Seismic survey probes urban earthquake ... - Wiley Online Library

Eos, Vol. 80, No. 2, January 12, 1999

VOLUME 80

NUMBER 2

JANUARY 12,1999 PAGES 13—24

Seismic Survey Probes Urban Earthquake Hazards in Pacific Northwest PAGES 13,16-17

A Number of Cities

A multidisciplinary seismic survey earlier this year in the Pacific Northwest is expected to reveal much new information about the earthquake threat to U.S. and Canadian urban areas there. A disastrous earthquake is a very real possibility in the region. The survey, known as the Seismic Hazards Investigation in Puget Sound (SHIPS), engendered close cooperation among geologists, biologists, environmental groups, and government agencies. It also succeeded in strik­ ing a fine balance between the need to prepare for a great earthquake and the requirement to protect a coveted marine environment while operating a large airgun array. Because of the perceived threat of airguns to the environment, public outreach and applica­ tions for permits began many months ahead. Expert oversight by biologists allowed safe use of an array of large airguns amid a dense marine mammal population. Analysis of survey data is now well under way and among the results will be a high-resolu­ tion, three-dimensional velocity model for Puget Sound. Meanwhile, a second phase of SHIPS is planned for May 1999. Explosive charges will be fired along lines through Seat­ tle and Tacoma, Washington, to clarify crustal structure and determine where earthquake shaking might be most intense. The first phase of SHIPS took place March 1024,1998, in northern Washington state and southern British Columbia.The geological sur­ veys of the United States and Canada as well as six universities pooled resources, making it one of the most complex seismic studies ever con­ ducted to investigate urban earthquake hazards.The study delineated active faults and probed structures that control the propagation and amplitude of earthquake waves. Collected were multichannel seismic reflection (MCS) data and wide-angle seismic reflection and refraction data, which were recorded on hundreds of land and seafloor seismometers scattered throughout the region (Figure 1).

A number of cities lie within the region, includ­ ing 01ympia,Tacoma, Seattle, Everett, and Bellingham, Washington, and Victoria and Vancouver, British Columbia.The region overlies the Cascadia subduction zone, where the oceanic Juan de Fuca plate is being thrust beneath the western margin of North America. Recent research indi­ cates that the entire region might suffer a disas­ trous, magnitude 9,subduction-zone earthquake, and very large crustal earthquakes might strike along structures, such as the Seattle fault, that extend directly beneath some urban centers [e.g.,Atwater and Moore, 1992; Clague and Bobrowsky, 1994; Satake,et ai, 1996].Also,Seattle and Tacoma lie atop two deep Cenozoic basins, and the southern part of the Vancouver urban area overlies the large, young Fraser delta; it is unknown how these features might affect the degree of earthquake shaking. To help assess the hazards, SHIPS was to identi­ fy and map major fault zones at depth and define the tectonics of deep structures that con­ trol earthquake occurrence, including the interplate thrust fault, where great earthquakes can originate.The study also was to characterize the three-dimensional structure of sedimentary basins beneath densely populated areas, thereby improving estimates of strong ground motion, and provide detailed seismic velocities to improve the analysis of earthquake locations and focal mechanisms. Analysis of SHIPS and other seismic data will give a hazards assessment from the toe of the offshore accretionary wedge eastward nearly to the magmatic arc.The other material includes Lithoprobe seismic reflection and refraction information across southern Vancou­ ver Island [e.g.,Hyndman et ai, 1988]; seismic data from within Puget Sound [e.g., Johnson et a/., \994; Pratt et al., 1997] and the Strait of Juan de Fuca; and seismic sections from the GEOMAR/U.S. Geological Survey (USGS) study across the northern Washington convergent margin [Flueh etai, 1997].

Wide-angle Seismology A strong inducement to conducting the survey in this region of rugged topography and irregular coastline was the interlaced waterways, where seismic information could be collected with a mobile airgun array rather than with expensive onshore techniques or operations using explo­ sives. SHIPS first collected wide-angle seismic data, firing airguns from the R/V Thomas G. Thompson and recording only with onshore and ocean-bottom seismometers.The 16-gun array was discharged every 40 seconds to maximize signals at large ranges. During the entire survey the airgun array was fired 33,000 times. The Integrated Research Institutions for Seismology Consortium provided 257 temporary seismographs, deployed on land adjacent to Puget Sound, the Strait of Juan de Fuca, and Georgia Strait (Figure l ) . A n additional 71 recorders are permanent earthquake seismographs operated by the University of Washington and by the Pacific Geoscience Cen­ tre of the Geological Survey of Canada.The USGS and Dalhousie University contributed 15 ocean-bottom seismometers, deployed near Seat­ tle and Vancouver. Airgun shots recorded along a multitude of travel paths will support analysis of many twodimensional seismic sections and three-dimen­ sional tomographic velocities. Seismographs along key transects were spaced 3-10 km apart to provide detailed structural information across important geologic features, such as the Seattle fault and the sedimentary basins beneath Seattle and Tacoma. A preliminary review of seismic data suggests that wide-angle arrivals were successfully record­ ed throughout the study area. Locally, largeamplitude shear-wave refractions from the upper crust were recorded, providing key information needed to assess the ground shaking from a major quake. Past tomographic studies using local earthquakes have yielded three-dimension­ al velocity models for the Puget Sound region [e.g., Symons and Crosson, 1998]; however, infre­ quent shallow earthquakes and wide instrument spacing make resolving the shallow velocity structure difficult. SHIPS seismographs recorded arrivals from 62 local earthquakes and quarry blasts. When all earthquake travel times are com­ bined with wide-angle data from the SHIPS airguns, a high-resolution, three-dimensional velocity model will result.

Eos, Vol. 80, No. 2, January 12, 1999 deployment. MCS data were then collected southward toward Puget Sound (Figure 1). Migrated MCS data (Figure 2) shows the Seattle fault, which is a reverse or thrust fault that has an apparent throw of about 8 km.This fault is one of the most worrisome earthquake threats to downtown Seattle. Crossing reflections from the deep basin might reveal south-dipping thrust faults. Stacked MCS data from the Strait of Juan de Fuca show a strong band of reflections between 10 s and 14 s. This band may correlate with the "E" reflection evident in data collected during the Lithoprobe crossing of Vancouver Island [e.g.,Hyndman, 1988; Calvert and Clowes, 1990]. This correlation is suggested because the reflec­ tions in SHIPS data and those in Lithoprobe data come from comparable depths, and have similar layering and structure. A concern before the cruise was that airguns fired in confined waterways would cause intense, waterborne reverberation that would mask deep reflections. Preliminary processing of MCS data indicates that most reverberation can be removed by prestack deconvolution. Most strong linear noise trains, caused by reflection from isolated bathymetric features, can be removed using frequency-wave number filtering before stack. Most difficult to remove is noise from passing ships and high-amplitude, broadbanded cultural noise in bays near large cities. The Tully towed two 300-m streamers that were used to record airgun shots during 10 constantmidpoint, expanding-spread profiles (ESPs). When collecting ESP data, the two ships approached each other on opposite courses, met in the middle of the profile, and then steamed apart. ESP data from the eastern Strait of Juan de Fuca (Figure 3) shows a reflection band from rocks as deep as 60 km (17 s). A similar band is apparent at 20-30 km depth in ESP data from the western mouth of this strait.This reflection band may correlate with the "E" reflection.

Environmental Issues

Fig. 1. Seismometer deployments and tracklines of the Seismic Hazards Investigation in Puget Sound. LW= Lake Washington. Dashed black line labeled "SF" indicates the Seattle fault. Top three items in leg­ end show the type of instrument at a particular site: REFTEK = temporary onshore seismograph; OBS = ocean-bottom seismometer; EQ Station = permanent earthquake seismograph. The middle two legend items show the type of data collected along tracklines: MCS = multichannel seismic data; WA = wideangle seismic data. The bottom four legend entries show colors that encode the institution that main­ tained the instruments: USGS = U.S. Geological Survey; GSQ UVic & UBC = Geological Survey of Canada, University of Victoria, and University of British Columbia; OSU = Oregon State University; UW = University of Washington. Original color image appears at the back of this volume.

MCS Survey SHIPS obtained MCS reflection data using the University of Washington's Thompson, under con­ tract to the USGS, and the Canadian Coast Guard ship John PTully, provided by the Geological Sur­ vey of Canada. USGS mounted an entire MCS recording system on the Thompson.The airgun array consisted of 16 guns deployed from a spe­

cially built gantry that extended transversely across the stern of the ship. During MCS survey­ ing, the array was fired every 20 s. DFS-V instru­ ments recorded 24-fold, 16-s data from the 96-channel streamer, which was 2.5 km long.The MCS streamer was deployed in the northernmost part of the survey area—the only part free of dense ship and boat traffic that elsewhere threat­ ened to destroy the streamer during

A serious issue for marine seismic investiga­ tions in U.S. waters, particularly along the west coast, is the making of loud underwater sounds near marine mammals, which are plentiful in the survey area. Litigation is based primarily on the Marine Mammal Protection Act, which makes harassment of marine mammals illegal."Harass­ ment" is defined as "...any act of pursuit, torment, or annoyance which...has the potential (our emphasis) to disturb a marine mammal or marine mammal stock in the wild by causing disruption of ["many subtle"] behavioral patterns, including, but not limited to, migration, breathing, nursing, breeding, feeding, or shelter­ ing." Individual scientists who violate this act can receive substantial fines and be imprisoned. Eighteen months were needed for SHIPS to enlist the cooperation of environmental groups and the public.A permit from the National Marine Fisheries Service (NMFS) and permission from the U.S. Fish and Wildlife Service (USFWS)

Eos, Vol. 80, No. 2, January 12, 1999 peak to peak) loud enough to damage the hear­ ing of a marine mammal that approached too closely Much opposition to using airguns in con­ fined waterways stemmed from a concern that sound would concentrate to injurious levels wherever water shallowed toward a beach; sound supposedly would concentrate as in a megaphone used backward. However, theoreti­ cal modeling of underwater sound propagation indicates that most sound is transmitted out of the water and into the subsurface [e.g., Jensen andTind/e, 1987],as shown by the blue and green streaks that slope downward and to the right in Figure 4. Acoustic modeling also indicates that sound pressure levels are low in an acoustic shadow zone right along a beach (Figure 4). Using airguns became less contentious once this shadow zone was acknowledged to be a potential refuge for mam­ mals seeking to escape airgun noise. Also, SHIPS personnel agreed not to shoot airguns into nar­ row, dead-ended waterways where marine mam­ mals might become trapped. Instead airguns were shot outbound from a dead end.

and the Canadian Department of Fisheries and Oceans (DFO) took 6 months. Stipulations in the NMFS permit held that safe distances for marine mammals from the airgun array varied from 100 m to 500 m, depending on mammal species.To implement these stipulations, the SHIPS consor­ tium employed six biologists on the Thompson and six more on the Tully.The permit also required airgun operations to cease whenever a marine mammal became stranded or died, com­ mon natural occurrences,so SHIPS arranged for a marine-mammal veterinarian to be on call to determine the cause of any fatality; however, no stranding or fatality occurred during the survey A biologist deployed a calibrated hydrophone near mammals to record airgun-sound levels and to monitor mammal behavior.The sound levels and behavior will be correlated to help establish sound levels to which mammals can be exposed during future airgun surveys. This is important since basic data are lacking. The USFWS agreed that SHIPS would have an insignificant impact on sea otters and marbled murrelets.Although the sea otter is an endangered species in its Alaskan range, quirky law holds that otters transported from Alaska into Washing­ ton State waters are not protected at all.The murrelet, however, is a threatened seabird that is an index species to the health of old-growth forests,so activities that conceivably could affect this bird attract minute scrutiny. Several months before the survey, the Canadian DFO authorized airgun use, but the day before the survey began, DFO raised new concerns. It then required SHIPS to have a Canadian observ­ er on the airgun vessel and directed that airguns be used far from known concentrations of marine mammals.

Discussions With Environmental Groups Early in discussions with environmental groups SHIPS representatives learned to emphasize that a goal of the survey was not earthquake predic­ tion, but rather mitigation'of the impact of a major quake. Many members of these groups believe that earthquake prediction is not realiz­ able; therefore, a survey with this as a main goal is not worth the environmental disruption. This study used an airgun array with a peak-topeak, sound-pressure level (260 dB re 1 uPa-m,

50

100

150

SHIPS contacted 75 scuba shops in the United States and Canada and the local marine-mam­ mal stranding network. Also notified were numerous police departments and 911emergency service districts so that when people felt unusual vibrations they would receive imme­ diate reassurance.A press conference 2 days before the survey alerted the general public to the operations.The U.S. and Canadian Coast Guards published local notices to mariners. The ship received numerous telephone com­ plaints from homeowners around Puget Sound who were concerned that the vibrations they felt would induce landslides. Complaints to the ship peaked when the study was blamed for widely

SHOT NO. 200

250

300

350

—L—

" E " Reflector # | A p p r o x . 60 km

'

IflBtt

-Directarrival

ISllllr

Fig. 3. Expanding-spread profile data from the eastern Strait of Juan de Fuca showing reflections from depths as great as 60 km. The two ships involved in this recording matched their speeds to keep a con­ stant midpoint between source and receiver.

Eos, Vol. 80, No. 2, January 12, 1999 Sound

500 m

Nearshore acoustic

Fig. 4. An acoustic model to illustrate that over a sedimented bottom, underwater sound beams into the subsurface instead of concentrating in the water layer that shallows toward a beach. An acoustic shadow zone occurs right along the beach. Red shows transmission loss less than 40 dB and purple shows losses greater than 110 dB. The sound has a frequency of 50 Hz, which is in the middle of the frequency spectrum of airgun sound. The model was calculated using the program "ram "by M. D. Collins, which is available at ftp://ram.nrl.navy.mil/pub/RAM/. Original color image appears at the back of this volume.

felt shaking from a minor earthquake. Some newspapers speculated that the airgun shots had triggered this quake. In the Georgia Strait, the survey ships encoun­ tered U.S. and Canadian submarines and surface warships that were testing torpedoes across the planned trackline.This forced a change in survey plans, but in general the U.S. Navy coop­ erated closely, allowing airguns to be fired near nuclear submarines docked in Hood Canal and providing an infrared scope to sight marine mammals at night. The survey began in Lake Washington, a large, 60-m-deep, freshwater lake just east of down­ town Seattle (LW in Figure 1), to complement later shooting west of downtown. Members of the media were invited so the public could be reassured about the planned use of airguns. In mid-afternoon, television crews onboard re­ ceived word that dead fish had appeared in the lake, and next morning bold newspaper head­ lines proclaimed that the airguns had caused massive fish kills at three locations in Lake Washington.The efforts spent building bridges to environmental groups averted a crisis when some of these groups calmed citizens who had organized to terminate the survey. Postcruise examination of detailed navigation data reveal­ ed that two of the three dead-fish groups were at least 8 km away from the active airguns when the fish reportedly were killed, and the third group of dead fish had already been dead a long time when we chanced upon them. No dead fish were reported during the remainder of the survey.

Acknowledgments The U.S. Geological Survey supplied major funding for the Seismic Hazards Investigation in the Puget Sound experiment.We thank Steven R. Bohlen and Michael D. Carr for guidance and support. The Canadian Geological Survey con­ tributed salaries for field personnel and ship time on the Tully and Vector.The latter was used in an important supporting role. The U.S. Miner­ als Management Service helped fund research into the effects of airguns on marine mammals. The Integrated Research Institutions for Seismol­ ogy Consortium provided portable seismographs and experienced field personnel. Lt. Kimberly Murphy coordinated our activities with the U.S. Navy's and arranged for an infrared scope. Biolo­ gists from the Cascadia Research Collective kept continuous watch aboard two ships to minimize the environmental impact of the airguns. We thank Capt.Alan McClenaghan of the Thompson and Capt. John Anderson of the Tully and the two ships' crews for outstanding cooperation during this field effort. SHIPS'Web site (http://walrus.wr.usgs.gov/docs/projects/ships) illustrates many aspects of survey planning, field operations, and data processing, and has links to related Web sites.

For more information, contact M.A. Fisher,U.S. Geological Survey, 345 Middlefield Rd.,Menlo Park, CA 94025 USA; E-mail: mfisher@octopus. wr.usgs.gov

References

Atwater, B.,and A. L. Moore, A tsunami about 1000 years ago in Puget Sound,Washington,Science, 258, pp. 1614-1616,1992. Calvert,A.J.,and R.M.Clowes, Deep,high-amplitude reflections from a major shear zone above the sub­ ducting Juan de Fuca plate, Geology, 18, pp. 10911094,1990. Clague,J.J.,and PT.Bobrowsky,Evidence fora large earthquake and tsunami 100-400 years ago on west­ ern Vancouver Island, British Columbia, Quat. Res., 41, pp. 176-184,1994. Flueh,E.,et al., Scientific teams analyze earthquake hazards of the Cascadia subduction zone, Eos Trans. AGU, 78, pp. 153-157,1997. Hyndman, R. D., Dipping seismic reflectors, electrically conductive zones, and trapped water in the crust over a subducting plate, J. Geophys. Res., 93, pp. 13,391-13,405,1988. Jensen,FB.,and C.T.Tindle,Numerical modeling results for mode propagation in a wedge, J.Acoust.Soc.Am., 82, pp. 211-216,1987. Johnson, S.Y., C. J. Potter, and J. M. Armentrout, Origin and evolution of the Seattle fault and Seattle basin, Washington, Geology, 22, pp. 71-74,1994. Pratt,T.L.,S.Johnson,C.Potter,WStephensen,and C. Finn,Seismic reflection images beneath Puget Sound, western Washington state: the Puget Lowland thrust sheet hypothesis, J. Geophys. Res., 102, pp. 27,469-27,489,1997. Authors Satake, K.,K.Shimazake,YTsuji,and K. Uyeda.Time and size of a giant earthquake in Cascadia inferred from M.A. Fisher, T.M. Brocher, R. D. Hyndman, A. M. Japanese tsunami records of January, \700,Nature, 379, pp. 246-249,1996. Trehu, C. S. Weaver, K. C. Creager, R. S. Crosson, T. Symons, N. P, and R. S. Crosson, Seismic velocity struc­ Parsons, A. K. Cooper, D. Mosher, G. Spence, B. C. ture of the Puget Sound region from three-dimension­ Zelt, PT. Hammer, U. ten Brink, TL. Pratt, K. C Miller, al nonlinear tomography, Geophys. Res. Lett., 24,pp. J. R. Childs, G. R. Cochrane, S. Chopra, and R. Walia 2593-2556,1997.