Francis Wu

Francis Wu

Professor of Geophysics, Emeritus, SUNY Binghamton

By David Zierler, Director of the Caltech Heritage Project

April 7, 2022

DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Thursday, April 7, 2022. I am delighted to be here with Dr. Francis T. Wu. Francis, it's great to be with you. Thank you for joining me today.

FRANCIS WU: Hi, David. It's good to have this opportunity to talk to you.

ZIERLER: To start, would you tell me your title and affiliation?

WU: I'm retired from the State University of New York at Binghamton as a Professor of Geophysics. I began teaching there in the Spring of 1970 and retired in 2012, having taught at Boston College for a year and half before that.

ZIERLER: Is this a true retirement for you? Are you still active or engaged in the literature at all?

WU: Yes, I'm actually reasonably active. What has happened is, my last project, a five year effort involving many research institutions in US and in Taiwan, for regional marine and land (active and passive) seismology as well as magntotellurics in Taiwan, aiming to contribute toward a better understanding of how plate tectonics has created Taiwan as we know. We digested most of the data by 2014, but we're kind of working through the implications. In fact, I'm on weekly Zoom meetings with people in Taiwan, my former student and a former colleague as well as other guests, and we're trying to dig deeper into more fundamental questions that I might talk about later.

ZIERLER: Tell me about seismology and geophysics at Binghamton. Is there a research center there, or is it a department of geophysics generally?

WU: It was the Department of Geological Sciences, and later added on Environmental Studies. I was the first geophysicist hire there in 1970. I was recruited to go to Binghamton as it had the result of a reasonably ambitious NSF development grant. We recruited Leon Thomsen, a Caltech undergraduate and Columbia PhD; also, Hartmut Spetzler, a Caltech PhD, a year before taking a position at U. of Colorado. There were staff changes throughout my tenure there, but we always managed to have three geophysics staff out of about 18 total in the Department.

ZIERLER: When you directed the Center for Study of Natural Hazards, was that specifically earthquake-related?

WU: No, actually we also had geomorphology and Caribbean geology as strengths in the Department when I got there. The Center was established before I joined the faculty. I took over a few years later. It was more broadly based than seismology. But the department emphasized basic science, so in the 1990's, when other interests developed, the center reached a natural end.

ZIERLER: Some nomenclature questions. Is geophysicist the best way to describe you?

WU: In a way. Seismologist is fine, too, I don't mind at all. Basically, geology largely deals with information that can be obtained from materials gathered at the surface or relatively shallow depths. In fact, that was my motivation for going into geophysics. I was a geology major at National Taiwan University, and as I was graduating, I felt that we knew the surface very well and could analyze the heck out of a piece of rock, but we need to connect to the interior of the Earth. One way to figure out was to probe downward with seismic energy; although I did not know too much about seismology and geophysics and there weren't many universities teaching the subject at that time, none in Taiwan and not that many in the United States.

ZIERLER: Do you see seismology as a subset of geophysics or a distinct branch of science?

WU: A subset, really. In fact, in the project I was talking about, we delve into electromagnetics, rock mechanics and gravity, so seismology is really a branch. Any application of physics to the study of the Earth is a branch and a part of the effort to understand the Earth.

ZIERLER: What have been some of the major fields of research that you have focused on over the course of your career?

WU: Near the end of my graduate studies, plate tectonics emerged as a major theory in Earth Sciences. The tools that I learned came in handy. I remembered looking back on Taiwan and thought, "Oh, by golly, that's right on a plate boundary." Seismic jolts were common as I can still recall vividly the strong earthquake one October morning in 1951, located about 150 miles from Taipei where I was attending junior high school. The weather bureau, in charge of seismic monitoring there, advised schools to let students go home, because a big one would occur again that afternoon. We happily went home, but the prediction did not come through - in fact, we still can't really predict earthquakes. But the global understanding of the outer 600 km or so of the earth that plate tectonics that began to emerge led me into fascinating studies. It led me to projects on South Island, New Zealand and in China (more specifically in Yunnan and Tibet). The camaraderie of the cooperative research projects with U.S. and the in-country scientists as well as the knowledge gained from on-the-ground data gathering and ensuing analyses made the research work fun and rewarding. After one year deployment of 11 broadband (STS-2) stations along the NS highway across Tibet in 1991-1992, and a year and half deployment of a cooperative array in southern Tibet and Nepal (jointly with researchers from U of Colorado, China and Nepal) in 2001-2002, I acquired a keen sense of the power of plate tectonic processes and subsurface probing using seismic means. Wishing to understand a system more completely I then organized an international, multidisciplinary project in 2004, after a major (Mw7.6) earthquake, to explore the tectonic processes of the island and in the ocean around the island, the so-called TAIGER project. Taiwan's mountain has been shown to grow at a very rapid rate of more than 1 cm/yr in the last couple millions of years. Geologically it is very young and the uplift is among the fastest in the world. For TAIGER, we marshaled hundreds of seismic stations across mountains to record local and distant earthquakes as well as artificial sources in order to image the subsurface structures, and the U.S. academic marine research vessel M/V Langseth deployed and retrieved ocean bottom seismometers and set off shot airguns in the ocean and recorded their signals on strings of towed hydrophones and we also mapped subsurface electrical conductivity with magnetotellurics. The field works took 4 years to complete and a few more years more data analysis. We had most of our basic results published by 2014 but further interpretation is still ongoing.

ZIERLER: What have been some of the most important theories for your work over the years?

WU: For seismology itself the mathematical theories of wave propagation saw key development since the 1960's and coupled with the development of computing, the remarkable progress in understanding the interior the Earth was made possible. Fellow graduate student Paul Richards was one of the main contributors in theoretical seismology. For the Earth sciences, the plate tectonics theory has stood the test since its elucidation and more and more details are being worked on with data from all sub-disciplines contributing. When we go to an active tectonics area, such as New Zealand, Tibet, Taiwan or California, we really see plate tectonics in operation.

ZIERLER: What about instruments? What have been some technological advances in instrumentation that have been really important for your research over the years?

WU: Prof. Benioff was the world-renowned designer of seismometers and strain meters and Prof. Frank Press was one of the designers for the newer long period seismometers. Before I became familiar with these instruments I learned about the "1-ton" Wiechert seismometers that were in use in Taiwan. In those seismometers the inertial component is a suspended mass and the motion of the earth with respect to it was being recorded photographically or later on ink-recorders. Benioff's strain meter was a 20-meter-long quartz rod, with transducers at one end and thus recording the change of distance between two piers. The electronics for all instruments were impressively massive with a low audible hum and they keep you warm in colder days. In the 1980's force-balanced sensors came in whereby the mass no longer moves with respect to the frame, instead the force required to keep it in place is converted back to acceleration (FBA), velocity or displacements. In the 1970's I deployed four station (vertical only) networks in Israel, with a Binghamton colleague who is an Israeli, using the old one component (vertical) seismometer trying to hunt for seismicity in the Dead Sea and Jordan Valley - it was a trick to smoke the recording papers in stiff wind! We did find some small events that our network wasn't dense enough to locate them well. I then used six FBA with tape recorders in the 1980's, in my effort to follow thee earthquake prediction efforts in SW China (Yunnan), whereby our team would send the instruments to the predicted location to catch the strong motion; we did catch good M5 aftershocks. Then beginning in the 1990's, our field projects in New Zealand, Tibet and Taiwan, were all benefited from a new generation of broadband sensors, as they are much more compact, and recording digitally on hard disks. The advanced hard disks we deployed on the Tibetan Plateau had a relatively puny capacity of about 150 mb, enough to store about 200 pictures in your phone! By the time of the TAIGER projects the number of instruments available for field work from the consortium of academic institutions, IRIS, we were able to mount expeditions with tens of the broadband and hundreds of other more portable type, on land and at ocean bottom. The convergence of technological advances made seismology much more powerful. In the last ten years or so, the instrumentation developed further along the same path, i.e., becoming smaller and with larger recording capacity. I think dense array is the key to major advances. But the recent advances at Seismo Lab to use fiber optics to achieve much higher density and virtually unlimited storage pushes us into a new age. Right now it seems existing cables are being utilized - you don't necessarily have them where you need them. But the potential is tremendous. The days of dense monitoring are perhaps coming. Soon we might be able to trace and locate little, subtle things before big ones that we have missed!

ZIERLER: You're saying that the more sensors, the better the prospects for early detection?

WU: An observation from my rock mechanics experiments early in my career showed that when a rock sample under stress experiences major episodes of weakening usually preceded by small acoustic events - foreshocks. Recent laboratory work on meter-scale rock samples investigating the effect of fault gouge, which is found along faults at the surface and in San Andreas fault drill-holes, show abundance of small foreshocks. We know that many major earthquakes have foreshocks - but we only know them as such after the mainshock had occurred. The magnitudes of these foreshocks were usually about 2 magnitudes lower than that of the main shock and well recorded. If we can see much smaller events, would they develop a time sequence and spatial pattern along the length of an active fault?

ZIERLER: Who have been some of the key institutional funders of your research over the years?

WU: Mostly NSF, 90%. I also had, in1970's and early '80's, interesting cooperative projects with the Chinese Seismological Bureau, under USGS funding. A cooperative study with Chinese colleagues of an unusual active volcano on the border between China and North Korea, far from the "Circum-Pacific ring of fire" was supported by DOE. While our project, in cooperation with Chinese scientists was aimed at mapping possible related crustal features, it turned out that the same data was useful for calibrating nuclear activities south of the border.

ZIERLER: What ways, if at all, have you interfaced in the public policy realm? Consulting for governments, that kind of thing.

WU: In the 1980's FEMA was formulating a national plan to deal with natural hazards. Perhaps the 1983 Adirondacks earthquake, with a magnitude of 5.1, fresh in mind, the state of New York organized a committee to look at seismic potential in the state as a whole. The group was active for a number of years. I did serve on four different NSF committees that participated in the granting processes. My main experience of involvement in public service was in Taiwan, with the organization in charge of seismic monitoring. It was gratifying that as a member of a board of advisors, I was involved in the establishing of strong motion network ahead of a very damaging earthquake in 1999. Also, for a number of years I worked fairly closely with scientists in the Institute of Earth Sciences, Academia Sinica, as well as served in various advisory roles; it has altogether been a very rewarding experience. Finally, I joined the US officials, from NSF and USGS, on the discussions of seismology protocol between China and the US many times; these were the occasions one can learn a lot about international scientific cooperation and different styles of doing science.

ZIERLER: On that basis, being based in the United States but having so many international collaborations, particularly in Asia. In what ways do you see seismology and geophysics really as an international effort?

WU: Earth sciences, of course, are global, as we need to assimilate all observations around the earth with their associated time signatures, to understand it more fully. Even in terms of so-called active tectonic plate interactions, they may generate large earthquakes, but very infrequently in human time. So, to broaden our perspective, it is better to cast a wide net to learn and digest as much as we can instances around the world and then try to synthesize. Naturally scientists in the region under question usually know the problem more deeply and the sharing of accumulated knowledge would benefit all concerned. In terms of earthquake research, areas known for recurring events would evidently be a good site for detailed monitoring. In fact, one well-populated part of Taiwan had a damaging M>6 event every 50 years or so - and that is certainly a good place, and there is a need, to put in a dense network to see whether there is a pre-shock process (including what we know as foreshocks) that would presage large events. At the same time we can image the 3-D structures around the fault. We can learn much quicker if we find similar sites around the world and monitor many of them. I often think that if the Italian seismologists deployed a dense local network in the L'Aquila area after a series of small earthquakes began to shake the area in December 2008 and reported the results everyday on the evening news it would have alerted the citizens even without a prediction. On April 6, 2009 the mainshock occurred and seismologists were taken to court for making light of unconstrained predictions based on radon fluctuations.

ZIERLER: I wonder what you've learned about culture and national perspectives as they relate to seismology. In other words, is there a uniquely Taiwanese approach to seismology or from other countries that you've seen?

WU: A good number of the Taiwanese seismologists were trained in the US or has worked in other international institutions so they have a very international outlook. But I can feel a palpable difference of style between US and Taiwan, much of it evidently due to different cultures in the organization of the scientific communities. Incidentally, an interesting trend in China of their geophysics program, at the University of Science and Technology, for example, to build on a classical physics background in the first few years in college, such that they're quantitatively very well-prepared. I see a good number of students from China with such background who found great success in the US and elsewhere in the world. An associated phenomenon is that contributions from Chinese authors in the US geophysical journals have increased notably in the last few decades.

ZIERLER: Let's go back to your undergraduate experience. Were you always interested in seismology and geophysics, and geology was a foundation to that?

WU: When I got into geology, seismology and geophysics weren't particularly known in Taiwan. I enjoyed exploring geology and quite happily went to the field. In fact, one of my motivations for going into this discipline was that I may get to travel around the world to do it. But I gradually felt that geology was great for understanding the surface outcrops. It would be useful to know more below the surface. By that time, I'd heard about geophysics, mostly in name, so with my classmate, Leon Teng, a classmate in Taiwan as well at Seismo Lab, I took a few additional math and physics courses. We didn't know what exactly we needed. Far from having a good physics background, we had to work hard to get into it.

ZIERLER: How did you become aware of the Seismo Lab at Caltech? Was this known in Taiwan? Was its reputation appreciated when you were an undergraduate?

WU: Caltech was very well-known from reading Time magazine - I think it was an issue with Dr. Dubridge on the cover. After I graduated from college, I went into military service first, but I think it was through Prof. Clarence Allen's visit to Taiwan that I learned about the Seismo Lab. But I actually applied to Berkeley first and was admitted there, as Leon Teng applied to Caltech and was admitted in early Spring 1961, I decided to try. I applied and very fortunately got accepted.

ZIERLER: How was your English when you arrived in the United States? Had you learned it in school?

WU: Yes, actually, I arrived back in Shanghai in 1946 from Northern China to begin 4th grade education; English class began in the 3rd grade there. So I had catch up fast. Shanghai was an international port, and my father worked for the Customs House, where British officers ran the organization. So for him English was used every day; we kids picked up phrases here and there. Years later my sister became an English major in college. She loved Victorian and other novels, and I got to read a few of them. Thus I had some extra help along the way. For my college classes, most of the textbooks were in English. That made the transition to Caltech a bit easier.

ZIERLER: What year did you arrive at Caltech?

WU: Summer, 1961, by boat. By the way, I still remember how impressive Mt Wilson was one November day after rain and wind - smog hid it until then.

ZIERLER: Was it the Seismo Lab specifically that you had applied to? That's where you first started your studies?

WU: No, it was to the Department of Geological Sciences. I recall that Bob Sharp was the chair then.

ZIERLER: When you arrived, of course, the Seismo Lab was in the mansion off of campus. Tell me about it. What did it look like? What was the environment like there?

WU: I started taking courses on campus, field camp, and all that stuff. Pretty soon, I was told, "OK, you're a research assistant, so your assignment first is to work with Dr. Benioff, and he's at the Lab." There was a mail truck going from campus to the Seismo (Donnelley) Lab every day, and I was told that I could hitch a ride. Coming back, there was a convenient bus down the hill. I did that for a few months before I got a scooter for the trip. The Seismo Lab was sort of a totally independent domain in those days. I think faculty came back to campus only to teach but were there most of the time. It was a very complete unit. Frank Press was the director, Drs. Benioff and Richter were the professors in my first year. Within one year, two young professors who had just graduated from Caltech, Bob Phinney and Stewart Smith, and then Don Anderson began to teach. Everybody at the lab, from the secretaries, to the draftsman, the seismogram readers, the engineers, graduate students, and to the professors, we all knew each other. The lab had all these stately bedrooms; Frank Press in a suite and so were Benioff and Richter. Then, there was the big living room downstairs with nice sofas where many of the graduate students congregated; some chose closets, or in my case, for the first year, in the passage way to the magazine stacks for my first year. Initially Dr. Benioff give me odd tasks, such as digitizing. But post-docs and senior grad students such Ari Ben-Menahem, Shelton Alexander, Charles Archambeau, David Harkrider and Nafi Toksoz were all ready to talk to new comers and answer inquiries. It was a really open system. Learning had no limits, and neither did topics. Any topic, you can start on it and you try to figure it out along the way.

ZIERLER: What were some of the major debates in seismology that were happening at the Seismo Lab when you got there?

WU: The Seismo Lab had an important tradition. Under the main floor of the stately Donnelley Lab with ornate bedrooms and grand living room there was, in the bowel of the building, among the heaters and boilers, the coffee break room. Held every mid-morning and mid-afternoon, this is the place where people came to talk about their ideas and works. In the early 1960's one of the major topics that came up was can we distinguish earthquakes and underground nuclear explosions. This came about especially because lab director Frank Press was a Science Advisor for the nuclear treaty being negotiated in Geneva in those days. He would come back from Geneva to the coffee break and talked about the key seismological issues at the meeting. One would think explosions generate essentially symmetric radiation in all directions, but gradually it was found out that the Nevada explosions was not spherically symmetric. Late on, Archambeau theoretically answered the question. Otherwise, plate tectonics and the rotation of the Pacific Basin - an idea that Dr. Benioff proposed after studying circum-Pacific structures - loomed bigger and bigger.

ZIERLER: What was the process for determining who your thesis advisor would be?

WU: When I first arrived at Seismo Lab, just about everybody was Frank Press's student. I do remember John Gardener, whom I never met after he graduated, was Richter's PhD. But when I started preparing for my PhD, Frank Press left to become the head of Department of Earth and Planetary Sciences at MIT; Don Anderson succeeded him as the director of the Lab. I was Don Anderson's advisee.

ZIERLER: What were you interested in once you got to the Lab and surveyed everything that could possibly be worked on? What resonated with you?

WU: It was a totally new world I entered. There were a couple of years of absorbing how seismic waves propagate and what we could get out of them. Then, I worked with Ari Ben-Menahem to apply what I learned about surface waves to decipher the faulting behavior of a M7.1 Iran earthquake and wrote my first research paper. The basic concept of my PhD thesis was simple. I wanted to apply all that had been learned in the past few years about wave radiation from the source and attenuation along the way to figure out what the total radiated energy of an earthquake was. It had to take into account what we knew about fault radiation pattern for different waves and fault growth and wave attenuation. Although the Worldwide Network of Standard Seismographs (WWNSS) was a great improvement of station coverage had improved greatly, the instruments were still relatively narrow in frequency band and of course every seismogram I used I got to hand-digitize. I learned a whole lot about seismic radiation and about faulting dynamics. But it was perhaps too ambitious, In retrospect I could have done so much more. But it's a PhD thesis and how far can you get? [Laugh]

ZIERLER: Did you have the opportunity to interact with Charles Richter at all?

WU: Oh, yeah.

ZIERLER: What was he like?

WU: There was a stately curved stairway on the main floor and in the crook of the stairway was a seismograph. Dr. Richter's office was at the head of the stairs. Any time there was a good sized shake somewhere around the world or in California, many of us would congregate around the ink-recorder drum of the seismograph. Dr. Richter would be there taking measurements on the drum and mumbling to himself. We all stood around and learned from him that way. He would answer any questions, but I don't remember he ever said anything personal. I did take a course with him based on his text. He wasn't a man of many words but he gets to the point in a discussion very quickly.

ZIERLER: Were you there when Beno Gutenberg passed in 1960? Was there a memorial service? Or that happened before your time?

WU: I never met him; he passed away before I arrived.

ZIERLER: What was Benioff working on when you connected with him?

WU: I think his main interest in 1961 was his relatively new strainmeter that had recorded the 1960 Chilean earthquake in Nana, Chile and in Isabella, Calif. They record earth tides as well as distant and local large earthquakes. There are not a great number of them, because it requires a tunnel 10's of meters long. Incidentally, I remember a thesis on the shelf in a reading room contains a series of cartoons. They show Benioff, bending over, back strained, as the Benioff strainmeter; others including Benioff in various activities as Benioff vertical, Benioff horizontal etc. Can't remember whose theses it is. Those cartoon perhaps reflected the fact that Benioff appreciates humor very much. Another subject to which Benioff made important contribution was music. I knew his interest in music by the presence of a music stand in his closet. I also learned that he was a consultant for Baldwin Piano. By googling, I found his invention of electronic cello and violin as well as an electronic piano (more harpsichord-like?) for the well-known baroque pianist/harpsichordist Rosalyn Tureck.

ZIERLER: Did he give you a problem to work on? How hands-on was he as a thesis advisor?

WU: Dr. Benioff was at the lab only during my first year at the lab. I worked for him as a research assistant. He fully retired, I think, in 1962. The first thing he asked me to do is to digitize the strain seismograms for earth tides (I think it was a joint project of Benioff and Steward Smith) and also one recorded in Nana, Peru; for the latter he said, "Go figure out what the record is telling us. Talk to Alexander, talk to Harkrider." That was his style. But he also thinks about global earthquake distribution from surface to depth. Later on, it was called the Benioff zone - something that later became a key element of the plate tectonics theory. His interest was broad, but his expertise in instrumentation really led to major advances.

ZIERLER: Who ended up being your thesis advisor?

WU: First, Frank Press, and after he left, Don Anderson took over as thesis advisor.

ZIERLER: Press left to go to MIT?

WU: Yeah, to become head of the Earth and Planetary Sciences at MIT, and later, Carter's presidential advisor.

ZIERLER: Did switching advisors change your thesis research?

WU: No, Frank Press's style was very open, and he wants students and colleagues to talk to everybody. I think Frank Press himself was doing model seismology in his days at Lamont. He used a physical model like a plate and studied waves inside, trying to understand how elastic waves propagate. He encouraged me to do that, and having learned something about the San Andreas fault, I began to wonder what happens when surface waves crosses San Andreas. For example, I was wondering how the San Andreas Fault would influence propagation, So the Lab machine shop made me a relatively simple model. I created a model with that, and he was advising me on the work. So here I was a new graduate student, talking to the practically the guy to invented that kind of study.

ZIERLER: What would you say were some of the key conclusions of your thesis research?

WU: My thesis included two parts. The first part dealt with the energy of earthquakes as I described earlier. My conclusion was, the energy estimate would be higher than the value obtained from classical energy formula relating magnitude and energy based on point source. Now that broadband digital data from much larger number of stations and we understand the attenuation of seismic waves better too, new estimates can be made much more accurately. The second part was mostly computational results, aiming at the use of reflected seismic waves (PP) reflected from the crust between the station and earthquakes at large distances to estimate the crustal thickness at the reflection point. This parameter was found to vary depending on the tectonics, but the usual way in those days are using explosions os local earthquakes. So we might be able to determine the crustal structure under Tibet without going there. The work was published later with James Hannon, a post-doc fellow.

ZIERLER: What were some of the bigger issues in the field, and how did you see your thesis research as contributing to them?

WU: What is the total energy radiated from earthquakes is a fundamental question in seismology as well as Earth's energy budget. Up to that point, the energy was estimated based on a point source model. My thought was so much advances were gained about source and propagation and I should be able to make an attempt to apply all the recent knowledge on seismic radiation and the see the results. I don't think I settled the question. The second part was aimed at mapping the crustal thickness without deploying stations to the desired area. For better results deep (>150 km, say) are needed for the source and new broadband seismograms would help greatly. Attempts are using the PP waves have been tried more recently.

ZIERLER: What sticks out in your memory in terms of the Seismo Lab being an intellectual center, a place that attracted scholars from beyond Pasadena?

WU: Certainly, throughout the period, Richter, Benioff, Frank Press were all leading seismologists of the world. Soon after the new additions of Bob Phinney, Steward Smith and Don Anderson all had their new directions and excelled in what they did. As I began to attend national meetings (AGU -American Geophysical Union- and SSA -Seismological Society of America) I noticed that all the senior graduate student, Dave Harkrider, Archambeau, Nafi Tokoz were all becoming a force in the field. That when people say, "Wow, look at that. Somehow, it works." Certainly, Frank Press, as chief advisor to the nuclear negotiation, attracted a lot of attention. Nuclear discrimination was a major focus in those days. It's interesting because plate tectonics had started to ferment in the world around that time. But by and large the lab folks were not particularly involved in global tectonics. It is interesting that at Lamont, where much ideas came from, geology, geophysics (including seismology), and marine geophysics were integrated. However, Lab folks were working on fundamental seismological methods and theory, and those came in handy later.

ZIERLER: There was no notion of a Pangaea at that time.

WU: I wasn't too aware about the progress on that subject made around the world on this subject in the early 1960's. Actually at most universities in the US, for that matter, the topics did not flourish. But I was already educated by my paleontology professor in Taiwan. He found that on different continents at high latitude had warm-water corals and vice versa, implying the movements of continents. I did not learn much about paleontology but, when I gradually learned about global tectonics around the time of my thesis research, I accepted it readily. But it is said that in the US academic world before 1960 or so, if you advocated for continental drift, you would be viewed as a person with less scientific credibility. But global tectonics and all that was comprehensive and was all data-based. So, as I was digging into details of earthquake mechanism and radiation of seismic energy, the global tectonic theories was very convincing. By the way, Dr. Benioff probably was excited about global tectonics as the Benioff zone was one the very foundation of the theory.

ZIERLER: What about instrumentation? Were there instruments at the Seismo Lab that were not available elsewhere that served also as a magnet to draw outside people in?

WU: Certainly, the strain meter is uniquely Benioff. I think another institution may have one or so, but Benioff's collection is still there. In those days, the Seismo Lab still had a instrument group, and they were helpful to visitors and graduate students. They experimented with different free periods of instruments and galvanometers to look for a "sweet spot". In those days the World-wide Standard Seismograph Network was the "new" (Since ~1960) data source and the basic instruments were the Benioff short period and the Press-Ewing long period seismometers, but the operation was totally independent of the Lab. Later on a shock wave research lab was built in the Kresge Lab (the "Lower" lab) - I think as a part of the Don Anderson's effort to understand the mantle; but that was for studying materials in the interior of the Earth and it attracted people.

ZIERLER: It's a hard thing to feel, but the origins of the Seismo Lab, did you feel that when you were a graduate student, why it was built, how it was built? Do you feel like you were part of a generation that was still connected to that?

WU: When I was in the environment I thought it was all routine, but having experienced a few different academic institutions throughout the years, I could look back and see how unique the environment was. Richter and Benioff provided the link to Gutenberg and Press was the clear symbol of the future. Press brought in a much larger group of students than before. When the seismology was moved to San Rafael Hills the solid granite/schist rock provided a suitable site for seismic recording. While the Lab was still clearly connected to the campus as all classes were conducted there, the lab is a self-contained unit. In the early 1960's Richter had his large seismogram reading room, with big tables and seismograms stacked in the cabinets around them. I have the wonderful memory of Richter reading the seismograms of a big earthquake on the drum recorder by the staircase. It was perhaps in 1965, Richter determined the approximate location and the magnitude. I have the image of him singing something like: "This one looks like one from 193-." He'd go downstairs and find the record. "See? That's history right there." Dr. Richter had two women assistants. He would have them find the seismograms and make more comparisons. We were all there helping, looking at waves. That kind of education was priceless. Hearing all the humming, the understatements, it was a wonderful experience. I and my colleague tried to replicate the seismogram reading experience with our seismology students in Binghamton.

ZIERLER: Tell me about staying on for the post-doc at Caltech. Was there unfinished work that you wanted to complete? Did you want to hang out for a year before you went on the job market? What happened?

WU: Actually, both, I guess. I could've apply for a job at a university somewhere teaching seismology, but I didn't go for that for two reasons. One, I took the post-doc position at the Lab under a new grant of Don Anderson and Stewart Smith for studies related to the San Andreas fault. It was an opportunity to learn more about the dominant structure. I thought electrical resistivity measurements using magnetotellurics may be useful for understanding the subsurface structures around the fault. I remember talking to Don Anderson and Stewart Smith about this method. They said, "Oh, sure, go ahead and try it." I remember wiring my own instruments for field measurements for tests and wrote a paper on the inversion of electromagnetic data. I tried to pursue it even after I went to Boston, and much later magnetotellurics was an integral part in my TAIGER project. In the 1990's when the drilling of San Andreas near Parkfield was in progress resistivity profiles across the fault was conducted by several groups; they added clues about fluid flow around the fault. Secondly, the Parkfield earthquake occurred in 1966. I did several studies on it. The post-doc experience was rich with exploration and did make me realize that research in Earth sciences is fun but persistence was necessary.

ZIERLER: Did you ever think about going back to Taiwan? Or at that point, you knew you wanted to make a career in the US?

WU: Taiwan wasn't ready to "take off" yet, when I finished my studies. I could have decided to become an organizer, to build a Taiwanese geophysics program from ground up, not a easy job for a freshly minted PhD in that environment. When Boston College came out to recruit, Father Skehan, the chairman of the department told me about the Jesuit seismological tradition and the New England Seismic Lab called Weston Observatory; that was very attractive so I went to Boston. At that time Frank Press was still head of the MIT department of Earth and Planetary Sciences and Nafi Toksoz as well as other Seismo Lab denizens were there. I would go to MIT two or three times a week for seminars and actually learned rock mechanics from Bill Brace and magnetotellurics from Ted Madden. Taiwan however was always on my mind. In fact, when I was working on my PhD, global tectonics theories flourished, and I started looking into Taiwan earthquakes as a product of plate tectonics. On one my MIT visits, a student from Taiwan, Yi-Ben Tsai and I talked about a new science initiative was being launched by the new Science Commissioner, Dr. Ta-You Wu, a professor of physics at SUNY Buffalo. So we called Leon Teng right away, who was teaching at USC, and decided to write a letter to Dr. Wu to propose a earthquake studies program. Dr. Ta-You Wu answered within the week, and within one year or so a program began to roll. That became the first act of the present-day Institute of Earth Sciences, Academia Sinica in Taipei. Actually, in those days voices from abroad were effective, as demonstrated by the fact that this proposal was made by two assistant professors and one senior graduate student.

ZIERLER: For the last part of our talk, I'd like to ask some broad questions about the impact to you personally from being at the Seismo Lab and the way you see the Seismo Lab influencing the field more generally. For you, the education you got, the approach to seismology and geophysics, how did that influence your work in the intervening years?

WU: In basic training, I had a chance to catch up quite a bit on classical physics and the dramatic arrival of computers. But in overall scientific approach, I think what really impressed me was that the professors in particular did not draw clear boundaries between topics. In seismology, sometimes one could go to grad school, learn a particular technique and stick to it throughout one's career. I think Caltech's style was to focus on the interesting problems first and let's figure out what was needed to solve them. Throughout my career, I delved into a few different fields thanks to that spirit I learned there. For a while, in Binghamton, my computing funds from a grant was running low and the computer center charged $700/hour on the IBM computer couldn't be reduced. I looked for free computing elsewhere (relief finally came when in the mid-1980's NSF funded my fellow seismologist and me a "mini-computer"). Also, I went back to the San Andreas fault that I became acquainted while a graduate student and studied the materials in the fault zone. I found various clays in the fault zone and wrote paper regarding the gouge and what might be at depth of the San Andreas zone. (I was gratified that recent papers on the subject sometimes refer to my effort.) In that period, I also delved into rock mechanics experiments monitoring ultrasounic pulse from small cracks before the final failure (foreshocks?) I think that's the Lab spirit working. Caltech gave me the license, the courage to just go and say, "Let's do it. The problem is there. I want to see what I can find out."

ZIERLER: What were some of the ideas or approaches to seismology that came out of the Seismo Lab that have continued to influence the field?

WU: It takes a bit of thinking to pick one among the many threads. The top one goes to Don Anderson's multiple fronts in the field of seismology and beyond. One thing I notice when I think about it is that for many years geophysicists, seismologists, geochemists etc. were trying to understand the significance of a large number of isolated volcanoes, many on the ocean floor and some are on land that seem to have long eruption history. For years a commonly accepted hypothesis viewed them as a worldwide plate position reference system, i.e., they are from deep (near core-mantle boundary) sources that are essentially fixed. Using the reference we can think about whether the subducting plate underthrusts or is being overthrusted in the sense of absolute motion. Well, the attempt to image the lower mantle plume hasn't confirmed its existence. The discussions of the Don provoked deep debate by pointing out that the earth is cooling and the idea of hotspots been in operation for hundred of millions years may violate second law of thermodynamics. The debate is not totally settled but Don has started and point out the need to test the hypothesis from all sides.

ZIERLER: I'd like to ask a few questions about theories and hypotheses that have stood the test of time and those that have been overturned based on new findings. If you look at the big ideas from the Seismo Lab when you were a graduate student, what has held up over the years?

WU: In the past 6 months, I've been in a discussion with Taiwanese colleagues about this exact question. What do we know about Taiwan now, what hypotheses have people proposed, do they stand up, and how do we test them? Geology is a really tough problem. Almost like studying outer space, way out beyond the solar system. Because we don't have the chance to really grab a sample from the deep interior and perform tests on it. For example, Don Anderson's phase transition in the mantle arises from lab experiments on minerals and seismological observations. Don's theory is quite convincing based on all the evidences we have. But the deepest we can drill is about 12 kilometers, so no sample to examine in the lab. Seismic evidence is strong, but we are limited by the available bandwidth of the waves - details of features with scales of less than a few hundred meters are hard to come by. But even at that scale a lot of hypothesis-testing of key hypotheses can be done. Regarding Taiwan, the hypotheses of how the young mountain ranges were built can be tested in many aspects. One example is the hypothesis that mountain-building starts on a shallow-dipping detachment fault - often associated with a shallow (<10O) east-dipping subduction and sediments were simply pushed up to build mountains. Incidentally few active subduction zones around the world lead to active orogeny. Since the mountain building there is still ongoing in Taiwan and very active (rising > 10 km/million years - one of the highest in the world), if the hypothetical process is ongoing, we should be able to see the effects, either as a very extensive fault zone (with mylonites?) or in terms of earthquakes located along the zone with very shallow dipping faults. So far not such features are found to support the hypothesis. For inactive structures we can try to test hypotheses in different ways. With dense local seismic and other networks and efficient processing we can try to subject many hypotheses to falsification.

ZIERLER: To flip it around, what were some of the accepted theories or orthodoxies that have been overturned in light of new evidence or data that's been discovered over the years?

WU: Remembering them makes me ancient; young students these days might go through an university curriculum and have never heard of mio-geosynclines. But, during my earlier career the rise of global tectonics pushed out the slew of theories large scale features of the world: permanence of continents, geosynclines, Benioff's rotation of the Pacific Basin… In seismology I experienced the triumph of double-couple over single-couple for the earthquake focal mechanism, the multiple "shaky" claims of successes in earthquake prediction and recognition of earthquake precursors…So there certainly have been progress in understanding our planet. ZIERLER: For my last question, we'll look to the future. I've asked you about ideas that have stood the test of time and ideas that have been overturned. What are the remaining question marks where we still don't know whether or not these ideas are right or wrong?

WU: Returning to the difficulty of earth sciences and the earthquake prediction problem; as far as society's concerned, the latter is the number-one problem. With regard to the former, while plate tectonics is still taking place now and clear signs that it has taken place around the globe has endured repeated tests, our knowledge of the smaller scale structures such as mountains is still limited. There have been various geological hypotheses, but in many instances they are hard to test. Part of it is the difficulty of not being able to see directly if the assumptions in the hypotheses are true, and especially when something occurred a few hundred millions ago where outcrops had undergone later modifications. Geophysics and seismology in particular provides some capability for properties at depth. Let me illustrate with the example of the assumption that a shallow detachment under a mountain on which sediments were pushed and compressed into mountains, e.g., the Appalachians. That involves the fundamental question of how earthquakes, especially large ones, evolve. We have the general understanding that an earthquake may develops as a sort of instability in the fault zone - after a long interval during which small slips along different parts of the zone, seismically or quietly, occurring along the fault zone, but at some point the system reaches an inflection point when small events begin to induce more and larger events and a cascade may happen. Is this how all earthquakes proceed? The only way to learn is to monitor potentially active zone intensely with all the instruments we can muster. It would be nice if a few such sites around the world can be chosen, with the purpose to quickly and efficiently learn about the mechanism. Toward such aim the "natural laboratory" may make sense, i.e., choose a region of very active tectonics, where significant events are relatively frequent and where research infrastructure exists. Taiwan is certainly one possibility for such effort. I admire meteorologists ability these days to predict the rare rains in California sometimes a week ahead of time. They have observations on the ground and in the air at all hours. It is much more difficult to see details inside the earth, and we're really blind until we get data. The earthquake process, how did it occur? How does it develop? It has everything to do with geology, with material property, with social benefits, and we need to do it.

ZIERLER: One way or another.

WU: The way is really to put down instruments. If we know something, maybe we can help.

ZIERLER: Francis, it's been a great pleasure spending this time with you. I'm so glad we were able to do this, for you to share your insights and recollections over the course of your career. I'd like to thank you so much.

WU: Thank you very much, David.

[END]