Director's Report

Figure 1.1: Map illustrating the distribution of stations in the BDSN, NHFN, HRSN, BARD, and Mini-PBO networks in northern and central California. A star indicates the location of the MOBB deployment.

Background and Facilities

The Berkeley Seismological Laboratory (BSL), formerly the Berkeley Seismographic Station (BSS), is the oldest Organized Research Unit (ORU) on the UC Berkeley campus. Its mission is unique in that, in addition to research and education in seismology and earthquake-related science, it is responsible for providing timely information on earthquakes (particularly those that occur in northern and central California) to the UC Berkeley constituency, the general public, and various local and state government and private organizations. The BSL is therefore both a research center and a facility/data resource, which sets it apart from most other ORUs. A major component of our activities is focused on developing and maintaining several regional observational networks, and participating, along with other agencies, in various aspects of the collection, analysis, archival, and distribution of data pertaining to earthquakes, while maintaining a vigorous research program on earthquake processes and Earth structure. In addition, the BSL staff spends considerable time with public relations activities, including tours, talks to public groups, responding to public inquiries about earthquakes, and, more recently, World-Wide-Web presence (

UC Berkeley installed the first seismograph in the Western Hemisphere at Mount Hamilton (MHC) in 1887. Since then, it has played a leading role in the operation of state-of-the-art seismic instruments and in the development of advanced methods for seismic data analysis and interpretation. Notably, the installation, starting in 1927, of Wood-Anderson seismographs at 4 locations in northern California (BKS, ARC, MIN and MHC) allowed the accurate determination of local earthquake magnitude ($M_{L}$) from which a unique historical catalog of regional earthquakes has been maintained to this day, providing crucial input to earthquake probabilities studies.

Over the years, the BSS continued to keep apace of technological improvements. The first centrally telemetered network using phone lines in an active seismic region was installed by BSS in 1960. The BSS was the first institution in California to operate a 3-component ``broadband" system (1963). Notably, the BSS played a major role in the early characterization of earthquake sources using ``moment tensors" and source-time functions, and made important contributions to the early definitions of detection/discrimination of underground nuclear tests and to earthquake hazards work, jointly with UCB Engineering. Starting in 1986, the BSS acquired 4 state-of-the-art broadband instruments (STS-1), while simultaneously developing PC-based digital telemetry, albeit with limited resources. As the telecommunication and computer technology made rapid progress, in parallel with broadband instrument development, paper record reading could be completely abandoned in favor of largely automated digital data analysis.

The current modern facilities of BSL have been progressively built over the last 14 years, initiated by significant ``upgrade" funding from UC Berkeley in 1991-1995. The BSL currently operates and acquires data, continuously and in real-time, from over 60 regional observatories, housing a combination of broadband and strong motion seismic instrumentation installed in vaults, borehole seismic instrumentation, permanent GPS stations of the Bay Area Regional Deformation (BARD) network, and electromagnetic sensors. The seismic data are fed into the BSL real-time processing and analysis system and are used in conjunction with data from the USGS NCSN network in the joint earthquake notification program for northern California, started in 1996. This program capitalizes on the complementary capabilities of the networks operated by each institution to provide rapid and reliable information on the location, size and other relevant source parameters of regional earthquakes. In recent years, a major emphasis in BSL instrumentation has been in densifying the state-of-the-art seismic and geodetic networks, while a major on-going emphasis in research has been the development of robust methods for quasi-real time automatic determination of earthquake source parameters and predicted strong ground motion, using a sparse network combining broadband and strong motion seismic sensors, as well as permanent geodetic GPS receivers.

The backbone of the BSL operations is a regional network of 25+ digital broadband and strong motion seismic stations, the Berkeley Digital Seismic Network (BDSN), with continuous telemetry to UC Berkeley. This network provides the basic regional data for the real-time estimation of location, size and rupture parameters for earthquakes of M 3 and larger in central and northern California, within our Rapid Earthquake Data Integration (REDI) program and is the Berkeley contribution to the California Integrated Seismic Network (CISN). It also provides a fundamental database for the investigation of three-dimensional crustal structure and its effects on regional seismic wave propagation, which is ultimately crucial for estimating ground shaking for future earthquakes. Most stations also record auxiliary temperature/pressure channels, valuable in particular for background noise quality control. Complementing this network is a $\sim 25$ station ``high-resolution" network of borehole seismic sensors located along the Hayward Fault (HFN) and under the Bay Area bridges, operated jointly with the USGS/Menlo Park and linked to the Bridge Safety Project of the California Department of Transportation (Caltrans). The latter has facilitated the installation of sensor packages at 15 bedrock boreholes along 5 east-bay bridges in collaboration with Lawrence Livermore National Laboratory (LLNL). A major science goal of this network is to collect high signal-to-noise data for micro-earthquakes along the Hayward Fault to gain insight into the physics that govern fault rupture and its nucleation. The BSL is also involved in the operation and maintenance of the 13 element Parkfield borehole seismic array (HRSN), which is yielding enlightening results on quasi-periodic behavior of micro-earthquake clusters and important new constraints on earthquake scaling laws and is currently playing an important role in the characterization of the site for the future San Andreas Fault Observatory at Depth (SAFOD). Since April 2002, the BSL is also involved in the operation of a permanent broadband ocean bottom station, MOBB, in collaboration with MBARI (Monterey Bay Aquarium Research Institute).

In addition to the seismic networks, the BSL is involved in data archival and distribution for the permanent geodetic BARD Network as well as the operation, maintenance, and data processing of 22 out of its 70+ sites. Whenever possible, BARD sites are collocated with BDSN sites in order to minimize telemetry costs. In particular, the development of analysis methods combining the seismic and geodetic data for the rapid estimation of source parameters of significant earthquakes has been one focus of BSL research.

Finally, two of the BDSN stations (PKD, SAO) also share data acquisition and telemetry with 5-component electromagnetic sensors installed with the goal of investigating the possibility of detection of tectonic signals. In 2002-2003, an automated quality control software was implemented to monitor the electromagnetic data.

Archival and distribution of data from these and other regional networks is performed at the Northern California Earthquake Data Center (NCEDC), operated at the BSL in collaboration with USGS/Menlo Park. The data reside on a mass-storage device (current holdings $\sim 10$ TerraBytes), and are accessible ``on-line" over the Internet ( Among others, data from the USGS Northern California Seismic Network (NCSN), are archived and distributed through the NCEDC. The NCEDC also maintains, archives and distributes the ANSS/CNSS earthquake catalog.

Core University funding to our ORU has suffered from permanent budget cuts to research programs from the State of California, and currently provides salary support for 2 field engineers, one computer expert, 2 data analysts, 1 staff scientist and 2 administrative staff. This supports a diminishing portion of the operations of the BDSN and provides seed funding for our other activities. All other programs are supported through extra-mural grants primarily from the USGS and NSF, and in the past five years, the Governor's Office of Emergency Services (OES). We acknowledge valuable recent contributions from other sources such as Caltrans, the CLC program and PEER, as well as our Earthquake Research Affiliates.

Highlights of 2005-2006

Research Accomplishments

Chapter 3 documents the main research contributions of the past year. Research at the BSL spans a broad range of topics, from the study of microseismicity at the local scale to global deep earth structure, and includes the use of seismological, geodetic, and remote sensing (InSAR) techniques.

Richard Allen and his students have continued to develop a methodology for earthquake early warning (4,12), documenting, in particular, that, for large earthquakes, the frequency content of radiated seismic energy within the first few seconds of rupture scales with the final magnitude of the event.

Bob Nadeau and collaborators continued to work on characteristically repeating micro-earthquakes (5,8). In particular, Nadeau has extended the study published in 2004 to investigate if the time varying patterns of periodic pulsing of deep slip, extends through an additional 5 years. The successful prediction of time periods with increased likelihood for larger earthquakes based on his earlier data suggest that there is significant potential in his approach to refine time dependent earthquake forecasts, at least for the central San Andreas Fault segment (5). Nadeau and collaborators are also studying repeating earthquakes in other tectonic settings, such as Taiwan (7).

Studies of data generated by the M6 2004 Parkfield earthquakes continue. Nadeau and collaborators have been further documenting the time and space distribution of non-volcanic tremors on the San Andreas fault near Parkfield (6). Graduate student Karl Kappler has been refining EM data analysis procedures (20, 37, 21). Doug Dreger and his student Ahyi Kim have developed a kinematic model of the Parkfield earthquake using jointly GPS and seismic data (9).

Doug Dreger and his students and collaborators have also worked on a variety of source and structure topics including the development of a promising method to identify seismic events with a strong isotropic component, very relevant to the recent nuclear test in North Korea (10), new earthquake ground motion simulations using the recently developed USGS SF06 velocity model (11), the determination of structure in the Santa Clara Valley using both teleseisms and microearthquake data (24), and evaluation of methodologies to constrain attenuation models in northern California (17).

Geodetic studies include work by Roland Bürgmann and his students and post-docs on the surface deformation in the San Francisco Bay Area using InSAR data (22) and on the motion of the Indian plate (18), as well as work by Nicolas Houlié and collaborators on the deformation of the Etna volcano, using GPS (19). In a collaborative effort involving three BSL faculty members (Dreger, Bürgmann and Romanowicz), separate tools developed by each of them were combined by graduate student Junkee Rhie to obtain a slip model for the great M9 2004 Sumatra earthquake, based on teleseismic and near field geodetic data (23).

Other regional studies include continuation by Bob Uhrhammer of the study of Bay Area historical seismicity (15) and, in particular, scanning of old seismograms with the help of undergraduates (16); the study by Peggy Hellweg of an unusual sequence of earthquakes near Orinda, CA (14); and the development of a methodology to reduce long period ocean noise at the broadband ocean bottom station MOBB(24).

On the topic of relation of oceanography and seismology, Romanowicz and her student Junkee Rhie have continued their study of the relation of ocean storms to the earth's long period noise known as "hum", and followed a particular sequence of storms across the Pacific (25), showing how the generation of the hum involves a three stage process (ocean waves, infragravity waves, seismic waves).

Deep earth structure studies start in the upper mantle, with work by graduate student Mei Xue with Richard Allen on a regional tomographic model beneath northwestern US, tracking the fate of the Juan de Fuca plate down to 400 km (27), while post-doc Federica Marone, working with Barbara Romanowicz, has developed a tomographic model of S velocity and anisotropy beneath the entire North American continent, documenting the presence of two layers of anisotropy beneath the stable continent, one in the lithosphere and one in the asthenosphere (29). Federica has also been working, with collaborators at Northwestern Univ., on a similar model for the Mediterranean region (32). Also on the regional scale, graduate student Ahyi Kim, working with post-doc Mark Panning and Barbara Romanowicz, has developed a regional upper mantle 3D model for southeast Asia (30). This model is the first step towards developing a higher resolution model of this region using more sophisticated tools for 3D wave propagation.

As in previous years, Romanowicz and her group have continued their work on refining tomographic models of mantle structure at the global scale (33,31) and have started to obtain results from a new approach, combining seismic and mineral physics constraints to map lateral variations of temperature and composition in the upper mantle (26). Also, graduate student Ved Lekic has started constructing a new global 3D model of attenuation in the upper mantle, using the spectral element method to compute synthetic seismograms in an elastic model (28). With post-doc Aimin Cao, Romanowicz also continues work on core structure, one example of which is given in this report, documenting the possible presence of short wavelength topography on the inner core boundary (34).

Finally, combining expertise in seismology and geodynamics, a study of the constraints on the structure of Jupiter's satellite Europa that could be obtained with seismic data has given rise to two publications (35), while Ved Lekic, working with Michael Manga, has been investigating tidal excitation of free oscillations on Saturn's moon Enceladus (36).

Infrastructure and Earthquake

Highlights in 2005-06 include the BSL's participation in the commemoration activities of the 1906 earthquake's centennial anniversary (see Chapter 46). These activities included a joint SSA/EERI/DRC conference which was held in San Francisco in April 2006. The BSL, and specifically Dr Peggy Hellweg, were heavily involved in its organization. The BSL also helped develop many exhibits, classes, and a series of public lectures on the UC Berkeley Campus, which were held between October 2005 and April 2006. Richard Allen and Barbara Romanowicz taught a Freshman Seminar on the subject of the 1906 earthquake and earthquake preparedness, which resulted in an informational movie, put together by the students, and targeted at incoming Berkeley freshmen. This movie is available through several websites on Campus, in particular:

Another new activity began in February 2006, with funding from NSF through the IRIS/GSN program (Global Seismic Network). It is a collaborative project with Tom VanZandt of Metrozet on the design and testing of new electronics for the STS-1 very broadband seismometer which occupies several hundred sites around the world, and in particular 10 of the Berkeley sites. The manufacturer of this sensor of unsurpassed quality has discontinued its production, but no equivalent new instrument of any design is available. Meanwhile the STS-1s, many of which have been around for 15-20 years, are deteriorating with age. In the Spring of 2006, significant effort was deployed at BSL to set up a testbed at the Byerly Vault to test successive iterations of new electronics developed by Metrozet.

As in the previous year, BSL's infrastructure development efforts have centered around several major projects:

The main goal of the CISN (see Chapter 39) is to ensure a more uniform system for earthquake monitoring and reporting in California. The highest priority, from the point of view of emergency responders in California, is to improve the robustness of statewide real-time notification and to achieve a uniform interface across the State to the California OES and other emergency responders. This represents a major challenge, as the CISN started as a heterogeneous collection of networks with disparate instrumentation, software systems and culture. Much effort has gone over the past few years to develop coordinated software between southern and northern California and in northern California, between Berkeley and USGS/Menlo Park. These two institutions are joined together in the Northern California Earthquake Management Center(NCEMC). Until now, earthquake processing responsibilities were divided into two components. Responsibility for running the association, location and duration magnitude modules has resided with Menlo Park, while local and moment magnitude, and finite-fault modules run in Berkeley. Redundancy is built into the system by operating two complete systems at all times, the second one as a "hot" backup. Recognizing potential problems associated with the separation of critical system elements across the San Francisco Bay, an effort was launched in 2001 to redesign the Northern California operations. In the new system, which is currently in advanced test mode, two complete systems, which provide processing from detection to location and computation of ground motion parameters and Shakemaps, will operate one in Berkeley and one in Menlo Park. Its implementation is awaiting the retirement of the CUSP real-time earthquake timing system in Menlo Park (see chapter 44). In the past year, CISN funds were also used to complete new broadband stations at Alder Springs, CA (GASB), and at the Marconi Conference Center near Point Reyes (MCCM). FEMA funds made available following the San Simeon earthquake of 2003 are being combined with CISN funds, to purchase equipment for 7 additional new stations that will fill gaps in the present distribution of broadband stations in northern California. This past year the equipment has been ordered and the sites identified among currently operating temporary northern California sites of the Earthscope/USArray program. BSL staff continue to spend considerable efforts in organizational activities for CISN, notably by participating in the CISN Project Management Group (Neuhauser and Hellweg), which includes weekly 2 hour phone conferences, and the Standards Committee (Neuhauser-chair, Hellweg, Lombard), which strives to define and coordinate software development tasks. Romanowicz and Hellweg serve on the CISN Steering Committee. Doug Neuhauser has also been serving on the CISN Steering Committee in the transition period following Lind Gee's departure in summer 2005. The CISN also represents California as a designated region of ANSS (Advanced National Seismic System) and the BSL is actively involved in planning activities for the ANSS.

The BSL concluded an agreement in June 2004 with IRIS to contribute 19 stations of the BDSN to USArray, while the experiment is deployed in California. This includes 17 existing stations and the two new sites mentioned above: GASB and MCCM. In the past year, BSL has continued to acquire telemetered data from these and other northern California USArray stations and to pay particular attention to the maintenance of those permanent sites which are part of USArray.

The BSL has completed the relocation of the critical operations of data acquisition, processing, archiving and data distribution to 2195 Hearst (``SRB-1"), a recently completed building on Oxford Tract, which was constructed to current seismic codes, with special attention to post-earthquakes operability of the campus computer facility. The computer center contains state-of-the-art seismic bracing, UPS power and air conditioning with generator back-up and extensive security and equipment monitoring. In the past year, BSL has finished moving all of its data acquisition, real-time earthquake processing computers, and data archive and distribution computers to the new facility, including telemetry equipment.

The Parkfield borehole network (HRSN, see chapter 41) undertook an effort to upgrade power modules in early 2005. This has now been completed and has effectively eliminated data drop outs and gaps that had plagued the network during the winter months. The HRSN continues to play a key role in support of the Earthscope SAFOD (San Andreas Fault Observatory at Depth) drilling project, by providing low noise waveforms for events in the vicinity of the target drilling zone. A procedure is underway to refine an automated similar event detection method based on cross-correlation and pattern scanning of the continuous HRSN data, to assist researchers working on repeating micro-earthquakes as well as those working on the target events in the SAFOD drilling zone.

In the past year, infrastructure development for the Northern Hayward Fault Network (NHFN, see chapter 40) has progressed notably. The coverage on the west side of the Bay has been augmented with the installation of the "Mini-PBO" station MHDL (Marin Headlands), which comprises borehole seismometers and a continuous GPS receiver. It is now part of the NHFN and of the BARD 42 network. Caltrans has provided funding and support for instrumentation of 3 land sites for the NHFN. Station VALB (Vallejo, CA) is operational, whereas PETB (Petaluma River Bridge) is instrumented and awaiting completion of telemetry infrastructure. The third hole has been drilled at St. Mary's College (SMCB) and is nearing completion. On-going network maintenance involves regular inspection of the collected seismic waveform data and spectra for nearby seismic events, and for noise samples, in order to assure that the instruments operate at maximum performance to capture the source spectrum of micro-earthquakes down to negative magnitudes.

The last year has been a transition period for the BARD continuous GPS (C-GPS) network, with the departure of Mark Murray and arrival of Nicolas Houlié. Also, the landscape has been changing in northern California with the construction of the Plate Boundary Observatory (PBO) of Earthscope, which is installing many new C-GPS stations. We are working with PBO to transfer several of our stations to PBO, those that are either not accessible by continuous telemetry or collocated with broadband seismic stations. We are also refocusing BARD on real-time continuous data acquisition at high sampling rates for earthquake hazards applications, responding to a growing interest of our own researchers and of the community. Data from the BARD stations have been traditionally acquired at 15 or 30 s sampling rates, but, for these applications, 1 s or higher sampling rates are needed. Conversion to these higher rates is limited by the telemetry capabilities and nature of receivers at each individual station. We started collecting 1 Hz data at 2 stations in 2003. In the last year, we progressively upgraded nine additional stations to continuous 1 Hz telemetry. These data are made publicly available at the NCEDC, while we develop methods to combine them with seismic data in our real-time earthquake notification system. We have recently acquired 5 new receivers (Ashtech $\mu$Z) which will enable us to upgrade several of the sites collocated with BDSN stations.

The NCEDC (see Chapter 43), continues archival and on-line distribution of data from expanding BDSN, NHFN, HRSN, BARD, Mini-PBO, and other networks and data collections in northern California and Nevada, including telemetered continuous data from USArray stations in northern California and vicinity. We are continuing to receive data from the SAFOD pilot hole and main hole, and data from 15 SCSN (southern California) broadband sites as part of the CISN robust ``backbone". Previously, only event data were archived from the 1000+ components of the Northern California Seismic Network (NCSN) operated by USGS/Menlo Park. As a major accomplishement in the past year, we have developed the necessary software and procedures and are now archiving continuous NCSN seismograms at NCEDC. We have developed software and are retroactively working our way to transfer older data from tapes to complete the on-line collection. Having easy access to these continuous data is important, in particular, for scientists working on non-volcanic tremors. It is also of interest to global seismologists working on body waves from large teleseisms. The change from event archiving to continuous archiving has significantly increased the amount of data archived (from 35GB/year to $\sim $1750 GB/year). This was accomplished with no additional NCEDC staff, and was made possible by our work to automate the data delivery and archival process, and by the significant decrease in the cost of RAID disk systems.

Finally, with seed funding from the USGS to Prof. Richard Allen, we have started work on the establishment and testing of a prototype earthquake early warning system.

BSL staff news

This past year has seen many changes in BSL personnel. Following Eleanor Blair's retirement, Kristen Jensen joined BSL as our new Manager in February 2006. Kristen previously worked as senior analyst in the office of the VCR. We are very fortunate to have her. Jenny Pehl left BSL in July 2006 to follow her husband to a new job in Las Vegas. Nicolas Houlié was hired as a post-doc in February 2006 to take responsibility for our continuous GPS program.

Four graduate students associated with BSL completed their PhD's in the past year: Aimin Cao, Junkee Rhie, David Dolenc and Ingrid Johannson. Ingrid joined the USGS/Menlo Park on a Mendenhall Post-doc, David is in Minnesota (Univ. of Minnesota in Duluth), while Aimin and Junkee are staying on as post-doc's at BSL. Mark Panning, who had stayed as post-doc at BSL after completing his PhD in Fall 2004, left for Princeton in July 2006, where he has a post-doctoral fellowship.

New arrivals have continued through the summer and early Fall of 2006. Kevin Mayeda transferred from LLNL to join BSL research staff in July, while Cyndy Bresloff has been helping with routine processing and data quality control since mid-August. Tina Barber-Riggins joined the business office in September. Alexei Kireev and Mario Aranha joined the programming team in September and October, respectively. They will assist Doug Neuhauser with tasks related to the CISN, the Early Warning Project and the NCEDC.

This year has also been marked by very sad news: Professor Bruce Bolt, who was Director of the Seismographic Stations from 1963 to 1989, passed away after a brief illness on July 21, 2005, just a few months before the start of the celebrations of the 1906 San Francisco earthquake centennial. A memorial, attended by over 500 people, was held for him at the Faculty Club on July 29, 2005. We include in this report a copy of the obituary (draft version as of 10/17/06) which we have been putting together for UC Berkeley.


I wish to thank our technical and administrative staff, scientists and students for their efforts throughout the year and their contributions to this Annual Report. Individual contributions to activities and report preparation are mentioned in the corresponding sections, except for the Appendix section, prepared by Kate Conner and Kristen Jensen.

I also wish to specially thank the individuals who have regularly contributed to the smooth operation of the BSL facilities: Rich Clymer, Doug Dreger, John Friday, Jarrett Gardner, Peggy Hellweg, Nicolas Houlié, Bill Karavas, Rick Lellinger, Pete Lombard, Rick McKenzie, Mark Murray, Bob Nadeau, Doug Neuhauser, Charley Paffenbarger, Bob Uhrhammer, and Stephane Zuzlewski, and in the administrative office, Kristen Jensen, Kate Conner, Jenny Pehl and Yolanda Andrade. I particularly wish to thank Doug Dreger for serving as Associate Director of the BSL, Peggy Hellweg for taking on responsibilities for CISN and data analysis coordination, Doug Neuhauser for his efforts to build the Earthscope related archival system, and Bill Karavas for organizing an effective test bed for the refurbishment of the STS-1 electronics.

I also wish to thank our undergraduate assistants, Brian Castriota, Chi Sum Chan, Mehershad Dahmubed, Josh Hunt, Kevin Lee, Rose Li, Tomasz Matlak, Rob Porritt Jake Siegal, Gretchen Sites and Noi Valera for their contributions to our research and operational activities.

I am particularly thankful to Kate Conner and Peggy Hellweg, for their help in putting together this Annual Report.

The Annual Report of the Berkeley Seismological Laboratory is available on the WWW at

Barbara Romanowicz

October 15, 2006

Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail:
© 2006, The Regents of the University of California