November/December 1998


In October 1997, the U.S. Congress charged the Director of the U.S. Geological Survey to "provide for an assessment of regional seismic monitoring networks in the United States" under Public Law 105-47. While a report is being duly prepared by the USGS, with broad input from the seismological community, some critics may ask, "Do regional seismic networks (RSN's) really have a future? Or are they just a dying branch on the evolutionary tree of observational seismology? Should they functionally be superseded, say, by a combination of widely spaced, super-quality broad-band stations and temporary dense arrays? Or organizationally by a national information outlet?" (Few, if any, would seriously apply these questions to California, but bear with me as I attempt a national overview.)

I believe that RSN's are vital elements both for building a next-generation national seismic system and for directly serving the 75 million people, including 46 million outside of California, who live in metropolitan areas in the U.S. at moderate to high earthquake risk (data from A Plan for Implementing a Real-time Seismic Hazard Warning System, A Report to Congress Required by Public Law 105-47, U.S. Geological Survey, March 1998, Table 1; note that the column total there is in error). Further, I believe RSN's have a key role to play in advancing the science of earthquakes.

The RSN's to which I am referring operate continuously on a scale of hundreds of kilometers in seismically active parts of the U.S., are generally operated by government agencies or universities, and have a multipurpose mission that relates to "earthquake monitoring and rapid emergency response; scientific research; and the acquisition of information required for earthquake hazard and risk analyses as well as for earthquake engineering" (Assessing the Nation's Earthquakes, National Research Council, 1990). Approximately two dozen RSN's in the U.S. fit this definition. From a recent survey by the Council of the National Seismic System (see http://www.cnss.org/NETS), such RSN's now operate about 90 percent of an estimated 1,776 stations in the U.S. with conventional weak-motion seismic instruments (roughly 1,400 short-period and 150 broad-band) and about 15 percent of the nation's 1,388 stations with strong-motion recorders.

For clarification, my use of the term RSN's includes both their physical infrastructure and the organizations that operate them. The infrastructure, in the form of regional elements of an evolving national seismic system, provides essential long-term focus on discrete seismically active parts of the nation. The network operators provide sustained attention to problems of regional and local importance, and their centers serve as "direct outlets for public information and for expert assistance to public policy makers, planners, designers, and safety officials" (National Seismic System Science Plan, U.S. Geol. Survey Circular 1031, 1989). Often overlooked is the important role RSN's play in the graduate education and training of this country's earthquake seismologists.

The history of RSN's in the U.S. can straightforwardly be traced from a proto-Southern California seismic network of Wood-Anderson seismographs in the 1920's to the telemetered networks of chiefly short-period instruments that grew significantly in the 1960's and 1970's. The original scientific motivation was clear--systematically to record and study local earthquakes over broad regions. By the 1980's, the view of RSN's was that their general role was "to delineate the time and space distribution of earthquakes on a fine enough scale to contribute to our scientific understanding of earthquake occurrence and related tectonic processes and to provide important baseline data for engineering investigations ..." (Seismograph Networks: Problems and Outlook for the 1980's, National Research Council, 1983).

Through the 1960's and 1970's, RSN's were a logical extension of observatory seismology, notably in university settings where there was a rich tradition of seismological science. By 1980 it was evident to all that advances in seismology demanded new digital instrumentation for earthquake science. The university community grasped a unique opportunity to move forward in 1984 in forming the Incorporated Research Institutions for Seismology (IRIS). But a fateful decision was made purposely to exclude RSN's as a program component as IRIS set out to promote "a national focus for the development, deployment, and support of modern digital seismic instrumentation"--for a global seismographic network and for array seismic studies of the continental lithosphere.

One legacy of this action, which still affects earthquake seismology in the U.S., is the dynamics of what can be called an IRIS-NSF arena vis-a-vis a USGS-NEHRP arena. In 1984, university RSN's were left squarely in the latter, and the majority of them became doubly disadvantaged, first by years of no-growth or reduced NEHRP funding and second by NSF program directions and influences on funding that (a) effectively handicapped attempts by RSN researchers to get NSF grants for network-related science and (b) generally precluded help to RSN's from NSF in improving instrumentation. Since the mid-1980's, RSN's have been seriously strained to meet the simultaneous challenges of modernization (if not sheer survival), increased expectations to meet practical user needs for public safety and earthquake loss reduction, delivery of all data into an open information system for community use, and scientific research. Present support for university network seismologists to perform many basic RSN functions depends unstably on competition for research funding, separate from operational funding. Some may reason that Darwinian survival of RSN's is appropriate, but this begs the question of how the real needs of populations at risk in the affected regions are to be met. (Recall the 46 million outside of California.)

H. Rishbeth and others, writing in a 1993 article in Eos on long-term solar-terrestrial monitoring, wrote: "Monitoring programs are often undervalued because of a false view that they have little to do with innovative research. In reality, the data derived from long-term monitoring operations often add to new scientific results and underpin other research." In a related vein, R. L. Vogel observed in a recent 1998 issue of Eos that "making data available for interdisciplinary or future uses has not traditionally been a valued activity in the scientific community." He goes on to say, "The information revolution is causing the values of the scientific community to be assessed." In the July/August 1998 issue of Seismological Research Letters, Lowell Whiteside made the point that "[t]he history of science shows that great breakthroughs occur when new tools are combined with available data and innovative scientific thinking." What data will be valuable fifty years from now? Fundamental data for advancing the science of earthquakes surely must include continuous temporal monitoring of key source zones from the microearthquake level (below magnitude 3) to the level of strong motion. Continuity of data, which may become critical to earthquake forecasting or prediction, cannot be provided by temporary dense arrays. And widely spaced broad-band stations cannot yield either a sufficiently low magnitude threshold of space-time sampling or data needed for advancing our understanding of main-shock ground motions in extensional and intraplate regions.

The importance of continuously monitoring specific seismic source zones in detail sensibly varies according to hazard and risk, but balanced nationwide attention is essential because risk is markedly increasing in many regions. Consider Utah's densely populated Wasatch Front area, where more than three quarters of the state's population and economy literally sits astride the most active segments of the Wasatch Fault. Population in this dramatically growing megalopolis is projected to increase from a 1995 base of 1.6 million to 2.7 million by 2020 and to 5 million by 2050. Many other areas warrant similar concern, such as San Juan, Puerto Rico, and Las Vegas, Nevada, the fastest growing metropolis in the U.S.

My experience tells me that--like politics and history--earthquake problems are ultimately "local", and the social dynamics vary widely. How user needs are defined and served in California may not be a desirable or cost-effective model for Alaska, Montana, or New England. RSN's have played an important role historically, especially in regions of moderate to low seismicity, in creating public awareness of earthquake dangers and in marshaling much of the relevant information to deal with those dangers. This special attunement to region-specific earthquake problems and social dynamics leads me to believe that RSN's have a unique potential in the seismological community to meet long-term user needs, on a region-by-region basis, for earthquake data, information (what users want or need to know), and related services. Anyone who has ever worked in a regional network center understands that RSN's are where "the rubber hits the road" when it comes to "societally relevant" earthquake seismology.

Where do we go from here? I see at least three possibilities for major progress in dealing with the issues I have raised: (1) creation of a next-generation national seismic system, (2) securing congressional line-item funding for ongoing operation of such a system, and (3) promoting better synergy between the USGS-NEHRP arena and the IRIS-NSF arena. Since 1993 there has been clear progress in creating what I will call a first-generation national seismic system by integrating RSN's with the U.S. National Seismograph Network. This system is inherently cooperative, multijurisdictional, and with resources unevenly spread. In my view, the challenge is to take a systems engineering approach to create the physical and informational infrastructure of a new technologically advanced national seismic system, consisting of national, regional, and local elements. The process now underway of writing the report to Congress on seismic monitoring in the U.S. that I referred to in my first paragraph has brought seismologists and earthquake engineers together in an unprecedented way to create a vision of such a next-generation national seismic system. The vision combines integrated seismographic monitoring on national, regional, and local scales with strong-motion recording and structural-response monitoring focused on urban areas at risk. It emphasizes services and information products, including rapid and timely information for emergency management. The evolution of real-time earthquake information systems and the great practical value of rapidly broadcasting maps of actual ground shaking have created a powerful new framework for unifying strong- and weak-motion recording for emergency management, engineering, and science.

In the report Assessing the Nation's Earthquakes (National Research Council, 1990), the foremost recommendation was that "[t]he federal government should establish a more rational, coordinated, and stable means of support for the seismic networks of the United States ...." The most effective way to do this, it seems to me, would be through congressional line-item funding for a national seismic system of the type described above. For decades, the existence of infrastructure for fine-scale regional monitoring in the U.S. has depended on the sacrifices and career-long dedication of key individuals. I do not think this framework is sustainable. Nor do I think this burden should be perpetuated for future generations of seismologists. Decades of struggling to establish stable support for national and regional seismic monitoring, as well as strong-motion recording, could be decisively resolved by persuading Congress to recognize the fundamental importance of a national seismic system and to finance it under a line-item appropriation.

Finally, what about synergy between the USGS-NEHRP and IRIS-NSF arenas? The seismological community is not immune from internal tensions, some of which are implicit, for example, in Tom Jordan's opinion piece in the January/February 1997 issue of Seismological Research Letters, in which he took a balanced look at the value of both applied and basic science in the study of earthquakes. Perhaps a timely lesson can be drawn from well known observations of hemispheric differences--and conflicts--in the human brain. We know that both sides of the brain process the same information in different ways, that they are motivated differently, and that one side tends to seek dominance and inhibit the other. In art and writing, the solution to dealing with these tensions between the left-brain and right-brain lies in recognizing the special capabilities of each side and in finessing a complementary division of labor. (See, for example, Gabriele Lusser Rico's book, Writing the Natural Way, or Betty Edwards' classic text on rawing on the Right Side of the Brain.) In helping the nation deal with its earthquake problems, are there special capabilities on each side and possibilities for a complementary division of labor between the USGS-NEHRP arena and the IRIS-NSF arena? I think any outside observer can readily see them. Can we?

Peter Bowen, a contemporary Montana writer, observed: "The cowboy won't ever be replaced--no mere machine could ever stand the abuse." Will RSN's be replaced? Part of me rushes to say, "Not soon--for the same reason as cowboys." Another part of me, seeing the vision of an evolving national seismic system, adds, "No, RSN's have a unique role to play--and better things are coming."

Walter Arabasz
Department of Geology and Geophysics
University of Utah

To send a letter to the editor regarding this opinion or to write your own opinion, contact Editor John Ebel by email or telephone him at (617) 552-8300.

Posted: 19 November 1998
URL's updated: 21 January 2003