Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Contributions to Date of Regional Seismic Networks Most regional seismic networks currently in operation in this country have been sited to monitor active seismic zones. Because they consist of multiple sensors distributed over relatively small areas, they are essentially telescopes focused downward into the earth to "see" the seismic source. Such networks have been in operation for only about two decades but have made extensive contributions to our knowledge of the spatial, temporal, and physical characteristics of earthquake occurrences. Heaton et al. (1989) recently reviewed these contributions and discussed the future of networks in the context of a science plan for a National Seismic System. Briefly, the contributions include the improved detection and more accurate location of earthquakes, especially those of lower energy levels; greater precision in focal depth determinations; enhanced monitoring of seismic energy release as a function of space and time; refined determinations of the attenuation characteristics of seismic waves; three-dimensional descriptions of the seismic velocity structure of the interior of the earth; and more reliable specification of the earthquake faulting process. Thus the fundamental contributions from seismic networks are intrinsically observational, and these observational data make possible a wide range of derived contributions that are of direct benefit to both science and society. Recent examples of such contributions with direct societal benefits are described in two earthquake case studies in Chapter 4. Other examples include contributions from networks associated with active volcanoes such as Mount St. Helens in Washington and Kilauea in Hawaii. The networks there track the subsurface motions of magma bodies and their associated 1 1
12 ASSESSING THE NATION'S EARTHQUAKES mechanical deformation and thereby provide invaluable data to emergency preparedness agencies. Similar contributions are made when damaging earthquakes occur in populated areas, such as happened in one of the case study events, the Whittier Narrows shock in the Los Angeles basin. In addition, data from regional seismic networks are essential to the safe siting of nuclear and other hazardous waste repositories as well as large, unique engineering structures such as the proposed Superconducting Super Collider. Siting such structures safely requires an already-developed adequate seismicity data base; once a site has been proposed, it is not possible to wait for data to be gathered. Seismic networks provide a major contribution to the estimation of U.S. seismic hazards, which vary greatly across the nation: seismicity is highest on the West Coast, but 37 states are in the two highest (out of four) risk zones. The current federal National Earthquake Hazard Reduction Program (NEHRP) recognizes this pervasive threat and seeks to mitigate it. The program was created by the Earthquake Hazards Reduction Act of 1977; its principal agencies are the Federal Emergency Management Agency, the U.S. Geological Survey, the National Science Foundation, and the National Institute of Standards and Technology. One of the major NEHRP elements is "hazards delineation and assessment" (FEMA, 1988~. In particular, the estimation of seismic hazard requires as input (1) spatial definitions of the seismic source zones (especially important is the accurate definition of currently active geologic structures as well as their seismotectonic host environment, e.g., the thickness of the crustal seismogenic zone); (2) rates of occurrence of earthquakes of various magnitudes for each zone; and (3) ground motion estimation for seismic vibrations from earthquakes of varying magnitudes and at varying distances. Clearly, only the highly accurate data from dense regional seismic networks that are dedicated to the investigation of specific seismic zones or regions can provide adequately for such specific requirements. This is especially true for the eastern United States, where the seismic station density before networks were established in the 1960s and 1970s was lower than one per state. It is important to reemphasize that the required input data from the regional networks cannot be obtained as the need arises for their use; rather they must be obtained before such needs arise. It is also important to note that the determination of seismic riski.e., the esti- mation of probable consequences of earthquakes for life and property- depends directly on the availability of accurate seismic hazard estimations, which in turn are based largely on data from regional networks. The technological growth of industry in this country, in concert with increased land use during the past several decades, has resulted in a dramatic increase in the elements of society at risk from earthquakes. Engineers have constructed larger and more complex facilities, such as long bridges, high dams, high-rise buildings, nuclear reactors, large computer centers,
CONTRIBUTIONS OF REGIONAL SEISMIC NETWORKS 13 offshore drilling platforms, and concentrations of high-technology industry. These and other critical facilities are often sited in areas of high population density that are also earthquake-prone, e.g., the computer chip industry in California. In addition to estimating the seismic hazard for such facilities, it is also necessary to thoroughly evaluate the probable responses of the structures themselves to seismic disturbances. Such studies are based directly on the best possible estimations of the amplitudes and frequencies of ground motions from both moderate and large earthquakes at distances ranging from nearby to regional. Seismic networks, especially those that include strong-moiion seismographs sited in He structures themselves as well as in the free field, are the only source of the input data required for the necessary estimations (Committee on Seismology, 1980~. Clearly, a lack of such monitoring efforts exposes our society to increasingly unacceptable and unspecified risks from future earthquakes. One of the current frontiers of research in seismology involves the pre- diction of earthquakes. For example, the U.S. Geological Survey has predicted that a magnitude* 6 earthquake will occur on the Parkfield section of the San Andreas fault in 1988 +5 years. In general, however, the present stage of development of this research field is such that estimations of future earthquake occurrences are generally derived in more probabilistic terms and are based on detailed analyses of the spatial and temporal patterns of earthquake activity in the forecast area. Both probabilistic and determinis- tic analyses include the recognition of (1) "seismic gaps," i.e., locales that are known from prior activity to be earthquake-prone but currently are seismically quiescent, (2) repetitive "characteristic" earthquakes from a given fault segment, and (3) "slip-deficient" fault segments. Only the resolving power of the inward-looking regional seismic network "telescope" can provide data of adequate precision, detail, and completeness to satisfy the requirements of this most difficult and demanding seismological taskthat of predicting earthquakes in a quantitative manner. However, the benefits to society that would result from this ability are so enormous that we must continue these efforts. The dense spatial coverage provided by regional seismic networks has been directly exploited in recent studies of crustal velocity structure. Some of these studies are similar in concept to computer-assisted X-ray tomography, the CAT-scan in medical technology, which yields three-dimensional, com- puter-generated "images" of the interior of a body without directly accessing the volume being investigated. For example, Hearn and Clayton (1986) have presented detailed images of lateral variations in the shallow crustal *"Magnitude" as used throughout this report is a generic term for the relative size of the earthquakes discussed. The term may refer variously to a body wave, surface wave, moment, leg, local, or Richter scale magnitude.
14 ASSESSING THE NATION'S EARTHQUAKES velocity structure in southern California that they obtained from data de- nved from the seismic network stations located there. These velocity variations are associated with surface tectonic features such as the San Andreas fault. Also in southern California, Humphreys et al. (1984) studied the deeper mantle structure beneath the Transverse Ranges to image a spectacular, high-velocity tabular root extending several hundred kilometers into the mantle (see also Heaton et al., 1989~. In the Midwest, Al-Shukn and Mitchell (1988) mapped a three-dimensional pattern of low velocities in the crustal rocks of the active portions of the New Madrid fault system in southeastern Missouri. The seismic velocities there are lowest in those subsurface volumes exhibiting the greatest concentration of earthquake activity. The observed several percent decrease in compressional wave velocity is consistent with a source zone containing fluid-filled cracks. Studies such as the three mentioned here have led to a markedly improved understanding of the physics and architecture of the earth's crust. Again, the many stations of the regional seismic networks are required to achieve the detail and resolution necessary to accomplish such CAT-scans of the earth. When large fault motions occur on the floors of oceans, they produce not only earthquake vibrations but also energetic water waves, called tsunamis, that travel across the oceans and run up on distant coastlines. Between 500,000 and 1 million residents along the coastlines of Hawaii, California, Oregon, Washington, Alaska, and the U.S. Pacific Territories are at risk from these rare but devastating waves. For example, the 1964 Alaskan earthquake (magnitude 9.2) generated a tsunami that caused 122 fatalities in Alaska, California, and Oregon and $100 million in damage in Alaska, Hawaii, California, Oregon, and Canada. Tsunamis are predominantly, but not exclusively, a Pacific hazard: in the Atlantic Ocean in 1929, the Grand Banks earthquake off the coast of Newfoundland (magnitude 7.2) also resulted in damage and fatalities (Committee on Seismology, 1980; Lander and Lockridge, 1989~. Additionally, submarine facilities, such as communications cables, are at risk from these earthquakes as well as from submarine landslides triggered by earthquakes. The Pacific Tsunami Warning Center at Honolulu, Hawaii, is an international cooperative effort to provide tsunami watches and warnings to the Pacific region. Onshore regional seismic networks contribute to the detection and location of submarine earthquakes that are potentially tsunamigenic. Needed, but not currently in place, are networks of ocean-bottom seismographs on U.S. continental shelves to improve detection and location capabilities in those nearshore areas. The combined onshore and ocean-bottom seismic networks would allow for a more rapid determination of focal mechanism and thus a more reliable assessment of the tsunami- generating potential of shallow offshore events. Earthquakes are common in volcanic areas, and seismic networks are the fundamental tool for their study. Data from networks have shown that
CONTRIBUTIONS OF REGIONAL SEISMIC NETWORKS 15 "volcanic earthquakes," those that result from the thermal and mechanical forces of volcanic processes (volcanic a-type, high-frequency earthquakes), are indistinguishable from tectonic earthquakes, which result from the mechanical fracturing of rock due to tectonic forces. Other volcanic earthquakes (vol- canic lo-type, low-frequency events) and harmonic tremor (vibrations due to the shallow movements of magma) have distinctive properties. For example, studies at Mount St. Helens indicate that harmonic tremor there consists of a persistent sequence of lo-type earthquakes. Studies in Hawaii and Alaska have resulted in the development of new models of the sources for the volcanic shocks that include reverberations within the magma body triggered by brittle failure of the adjacent rock as well as a point-force reaction to an explosive eruption. The swarm-like series of magnitude 5.5-6.0 earthquakes that occurred in 1978 near the Long Valley caldera in eastern California raises the possibility of yet another type of volcanic earthquake, one due either to the abrupt injection of magma into a dike or to a complex shear failure on fault planes of differing orientations (Hill, 1987~. Clearly, much work remains to be done to understand what the various types of volcanic earthquakes imply about the volcanic processes that affect the westernmost states. The core of the earth has long held a particular fascination and position of importance because of its inaccessibility and because it is the source of the earth's magnetic field. Regional networks, when integrated within a continent-wide National Seismic System, can contribute to its study. Recent studies of the structure of the core and of its boundary with the mantle using compressional waves that penetrate through the deep interior of the earth suggest considerable complexity that could have important geodynamical and geochemical consequences. It appears that topography of +8 km or so may be present on the core-mantle boundary. Establishing whether that boundary is thermal or chemical in nature is important for thermal modeling of the earth's interior. Also, although the velocity gradients in the outermost core appear not to be anomalous (as was once thought), and although the inner core-outer core boundary may indeed be a simple discontinuity, the first-generation three-dimensional core models indicate greater, not less, complexity for core structure (Lay, 1987~. The rapid progress made in imaging these most inaccessible regions testifies to the benefits that can be reaped from the high-quality data derivable from the larger regional and global networks. Finally, the importance of seismological facilities for education deserves emphasis. This includes not only the training of the nation's seismologists but also the general education of a broad student population. Terminating funding for some seismic networks will cause a certain number of research- ers to seek new avenues of funding in more adequately supported areas of research. Once these scientists are lost to other research fields, they cannot
16 ASSESSING THE NATION'S EARTHQUAKES easily be reclaimed for seismic network studies even if funding priorities change. Thus, given the small number of network seismologists to begin with, a short-term reduction of support will have long-lasting consequences. Students at universities that operate regional seismic networks unques- tionably have an enhanced educational experience. The incoming digital data stream from multiple sensors provides hands-on opportunities to apply and develop the seismological theories developed in the lecture hall and the laboratory. Not only can near-real-time analyses be performed, but the presence of a continually expanding archival digital data base also permits a full range of thesis and dissertation investigations. The day-by-day, real- time acquisition of seismic data provides an earth surveillance setting and format that are particularly dynamic and impart to students an excitement about earth processes that often lasts a lifetime. In summary, regional seismic networks have made fundamental contribu- tions to the estimation of national seismic hazards and strong earthquake ground motions, the prediction and forecasting of earthquakes, the specification of the three-dimensional internal structure of the earth, the surveillance for tsunamis, the study of volcanic earthquakes, and the training of students. Such worthwhile efforts should be continued and enhanced.
