-30 sites.
In addition to the seismological instrumentation, the BSL is participating in a cooperative regional network of continuously operating Global Positioning System (GPS) receivers in northern and central California. The BSL currently contributes ~20 stations to the Bay Area Regional Deformation Network, and many of them are co-located with seismic instrumentation. These sensors use the same technology that helps MUNI know where its buses are to measure how strain is accumulating on faults in northern California.
The seismic and geodetic instrumentation described above serve dual purposes. On one hand, they provide critical data for rapid response to earthquakes. One the other, they provide essential data for research on fundamental problems related to how and why earthquakes occur. Both aspects contribute, on different time scales, to reducing the loss of life and damage to property from earthquakes.
In conjunction with the expanded instrumentation, the BSL has upgraded its capabilities for earthquake monitoring in the last 10 years. This has included a complete overhaul of the computer facilities at the BSL, the development of new software for data processing, and the establishment of a backup processing facility.
At the time of the Loma Prieta earthquake, all earthquake analysis was performed "by hand", i.e., a seismologist was required to locate and determine the magnitude of an event. This generally required 15-20 minutes - if a seismologist was on-site at the BSL. Then the seismologist would proceed through a call-down list, contacting organizations with a known interest in earthquake information.
Today, most earthquake analysis is performed automatically. The USGS and the BSL work together to provide information rapidly after an event, sharing data across a dedicated circuit connecting their facilities. Preliminary locations are generally available within a minute, with final locations and magnitudes within 4-6 minutes. Information about the earthquake's location and magnitude is provided by WWW (quake.wr.usgs.gov/recenteqs), by email, and by pager. For example, a computer monitor at the OES headquarters in Sacramento provides a map of California earthquakes. When an event of magnitude 3.0 and higher is registered, the display will flash and alert the person on duty that an event of interest has occurred. The USGS is developing a system to distribute information over the Internet and we are participating in this process.
While information on the location and magnitude of an earthquake is important, emergency response operators are most interested in the questions of how much shaking was experienced - and where? This information, generally presented in a map (often referred to as "Shake Maps"), requires direct observation of ground motion (made by strong-motion instruments). In areas where instrumentation is sparse (most of northern California), seismologists use models to predict the amount of ground shaking. The development and implementation of these maps is a major effort today at the BSL, the USGS, and the California Division of Mines and Geology. We are working together to share data and models in order to produce the best information for post-earthquake response. Prototype maps are available on the Web and the USGS, BSL, and CDMG are working to improve the system.
In addition to our efforts to improve and enhance the quality of earthquake information available following an event, we are also working to "harden" our facility in order to insure that we remain operational during a Bay Area earthquake disaster. The BSL is equipped with "uninterruptible power supply" systems and a generator in order to provide power to operate our computer systems, lights, and other critical systems. Our building is currently being renovated and retrofit, but our proximity to the Hayward fault remains a concern. Several years ago, we concluded that the best way to insure that earthquake information is available to agencies such as OES was through the establishment of a redundant processing facility outside of the Bay Area. We recently installed a computer system in the Sacramento office of OES for this purpose. The system receives data from a subset of our network and operates independently of our facility at UC Berkeley, detecting and locating the larger earthquakes in northern and central California. We are working on this system with the goal of making it a better resource for the OES personnel.
For more information on the BSL's role in earthquake monitoring, see the contribution on Network Data Analysis from the BSL 1997-1998 Annual Report.
At the time of the Loma Prieta earthquake, data from the BSL instruments were primarily archived on paper and film. Access to these data was limited, especially for users outside of the Berkeley community. Today, the BSL and the USGS operate the Northern California Earthquake Data Center. This joint facility is located at UC Berkeley and provides a long-term archive for earthquake data. Earthquake catalogs, seismic waveforms, GPS data, electromagnetic data and others are stored here and available over the Internet (www.quake.geo.berkeley.edu). In the case of the Berkeley networks, data are available almost immediately. This means that researchers around the world may begin to study an earthquake as soon as it occurs - and that it will be available for future researchers in the years to come. Data are available to both the general public as well as the scientific and engineering communities, such as earthquake lists, seismocams, earthquake catalogs, and access to waveform data.
In addition to earthquake monitoring, the Berkeley Seismological Laboratory plays an important role in research studies relevant to Bay Area seismic hazard.
Finite Fault Studies and Rapid Determination of Strong Shaking
Doug Dreger, Asya Kaverina
At the time of the October 17, 1989 Loma Prieta earthquake there were two regional distance broadband instruments that had digital recording with dialup capability. One of the instruments operated by UC Berkeley was located at Columbia, CA and the other operated by Caltech was located in Pasadena, CA. These instruments provided valuable recordings of the earthquake but the network coverage was not sufficient for a regional distance assessment of the earthquake source. Furthermore the CMB station clipped following the regional P wave arrival and at the time there was no co-located strong motion station. In the ten years since the Loma Prieta earthquake the numbers of broadband stations operated by UC Berkeley in northern California and by Caltech in southern California have increased many fold, and the dynamic range of these stations was widened with the addition of co-located strong motion instruments to ensure that the systems stayed on line during the large events of interest. With the improved network coverage it has become possible to analyze the broadband wavefield recorded by these instruments in near-realtime to determine the finiteness of the earthquake source. In other words it has become possible to determine the length and width of the fault and the amount of fault slip in 10s of minutes of the earthquake. This information is of direct importance in the estimation of the level of strong shaking in areas close to the earthquake source. Thus it has become possible to provide estimates of the level of strong shaking close to the earthquake based on information carried in regional distance seismograms. We have recently completed a study, which analyzed the 1992 Landers (MW7.3) and the 1994 Northridge (MW6.7) earthquakes to assess the feasibility of such a near-realtime hazard characterization. The results of this study indicate that it is possible to determine the finite source parameters of earthquakes within the coverage region of the broadband networks and to calculate near-source strong shaking maps within approximately 10-30 minutes following the shock depending upon the level of approximation. Shaking maps provided in a timely manner are of use to state and national emergency response agencies, operators of public utilities, rail, air and highway transportation and freight providers who must quickly assess the level of damage to begin post-earthquake remediation efforts.
For more information, see the slide show or the PGE-PEER Final Report.
Analyzing the Effects of 3D Earth Structure
Doug Dreger, Christiane Stidham, Shawn Larsen (LLNL)
At the time of the Loma Prieta earthquake ten years ago it was possible to investigate the nature of the Earth's crust through the interrogation and modeling of recorded seismic waves, however those studies were mostly limited to 1D models of the Earth's structure. The detailed interrogation of the Earth's 3D structure has only recently become possible due primarily to advances in computer technology and in particular the development of large parallel processing super computers. Our group has collaborated over the past four years with Dr. Larsen of the Lawrence Livermore National Laboratory to investigate the geologic structures in the greater San Francisco Bay Area that are most relevant to the generation and amplification of strong ground motions. Of particular interest is the estimation of strong ground motions in future earthquakes so that an educated assessment of the shaking and resulting damage may be made in advance of the earthquake to reduce losses. Towards this end our group has developed a 3D model of the Earth's seismic velocity and density structure of the crust in the greater San Francisco Bay Area and have used a numerical algorithm developed by Dr. Larsen to simulate ground motions for earthquakes. This algorithm is arguably the most advanced in use today and includes a suite of parameters that are known to be important in the generation of strong ground motion. Ironically the 1989 Loma Prieta earthquake remains the best data set to date with which to compare model predictions to observations due to the scope of strong motion instrument coverage. We have modeled the peak ground motions and seismograms recorded at these stations using realistic representations of the earthquake source and Earth structure. We have found that our model predicts the distribution of peak ground motions quite well and have discovered that the difference in material properties of rocks on either side of the San Andreas Fault plays a significant role in the level of strong ground motion throughout the San Francisco Bay Area. In addition the sedimentary basins of the Santa Clara Valley, Livermore Valley and San Pablo Bay are found to amplify ground motions by factors of between 3 to 5 over estimates determined from standard one-dimensional models. Understanding the three dimensionality of the Earth's crust in the San Francisco Bay Area, the physics of wave propagation through realistic geologic structures and associated amplification effects will aide us in characterizing what the levels of shaking may be like in the next San Andreas or Hayward fault earthquake.
For more information, see the 3-D Earthquake Movies
Regional Seismic Moment Tensor Studies in Central and Northern California
Doug Dreger, Hrvoje Tkalcic, Wu-Cheng Chi
The regional broadband stations operated by UC Berkeley have been used to routinely estimate the seismic moment tensor of magnitude 3.5 and larger earthquakes throughout central and northern California. These analyses are done automatically and are subsequently reviewed by Berkeley Seismological Laboratory analysts to provide a reference catalog of earthquake focal mechanisms. Earthquake moment tensors provide a robust estimate of earthquake size, describe the orientation of the fault in space, are used to infer the geometry of faults at depth, and to study the regional tectonics. More information about the moment tensor project, a catalog of moment tensor solutions and maps showing solutions obtained for recent earthquakes can be found on the Web
Simultaneous Determination of Earthquake Location and
Moment Tensor Using Real-Time Waveforms from Sparse Regional Network
Fumiko Tajima, Charles Mégnin, Barbara Romanowicz, Doug Dreger
Currently the REDI system automatically determines moment-tensor solutions for events with a magnitude M > ~3.5 when it receives a notice of event information, i.e., location and origin time which are determined using travel time picks from the dense short-period NCSN. However, if a large earthquake should hit a crowded metropolitan area such as the San Francisco Bay area, many life lines may be destroyed, causing the loss of critical data for the earthquake and accordingly delaying the rescue efforts which may result in an avoidable death toll (ex. in the 1995 Kobe earthquake). As BSL operates the high quality network of broadband and strong motion instruments colocated, we are obliged to develop a stand alone capability to immediately estimate earthquake source parameters (location, size and mechanism) for significant events using the sparse network data. With the quality of the waveform data from the network and the methodological development, it is now feasible to configure such a system.
At present we are developing and testing a method to simultaneously determine centroid source location and moment-tensor solutions for earthquakes with a magnitude 4.5 or greater based on a grid search method. The test results are promising in comparison with the ground truth location and moment-tensor solutions depending on the constraints of structural models. In the near future we plan to model the real-time waveforms with a short time interval (~10 sec) by assigning a powerful PC to each station and relying on a parallel computational scheme to continuously monitor the seismic activity field in the region. We are in the process of developing such a system which would enable us to provide urgent information to general public should a catastrophic event like the Loma Prieta earthquake take place in the network coverage.
Follow this link for further explanation and a movie simulation of the determination of earthquake parameters in real-time based on this scheme.
High-Resolution Monitoring of the Hayward Fault
Tom McEvilly, Bob Uhrhammer, Rich Clymer, Larry Hutchings (LLNL)
In the decade since the 1989 Loma Prieta earthquake, awareness has evolved steadily within the earthquake hazards community as to the Hayward fault threat in the east San Francisco Bay region. The Northridge (1994) and Kobe (1995) events served to heighten this awareness, as did loss estimates reaching the $20-50 billion range and thousands of deaths for a major Hayward fault earthquake. Arguably one of the most hazardous faults in the world, considering the likelihood of the earthquake and the high level of urban development along it, the Hayward fault has been assigned the highest probability for a destructive earthquake in the Bay Area in next 30 years, with an estimated recurrence interval of ~167 years. Long-term Hayward fault slip rate estimates of about 10 mm/yr suggest that more than a meter of slip potential has accumulated since the most recent events, making the Hayward fault capable of M6.5 or larger events in the near future [Lienkaemper et al., 1991; Savage and Lisowski, 1992]. A number of projects were initiated after Loma Prieta to improve our understanding of Bay Area earthquakes - one was the Hayward Fault Network of wide dynamic range, borehole- emplaced, digitally-telemetered seismometers and dilatometers. This developing facility provides the only high-resolution monitoring capability for microseismicity at the M~0 threshold that is required to define the detailed spatio-temporal seismic signature of the fault. The goal of this joint UCB/USGS project is to provide the seismological community with state-of-art high- frequency and wide dynamic range close-in data on this intensely studied fault zone, to feed these new data into real-time monitoring streams, and to conduct a research effort in defining the ongoing seismicity at maximum spatial and temporal resolution. The resulting view of fault-zone process will complement intense mapping efforts on the fault and data from networks of dilatometers, tensor-strain meters, GPS systems, InSAR images, creepmeters and surface-installed seismographic stations.
For more information, see the contribution on Deep Bore Hole Instrumentation Along San Francisco Bay Bridges from the BSL 1997-1998 Annual Report.
Hayward Fault Subsurface Slip from Joint
Analysis of Microearthquake Recurrence and Space Geodesy
Tom McEvilly, Roland Burgmann, Bob Nadeau, Bob Uhrhammer
This project addresses the large natural hazard presented by the Hayward fault to the San Francisco Bay area and the national economy through the use of space based technology, namely GPS and differential radar interferometry, combined with new high-resolution borehole-based observation of the microearthquake activity on the fault. The results will improve understanding of the dynamics of the Hayward fault deformation and slip process, and will be directly applicable to the development of much improved earthquake hazard models and strong ground motion source models through precise information of the size, segmentation, and loading rate of future ruptures of the Hayward fault. Imaging distributed slip on subsurface faults from surface displacements is a difficult task, plagued by limits in the spatial resolution of the surface displacements and ambiguities in the depth resolution of slip variations. Even an optimal distribution of geodetic measurements will not be able to resolve details of subsurface creep at depths greater than ~5 km. Unambiguous surface creep measurements provide powerful constraints for the lateral distribution of slip in inversion studies (Harris and Segall, 1987). The ability to derive even just a few point measurements of fault slip rate would add invaluable constraints on the spatial distribution and magnitude of aseismic slip at depth. A new way to obtain the subsurface slip rate has been developed by Nadeau and McEvilly (1999) using recurrence histories of repeating characteristic microearthquakes and the source scaling laws of (Nadeau and Johnson, 1998) for the Parkfield, CA experiment. With the method, variations in the estimated slip rate can be determined independently of the scaling relations. The method requires complete detection at very low magnitude levels (M<2) in order to observe the recurrence intervals in the range of a few months to 2-3 years. Few regions have such high- resolution detection capability, but it is being achieved for the Hayward fault today with the developing Hayward fault borehole seismic network. We have just begun to apply this technique to the Hayward fault in this new project, and initial results are very encouraging.
For more information, see the contribution on Seismic potential of the northern Hayward fault from space-based SAR interferometry and GPS measurements from the BSL 1997-1998 Annual Report.
Bay Area Geodetic Studies
Mark Murray, Ray Baxter
The Berkeley Seismological Laboratory has several ongoing studies that measure how the Earth's crust deforms along the San Andreas fault system in the San Francisco Bay area. This crustal deformation is caused primarily by plate tectonic forces that drive the Pacific plate northwest relative to North America at about 50 mm/yr in central California. In the long-term average, the San Andreas fault and other nearby parallel faults, such as the Hayward and Calaveras, accommodate about 35 mm/yr of the total relative plate motion, with the remainder being distributed over a broad zone that includes the Sierra Nevada range and the Basin and Range province in Nevada and Utah. In the short term, many of the faults in the San Andreas system are locked in the brittle upper crustal as the more fluid crustal below about 15 km continues to freely slip. This differential motion causes elastic strain energy to accumulate in the upper crust, which eventually becomes large enough to overcome the friction on the fault and is released in earthquakes. Studies of how the surface of the crust deforms as the strain accumulates between earthquakes can reveal the location of the locked zones on the fault and the rate of slip at deeper depths, and thus provide constraints on the seismic hazards of the faults.
To make these geodetic measurements, the BSL has installed a network of about 20 continuously operating Global Positioning System (GPS) instruments in Bay Area and northern California. The Bay Area Regional Deformation network, which also includes another 20 GPS instruments maintained by the U.S. Geological Survey and other institutions, provides precise (several mm), daily measurements of the instrument positions. Changes in these positions over time with respect to other GPS stations located throughout North America and around the globe allow us to map the distribution of crustal deformation across the broad Pacific-North America plate boundary and in the vicinity of active faults. Many of these sites have been operating since 1994 and have average relative motions determined at the 1 mm/yr precision.
Although most of the changes are fairly constant over time, the continuous measurements from these stations also allow us to look for unusual variations, especially displacements caused by earthquakes. These displacements can occur both during (coseismic) and after (postseismic) earthquakes, and each type provides different information about the earthquake. The coseismic displacements can be used to map out the location, dimensions, and magnitude of the earthquake rupture. We are currently developing methods to quickly analyze the GPS data to determine these fault rupture characteristics and to combine it with the seismic data in order to improve rapid dissemination of earthquake information in the immediate aftermath of a major earthquake for emergency response purposes. Postseismic movements after the earthquake, like aftershocks, indicate how the crust responds to the new distribution of strain energy and may reveal higher rates of slip in the deeper crust that in turn cause the upper crust to once again begin accumulating strain. To date, no convincing geodetic displacement has been measured prior an earthquake. Although our monitoring efforts are not predicated on detecting earthquake precursors, we cannot rule out the possibility that they exist and remain hopeful of someday "catching" one. In the next few years, additional GPS instruments and other geodetic instruments, such as even more precise borehole strainmeters, will be installed in the Bay Area to improve our ability to detect these and other types of temporal variations in crustal deformation.
For more information, see the contribution on Geodetic Studies in the San Francisco Bay area from the BSL 1997-1998 Annual Report.
High-Resolution Monitoring at Parkfield
Tom McEvilly, Bob Nadeau, Rich Clymer
The data acquired in this experiment are unique and they are producing results that force a new look at some conventional concepts and models for earthquake occurrence and fault-zone dynamics. This research began in 1986 as a proposed direct test with proven and modern technology of two hypotheses critical to our understanding of the physics of the earthquake process, implications for earthquake hazard reduction, and the possibilities for short-term earthquake prediction:
The high quality borehole seismometer network installed at Parkfield in this project has produced unique earthquake and controlled-source (vibrator) data sets. After a major investment of time and money, there now is a unique baseline of fault-zone behavior with distinct features based on hard observations rather than on theories, forcing new thinking on the dynamic processes and conditions within the fault zone at the sites of recurring small earthquakes that must be incorporated in developing new models for fault-zone deformation. In a series of journal articles and Ph. D. theses since 1991 we have presented the evolution of a new and exciting picture of the San Andreas fault zone responding to its plate-boundary loading. Compelling evidence exists for changes with time both in seismicity and in wave propagation (from microearthquakes and from the vibrator) that appear to be coupled, and the region of the fault zone involved is the presumed M6 nucleation volume SE from Middle Mountain. Synchronous changes well above noise levels have been seen among several parameters including seismicity rate, average focal depth, S-wave coda velocities, characteristic sequence recurrence intervals, fault creep and water levels in monitoring wells. The same data demonstrate conclusively the existence of an extremely regular and localized process of ongoing earthquake-accommodated slip in the fault zone. Plausible assumptions lead to estimates for the spatial distribution of variations in slip-rate on the fault surface from changes in recurrence of these 2000+ characteristic microearthquakes. We do not (yet) know the relationship of these variations to the M6 nucleation process but the unique findings so far have significant implications for source dynamics, for earthquake forecasting, and for scaling relations among source parameters such as fault slip, rupture dimension, stress drop and seismic moment. These scaling laws developed from the Parkfield earthquakes can be projected to fit earthquakes up to M6, and they predict unprecedented high stress drops and melting on the fault surface for the smallest events. Exhumed fault-zone rocks provide independent evidence for such condition. Recurrence interval variations in the characteristic event sequences (>60% of the microearthquake population) have been used to map fault slip rate at depth on the fault surface. If this phenomenon is found through other high-resolution studies to be generally common behavior in active faults, we have the basis for a new method to monitor the changing strain field throughout the seismogenic zone. Along the way in this exciting discovery process we have challenged the conventional earthquake source model, affirmed characteristic earthquake occurrence and developed four- dimensional maps of fault-zone microearthquake processes at the unprecedented scale of a few meters. The significance of these findings lies in their apparent coupling and inter-relationships, from which models for fault-zone process can be fabricated and tested with time. The more general significance of the project is its production of a truly unique continuous baseline, at very high resolution, of both the microearthquake pathology and the subtle changes in wave propagation, providing to the seismological community an earthquake laboratory available nowhere else. This unique body of observations and analyses has also provided much of the impetus for Parkfield as the preferred site for deep drilling into an active seismogenic fault zone in Project SAFOD.
For more information, see the contribution on Seismological Studies at Parkfield, California from the BSL 1997-1998 Annual Report.
Although the BSL has come a long way in the last 10 years, there is still a long way to go toward the goals of preventing death, injury, and property loss from earthquakes. We have made progress in our ability to record even the largest earthquakes with fidelity and our datalogging system should ensure that no data are lost due to power or communication failures. This means that data will be available for post-earthquake studies in earth science and engineering design. However, our ability to provide information immediately following a major earthquake depends on the performance of the telecommunications systems. At this point, the BSL depends primarily on landlines both to bring data in and to push information out. We have some experience with satellite systems (4 stations have transmit data by satellite and phone) but currently lack the funding to replace all of our landlines. The satellite communication market is quite volatile at the present and it may be possible for us do this in the future. But for now, we have to live with the limitations of landlines.
A second major problem is the lack of modern instrumentation in northern California. Our ~30 stations are a small drop in the bucket. Only a limited number of the current USGS and CDMG sites are capable of providing the necessary on-scale data needed for rapid response - and for future research in earthquake science and engineering. The BSL, USGS, and CDMG have a shared vision for earthquake monitoring that includes 350 sites similar to the current BSL stations, plus an additional 1200 stations for monitoring strong ground shaking. This sort of densification is necessary to produce accurate shake maps.
If we make progress in these two areas in the next 10 years, then we will have come a long way toward the goals of saving lives and reducing property loss.
An important legacy of the Loma Prieta earthquake is the expanded collaboration among agencies in earthquake monitoring. The BSL, the USGS, and CDMG are working together toward unified earthquake monitoring. Each agency has different missions -- and different goals -- but are working together to reduce earthquake losses. The groups have organized themselves at TriNet North with the following goals:
The TriNet North group is currently drafting a plan to improve earthquake monitoring in northern California. There is no funding for this effort at this time, but we hope to attract interest in this initiative at the state and federal level.
Rick McKenzie
It was just after 5:00 PM. I had just returned from Room 562, our telemetry and recording room where I had checked to make sure all of the instruments were running. Now I had sat down to read and analyse the last earthquake of the day before leaving to pick up my wife, Bea, at Montclair and go home for the night. I happened to have KQED tuned in on the radio when at 5:04 PM the person at their studio said "Wow, that's a big one, the whole studio's shaking!" A second later it hit Berkeley. It wasn't a strong rapid shaking, but rather a rumbling while the whole Earth Sciences Building slowly swayed back and forth. At the time I compared it to the Alum Rock Earthquake east of San Jose in June of 1988: the motion was about the same but Alum Rock was brief, less than five seconds, while this latest event kept on going for fifteen to twenty seconds. I started to walk across the hall to Room 562 again, to see what this earthquake looked like on the recorders, but I noticed that all of the light fixtures in the hall were swaying in long arcs and decided to wait. So...I had not yet started to read the last earthquake of the day.
Shortly several of our graduate students joined me in the analysis room and I suddenly felt like a general. Usually I would have taken some one of the necessary tasks and followed it to completion...not this time. "Greg, go down to the ground floor, change the paper on the 100X torsion, and bring it up and develop it." "Nick, when the film clears, make the phase picks on the develocorder film...then get them in the computer." "Ann, answer the phones." Everything seemed to be flowing smoothly in an atmosphere of total chaos. About the time we were getting the hypocenter out of the computer, one of our Professors came in and made an educated guess about the amplitude on one component of the 100X record. Nick entered this information into our computer and there is was...a magnitude 7.0 earthquake, 19 km deep, about 14 km NE of Santa Cruz. (not bad for what we had; the worldwide average magnitude was 7.1).
This was the start of what proved to be a very long week. I first heard from Bea a few hours later. All of the books had come off the shelves where she worked. Just like me she couldn't get a line through on the phone. She had finally taken a ride home with another teacher at work, and managed to reach me from home.
We had a meeting at our office after 8:00 PM and plans were made to deploy portable instruments around the epicenter the following day, if the vans could get through. We decided what special procedures we would follow in seismic record keeping for this special event. Everyone daparted about 9:00 PM...except me. I decided to stay on and read aftershocks to see if I could get a feel for the total length of the fault rupture. Unfortunately the traces on the film became jumbled by the event. My hypocenters were landing all over the map southwest of San Jose. The following day I determined that the traces for our stations LLA and PRS and been reversed on the film. Interchanging these readings for the twenty-some-odd events that I had read, things looked much better with events lining up roughly along the San Andreas Fault from Los Altos to Watsonville.
And so things went. I worked on until evening of the second day, and then went home for some rest. I came back the following morning, worked all night, and went home the evening of the following day. Six days, three nights at home, a very long week indeed.