U.S. Geological Survey Open-File Report 96-260

Scott: Would you discuss the development of policies and programs aimed at reducing the disastrous effects of earthquakes? You participated in many of those developments, and I would appreciate your giving your recollections of some of the things you consider most relevant.

Wallace: Yes. I think it is important to record how some of those developments came about. Also, some of the things I can talk about represent important general tendencies, even generic lessons. We should not lose sight of those. I see younger members of the USGS staff who seem to think that by some magical process earthquake issues and geologic hazards were always on national agendas. They seem to think that the justification for their work has always been clear and well-understood. Nothing could be further from the truth.

In describing the process as I saw it, I would first like to review a trail of selected activities and documents that led to the Earthquake Hazard Reduction Act of 1977. It took some thirteen years of struggle, debate, and in-fighting to formulate and enact a National Earthquake Hazards Reduction Program (NEHRP). These battles--and they were battles--took large investments in manpower, time and funds, which were diverted from other programs in anticipation of future authorizations and augmented funding for earthquake programs. The process, of course, is never-ending, and the balance of program elements changes and evolves. But that major step of getting the 1977 Act in place was a big victory.

Scott: Most of us agree that the great Alaskan earthquake of 1964 started the modern era of earthquake awareness in the United States, and later other serious earthquakes emphasized the danger to life and the seriousness of economic consequences.

Wallace: Yes. In the early 1960s, before the Alaskan earthquake, several of us here at USGS discussed new big programs we should begin to push. One idea was a complete study of the San Andreas fault. We knew the fault had had big displacements, events that created earthquakes, and, thus, major public policy issues were involved. So in 1961 or 1962, several of us started pushing that idea. We proposed about a $2 million program with geology, geophysics, and allied topics, and sent it off to headquarters in Washington, D. C. for consideration. It got nowhere.

In seeking funds, the management at USGS headquarters had to make judgments on what could be "sold" to the Department of the Interior, and to Congress. They decided that our proposal would not sell, or perhaps they just did not want to put it high on the priority list. Also, we probably did not present it properly. But we made a noble effort back in 1961 and 1962. Parke Snavely, Jr., and I did most of the work on the proposal, and Earl Brabb was also involved. But we got nowhere.

Scott: But even before 1964 there had been some significant actions regarding earthquake hazards.

Wallace: Oh, yes indeed, there were spectacular accomplishments from the late 1880s on. Much of that early history is nicely summarized in publication that you and Bob Olson prepared. (Scott, Stanley and Olson, Robert. A., (ed.) California's Earthquake Safety Policy; A Twentieth Anniversary Retrospective, 1969 - 1989, Earthquake Engineering Research Center, University of California, Berkeley, 1993.)

Within the USGS, for example, G. K. Gilbert in 1884 wrote a paper which was published in Science titled "A Theory of the Earthquakes in the Great Basin, With a Practical Application." Also in a 1909 issue of Science, Gilbert published a paper titled "Earthquake Forecasting," the first reference I know of that has "forecasting" or "prediction" in the title. An early USGS publication pertaining to earthquakes was by C.E. Dutton. It appeared in 1889, reporting his findings on the Charleston earthquake of 1886.

Scott: In your view, was USGS pretty well-positioned to take a lead role in the more recent developments?

Wallace: By tradition, yes. But in fact, as late as the 1950s only a handful of investigators were active, and only meager funds were available for seismologic, geologic and engineering studies.

Wallace: The great Alaskan earthquake on Good Friday 1964 was a major turning point and a trigger for new programs. To my thinking, and most people's thinking, I believe, the Alaskan earthquake was the beginning of, and stimulus for, our whole modern earthquake program. That earthquake showed what great seismic events could do very close to home.

Scott: That earthquake did generate a lot of concern. Did you get involved in the Alaskan earthquake studies?

Wallace: I did not go to Alaska. That year I was trying to get out of branch chiefing--I was not cut out for administration. The Alaskan earthquake happened in March, but I was still deeply involved in Nevada, Utah and Arizona, and could not even get away to go to Alaska. I probably should have just gone, but I could not get away because of my other responsibilities.

Although I did not go to Alaska, I was involved in a lot of discussion about what should be done as a result of the Alaskan earthquake. Dozens of other USGS people did participate, however, and a wealth of information came out of the Survey's efforts. The USGS produced twenty-eight volumes in the USGS Professional Paper series, plus numerous others in the USGS Circular series and outside journals. I shall fully reference only two as examples. (Hansen, W.R., Eckel, E.B., Schaem, W.E., Lyle, R.E., George, Warren, and Chance Genie, The Alaskan Earthquake March 27, 1964: Field Investigations and Reconstruction Effort, U.S. Geological Survey Professional Paper 541, 1966. Plafker, George, "Tectonic Deformation Associated with the 1964 Alaska Earthquake," Science, v.148, n.3678, 1965, pp.1675-1687.)

Scott: Could you give us an overview of some of the more important other reports that came out, in addition to the USGS reports?

Wallace: Two types of reports began to emerge. The first concerned reporting on the earthquake itself, the second were reports by committees established to recommend future federal responses to the earthquake hazard. One of the first big reports on the earthquake itself came from The National Academy of Science, which established a committee to do that, chaired by Konrad B. Krauskopf of Stanford. A 596-page report was prepared and published based on the work of 48 contributors, including several from the USGS.

The U.S. Coast and Geodetic Survey, an agency under the Environmental Science Services Administration in the Department of Commerce also put out a three-volume report on the earthquake. This great federal effort to investigate the Alaska earthquake led to follow-up proposals for federal actions. But then it took years of long and arduous debate over what elements should be included and what the funding levels should be. As I noted, eventually the National Earthquake Hazards Reduction Act of 1977 was enacted.

Scott: Review some of these other reports and recommendations, and if possible, say something about what you saw as the effects of each.

Wallace: In September 1965 an ad hoc panel on earthquake prediction chaired by Frank Press issued a report, which was prepared for the Office of Science and Technology. Press was a prominent seismologist at the Massachusetts Institute of Technology, as well as one of the most skillful scientist-politicians to come along. Press later served as Special Assistant to the President for Science and Technology in 1979-1981, during the Carter Administration, and as President of the National Academy of Sciences (1981-1993). After retiring from the presidency of the National Academy of Sciences, Press moved to the Carnegie Institute of Washington, D.C.

The Press report was slanted toward earthquake prediction, and used the term in its title. (Ad Hoc Panel on Earthquake Prediction, Press, Frank, (ch.), Earthquake Prediction: A Proposal for a Ten-Year Program of Research, Administrative Report to the Office of Science and Technology, Washington, D. C., September, 1965.) But many people did not think earthquake prediction was a reasonable goal. Charles Richter, after whom the Richter Scale was named, characterized as "charlatans" those who claimed that prediction was plausible.

Moreover the "ten years" in the title led to a misperception by some, especially the media, that scientists were promising to predict earthquakes routinely within ten years. That misunderstanding has plagued us ever since 1965. We heard the recurring comment: "Well, you said you would be predicting earthquakes in 10 years." To my knowledge, no responsible scientist ever said that in print. Admittedly, however, there were periods of great optimism and euphoria, when many scientists thought prediction breakthroughs were just around the corner.

Scott: The "ten years" probably referred to the proposed research program. That is, the program was to continue for ten years.

Wallace: Yes, that is correct. The Frank Press report anticipated a lot of the principles, such as the "gap theory" of earthquake succession, with maps of some areas, showing recent earthquakes, and gaps where others might be expected in the next couple of decades. The gap theory became very popular and still is a reasonable idea.

Scott: My impression is that the gap theory is seen as a good deal more than a reasonable idea. I have heard Bruce Bolt and some others refer to it very favorably, even quite recently.

Wallace: Yes. The Press report had diagrams showing how we should test strain at depth using deep wells, but only one deep well has ever been drilled. It was located at Cajon Pass, and was stopped long before completion. Now people are again looking at the idea of deep-hole drilling along the San Andreas fault. The Press report also proposed a full array of instruments to measure crustal changes, including seismometers, magnetometers, and strain meters.

That gives a pretty good picture of the Press report, an exceptionally good report that was very influential. A lot of what it recommended has come to pass.

Scott: I have long been bemused by the prediction discussion and debate. Some of the controversy hinged around the question of how quickly prediction would be achieved, and what degree of precision might be hoped for in location and timing of predicted events. Another issue was whether prediction would be possible generally, or only on certain special faults.

There seemed to be a lot of differences in interpretation of what prediction is, as well as its "validity" as a major research goal. Maybe you could start by defining what you see as constituting an earthquake prediction?

Wallace: In 1976 a panel of the National Academy of Sciences, chaired by Clarence Allen, defined earthquake prediction as a statement of magnitude range, geographic area, and time interval of occurrence. A confidence level was to be given for each prediction. (National Academy of Sciences, Panel on Earthquake Prediction, Allen, C.R., ch., Predicting Earthquakes; A Scientific and Technical Evaluation--With Implications for Society: National Academy Press, 1976.)

In the early 1980s, three of us, Karen McNally, Jim Davis and I, who were serving on an advisory panel to the Southern California Earthquake Preparedness Project (SCEPP), found that planners needed more specific statements of the time element to help design hazard mitigation plans. At the time, the presumption, incorrectly, had grown that routine predictions would be forthcoming at any time, and, naturally public officials wanted to have response plans ready on the shelf.

We three prepared a report for SCEPP which defined more precisely what short, intermediate and long-term predictions were and how they all differed from the more common general statements such as "an earthquake is coming", or that "California is earthquake country." Planners made good use of it. Los Angeles, for example, designed rather different mitigation response plans for short-term predictions than were to be taken for long-term predictions. (Wallace, R.E., Davis, J.F., and McNally, K.C., "Terms for Expressing Earthquake Potential, Prediction and Probability," Bulletin of the Seismological Society of America, v.74, n.5, 1984, pp.1819-1825.)

I would like to review those controversies, but it is pretty complicated. There have always been strong and quite honest differences of opinion about where emphasis should be placed in the earthquake hazard reduction program. Over the years I have talked about three lines of defense against earthquakes.

Scott: What do you mean by "three lines of defense"?

Wallace: The first line of defense is earthquake engineering, that is, building earthquake-resistant structures. Over and over it is said, "The collapse of buildings is what kills." The Long Beach earthquake of 1933 triggered many innovations in earthquake engineering.

The second line of defense--which is just now beginning to be used effectively--relates to the prudent use of land--not building on active faults or areas susceptible to landsliding or earth movement. California's 1972 Alquist-Priolo Special-Study Act requires special examinations of areas near active faults before major structures are put up. Such an act was conceivable only after the USGS had produced several rather detailed strip maps showing locations of the most-recently active strands of the San Andreas fault.

Furthermore, some of us were finding evidence that movement on a given strand of the fault tended to occur again and again in the same place. Consequently, avoidance of those strands seemed imperative. Previously, until about 1960, we had no proof that movement on one narrow strand of the fault was more likely to repeat, rather than shift to other parts of a mile-wide band of highly sheared rock found along the fault zone. Of course, we all knew that it took repeated movements to create the broad fault zone itself. In 1967, however, I published a note reporting repeated movement on some strands along the San Andreas fault in the Carrizo Plain of central California. Since then, thousands of examples have been found.

The third line of defense should be prediction, but short-term and intermediate-term prediction are still in the research phase, and still to be realized, but long-term prediction (forecasting) is with us already. Many structures around the world do not conform even to minimum standards of engineering practice. Prediction, when achieved, could help us know when to evacuate collapse-hazard structures for short periods of time. In less-developed countries many communities cannot afford sophisticated engineering methods nor do they have suitable natural materials nearby with which to build resistant structures. That is a story in itself.

Wallace: Stan, you referred earlier to some skepticism about prediction--and, indeed, prediction is not here yet, except in this broad way of probabilistic forecasting. For example, the USGS estimates a probability of about 67 percent that a magnitude 7 earthquake will occur within 30 years on one of the faults in the San Francisco Bay area. We can also do pretty much the same thing in southern California, where the chance is about 80 percent in 30 years, I believe, but we cannot yet do it most places in the world. We are starting a whole new process here in California. It is exciting. (Working Group on California Earthquake Probabilities, Probabilities of Large Earthquakes in the San Francisco Bay Region, U.S. Geological Survey Circular 1053, 1990.)

We must, however, be wary of those who say, "The less data you have, the more you need probability." The weakest part of the probabilistic approach is the tendency to give a definite numerical answer, despite the enormous unknowns. We must continue to try to understand the fundamentals of how the earth works, and not settle for quick answers that may be misleading or meaningless.

I believe that for a long time, we won't be able to predict the hour, day and minute, although there were times when some of us thought that sort of prediction was just around the corner. In earthquake prediction there have been periods of excitement that; "By golly, we might be at the threshold of a technique for making short-term predictions."

Scott: Discuss some of the techniques that have seemed promising.

Wallace: In the late 60s, for example, the Soviets told us about Vp/Vs, a ratio of the velocities of the "P" wave and the "S" wave. They had beautiful diagrams showing how this ratio changes before a big earthquake. In 1969 Karl Steinbrugge led a group to the Soviet Union, and I was privileged to go along. I photographed these diagrams that were on the walls of Soviet labs, and we were really excited--"Here we are!"

The U.S. picked up the technique and followed it. Some people thought, in retrospect, that the 1971 San Fernando earthquake could have been predicted by this technique. US scientists developed theories about why this ratio change should happen. So there was euphoria in the prediction world for a couple of years after 1970. But it sort of fell apart. It did not prove out, although I think one small earthquake was reasonably predicted in the Adirondack Mountain region by the Lamont group using this technique.

About then the Chinese were predicting earthquakes by observations of animal behavior. They made a strong case that they had predicted earthquakes by combining animal behavior and what they called macroscopic events; observations of changes in water wells, gas wells that had exploded and started burning, and the like.

Scott: Those seem like rather different things--predictions based on large-scale events and predictions based on animal behavior.

Wallace: They are, but they were all woven into a mish-mash in the Chinese literature. In 1978 Ta-Liang ("Leon") Teng and I were with a delegation to China headed by George Housner. Leon and I represented earth science on an otherwise engineering-oriented mission. At one point, Leon and I left the engineering discussions and interviewed the Chinese about their prediction program, focussing on animal behavior and their so-called macroscopic precursors. Having been born in China, Leon is fluent in the language, including several dialects, so the interviews were especially productive, and Leon and I prepared a report on the subject. (Wallace, R.E., and Teng, Ta-Liang, 1980, "Prediction of the Sungpan-Pingwu Earthquakes," Bulletin of the Seismological Society of America, v.70, n.4, August 1976, pp. 1199-1233.)

Scott: Do you have any other examples of prediction?

Wallace: One of the most exciting and significant was the prediction by the Chinese of the great Haicheng earthquake of 1975. A special group headed by Barry Raleigh of the USGS went over from here to try to learn how the prediction had been accomplished. Apparently that prediction had involved a wholly different technique. There had been many, many foreshocks in the few days before the quake.

Scott: That is the earthquake where they warned people in advance to go outside.

Wallace: Yes, they claimed they saved more than 100,000 lives with that prediction and warning. The people got out of their houses into the open spaces. (Haicheng Earthquake Study Delegation, Barry Raleigh, ch., "Prediction of the Haicheng Earthquake," EOS, Transactions of the American Geophysical Union, v. 58, n. 5, 1978, pp.236-272.)

Scott: The literature on earthquake prediction often refers to the "probabilistic approach" and the "deterministic approach." You also see that discussed in earthquake engineering generally, as well as in many other fields. Would you digress a moment to discuss how you see those approaches?

Wallace: Yes. During the past several years a push has been made to develop quantitative means of expressing the future likelihood of strong ground motions and other hazards related to earthquakes. A major driving force has been the needs of the nuclear power industry to develop plants that were safe, and could be demonstrated to be safe. Unless the general public, public administrators, and politicians could be satisfied that nuclear plants could be made acceptably safe from earthquakes, nuclear power would be a questionable source for the ever-increasing demand for power. Back in the era when many saw nuclear power as the wave of the future, a lot of effort was put into geologic and seismic studies, and into methods of analyzing risk.

Both deterministic and probabilistic methods of seismic hazard analysis try to quantify the earthquake hazard so as to satisfy a variety of needs, ranging from figuring the kinds and amounts of reinforcing used in concrete, to estimating the relative value of different regional planning methods and emergency response techniques. Without such values, designers, decisionmakers and others concerned with reducing the hazards have no common starting point.

In 1988 the National Research Council of the National Academy of Sciences issued a key publication on the topic of seismic hazard analysis. Keiiti Aki served as chairman of the panel of ten, including me, to review the state of the science. (National Research Council, Panel on Seismic Hazard Analysis, Aki, Keiiti, Chairman, Probabilistic Seismic Hazard Analysis, National Academy Press, 1988)

This review recognized five types of seismic hazard analyses, including "deterministic seismic hazard analysis" (DSHA), single- to multiple-model "probabilistic seismic hazard analyses" (PSHA), and hybrid procedures. Tom Hanks and Allin Cornell presented the idea that PSHA and DSHA have far more in common than they have differences. The main difference is that PSHA considers time. (Hanks, T. C., and Cornell, C.A., "Probabilistic Seismic Hazard Analysis: A Beginner's Guide," Proceedings of the Fifth Symposium on

Current Issues Related to Nuclear Power Plant Structures, Equipment and Piping, Center for Nuclear Power Structures, Equipment and Piping, North Carolina University, 1994, pp. I/1-1-I/2-17.)

Advocates of the probabilistic approach champion its use in engineering and planning, whereas the deterministic advocates decry the probabilistic approach. I believe that I was invited to be on the Aki panel because, as a geologist, I was considered able to help balance what many saw as a band-wagon charge to standardize on a probabilistic approach.

After the Aki report was published, I was sorely criticized by a good friend and staunch deterministic advocate, Ellis Krinitzsky, who insisted that I had sold out to the probability advocates. (Krinitzsky, E.L., 1993, "The Hazard of Using Probabilistic Seismic Hazard Analysis," Civil Engineering, November 1993, p.1-52.) In fact, however, I embraced neither the PSHA nor the DSHA approach entirely, but believe that a combination will most likely serve best. I do advocate the need for much more effort on learning more precisely how the earth works, so that the data, rather than the mathematical process, will control the outcome. The mathematical process, in principle, quantifies the exercise, but too often tends to obfuscate it.

My first exposure to PSHS had been several years before the Aki committee met, and it horrified me. I saw probability used to create numbers, where basic data on earthquake process was scant or lacking. A mathematical friend of mine actually said, "The less data you have, the more you need probability." In my opinion such a philosophy must be terribly wrong. The process could be another example of what I call "implementing ignorance"--that is, give the client a number, even though based on incomplete or inaccurate basic data.

Scott: Yes. While your math friend was right about having to rely more on probability when you lack adequate information, there is a serious down-side. Relying more on probability than on good information can give some exceedingly questionable results.

Wallace: Yes. But of course many similar types of inadequacies go along with the deterministic approach, which is generally molded into a final report by using "engineering and geologic judgement." Here again, the lack of data on how earthquakes actually operate may overwhelm the whole process. In short, both the probabilistic and deterministic approaches have to operate "out in the blue" to the extent that they lack good data and a firm understanding of earthquake processes.

Fortunately, the wealth of data has grown enormously in the past several decades, and the geologist who has followed this development may well have the best intuitive grasp of the hazards. On the other hand you probably will not find two geologists--or two seismologist--in full agreement. So where does this all leave the decision-maker, who probably has little personal background in the subject?

Wallace: The Frank Press committee report on earthquake prediction, which I discussed earlier, apparently made many engineers furious. The engineering community did not consider it the right approach for a national program to counter the effects of great earthquakes. The National Academy of Engineering, through a group of engineers chaired by George Housner, developed a committee on earthquake engineering research just to prepare an alternate agenda. This was under the Division of Engineering, National Research Council. Before this committee was created, I believe they had no formal group considering earthquake engineering, but I may be wrong on that point.

The Committee on Earthquake Engineering Research, Division of Engineering, National Research Council, National Academy of Engineering, with George Housner as chairman, prepared a report to the National Science Foundation. Their excellent report developed the engineering viewpoint on what should be done to reduce earthquake hazards. The report represented a major difference in emphasis between (1) earthquake prediction, seismology and earth science, on one hand, and (2) earthquake engineering, on the other. (National Research Council, National Academy of Engineering, Committee on Earthquake Engineering Research, George Housner, (ch.), Earthquake Engineering Research, National Academy of Sciences, 1969.)

These two documents--the Press report and the Housner report--were the beginning of well over a decade of consensus-developing reports and discussions--they might be called battles. These two documents really started the process of identifying what the nation should do to reduce the hazards of earthquakes. While both reports were very important contributions, they did not in themselves develop consensus at all. Quite the opposite--they sparked controversy. But they both brought out very important elements that had to be part of any comprehensive program.

Scott: Yes, these were two key documents treating the idea of a national program. What about the others that followed?

Wallace: Indeed there were others--these two were just the beginning.

Wallace: Just before the 1969 Housner committee report came out, a proposal for a ten-year national earthquake hazards program hit the streets. It was put together by a federal interagency working group for earthquake research, of the Federal Council for Science and Technology. William Pecora, director of the U.S. Geological Survey, was chairman of that interagency committee, which was made up of government people, with none from the outside. (Federal Council for Science and Technology, Ad Hoc Interagency Working Group for Earthquake Research, Pecora, W.T., ch., Proposal for a Ten-Year National Earthquake Hazards Program; A Partnership of Science and the Community, Prepared for the Office of Science and Technology and the Federal Council for Science and Technology, 1968/1969.)

That report emphasized primarily earth sciences, which left large sectors out in the cold. In addition I will say that they made the mistake of starting to talk about money too soon. The report proposed budgets for a variety of rather detailed studies. Admittedly, at some point that had to be done, and their doing so helped shape the later plans. But the narrow focus and premature budgets alienated many serious players.

Wallace: About the time these three reports were produced, a fourth panel was created under the Executive Office of the President, Office of Science and Technology, and was chaired by Karl Steinbrugge. Others on the panel were Clarence Allen, Henry Degenkolb, Richard Jahns, Nathan Newmark, Jack Schoop, Robert Shea, James Stearns and James Wilson. This committee produced the Report of the Task Force on Earthquake Hazard Reduction. I think an earlier version was issued in 1968 or 1969, but the copy I have here is dated August 1970. (Task Force on Earthquake Hazard Reduction, Steinbrugge, K.V., (ch.), Earthquake Hazard Reduction, Executive Office of the President, Office of Science and Technology, September 1970.)

This task force benefitted from all the earlier reports on the elements that should be included, but used an entirely different tack, and a very smart one. They started out, for example, with "engineered earthquake resistance for new governmental facilities," through "local seismic networks", "basic research in seismology", and so on; A-1 through C-8. They did the smart thing of classifying these according to long-term (C), intermediate-term (B), and short-term (A) payoffs. Everybody--each of the competing groups--had something in there ranked "high priority," which helped make everyone feel good.

What a very skillful move that was! And the report did not mention one word about money. This is my interpretation of how it worked. Everybody's pet project was mentioned, and everybody had a high priority in short intermediate or long-term payoff. They all had something.

The report had a very simple format with a series of recommendations, given as a short underlined item for each recommendation. I will open it up to Recommendation No. A-9: "It is a federal responsibility to provide a realistic total plan for earthquake disaster response, and the plan must involve state and local governments where the hazard warrants." There is another page or so of elaboration, but that, as a policy statement of a priority--of something important to do--really rang a bell.

I give the Steinbrugge report credit for beginning to pull a consensus together. It is short, whereas the others are all rather long and ponderous. With that, people started talking the same language and thinking in more nearly parallel ways.

Scott: In your view, was it the deliberate intent of Steinbrugge and the others involved to do precisely that--to side-step controversy over dollars and promote consensus on set a of recommendations that could move ahead?

Wallace: Yes, absolutely. They were very much aware of all the dissension that had accompanied the other reports. I understood from Karl that it was Dick Jahns who came up with this format for the 1970 Office of Science and Technology report, as a way to avoid the conflicts of money priorities that Bill Pecora had fallen into. I think it is a real lesson for any groups that are trying to get consensus. Don't speak of money, too soon, and speak of priorities in some general way so that everybody is included.

Scott: Do you know any of the background of the formation of those various groups, and especially of the Steinbrugge group?

Wallace: No, I know very little. I see that you are listed as an advisor to the Steinbrugge group--do you know of the background?

Scott: I never really knew much of the background, but sometime after the group was already established he called me up one day and asked if I would help out. I do, however, recall some discussions with Karl, and attending one meeting of the group in Washington. You are quite right about Karl trying to avoid discussions of money and budgets. While the goal was to include something that each group considered valuable, he and the others also very much wanted an end that would truly be effective in reducing hazards. He wasn't just trying to throw everybody a sop--he was trying to build a workable program that would be accepted and that would make some progress in earthquake safety.

For a decade or more, Karl seemed to be participating in high level, federal executive and policy roles--the Executive Office of the President, the National Science Foundation, with the President's Science Advisor, etc. Karl was a leader--he was president of the Seismological Society of America, and of the Earthquake Engineering Research Institute along about then. He played a leadership role in many ways.

Later on in the 1970s, Karl and Frank Press were two advisors to the Office of Management and Budget (OMB) on earthquake programs. As I understand it from Karl, he sort of fought against the emphasis on earthquake prediction idea at first, but finally about 1973 or 1974 decided that prediction was here to stay.

OMB was looking into consolidating the earthquake programs in one agency--they were worried about duplication. Karl told me that he had a drinking-buddy friendship with Coast and Geodetic Survey people such as Bill Cloud. He was more for having the whole earthquake program go to the Coast and Geodetic Survey, which then was part of NOAA (National Oceanic and Atmospheric Administration). In the end, however, he joined Frank Press in recommending to OMB that the earthquake program come to the USGS, which it did in 1973. That is an activity that I should recount at some point.

Scott: With all the controversy about earthquake prediction, how was the idea of prediction itself evaluated?

Wallace: Karl Steinbrugge's 1970 report, activities of the Committee on Seismology, NAS, and the continuing controversy about prediction, prompted establishment of a panel to evaluate prediction. Clarence Allen chaired the panel on earthquake prediction of the Committee on Seismology of the National Research Council, National Academy of Sciences. (Committee on Seismology, National Research Council, Allen, C.R., (ch.), Predicting Earthquakes: A Scientific and Technical Evaluation--With Implications for Society, National Academy of Sciences, 1976.)

The panel addressed the question: Is prediction a credible goal for scientific research, or are the people who talk prediction, indeed, "charlatans," as Charles Richter suggested? In brief, the report concluded that prediction is not only a credible area of study, but, in fact, very important. The report did not get published right away. They had a series of meetings, and then it took time to get the contributor's papers in and edited, I suppose. It is dated 1976.

I remember Clarence making a preliminary report to the Committee on Seismology, probably in 1972 or 1973, indicating that the panel would say that prediction is a credible area of research. A representative of Federal Emergency Management Agency (FEMA) was there. It may have been Ugo Morelli, who has had a lot of influence over many years, in a consistent, low-key way. Anyway the comment was made, "If you are finding prediction credible and recommending that it should be pushed as an area of research, this has such important public policy implications that we should start a parallel review of the public policy implications of earthquake prediction." There was immediate agreement on that.

Scott: As I recall it, there was a good deal of concern that a prediction could cause panic, and thus be a disaster in its own right.

Wallace: Yes, that was the essence of concern about even pursuing prediction, so it was important to look at that problem in an objective and rational way. The National Research Council selected Ralph Turner, sociology professor at UCLA, and a committee got under way. I think FEMA was the principal sponsor. I was a liaison to the study and followed it with great interest.

As an aside, in my opinion, FEMA itself was a disaster area for so many years. And before FEMA, it had been like that with civil defense, emergency services, etc. Those efforts themselves tended to be a series of disasters. At one time I had the horror stories about them all in mind, but thank goodness I have forgotten the details. There was the big concern about nuclear attacks.

Scott: My own impression at the time was that much of the nuclear preparedness was quite unrealistic. I suspect the priority given to nuclear attack diverted attention from realistic preparation for the more conventional disasters that occur all the time.

Wallace: I agree. Anyway, President Carter tried to improve things by dividing up the activity differently. He also tried to put more preparedness into what they did, rather than just responding to disasters after the fact. Also in 1975 Ralph Turner's group issued Earthquake Prediction and Public Policy, even before Clarence Allen's panel report came out in July of 1976. Essentially the Turner report said that prediction is difficult, but if done right, it can have enormous benefits. I certainly agree that public policy and response strategies should be developed simultaneously with predictive capability. (National Research Council, Panel on the Public Policy Implications of Earthquake Prediction,

Turner, R.H., (ch.), Earthquake Prediction and Public Policy, National Academy of Sciences, 1975.)

Scott: You talked earlier about how tricky it was to address questions of fund allocation. Say something about how these money matters were resolved.

Wallace: In September 1976, the National Science Foundation and the U.S. Geological Survey produced a report on NSF and USGS program options that came to be called the Newmark-Stever report (often referred to as the Stever-Newmark report). Guy Stever, President Gerald Ford's science adviser (1974-1977), and head of the National Science Foundation, along with Nate Newmark, a renowned engineer at the University of Illinois, oversaw the preparation of this report. Newmark served as chairman of the Science Adviser's Advisory Group on Earthquake Prediction and Hazard Mitigation. Thirty or more people attended a planning session. Stever attended the meeting, which meant that the issue of earthquake hazard reduction automatically reached the Presidential level, and thus gave the report a lot of leverage. (National Science Foundation, Research Applications Directorate (RANN), and United States Geological Survey, Earthquake Prediction and Hazard Mitigation Options for USGS and NSF Programs, 1976.)

The panel decided that it was time to get busy with actual programs to do something, and so made recommendations on fund allocations. The report recommended money for the earthquake process, global seismology, engineering, and on and on. Meanwhile a good deal of consensus had been developed. USGS had taken over the old Coast and Geodetic Survey (NOAA) earthquake program. The Earthquake Engineering Research Institute (EERI) was going well, and was influential. By now, everybody felt fairly comfortable about dividing into a national program. But the consensus-building had taken time--it was now twelve years after the 1964 Alaska earthquake had created the concern, and eleven years after the Press report had made the first recommendations for a national program.

Scott: The Newmark-Stever report was the last big planning effort before the 1977 Act was passed. Why don't you discuss NEHRP next?

Wallace: Yes. It has been fascinating to me to reflect on all the pushing and pulling, and thousands and thousands of hours of effort, that it took to develop some sort of a working consensus. It was a slow, slow, difficult process.

Wallace: I referred to "developing consensus," which is really a polite way of saying that agreement was reached on a "power structure" and a "money distribution plan." Many thought the wrong elements were emphasized, but that was the way the politics of personalities and power of the time worked it out.

Scott: Do you mean that many who more or less went along with the consensus nevertheless really thought that the Newmark-Stever report, and the NEHRP program, emphasized the wrong elements?

Wallace: Yes. There was a battle between the engineers and earth scientists for dominance in the program, and within the earth sciences there was a battle between seismological strategies and geologic strategies. The sociologists and a few public policy people had worked within a National Academy of Sciences committee on socioeconomic effects of prediction, but they had no significant place within the program, at least until much later.

Moreover, from the moment that the Press report championed "prediction," that idea seemed to dominate. Even the sociologists hung their program proposals on prediction. The seismological community, strong proponents of prediction, dominated within the earth sciences in terms of budget distributions. An important factor, I believe, was that Congress, at least some members, were turned on by the idea of earthquake prediction--a hightech approach in a hightech era. In my opinion the whole earthquake program might not have gotten off the ground without the excitement generated by the idea of prediction.

At long last, in 1977 the national act was passed--the National Earthquake Hazards Reduction Act of 1977, Public Law 95-124. Congress essentially said, "We have passed this act defining goals for earthquake hazard reduction, now we must have an implementation plan." Responsibility for preparing the implementation plan was assigned to the Office of Science and Technology Policy, which was headed by Frank Press, who was also by then, science adviser to President Carter.

Scott: Regarding the consensus-building, I guess what we did was achieve enough of a consensus to get an act passed. But the battles continued after NEHRP was in place, and I believe the expectations of some of those who went along with the consensus were not realized. Also I believe the engineers saw themselves as having a rather limited in-put into the actual process of implementing and administering NEHRP, and allocating the money. They saw USGS and NSF as dominating the spending, and they--the engineers--felt left out. Those are my impressions, admittedly gained more from the engineering side.

Wallace: On that point, let me digress for a moment to emphasize and characterize again the Press and Housner reports. The Press report championed the topic of earthquake prediction but made no attempt to push earthquake engineering even though George Housner was on the panel. The Housner report emphasized engineering to the exclusion of earth science, except for soils engineering, but it seems to be presented as an appropriate total program. Efforts to agree on a national program, thus, started with a standoff of two major concepts divided along discipline lines. One might argue that this standoff stalled progress toward a National program.

In contrast the Pecora and Steinbrugge reports clearly were developed with the idea of bringing many disciplines together in a National program. Later the Newmark-Stever report also considered a broad range of disciplines to tackle earthquake problems. In my estimation, the Newmark-Stever report was very effective. But there never has been a way to divide up money so as to please everyone, nor did the contest end with the passage of the National Earthquake Hazard Redction.

For each of the reports it is illuminating to note the membership of panels and advisory committees. The advice received from any committee or advisory panel very naturally turns on the backgrounds, disciplines and institutional attachments of members, and particularly who prepares the final report.

Anyway, Karl Steinbrugge was selected to head the effort to put the implementation plan together. He assembled a working group and invited me to participate. He wanted me to move to Washington, D.C., for six months, but I ended up commuting--I would be back there for a couple of weeks, then return here for a couple of weeks. I worked very closely with Karl and Ugo Morelli. Others on the Working Group included Bill Anderson, Charles Culver, Henry Hyatt, James Lefter, and Don Nichols.

Scott: That must have taken a lot of effort.

Wallace: Yes, it did. I was really impressed by how Karl approached it. When we got back there to Washington, he said; "First of all, let's make a spread sheet." Along the top of the sheet we listed all the different disciplines and expertise needed. Along the side, we listed all the different regions of the U.S. We filled out the resulting matrix with names of people and organizations we needed to contact, people who should be involved.

Scott: You wanted to involve all the appropriate disciplines and also to ensure geographic coverage of all the main regions?

Wallace: We wanted disciplinary and organizational involvement, and geographic coverage--the things that work in grass-roots politics. It was really an exercise in the grass-roots politics of earthquake hazard reduction. In addition to the members of the working group members, there was a steering group, including Phil Smith, deputy to Frank Press, Bob Hamilton, who headed the USGS earthquake program, and Charles Thiel of FEMA. Rob Wesson of USGS also drafted important documents.

In addition there was an advisory group consisting of Clarence Allen, Henry Degenkolb, Charles Fritz, Paul Jennings, Karl Kisslinger, Henry Lagorio, George Mader, Nate Newmark, Bob Rigney, Nafi Toksoz, Ralph Turner and Bob Whitman. All of them had some important input into the effort. Phil Smith, for example, part of the President's science advisory staff, has a superb intellect, is skillful as a politician, and has an intense interest in the earthquake program. In the top levels of government the project was highly visible and enjoyed instant communication. It got the kind of attention that bigger programs often do not receive. No wonder it was successful!

Scott: There must have been regular feedback to the President's inner staff, and maybe to Carter himself?

Wallace: Yes, I believe so. First the report discussed more general matters, such as background and policies, public and private financial institutions, earthquake hazard reduction through construction programs, and things like that. Then the report presented a set of 37 issues. I will give some examples here: Issue no. 1, "Lack of contingency response planning," Issue no. 13, "Fire following earthquake," No. 14, "Lifelines," No. 17, "Risk map development," No. 20, "Critical facilities," No. 23, "Lack of earthquake hazard reduction criteria in federal grant programs," No. 36, "Emergency health services," No. 37, "International cooperation", and on and on. We tried to cover everything. The new program was supposed to deal with each of these issues, which were phrased in terms to facilitate this. (Office of Science and Technology Policy, Executive Office of the President, Working Group on Earthquake Hazards Reduction, Steinbrugge, K.V. (ch.), Earthquake Hazards Reduction: Issues for an Implementation Plan, 1978.)

Scott: There were all kinds of debates in the course of trying to formulate a national plan, and you were in a ringside seat at the crucial stage of agreeing on the implementation plan. Do you want to say anything more about the process and the often conflicting interests involved?

Wallace: Yes, throughout the preparation of all the documents, and especially in the NEHRP phase, ideas on how to prepare for disasters covered a range of approaches. For one thing, they seemed to divide between (1) pre-disaster efforts at prevention and mitigation, and (2) planning for faster and more effective post-disaster response. There were and still are debates about the roles of the federal vs. state governments, and the public sector vs. the private sector.

Other important matters include the roles individual citizens can play, how the public can be kept informed, and the ways education can help. Each major idea has its protagonists, who fight for recognition and especially for financial support.

Scott: I believe there were some parts that you especially worked on.

Wallace: I was involved in things said about prediction--for example, I told Karl that I thought we should have a section about ethics in earthquake prediction. After the 1971 San Fernando earthquake, policies regarding the issuance of statements that might be interpreted as earthquake predictions became more of an issue. Question were raised about how such institutions as Caltech, U.C. Berkeley, or USGS, should handle the problem.

Should individual scientists issue predictions, or should all predictions be issued through institutions? We knew that if a major prediction were even contemplated, word would always get out. We also knew that a prediction of a major earthquake, if picked up by the media, would have a big impact and cause major public reactions. So, what were the ethics of making a prediction? How could prediction be made beneficial? Institutions also began to wonder about legal liabilities. The Turner report of 1975 had considered many of these problems, but by 1978 other concerns had surfaced.

About that time, the USGS formed a National Earthquake Prediction Evaluation Council, and California developed its own evaluation council. There was much concern about how to deal with, neutralize or counteract soothsayer-type predictions that lacked scientific merit. A few scientists--or at least people possessing some scientific credentials--clearly had elements of fortune-telling in their predictions. People were making predictions based on all sorts of wild ideas. I urged that we have a section on the ethics of prediction, and Karl said, "I wouldn't touch that myself with a ten-foot pole, but if you want to write it up, go ahead." I did write a very brief recommendation to professional societies that they develop guidelines.

Not long afterward, the Seismological Society of America took up the issue, in response to this report. SSA publishes many scientific papers which could be seen as containing predictions. What was their responsibility to the public, and legally what was their liability to lawsuits? They wrote up a policy for SSA.

Scott: Those were early efforts to deal with a very difficult subject that is still with us. A lot of progress has been made, but problems still come up regularly over the issuance of predictions or forecasts.

Wallace: They certainly do--people are making predictions around the world. At one point the Soviets predicted big earthquakes for southern California. Keilis Borok, an impressive scientist and mathematician, member of the Russian Academy (formerly Soviet Academy) of Sciences, and with an institute of his own, made such a prediction, which led to a review by the National Earthquake Prediction Evaluation Council (NEPEC) and by the California council. They did not validate the predictions in question.

One might question these negative evaluations, however, because the predictions seemed to be fulfilled by the big Landers earthquake, which occurred east of Los Angeles in 1992, and the Northridge earthquake of 1994 in western Los Angeles basin! The Russian (Soviet) methodology is still being discussed here. Jack Healy of the USGS worked with them to find out exactly what they do mathematically, and to suggest how the U.S. might proceed with their prediction method. Jack's efforts got little support, however, and have currently ceased because of his retirement. Also our Russian colleagues have fallen on very hard times.

Scott: You found dealing with predictions, forecasts and other such projections a complicated, tricky business.

Wallace: Yes, and I hope this discussion brings out some of the complexities, but I can't cover them all. How does one cope with the uncertainties of thought and concepts that are bound to surround an emerging science? How should California engineers translate a Russian prediction into what ought to be done about constructing buildings in southern California? Say the prediction shows up in the newspapers, and is discussed by the seismologic community at various meetings. What does one do about it?

The federal and state councils do provide for evaluations, but with prediction in its infancy, a new type of prediction or forecast has to mature as an idea. A scientist gets an idea, which may prove correct and useful, but maybe not. First, a new idea needs to be tested and evaluated. If it begins to prove out, an institutional and application phase must follow. That is true of prediction, and really of all technological research. In any event, I did write up a section on it in the Issues volume. I will also discuss it more when I take up various USGS projects, especially the regional planning effort that went on in the San Francisco Bay Area.

Scott: What was the next step in developing the NEHRP program?

Wallace: In 1978 President Jimmy Carter transmitted to Congress a 30-page statement about the program, summarizing the things that needed to be done. (Executive Office of the President, The National Earthquake Hazards Reduction Program, June 22, 1978.) This was the document that finally set the National Earthquake Hazards Reduction Program into action. At that point, we were really under way as a program. That is fairly recent history, although by now it has had numerous reviews. How will it be modified? I am sure it will not look exactly the same, except in its overall concepts.

Scott: Yes, quite a few years have gone by since the 1977 Act was passed, and there have been several critical reviews that call for changes in the program, and especially in its management.

Wallace: True. On the other hand, the next big changes will probably come as a result of the next big earthquake, whenever that is. The 7.3 Landers earthquake occurred in 1992--the world's biggest in that calendar year. It hit in a relatively remote, unpopulated area. An earthquake that size would have had a huge impact if it had occurred in the Los Angeles metropolitan area. Or if we had an extension of the 1989 Loma Prieta earthquake nearer San Francisco, the damage would be double or triple what it was in 1989. Events like that would set major changes in motion.

The 1994 Northridge earthquake occurred after I wrote the previous paragraph. As perhaps the most costly earthquake disaster in US history, it will certainly have an impact on NEHRP programs. This is a perfect example of the influence that special earthquakes exert on policy. Early data suggest that strong-motion seismology and hazard delineation of finely defined areas should carry higher priorities. Furthermore, the engineering community seems already to view strong ground motion somewhat differently, inasmuch as unusually high values of acceleration were recorded and records were promptly and widely distributed. The performance of welded steel is also being reviewed.

Similarly, the earthquake of 1995 in Kobe, Japan, I am sure, just two weeks after it happened, will stimulate additional attention among engineers on the role of unconsolidated sediments in influencing both ground shaking as well as liquefaction and ground failure.

Scott: Yes, that has been the history of earthquakes--the big policy changes result from major damaging urban earthquakes.

Wallace: I made up a graph once to show this to audiences. Each earthquake is followed by a spike of interest and activity, then comes sort of a typical decay curve; then another spike at the time of another earthquake, and again it tapers off. So the intense concern immediately after a damaging earthquake falls off sharply a little later, nevertheless the baseline of interest and awareness has gradually risen. That has been true here and elsewhere in the world.

When I started my Ph.D. work on the San Andreas fault in 1940, the fault was relatively unknown. At least, its name certainly was not a household word. Now almost anywhere in the world that I travel, I find that people have heard of the San Andreas fault. Both here and elsewhere, earthquakes are much better understood now than they were a half-century ago. While it seems like a slow process, cumulatively major progress has been made. Each decade has seen significant advances in understanding.

Scott: The passage and funding of the 1977 NEHRP Act (also called the Cranston Act) resulted in a big boost in financing for earthquake programs generally, but especially for USGS. Would you say a few words about that?

Wallace: Certainly the Newmark-Stever report had the greatest effect on funding for the whole NEHRP activity. Indeed, it was the final step needed to make NEHRP feasible. Newmark-Stever suggested A, B, and C levels of funding. The A level--the lowest level--was essentially the one chosen, and it worked out to about $50 million total for the whole program. The figures that had been suggested for the higher levels of funding were of the order of $100 million. The funds budgeted in a direct appropriation continued to be in the ballpark of $50 million for quite a few years.

In addition to the $50 million in NEHRP funds in the form of a direct appropriation, and there were special funds from other sources. So when the NEHRP program started and for quite a while afterward funding actually went on at the $60 or $70 million level. That was for everything, including earthquake prediction, earthquake engineering, and social/economic issues.

Something like $30 million was allotted for the USGS program, with about half of that for prediction, and about one-fourth for non-USGS research (e.g., university and private industry research). There was about $20 million for earthquake engineering, although it was more in some years. So if I talk about a $70 million funding level in those years, that figure includes monies transferred from other agencies for special studies. In summary, however, earthquakes became a much bigger program after passage of the 1977 act. For a while I would say the program was fairly well-funded, although since NEHRP's early days inflation has eaten into the program.

Scott: The NEHRP program was really a major new development for the earthquake field and for USGS.

Wallace: Yes. Through the years the USGS program had been driven by the original authorization creating the USGS in 1879, and subsequent legislation and budget authorizations. A lot of the early earthquake work by the USGS was accomplished under general budgets know as "Surveys, Investigations and Research, (SIR)." But the new NEHRP earthquake program represented a separate and major funding initiative, which at present is coordinated by the Federal Emergency Management Agency.

Within the Geologic Division the NEHRP budgeting has given very strong emphasis to the earthquake program. The U.S. Geological Survey is divided into three divisions, (1) National Mapping, (2) Water Resources, and (3) Geologic. Earthquake work comprises about one-quarter of the geologic division's entire program, which represents a remarkable change. I noted earlier that back in 1962 the Survey's directorate did not consider an earthquake program was salable at the $1 or $2 million level. Yet in 1977 it was salable at a $30 to $35 million level.

Scott: To what do you attribute the big change? I'd guess it was partly all the work that led up to NEHRP, and also it was partly due to a series of earthquakes.

Wallace: Yes, a series of well-chanced earthquakes, plus a lot of tender loving care. In the earthquake business, you need to keep up-to-date plans on the shelf or in your back pocket, ready for a time when things happen without warning. Then, when the opportunity strikes and interest is high, "We just happen to have a plan that meets the need."

Back to contents -- On to next section

U.S. Department of the Interior | U.S. Geological Survey
URL: https://pubs.usgs.gov/of/1996/of96-260/wallace-VI.html
Page Contact Information: Michael Diggles
Page Last Modified: March 9, 2007