Home  /  Interviews  /  John Vidale

John Vidale

Dean's Professor of Earth Sciences, University of Southern California

By David Zierler, Director of the Caltech Heritage Project
April 19, 2022

DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Tuesday, April 19, 2022. I am delighted to be here with Professor John Vidale. John, great to be with you. Thank you for joining me today.

JOHN VIDALE: My pleasure!

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

VIDALE: Dean's Professor of Earth Sciences at the University of Southern California.

ZIERLER: What does dean's professor signify? Is it just an honorific?

VIDALE: Yeah, it's an honorific, and it comes with funds to buy things that otherwise are hard to pay for.

ZIERLER: Some overall questions about your research. What are the things you're focused on currently?

VIDALE: I'm interested in a lot of things, but the last couple years, I've been working mostly on the inner core, the motion and the structure.

ZIERLER: What do we know about the inner core, and what remains a big question mark?

VIDALE: Well, we're mapping out kind of a general idea of a lot of its properties. There are maps of the seismic velocity, the attenuation, the anisotropy. People have layered models, three-dimensional models. The point is to sharpen the models and try to understand why it has taken the form it has today.

ZIERLER: Where is this related to your interest in earthquakes? Or is it a separate field of inquiry?

VIDALE: It's entirely separate. One of the things one can do with seismic waves. Through my career, I've just approached probably a dozen different problems with seismic waves.

ZIERLER: Why is it unrelated?

VIDALE: It's the same data. We're looking at the vibrations that come in on the seismometers, and often, we do what we call array analysis for earthquakes just as we do for deep-earth structure, but the goals are different. The structure in the inner core has very little to do with the earthquakes we see at the surface. We're finding conclusions that depend on observations of earthquakes and observations of the inner core. But actually, technically, that's not true. We need to understand all parts of the signal we're recording in order to interpret any one part correctly. One thing I'm working on is how the inner core is moving over time. To do that, I have to understand how the sources, which are all on the surface, are changing over time. It's all tied together in a technical sense, but in terms of the science, it's entirely separate.

ZIERLER: When you say moving over time, what does that mean? How and where does it move?

VIDALE: That's something we're arguing about. Originally, some Caltech guys, Xiadong Song and Paul Richards, in the late 1990s, argued they could see the inner core rotating. And they believed that, and they still believe that. But other people have other ideas. My initial work supported what they were saying, although at a much slower rate than they initially suggested. But the latest paper we have that's coming out in a couple weeks argues the inner core's actually oscillating back and forth, although I'm not sure I believe my own paper. I think it's the best guess, but I wouldn't swear that it's right.

ZIERLER: In the way that there are so many obvious social or societal implications for earthquake research, what about research on the inner core? Is there anything going on there that would affect us here on the earth's surface?

VIDALE: Some of it is understanding how we got to where we are today. The inner core is part of the structure that generates the geodynamo, the outer core that circulates to make the magnetic field. If we track the history of the inner core, that tells us something about when the dynamo might've been present, when it was strong. We need the dynamo to kind of shield Earth from rays from space. In a vague sense, the core is tied to life developing on the surface of the earth, but that's kind of stretching it. Loosely, it's just a curiosity.

ZIERLER: Tell me about your work on earthquake hazard mitigation. First of all, what does that phrase mean?

VIDALE: Hazard mitigation just means trying to reduce the cost and loss of lives in earthquakes. There's an element of rebuilding more quickly afterwards, which I guess is really part of reducing the cost of earthquakes. I've kind of worked on a hodgepodge of topics. I can't really give you a long narrative of how it all ties together, but the most direct thing I did was, I was director of the Pacific Northwest Seismic Network for a decade, and that was motivated by trying to lessen the effects of the earthquakes. We were trying to see the patterns in faults, the patterns in plate tectonics, study the volcanoes, get an idea of what tsunamis might hit the coast. That was directly trying to mitigate earthquake hazard.

ZIERLER: Does your work on volcanoes exist at the interface between your interest in the inner core and earthquakes and plate tectonics?

VIDALE: I would be hard-pressed to tie the volcano work to the inner core. It has a bit of a tie to the plate tectonics and the hazard, but it's really a separate item. I looked at volcanoes because my job in Seattle was to run a lot of the instruments that were recording when the volcanoes were active and interpret that. But the research was mostly when I noticed patterns in the earthquakes that were something we didn't already understand. It's all very indirect. The last paper I wrote on volcanoes was about what might cause the earthquakes when we see deep under volcanoes that are surprising because they have kind of a different sort of pop to them than the regular earthquakes. There was an idea in the paper that it's a certain structure underneath the crust that might be responsible. It's quite a stretch to argue that that makes it safer living around volcanoes, it's mostly just trying to understand the system to kind of fortify the basic science of Earth's volcanoes.

ZIERLER: With all the advances in computational power, what opportunities has that afforded with your interest in modeling in seismology?

VIDALE: Seismologists in general have benefitted greatly from high-powered computers. I've used them some in my research, although I don't usually use the biggest computers. I ran the Southern California Earthquake Center for a couple years, and they're probably the biggest user of supercomputers in the seismology community, so I've been connected with some of the most powerful things we can do with the big computers. It really allows us to consider complicated geological structures and the great variety of earthquakes that could happen, and to get a comprehensive picture of what threats all these earthquakes cause to all the places we can fit into our models, which in the case here, was Southern California.

ZIERLER: I'm curious specifically in Southern California where your research is localized, in other words, happening in a very specific place, and where it might be localized, but you can extrapolate on a global level because there are similarities throughout the globe?

VIDALE: None of the research is interesting unless it can generalize to all around the earth. But the power is in making specific predictions or discoveries in the area where the data comes from. Most of my work has been on a general problem, trying to collect the best data in the world, the biggest dataset I can find. Sometimes, though, it's been focusing on a local dataset to find a local instance of something we need to understand better on a global scale.

ZIERLER: I'm curious what interface you might have with policymakers in the course of your work. For example, directing the Southern California Earthquake Center, did that give you opportunity to work with officials at the local, state, or even federal level?

VIDALE: It did a little bit. But SCEC is really an enterprise to understand earthquakes. There are elements of SCEC that work with the state of California closely to map the faults and produce the hazard map. But the closest I worked with the government agencies was up in Seattle. Because there, we were the seismic interface. I was actually the state seismologist, as far as my job at the University of Washington. There, every couple months, I would go to the emergency management meetings for earthquakes, and we'd talk to our senators and university representatives, who were liaisons with the government, in order to argue for the programs we needed funded. In Seattle, we regularly talked to government officials at the federal, state, and local levels.

ZIERLER: A nomenclature question. Where is there overlap between earthquake prediction and earthquake early warning? Or is there none?

VIDALE: It's nomenclature. Really, you can use the words for the same thing. Earthquake early warning is predicting the shaking, most often, the shaking coming in the next five seconds. If you're really lucky, you can anticipate motion a minute or two ahead of time if the biggest faults break starting far away from you so that it takes time for the rupture and the waves to approach. But prediction's generally taken to mean seeing earthquakes before they happen. And we're doing terribly at that. There's basically almost no progress in the last 100 years on predicting earthquakes. We're able to say that if there's a big earthquake, there's an increased chance of an aftershock or even another earthquake. But it's not enough of a certainty that there's much even to do with that warning. We'd love to predict earthquakes. There are actually a lot of efforts here at USC and other places to still try to predict earthquakes. I saw a talk during lunchtime today in which they're putting out arrays of arrays to try to get the sensitivity in the vague hope that there's something to see. It's an active topic, but frankly, I think most of us are pessimistic, yet it's one of the drivers for funding, so we'd love to see it succeed. But with each passing year, the chance of something being missed is fading. And Caltech was a center of earthquake prediction back in the 1970s. But it just didn't work out.

ZIERLER: It begs the question, when you look historically, if there's basically been no progress in earthquake prediction over the past 100 years, that leaves us with two possible conclusions. One is that earthquakes simply are not predictable, and the other is that we have not yet mastered the tools and theory so that we understand the science of prediction. Where do you fall on that continuum?

VIDALE: We know that earthquakes are intrinsically predictable. It's a physical system, and it obeys the laws of physics. We can see the tectonic plates moving. We have a decent idea where the faults are. It's just a question of if we can ever understand enough detail to see what's coming next. It's entirely possible that one needs to know everything with an impossible precision in order to know where the fault's going to break in the next big earthquake. But it's also possible that there's something to see that we haven't figured out.

ZIERLER: Maybe to make the question a little more specific, for those who say that, for example, Los Angeles is due for the big one, that it's only a matter of time, versus other people who say the very basis of us being due for the big one presupposes that there's a cyclical nature of these things, where would you answer in that regard?

VIDALE: It's certainly true that there are going to be earthquakes in the next thousand years. The tectonic plates move, the energy's going to the strain. It's inevitable. The question is just how specific we can be about when the next earthquake's going to strike. We know it's going to happen. The question is about the distribution of possible times. Is it anywhere between now and 500 years from now? Or can we say it's somewhere between 50 and 60 years from now? Can we recognize it a week ahead of time? It's all a question of specificity.

ZIERLER: How much is part of it simply not having enough sensors in place?

VIDALE: More data is good. The issue is, how much do you need? Until we see something before an earthquake, it could be it's just that you'd need an instrument every square meter or more on the fault. It's just undefined how much you'd need. The issue is, we don't know on what scale the nucleation of earthquakes occurs. It's entirely possible the nucleation of an earthquake happens within a square inch. Then, we'll never be able to look at every square inch and know how a rupture would spread from there. Or it might be that there's a five-kilometer zone that has some noticeable change of properties in a way that has a good likelihood of leading to a big earthquake, then we're on the verge of seeing it.

ZIERLER: For your own research, where are you more on the theory side, and where are you more on the observational experiment side?

VIDALE: I'm kind of a mix. I don't go out in the field much, although I have been involved in experiments that have collected data, usually some grad students or collaborators of mine have gone out in the field. On the other hand, I say I'm not a theorist, but I've actually come up with a number of theoretical developments. But they're simple things. I'm not a very sophisticated theorist, but I have made a number of fairly widely used simple theories.

ZIERLER: What are some of the theories you've developed?

VIDALE: One is embarrassingly simple. It's a formula where, if you have a square, and you know when a seismic wave got there on three of the corners, it tells you when the wave gets to the fourth corner. It's like a solution of the quadratic equation, almost. It's just got one square root in it. But I described it in a paper. Basically, you solve that equation on a grid. And it's gotten, I think, close to a thousand citations. I guess the idea is that people didn't take a formula like that and put it on the grid, then find all the travel times at once. People had been, prior to that, shooting rays. They'd send off some energy in a particular direction and track where it went. They'd send out 100,000 rays, and hopefully it would hit everywhere they were interested in. But when you do it on a rectangular grid, it just marches very quickly, and you get the whole solution. That's something a lot of people have used. But the theory that underlies it is no more mathematical than what I just said.

ZIERLER: What is the underlying theory?

VIDALE: Any time you write the simplest equation, I guess you can call it a theory. The theory is just that energy propagates as a wave, has a wavefront that advances over time. That's not a very controversial theory.

ZIERLER: I'm curious where you've seen opportunity to look at applications, either with sensors or tools. What are some of the real-world applications that might have value beyond geophysics and seismology?

VIDALE: I think that simple thing I just mentioned about how to calculate the arrival time of energy in a grid has uses beyond seismology. But generally, most of my work has been in the confines of seismology. Some of the things we can do with seismometers have a general purpose. We've looked at signals from space shuttles, signals from exploding houses, submarines in the Puget Sound. I saw a couple of explosions offshore Washington, and I never did figure out what blew up. We can see exotic signals in our arrays that come from power plants. We recorded a signal once to help a guy argue that drilling for the local subway was unreasonably noisy and got them to change their construction practices. Seismometers are pretty versatile. Whenever we put out a lot of instruments, the first thing we have to do is make sure we understand all the signals, so we often see a lot of exotic things that had nothing to do with our original experiment.

ZIERLER: Because you have not found the need to do a lot of field work, what does that tell us about remote data collection and analysis? How much of a seismologist can you be simply having access to the internet?

VIDALE: Seismology's unique in that we have a very good data archive, the data management center, which is there in Seattle. It's part of IRIS, Incorporated Research Institutions for Seismology. A lot of fields are envious, and a lot of other countries are envious. We collect data from a lot of other countries, and we can just log in and collect all the data at once if we like. In seismology, it's easy to get a lot of the data. On the other hand, a lot of times, to get a good look at something, you have to go out and get some fresh data. I guess I said I wasn't a field seismologist, but some of the experiments I've been in have collected tremendous amounts of data. I was part of the Array of Arrays array that we put up in Washington. We put out eight dense arrays, all aimed at listening to the tremor coming from the subduction zone. That was, at the time, one of the more intensive instrumental operations. Of course, running the Seismic Network in Washington, we had close to 1,000 instruments across the state, so we actually had a team of 10 or 15 people dealing with the software, installation, and maintenance of equipment. I guess when I say I don't do field work, it doesn't mean that I'm not involved in projects that have a lot of field work.

ZIERLER: In the way that, for example, the state of Washington is saturated in instrumentation and sensors, what areas of the world are really deserts in this regard where you'd like to see more data analysis and collection?

VIDALE: I tend to be science-driven. To me, places that need instruments are places where there's something we don't understand. The topic of the seafloor always comes up because it's far harder to put instruments on the seafloor than it is on land. We scoped out the price of monitoring offshore of the Pacific Northwest, and we came back with the price tag of half a billion dollars. It would be useful to understand the prospects of the next giant earthquake offshore the Northwest, but it's not so useful that it justified half a billion dollars to put instruments on the seafloor. The seafloor's probably the most under-monitored place on earth. Of course, the move now is to put instruments on other planets. As far as I know, the Moon had a few, Mars has one seismometer. Those are very under-covered areas.

ZIERLER: Beyond the funding limitations, what are some of the technical limitations in getting sensors to the seafloor?

VIDALE: The technical problems are twofold. One is getting the instruments there, the other is getting the data back because you can't transmit data through water with any speed. To put instruments on the seafloor for monitoring, you have to run a cable out to them. And if you put 50 instruments on the seafloor, that means you're running thousands of km of underwater cable. That's where it gets very expensive.

ZIERLER: Let's now go back to the historical narrative. To set the stage for your time at the Seismo Lab, did you go into Yale having an interest in geophysics already? Or you developed that over the course of your undergrad experience?

VIDALE: I went in there interested in physics. Physics just seemed like the way to solve things. And I had relatives who were physicists, and they all seemed to have good careers. But in physics in undergraduate school, each class gave you a problem set every week, and there was no end in sight. You go to grad school and get more problem sets. I took a geology class junior year, actually with a friend of my mother's, Brian Skinner, and it was so much more fun. For example, I came up with that simple theory that was useful. You don't have to be a rocket scientist to get to the frontier of research in geology. We could get right to the questions that people were working on for research. Senior year, I added a geology major. I took 11 geology classes out of my 12 classes. I wound up with both geology and a physics B.S. degrees. I went from there into seismology, stepping into the Seismo Lab at Caltech.

ZIERLER: At Yale, is there an analog to the Seismo Lab? Is there a specific institute devoted to seismology?

VIDALE: No, there was just one seismologist I remember, who you talked to, Emile Okal. He was there before he went to Northwestern. There was another guy in oil seismology, Mike Oristaglio, but he shortly left for industry. Nobody else was in seismology there. It was before geodesy was a field. It was really just me and Emile. In fact, one of the inducements was that they were changing the curriculum of the geology department. There was a course they were phasing out, but I needed it for my major. Emile gave me a class where I was the only student. He'd lecture me for, like, two hours once a week, and I'd just sit there in the class and talk to him. It was fun. He's a fun guy.

ZIERLER: Was Emile the one who made you start thinking about the Seismo Lab in Caltech?

VIDALE: Yeah, actually, I was systematic as an undergrad. I went down the corridors at Yale and said, "I'm interested in seismology. Where should I apply?" They gave me a list of schools. Caltech, Stanford, MIT, Scripps, and I think Northwestern for a safety school. I went out and visited all those places, and what I noticed was that everywhere had a Caltech seismologist. Everyone I was talking to was from Caltech. I thought, "I guess I'll go to the source," and that's where I went.

ZIERLER: What year did you arrive in Pasadena?

VIDALE: 1981.

ZIERLER: What were your impressions when you first got there?

VIDALE: That's hard to remember.

ZIERLER: Are you an East Coast guy? Had you ever spent time in California?

VIDALE: I went to high school in New Mexico, so I wasn't entirely East Coast, but I'd spent 10th grade in Bethesda, 11th and 12 in New Mexico. Before that, I was on the East Coast, New York, New Haven, briefly in Philadelphia. California was great. The weather is fabulous, the Caltech campus is like a small town, manicured grounds. The Seismo Lab was kind of incredible, a big building full of seismologists, that giant entryway. A lot of other students who were all very capable. I was quite impressed. I'd already visited there. I thought about going to Caltech for undergraduate school as well. But the grad students, at least then, were much more normal than the undergrads. Yale had a much more appealing student population and curricular variety than did Caltech back in the 1980s for undergraduates.

ZIERLER: Once you got a sense of what was happening at the Seismo Lab, what were the big debates? What were the professors working on at that point?

VIDALE: That's a good question. The big one I remember was whether the mantle was layered or the whole mantle convected. It was pretty lively in the sense that I don't think Don Anderson ever gave up on a layered mantle. Brad Hager, who went to MIT a long time ago, was in favor of whole mantle convection. And Mark Richards, who's now the provost of the University of Washington, gathered some key evidence to argue that the mantle convected as a whole. That was certainly one of the issues. There was always the struggle to build the Southern California network. That wasn't really a science question, that was finding the resources to install the latest instrumentation in sufficient density to both monitor the region and have the tools for the science. That's been progressing ever since the Seismo Lab was formed.

ZIERLER: What were some of the most important instruments at that point at the Seismo Lab?

VIDALE: I didn't have anything to do with it, but the light gas guns in the sub-basement were very active. Tom Ahrens had an incredible number of projects, trying to understand what happened when you bashed two rocks onto each other. I guess it was really bashing a piece of metal on a hunk of plastic into a rock. That was very active. I think for the most part, we gathered worldwide data. The best instruments were the ones someone else was running. The network was great, but back then, we recognized that a lot of the problems were global and weren't so focused on the local network. In fact, they kind of discouraged use of the local network in those years as being more parochial problems. Another debate that was ongoing was how deep the continents were. There was the idea that the continents are maybe 100 or 200 kilometers deep, and the mantle flows underneath that, or maybe the continents were 500 kilometers deep with these big keels as they move through the mantle. Don Anderson and Tom Jordan, who was next door to me until I moved offices last month, argued, I think rightly, that these keels do often go quite deep. Of course, by the time I got there, earthquake prediction was pretty much over. The fiasco of the Palmdale bulge, just embarrassing. People were staying away from earthquake prediction.

ZIERLER: What was embarrassing about it?

VIDALE: It was most likely just wrong., although there are some divergent points of view. They did some surveying out in the Mojave Desert. They thought the whole desert was puffing up by a couple of inches. It made the cover of Time Magazine, it drove big funding programs, it was a competition with the Russians and the Chinese for who could predict earthquakes. But it turned out they just changed the way they leveled and forgot to take into account that light bends in a gradient. The Mojave Desert hadn't gone anywhere, and it took a while to realize that.

ZIERLER: What about computers? Was your sense that the computers were becoming more modern than what was available a decade before?

VIDALE: Yeah, that was a big struggle, getting the latest computers. In fact, I'm not sure if you've heard this from other people, but my impression was that it was a crisis when I first got there. There was this old Prime computer that not only was slow, but it wasn't clear who was going to pay for it. As I recall, Rob Clayton came in and worked a deal where he sold off the computer, actually got some money in exchange for that old computer, and put in a new computer that was more capable and affordable. The first in a line of computers that Caltech's acquired. Of course now, at least last I heard, there was a giant computer in the basement using some fraction of all of Pasadena's power with an incredible capacity. I'm not sure that's still running. It almost sounded unsustainable.

ZIERLER: What was the process of figuring out who your thesis advisor would be?

VIDALE: It was a little bit complicated. When I visited all those colleges, a lot of the best people had been students of Hiroo Kanamori, so he was my first choice. He's just incredible.

ZIERLER: What do you like so much about Hiroo?

VIDALE: He's always cheerful, he knows everything, there were just no problems. I started with him, and he gave me a nice little project, and I think I wrote a paper there my first summer, a small paper. But eventually, I ran into trouble because he was so cheerful, nice, and vague that I think he kind of knew how everything was going to work but left it for the student to figure out. He either looked up some 50-year-old paper that answered everything, or else he'd go, "I don't know, let's see what happens." I don't know if I wanted more guidance, but I was talking to a lot of different people because the doors were always open, I wandered up and down the hall. The way it worked was, you needed to have a couple of projects. I talked to Rob Clayton, and he said, "I've got a project for you. We've got this code. If you step up the numerics from second order to fourth order, it'll run faster. Why don't you do that?"

I did that, and that involved taking the code and reducing it to assembly language. You take a code, then the computer breaks it into these tiny steps so it can run on the simple circuits of the computer. If you step in and look at that very detailed breakdown into simple steps that the computer makes, you could make it more efficient. I upped it to fourth order, I reprogrammed the assembler to get another factor or two in speed, then we had a very fast numerical simulation. Technically, it's a fourth-order finite-difference equation. We did that, and I think we wrote a paper about that, not sure. I then tried to use our new methods to solve core issues.

ZIERLER: When you say the core issues, what were they?

VIDALE: Basically, we were trying to push back the frontiers in a lot of areas, but some were more central than others. I had the code that worked, and I was wandering down the hall, and I often wound up in Don Helmberger's office. Don was another cheerful guy, and he just loved spreading out plots and looking at them. He was engaged in lots of projects because he was so fun to work with. He generally had, like, five students, and they tended to be working on different things. He had all kinds of datasets and questions. That's how I found my thesis advisor. I just wound up in his office the most. I think in the end, I was technically joint between Rob and Don because Rob had provided a lot of the methods and was interested in the problems, but Don was the one really kind of driving what it was we were figuring out with the research.

ZIERLER: Was there overlap in the research agendas that informed ultimately what your thesis was on?

VIDALE: My thesis was a grab bag. It was a little bit of nuclear tests in the Aleutian Islands, a little bit of how far down slabs go and how easy it is for us to see them, partly what kinds of patterns of shaking develop across the basins in Los Angeles. Then, it was partly a chapter of just equations and theory for how to make a two-dimensional calculation look like a three-dimensional calculation. I just stapled them together. It was not one of these theses that had a compelling narrative from beginning to end.

ZIERLER: What were some of the conclusions or connecting points between the four discrete areas?

VIDALE: I think the only connecting point was that they all used that finite difference technique. In fact, that was a little bit of a grievance I had and still have. I went to Caltech, and what I most came out of there with was developing four or five numerical methods rather than picking a problem and just going to find all the methods that addressed that problem. It took me probably a decade to figure out that I really didn't want to be just writing numerical codes, I wanted to try to get at some core science issues.

ZIERLER: The emphasis was breadth, not depth, at the Seismo Lab?

VIDALE: Well, everyone was different. Don Anderson really wanted to know about the earth, although he went off into some wild theories. By the time I was there, he was focused on contradicting deep plumes and whole mantle convection. And he continued to do that for the rest of his career. But he had his eye on real issues. And he had combined geochemistry and geophysics, so that was pure science. Then, Hiroo wanted to know all about earthquakes, so he was well-focused on a very serious problem, although early in his career, he'd looked at the deep Earth with tremendous effectiveness. Then, Tom Heaton, who was with the Survey when I was there, before he went to the Caltech faculty over in engineering, was all about figuring out strong motions, earthquakes, and the hazard side of seismology. Rob Clayton was strongly into techniques, he liked really nifty techniques and datasets.

ZIERLER: Where did you gravitate, with all of those approaches to emulate?

VIDALE: I just bounced around the hallways. I think I most wound up with Don. I know seismology a lot better than I know anything else, but it does apply in a lot of areas, so I'm forever looking for places to apply the seismology. If that means making some new technique, I have to dive in and do that. If that means digging up some different data, I can pester my friends until I can find the codes to use it or find the colleagues to work with. I guess it would really be Don Helmberger because he had methods, datasets, and he was interested in a range of problems. I'd say it would have to be that Don. Don Anderson was difficult. Our first year here, there was an upper level class that was, I guess, whatever he wanted to teach. But he taught geochemistry, and most of us either didn't understand it, or he didn't express it clearly. But he seemed to figure that none of us were worth investing energy in and went off to find other collaborators, I think.

ZIERLER: This is more of a generational question. Obviously, you wouldn't have been around to make the comparison. But by the time you got to the Seismo Lab–obviously, when it was at the mansion off campus, it was physically, and in some ways, intellectually, an island. How well-integrated did you feel with GPS and the campus in general with the Seismo Lab right there on the corner?

VIDALE: We were integrated well enough with the geologists and planetary scientists, but not that well with the campus. We took some math courses, and there were the campus activities, but we mainly hung out with the other Seismo Lab and geologists, somewhat with the planetary scientists. We didn't see that much of the physicists, chemists, and biologists. We still were kind of off on the corner there on the campus.

ZIERLER: Was there any interface with JPL?

VIDALE: There was some, but I didn't really participate. I think just as I was there, people were starting to get into GPS, and I think that was through JPL. I took a class once with Elachi, looking at remote-sensing photographs. But for the most part, we didn't see the JPL people that much. I think Brad Hager worked with them, but not so much the rest of us.

ZIERLER: After you defended, what opportunities were available to you? You stuck around Caltech for your post-doc?

VIDALE: For a month. I only did the post-doc because we all had to pick a date when we were going to graduate, so I picked the date, then I went back to my office and stapled my papers together, and I went, "I'm done. Can I just move this date up?" That's the only reason I did a post-doc. I had an offer to do a post-doc with Durk Doornbos with NORSAR, up in Norway. But then, my wife and I got an offer to go to Santa Cruz. It was a fairly nice position, we were splitting 12 months of research funding, so six months hard money, six months soft money, and we didn't have to teach at all. That's where I went from there.

ZIERLER: How long were you at Santa Cruz?

VIDALE: I was there for about five years, my wife for about ten.

ZIERLER: That was a faculty position?

VIDALE: It was a research position, not tenure track.

ZIERLER: What were some of the main areas you were focused on at that point?

VIDALE: I think I did more with numerical methods. I still thought numerical methods were more worthwhile than they are. I also kind of applied my methods to whatever they would work for. "Everything looks like a nail when you have a hammer," whatever the saying is. Some earthquake location, more modeling of slabs. I wrote up some of those numerical papers. But I didn't really get onto more interesting research until I got out of there.

ZIERLER: When did you get involved with the US Geological Survey?

VIDALE: That was the next part of my career. I went up there for about five years. I lived in Santa Cruz and just drove up. That was kind of doomed once we had our daughter because I couldn't commute two or three hours a day with an infant. Also, the Survey was getting kind of too applied for my tastes. It's all good work, but getting to a university brings a lot more interesting ways to spend one's time.

ZIERLER: How much freedom did you have to pursue what you wanted to with the USGS?

VIDALE: The freedom was sort of the issue because they let me do whatever I wanted for a few years, and I was like a kid in a candy shop. They had all these arrays, a thousand stations on the West Coast, that they were doing kind of routine earthquake catalogue work with. I took them and started to use them to look at mantle structure. But eventually, it was clear that I was going to get drawn into more operational aspects of the USGS, and that also contributed to my going to UCLA.

ZIERLER: Tell me about the position at UCLA. What was exciting about that for you?

VIDALE: Well, it was great fun. UCLA had a nice department, kind of a cast of characters with Leon Knopoff, Paul Davis, Dave Jackson, Mark Harrison. They're all very strong personalities. Keilis-Borok showed up there, against my wishes, I should say. It had a rich variety. Good students. Although, what kind of doomed my stay there was, I kind of backed into running an institute, the Institute of Geophysics and Planetary Physics. It's a very strange story. The guy who was running it got annoyed that his wife wasn't getting a faculty position, and he came to me and said, "I'm going to hand this off to you. You're going to run this next," even though I was one of the junior people in there, and there were, like, six National Academy guys all 20 years older than me in IGPP. I was supposed to run it, but he wasn't going to tell anyone until he left town. It was the strangest thing. I said, "OK, sure, why?" But then, I had to run this institute that had 20 faculty, just fractional appointments. Most of the FTE were in a department. But biologists, astrophysicists, mathematicians, geologists. Andrea Ghez, who is in there, got the Nobel Prize recently. It was surreal. I could go on for a long time, but basically, IGPP was something that started in the 1940s and that UC Systemwide had been trying to claw back ever since. They were slowly chipping away at the prerogatives of the Institute. It was kind of a nightmare. I was happy to get out to the University of Washington, although that was another funny story. I was sitting in my office one day, and Emily Brodsky came by and said, "I sent off the recommendation for your wife to the University of Washington." I said, "What? What do you mean?" And she said, "Oops." I went down to my wife and said, "You applied to the University of Washington?" She said, "Yeah, I figured if they hired me, they'd hire you, too." And that's actually what happened. I got a call from there when they got her application. They said, "Well, you'd do better as the director of the Seismic Network, and she'd do better as a regular faculty member." We moved up to Seattle. It was great up there, I liked it a lot.

ZIERLER: Was that your escape hatch from the Institute also, all the admin duties?

VIDALE: That made it much more appealing. Scientifically, I was more active down at UCLA, but at the University of Washington, they do have excellent geophysics and flowing data from the seismic network, so there are just unlimited projects one could do, even if it's kind of a sleepy subduction zone compared to Japan or New Zealand.

ZIERLER: Was there a similar institute there? Or what department were you housed in?

VIDALE: It was very similar to UCLA. It was an earth and space sciences department. It concentrated on seismology, but also glaciology, geomorphology, and it had a component of space physics. It was a nice department, actually one of the better ones at the University of Washington.

ZIERLER: How did being up in the Northwest change your research?

VIDALE: I shifted entirely to the subduction zone because I was running the Network that was monitoring the subduction zone with its volcanoes, so I shifted gears. To me, that's one of the things that keeps me a well-rounded seismologist. Sometimes I work on earthquakes, sometimes I work on volcanoes, whatever comes in front of me is interesting. It's that much easier if I already know the other sides of what's in the data.

VIDALE: Seattle's beautiful. Running the Network was great. Again, the detail that made it best was that my network manager, Paul Bodin, did the hard work and just left me to do the things I could do well. I think the fun part, though, was just being able to argue for what we wanted with the government for having the flow of data from all those instruments, having interest from the public in what our research was finding. It was all very gratifying.

ZIERLER: As the Washington State seismologist, what exactly were the responsibilities?

VIDALE: Oh, nothing. We made up the title. Actually, somebody made up the title before we got there, and I just used it. But I was diligent. I went to meetings at Camp Murray, engaged with the city and the state as best we were able to, trying to push them toward what they should be doing. Seattle has a real problem with trying to figure out how to fix their old buildings. We were pretty active.

ZIERLER: Was there state funding? What made the position official at all?

VIDALE: That's exactly right. It's a detail of the budgets, but when I went there, there was $60,000 a year that came to the Network as part of having the State Seismologist. Actually, the day I took the job, the previous person holding that, Tony Qamar, was killed in a logging truck accident. Very sad, but when I showed up, there was this line of funding that was uncommitted. But the first thing we did was bump up the budget, by going to the legislature, to$400,000 a year. Then, I could actually hire people to do all the things I didn't want to do and couldn't do well myself - carry the batteries up the mountains, wiring the computers, etc. It was very fortunate that after I showed up, we talked the state into upping the support for the Network.

ZIERLER: Directing the Pacific Northwest Seismic Network, what opportunities were there just in terms of interacting with new datasets?

VIDALE: Kind of limitless. The permanent instruments were in the ground, so those just flow. Then, if some interesting geophysical phenomena comes up, we can look into it. One of the last things I did was, I looked in a catalogue, and there was this cloud of earthquakes deep under Washington. There were few enough that I could look one-by-one to check if they really existed, and it turned out that three-quarters of them didn't. I figured out which events really happened and wrote a paper about the patterns that were revealed when we cleaned up that cloud of seismicity. We wrote a lot of miscellaneous papers. We wrote papers about how the Seattle basins respond to shaking, how much they amplify the motions, how much of a problem that is for bridges and tall buildings. Probably, the second-best thing about being in Seattle was having Art Frankel. He's another seismologist, and he's one of these guys who's pretty much always right, and everyone respects him because they know he's right. And he's grumpy, not complacent. So he was fun. He made a good colleague to make sure that when I worked on the basins out there, we were working on real problems, not just numerical fun.

ZIERLER: Were you interested at all in Mount St. Helens when you were up in Washington?

VIDALE: Yeah, I was interested, but it's kind of over-studied. It's so active that it warrants a lot of work. But most of the obvious things have been done, and it's covered with instruments. In fact, it is still somewhat of a mystery in terms of the configuration of the earthquakes under the volcano. The structure of the volcano's complicated enough that we get unreasonable answers whenever we try to look at the cloud of seismicity. We just haven't figured out how to focus that set of earthquakes. And actually, I did start an experiment, the iMUSH experiment that Ken Creager wound up doing. First, Olivier Bachman took over, then Ken took over from him when Olivier went to Zurich. We just put out four or five different kinds of experiments across Mount St. Helens to try to get as good a look at the plumbing of the volcano as we could. I'm not actually sure what we saw. It's messy under there. But we did spend several million NSF dollars with five institutions over the course of about five years.

ZIERLER: Between emergency management officials and property owners, how much worry was there about the next big eruption from Mount St. Helens?

VIDALE: There was a big one in 1980, and it erupted again in 2004 in a minor way. The thinking is, it takes 100 years to recharge St. Helens, although we don't really know that because we can't really track how the magma's refilling over time. There wasn't so much concern for St. Helens. Also, with the blast from 1980, there weren't that many people living nearby. That 1980 explosion was the worst case. They say it's a 2,000-year eruption. Technically, it blew out the side, so it cleared out everything about twice as far as they had previously thought was dangerous, which is why it killed 50 or 60 people. St. Helens wasn't so much the worry. There's a little bit of worry with the Newberry Volcano, where, if I remember right, they were putting in efforts to get out geothermal energy. It wasn't a real danger, but local residents are always concerned when you're pumping water through their local volcano since, for all they know, we're being irresponsible. The pumpers weren't us, that was actually a company, but we had to monitor the activity, so we heard about the concerns

ZIERLER: It sounds like you were having a good time in Seattle and weren't necessarily looking for the next opportunity.

VIDALE: That's true. I was pretty well-settled up there. I still sort of wish I was up there, although I came down here to run the Earthquake Center, but that actually didn't work out that well. Partly because Tom Jordan, another Caltech grad, did such a good job running it that he kind of used up the opportunities. There wasn't that much left to propose to do. He had also been a tremendous diplomat. Turns out it was kind of fragile. I didn't see that I was going to be a good long-term solution, and some other people saw that, too.

ZIERLER: When you say diplomat, diplomat to whom?

VIDALE: I guess diplomat, salesman, fundraiser, rainmaker. In terms of just finding resources for a group of people that's already doing a bunch of the projects at which they are best. I did well in the Northwest, I think, partly because people had been under-funding the work up there, so it was time to focus a lot more attention on it. And it's still a place that's getting increasing attention. Whereas, Southern California is as well-studied as anywhere in the world, so it's getting increasingly hard to find the next thing to do. Then, the guy who took over for me, Yehuda Ben-Zion, is doing a great job. He wants to put out many thousands of instruments on the San Andreas Fault system, and he's one of the people betting on there being something to see before an earthquake. He's looking with ever-greater sensitivity to find out if there's anything to see before a big earthquake. And there's a lot of science that goes along with that, understanding the mechanics of fault zones, making better catalogues of the seismicity to see how one triggers the next. But at the core of it, to be the biggest success, it would have to see something we've been missing so far that tells us how close we are to a big earthquake.

ZIERLER: With this effort, where's the basic science, and where are the applications in terms of preparing ourselves?

VIDALE: The science really is trying to understand the earthquake cycle. You can do science on 100 different things in Southern California, and SCEC does. It funds 100 little grants every year, it has collaborations with about 50 universities. It's working on all aspects of earthquakes and many aspects of geology. As far as the core issue of understanding the fault system and predicting earthquakes in particular, then it's really aiming to see just how big an area nucleates before an earthquake. Can you see it in the seismicity patterns, in the velocity structure, in other things like the attenuation, anisotropy, or deformation? It's really focused on whether there's a big noticeable nucleation before a big earthquake.

ZIERLER: Did you stay affiliated with SCEC after you stepped down from the directorship?

VIDALE: Just like the other 500 people in SCEC, so I don't have any leadership role in it here, which is fine. I haven't stayed all that connected. It was a little complicated when I stepped aside. I think trying to run SCEC was a good move, but it was really gambling, and mattered that my daughter is and mom was in LA. They gambled to see if I could do it, and I partly took that job because I thought no one else was stepping up. It didn't look like I was going to do that great. I don't know if Yehuda has a better chance than I had, but he's certainly got some exciting new directions that could work.

ZIERLER: When did you get involved with the Earthquake Early Warning Working Group?

VIDALE: Basically, Berkeley spearheaded the early warning system for the West Coast. They had a workshop right after the huge earthquake in Japan in 2011. Richard Allen was savvy enough to know that it had a better chance if it was a West Coast-wide system than if it was a Northern California system. They invited Tom Heaton and Hiroo as well from Southern California, two of the progenitors of EEW, and me from the Pacific Northwest. We all presented to agencies and foundations that included the Moore Foundation at a workshop. Also, Mark Richards, now the UW provost, who was a dean at Berkeley at the time, helped set it up. The Moore Foundation decided that they would give us \$6 million to push this early warning system along for the next couple years, to develop it to the point when the USGS could invest in it for the long term. It dates back to 2011.

Then, from Seattle, I coordinated the Pacific Northwest effort, which only went online after I left. It was years in the building, getting the software running, instruments in the ground. The real challenge is telemetry, trying to get the signals from the instruments to the computers fast enough that you can get a warning out before your five-second window has passed. And they're still in charge of the Oregon-Washington part of the operation from Seattle. Another thing I did was add Oregon to the mix. Before I arrived in Seattle, Seattle ran the whole Northwest. But that wasn't a good arrangement because then the Oregon senators were not lobbying with the Washington senators to make sure we had healthy funding, so we gave the Oregonians their share of the network, and that's worked well. Both logistically, they can take care of their state better, and politically, we all have more of a push in DC.

ZIERLER: Did coming back to Southern California, to USC, did that refocus your concentration on Southern California at all?

VIDALE: It did, although I really tried to understand the entirety of SCEC, so I really spent the first year or two at SCEC trying to understand the range of research that was going on. It was actually just a little difficult to assimilate in two years. Those 100 projects that people had been working on for decades were pretty complicated. I didn't really start that much research until I left SCEC. After I left SCEC, I worked on the inner core, partly because people have been neglecting the inner core for a decade or two, and partly because I wasn't that eager to be in the middle of SCEC anymore.

ZIERLER: People specifically in inner core seismology or across the board?

VIDALE: Across the board, I think. In terms of the motion of the inner core, this idea came up in the late 1990s about it rotating. People actually believed that for a decade or two. They thought, "OK, we figured that out." The last few years, there's been a rekindling of interest in the inner core because we're finally getting the resolution to see some of the interesting details. At the same time, a fraction of us don't believe that steady rotation interpretation anymore. We're wondering, "Is it moving? Is it oscillating? Isn't it moving?" With that question open, it becomes a more interesting target.

ZIERLER: When you say, "The resolution to see," what are we looking at, absent actually drilling down and being there?

VIDALE: Well, we're essentially looking at the echoes. We look at the seismic waves to try to figure out the structure from how the waves are mangled as they go through the inner core, and that's a powerful tool. I'm not really sure why it wasn't until a few years ago that we got these improved pictures of the inner core. There might've just been not that much interest. Now, every month, I see another paper claiming structure in the inner core, so it's definitely picked up.

ZIERLER: To bring the conversation right up to the present, in the way that you like to switch up your research agenda, what's next after the inner core? What do you have your sights set on next?

VIDALE: I'm still interested in pretty much everything I've ever been interested in. I'm always trying to figure out what data hasn't been analyzed, either new data or data that has a new angle. Actually, I thought I had a new project last summer because I was looking at the seismicity catalogues, and I saw these strange columns of seismicity. I thought, "I don't know if those are real or not," so I got another catalogue, and I saw more columns of seismicity, I got a catalogue from Japan, and I saw these columns, so I thought, "This is pretty interesting." But after I think two months, I realized it was all an artifact. All these studies had it because they all use a similar technique. When the station coverage isn't good enough, things kind of smear vertically, and you see a column. I thought I had another topic, and I spent a couple months on it, but I didn't. I really don't know. Every time I get a new dataset, I don't know what's going to happen until it happens. And I'm wondering if I'm going to get left behind by this machine learning, this whole new toolbox out there that students are learning. Also, writing Python is something I'm learning, but a lot of people grew up with it, and it's second-nature to them. In instrumentation, fiber optic cables have new kinds of data. It's always a battle to find something where you have an edge. We'll see. I don't know what's going to come up next.

ZIERLER: Two last questions to wrap up this great conversation. Looking back at the Seismo Lab, either with specific mentors or just institutionally, is there a way of doing geophysics and seismology that you learned at the Lab that has stayed with you in terms of your approach, your intuition, the kinds of projects to work on? Anything that stands out over the years?

VIDALE: I think the strength of the Seismo Lab is that it's big and excellent. It's just better to know the range of possibilities whenever you look at the data. Also, to have a network of people who have gone through the Seismo Lab. Helmberger's students, we sort of consider ourselves a family. We had a retreat in Singapore a few years ago, and there were, like, 50 of us. The resource of having those connections as well as the broad seismological background, I think, is the best element of the Seismo Lab.

ZIERLER: Last question, looking to the future. You hit on exactly what I wanted to ask you because it's really the big question. Machine learning. For you, where are there concerns that we're going to outsource our minds to the machines and not understand the underlying science, which is a real concern for up-and-coming generations, and where is that productive, where there are things we do now that the machines can just do more efficiently, and it provides more bandwidth for us to do what we do best?

VIDALE: The machine learning's really just a toolbox. We still have to direct it toward what we're looking to find. It can be very efficient. Although, I think the real danger, I guess, in seismology is that we're going to learn much of what we wanted to learn. We can take it to other planets. That's going to be vibrant for a long time. But we have to be flexible, too. If seismology is losing its utility, we have to be able to move into geodesy or other satellite fields. I don't see machine learning as a threat. I do think the number of problems we can approach is finite, and I sort of think seismology arose because of oil industry and nuclear test treaty monitoring, then earthquake hazard mitigation. All those things, there's some level of precision beyond which added progress is not going to warrant paying to keep hundreds of seismologists employed. I think we have to keep our eye on our ultimate goals and adjust our techniques and even fields of interest to stay focused on where there's the most utility.

ZIERLER: You seem to indicate that all of the low-hanging fruit in the field has already been picked.

VIDALE: We never know until we get it, but some things are definitely getting harder. There's a question of just how worthwhile it is knowing where every fault in the ground is if you're just after a hazard map, or how much we want to know just how thick the continents everywhere in Earth are. We have to have our eye on discovering something useful. We can always spend our time doing research, but seismology's thrived from being singularly useful in helping society. But that doesn't mean it's going to be a preferred tool forever.

ZIERLER: We'll have to see, as you say. John, this has been a great conversation. I'm so glad we were able to do this. I'd like to thank you so much.

[END]