Charles Sammis (MS '68, PhD '71), Seismologist

Charlie Sammis has conducted important research in a wide range of seismological fields, including analysis of fractal fragmentation of the Earth's crust, examination of instabilities in fault zones, and using non-linear dynamics to understand seismicity. Beyond our planet, Sammis has studied tectonic patterns on Venus, and he is a pioneer in applying rheological methods to study Earth's interior.

Before arriving at Caltech, Sammis studied at Brown University. His thesis at the Seismology Laboratory was on the structure and properties of Earth's interior using seismic data. Sammis went on to faculty appointments at Penn State and USC, and he has maintained a long-term collaboration with the U.S. Air Force Research Lab to utilize seismology to better understand nuclear weapons testing and arms control treaty verification.

DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Thursday, May 5th, 2022. I am delighted to be here with Professor Charles Sammis. Charlie, it's great to be with you. Thank you for joining me today.

CHARLES SAMMIS: Sure thing.

ZIERLER: Charlie, to start, would you please tell me your current title and institutional affiliation?

SAMMIS: I'm currently an Honorary Professor at the University of British Columbia in Vancouver.

ZIERLER: Now, how long have you been at the University of British Columbia?

SAMMIS: About three years at this point.

ZIERLER: What were the circumstances of being named to this honorary appointment?

SAMMIS: I'd been a Professor at USC for over 30 years. Then I retired, and I tried to decide what to do during my retirement. Since I have no other talents, I decided that I would stick with this. But what really motivated the move was my younger son is a professor here at University of British Columbia in chemistry. Since we had no close family or close friends in Los Angeles, we decided we would sell our house in Los Angeles, and move to British Columbia. Also, it was getting hotter and hotter every year [laugh] so we decided we'd go for the cold and rainy rather than the heat.

ZIERLER: Charlie, besides having that nice honorary connection, does it allow you to still be connected in the field? Are you active in the literature? Are you doing research currently?

SAMMIS: Yeah, I just wrote a paper last year, and I'm lucky enough to have a seismology group here at UBC to interact with. There's a professor that I work with here, and with his students. This honorary professorship carries all the privileges of a professor. I can actually raise money and advise students if I want to. I don't quite have the energy for that anymore. It's nice. It's all the nice things about being a professor without having to raise money and support students.

ZIERLER: Charlie, a nomenclature question. There are so many areas to the field. There's geophysics. There's seismology. There's Earth science. There's mineral physics. What's the best sub-discipline to describe what you've done in your career?

SAMMIS: Well, I would say it's certainly geophysics. As a subfield of geophysics, it's a mixture of rock mechanics, materials science and seismology. It's really interpreting seismic velocities and structures in terms of physical models. I was also involved also in earthquake source mechanics. I've done a lot of work in that field. It's about what's going on in an earthquake, mechanically and so on. When I was at the University of Southern California, I had a joint appointment in geophysics and materials science. I've had a fairly substantial footprint in materials science as well.

ZIERLER: Charlie, where in your research are you more on the observational experimental side, and where are you more on the theoretical side?

SAMMIS: More theoretical. I do a lot of computer modeling, not serious modeling, but simple physics-based modeling of earthquake mechanics. It's a combination of macro- and micro-mechanics. I depend on seismologists for observations, which is nice. Here at UBC I have a colleague, Michael Bostock, who's a first-rate observational seismologist. He has lots of interesting results with his students and I sit on all their meetings and try to think - could I do anything with this? [laugh]

ZIERLER: How have some of the major advances in computational power allowed you to do new and innovative things in modeling?

SAMMIS: Not a lot. I work on small physical models. It's mostly MATLAB. I don't go into heavy computation. I do most things on my personal computer. I hook into the network, but I don't really rely on that. My colleague does that. Michael Bostock, the seismologist, does heavy analysis of large data sets with his students. I'm a bit like a parasite on that. What are their interesting results? How might I help with the interpretation?

Fractal Fragmentation

ZIERLER: Charlie, some technical questions as they relate to your research. First, fractal fragmentation, what is that? What does that mean?

SAMMIS: That was probably the one most creative and lasting thing that I did. The idea is that faults are full of ground up crushed rock. We did some observation measuring the particle size distributions. They turned out to be power-law. There was no peak value. It's just a pure power-law distribution. That was around the time that all this fractal analysis was getting popular. I made up a really simple model for why the crushed rock looked like that, and it all had to do with nearest neighbor interactions between the particles - how they interact with each other. It was a very simple model and, yet, it predicted the observed value of what they call fractal dimension, that is, the observed slope of the distribution, and it predicted a lot of things about it in a very, very simple way, which was satisfying. Most fragmentation models up to that point depended on assuming some initial distribution of the cracks in the rock, and then activating them, and then blah, blah. But none of them led to this fractal result. I got the fractal distribution directly from the geometry, how they were being loaded by each other, in a very simple way. I got a lot of mileage out of a very simple [laugh] idea.

ZIERLER: Charlie, we've all heard of fault zones, of course. What's a granular fault zone?

SAMMIS: At first glance, a fault is a planar interface between two rock masses that slide past each other. That sliding process produces a layer of crushed rock. In other words, as the fault surfaces slide past each other, they grind up the rock, and leave a layer up to several meters thick of crushed and ground up rock. The process I kind of figured out is how does that work, and why does it lead to the distribution of particles that we see?

ZIERLER: Charlie, a few questions about sister disciplines as they relate to your interest in geophysics. Let's start first with rheology. I haven't heard rheology talked about much in the world of geophysics. How do you apply it to what you do?

SAMMIS: Rheology is the study of how things deform. Things are either brittle and they break, or they're ductile and they flow, or some combination of the two. Another possibility is friction between sliding surfaces. That's what we call rheology. In other words, if you put a stress on something, how does it deform? In the Earth's interior, I started out looking at that, how would deformation occur in the Earth's interior? That's all ductile. Then I went to Penn State. I started out at Penn State, and I worked on that for a while. Then I came to USC and, of course, earthquakes were all around. I started working on brittle deformation, and then friction, and that's when this study of grinding of rock began. That's all a very brittle process. Rheology is all about understanding and then being able to quantify different deformation modes, and how they depend on temperature and pressure and so on.

ZIERLER: What about nonlinear mechanics? How is that relevant to your work?

SAMMIS: I did some work on nonlinear mechanics. That was the whole field of chaos - and fractals is part of that. The idea that most people up to about the end of the '80s or so approached physical problems by trying to simplify the problem, and then get a definite answer. For certain nonlinear systems, you can't do that. The interesting things are in the detail. Often, you can't even get a single answer. You can only get statistically answers, and you have unexpected emergent phenomena. There's a whole field that studies a wide range of such phenomena. One example people give is the shape of a cloud. How do you describe a cloud? Well, it's sort of an oval but then it has bumps on it. Then the bumps have bumps, and so on. It's sort of that idea that if you try to do the dynamics of cloud formation, you'd end up with nonlinear differential equations that don't have one answer. At the end of the day, you could say certain things about the way cloud features scale with size rather than the exact shape of a cloud. You have a statistical answer, in a sense. That's sort of the idea. It turns out most things that are the least bit complicated are like that. Even a simple pendulum is like that if you take into account all the things that perturb the motion. Even the motion of a pendulum is not predictable in the end. It's sort of that idea. Geophysics is full of that because we have a lot of very complicated systems, earthquakes being one. Meteorology's like that too. Obviously, you can't describe the weather for more than a couple of days, and then it goes haywire. [laugh] It's sort of that idea.

ZIERLER: Charlie, what about fractal geometry, specifically about how you study the instability of earthquakes?

SAMMIS: The question is, what is the instability that causes an earthquake? Most people agree it's a frictional instability. You have two surfaces, and they're trying to move relative to each other. Stress builds up and, finally, something happens, and it slips to produce an earthquake. That slip is triggered by a friction instability. Friction is fairly easy to describe, but not so easy to understand [laugh] Friction is a fairly complicated phenomenon even for two planar surfaces in contact. But what happens if you have two planar surfaces separated by some sort of a distribution of particles? Then it gets a little more complicated. What's the model for that? How does the distribution of particles affect this friction instability, and how do fluids enter in, and so on?

ZIERLER: Charlie, have you done consulting work with energy companies because of the obvious overlap between what they do in geophysics?

SAMMIS: No. I've done a lot of not necessarily consulting work but I've done a lot of work for the Air Force on monitoring nuclear explosions. That's probably the most applied thing I've done. I'm not very good at doing applied things and making money. [laugh] I would like to but I just haven't gotten into the consulting business very heavily. I've done a little bit but not much. The thing closest to that is this question of coupling underground explosions. That was a time when there were nuclear test ban treaties, and we were trying to make sure that the Russians didn't cheat on how big were their explosions, and so on.

ZIERLER: Charlie, at what point in your career did you become interested in plate tectonics and the planet Venus?

SAMMIS: That was because I had a graduate student, who's quite well known now, Bruce Banerdt, who is the head of the Mars program at JPL. While working on his PhD he did part-time work up at JPL. At the time, radar images of the surface of Venus came out which showed intricate fracture patterns, and we were all surprised. The surface of Venus is very hot, and everybody thought it would be characterized by ductile flow structures. In fact it appeared very brittle, and it had all these cracks and these crack patterns. The question was, what are they telling us? Banerdt and I worked on how could you interpret these fracture patterns to measure the thickness of basalt flows on Venus? The answer to why you have brittle fractures on Venus, is because the rocks are very, very dry. As soon as you take all the moisture away, the rock becomes brittle, even at fairly high temperatures on Venus. If it had the least little bit of water, it would flow. Like most of the problems I have worked on, something interesting showed up and I backed into it.

ZIERLER: Charlie, was that sort of a book-ended project? Was there anything that came out of the Venus research that was useful for your work here on Earth?

SAMMIS: No. [laugh] Not really. It was a one-off. We couldn't even find other places on Venus that were as good as the example we found where we first looked. It was interesting because the spacing of these parallel fractures set a thickness to the process that people couldn't measure other ways as easily. It was interesting in terms of the rheology of Venus, but I never found anything like that on Earth that's been that useful. The closest thing was a study by my first Ph.D. student, Mike Ryan at Penn State, who used the striations on basalt columns to estimate the cooling rates of lava flows on Earth.

ZIERLER: Charlie, a few questions as they relate to some of the big debates that have happened in seismology and geophysics over the course of your career. First, a generational question. By the time you had graduated, you finished up at Caltech, what aspects of plate tectonics were settled science, and what was still up for debate?

SAMMIS: I graduated in '71, and plate tectonics was pretty well established at that point. All the ocean floor evidence was in, and all the geological evidence. No one had found any evidence that contradicted it. There were still people who were questioning it, but the jury was really in by the time I graduated.

ZIERLER: Your work on the inner core, does that relate at all to some of the debates about the extent to which the inner core moves?

SAMMIS: No. [laugh] I didn't do a lot of work on the inner core. It turns out that that was Paul Richards, one of my classmates at the Seismo Lab (currently at Lamont), who figured out the different rotation rate between the inner and outer core. No, I followed that, but I didn't really contribute in any way.

ZIERLER: What about earthquake prediction? Was there a period of significant optimism that we would get to a place of earthquake prediction, and where are we now, from your perspective?

SAMMIS: There was a lot of optimism when they managed to predict the Haicheng earthquake in China based on precursors. The idea of prediction is that you look for something that happens before the earthquake. They saw water level changes, and this and that. I spent some time pursuing the possibility that the seismicity that comes before a big earthquake gives you some hint. If there were foreshocks that could obviously be identified, earthquake prediction would be easy. But there aren't. They're fairly rare. But there's a whole background low-level seismicity goes on all the time. Part of the application of nonlinear mechanics was the idea that you had sort of a critical state in the crust when it was ready to have a big earthquake. As you approached that critical state, you could see the approach to that state as a signal in the small low-level seismicity. It would build up to form a recognizable pattern. I did a lot of work on that, and that was very controversial. It was fun. The problem is with earthquake prediction, you've got to be really right. [laugh] It can't happen sometimes, and not at other times. If you're going to make public predictions, you better be right because, if you're wrong, people will be upset that they had to take steps, and that weren't necessary. The state now is, I think, there's still controversy. Some people think it's impossible. Some people think there are precursors. What I simply think is there are precursors some places but not others. They're not universal. In some places, the hydrology is just right, and you see a water level precursor. In some places, maybe the fault network is such that you'll see some sort of precursor in the seismicity, but it seems to be not very universal and, therefore, not terribly useful for public policy, other than being scientifically interesting. But in terms of society, we aren't close to making short-term predictions, weeks to months, which we'd like to have. We now make media predictions based on the detection of seismic waves inside of a few minutes. That works very well because it has a very sound basis. But this idea of something precursory that would give you a week or a month warning, we're still nowhere with that.

ZIERLER: Charlie, when you were at USC, were you involved at all in efforts to do earthquake early warning?

SAMMIS: No, I didn't get involved in that. We didn't do a lot of that. Caltech did most of that, setting up the network and so on. That was pretty well established before we even got involved. I don't even know the extent to which we are now, really. That wasn't a big thing at USC.

ZIERLER: Charlie, all of your work on nuclear monitoring, was any of that classified? Did you have to do any work in government facilities?

SAMMIS: No. The answer is I was on an Air Force contract for most of my career. But I couldn't take a contract and do any classified work. USC wouldn't let me do classified research. I've been to classified meetings occasionally, but the research, I couldn't do anything that was classified.

ZIERLER: Currently, what are you interested in right now? What's fascinating to you in the field?

SAMMIS: Well, since I've moved up here to Vancouver and I'm a sort of opportunist, I've shifted my research interest from big strike-slip earthquakes like those on the San Andreas network of faults in Los Angeles, to big subduction zone earthquakes that occurs every 400 years or so in the Pacific Northwest. There are a lot of very interesting phenomena that go on in these subduction zones like what they call "slow earthquakes" and "low-frequency earthquakes" and "tectonic tremor". The question is, what's going on there? What are the mechanism of those phenomena - how do they work? I published a paper last year on an interpretation of low-frequency earthquakes; why do they repeat, why do they occur in families? What's going on in those places where they repeat, and so on? I've gotten involved in the research that the seismology group here is doing. I have had a close look at their results in the way you can only do when you're retired. I have lots of time to [laugh] play around with this.

From Brown to Caltech

ZIERLER: Charlie, let's go back in the narrative. Let's go back to when you were an undergraduate at Brown. Obviously, you were focused on physics. Was geophysics or seismology on your radar at all as an undergraduate?

SAMMIS: No, not until the last minute. In my senior year, I was trying to decide what to do. I thought, well, physics, gosh, I was probably about the third-best undergraduate physics student at Brown, and that was a small program. I thought good grief, I'm probably nowhere in the world physics hierarchy. How can I use my physics training in a related field —so I looked at biophysics. I had a cousin who was a biophysicist up at MIT, and I went and saw him, and he was wearing a white lab coat and cooking things.

ZIERLER: [laugh]

SAMMIS: But then I looked across, out of his window, and here was a big building with antennas and propellors on the roof. What's that? That's the geophysics department. I thought, OK, and I ran over there, and got an application form. Then I came back to Brown, and I applied to MIT and Caltech and a few places. I thought, geophysics sounds about right. At Brown, I did a lot of acoustic work. I did a senior honors thesis in acoustic scattering and in poly-crystalline solids. I thought that's a pretty good background for this seismology stuff, so I'll try that.

ZIERLER: Now, were there geophysicists or seismologists on the faculty at Brown?

SAMMIS: No. There was one tectonicist that was one of Barclay Kamb's Ph.D. students, who I met after I'd already been accepted to Caltech, [laugh] and I went over and introduced myself. But geological sciences were not a big deal at Brown at the time. Now, of course, they have strong geophysics and planetary science programs.

ZIERLER: Charlie, when you got to Caltech and the Seismo Lab, how much catch-up did you have to do relative to your cohort, given that you were not a geophysics undergrad?

SAMMIS: Almost none. It turns out [laugh] that geophysics is not much of an undergraduate major. Almost everyone was either a physics or a mathematics undergraduate. None of my colleagues were geophysics undergrads, so we were all in the same boat. That's what Caltech looked for when I was there. You want somebody that really has a solid math and physics background, and then you can teach them geophysics. But it's harder to do it the other way around. I was sort of a drop-out physicist. [laugh]

ZIERLER: Charlie, when you got comfortable at the Seismo Lab, you got the lay of the land, what was your sense of the big debates? What were people excited about at that point?

SAMMIS: We completely missed out on plate tectonics. It was all Earth's interior and earthquake seismology.

ZIERLER: Yeah, I've heard that before. What's that about, this idea that the Seismo Lab missed out on plate tectonics? How do you understand that?

SAMMIS: I think we just didn't have anybody that had a real interest in it, because the real speciality at Caltech at the time were either earthquakes or Earth's interior. Don Anderson was a very, very strong leader. He was interested in Earth's interior, and so that's what I did my thesis on, the structure and properties of the Earth's interior. It's just that there was nobody pushing plate tectonics at the time.. We were not in direct contact with everybody. We lived in the Seismo Lab up on the hill. [laugh] People came through all the time but they were mostly people invited because they had something to do with the stuff we were working on, either earthquake mechanics or Earth's interior.

ZIERLER: Now, for you, did you settle on an interest and then pick an advisor, or did you have an advisor and that shaped your interests?

SAMMIS: I formed an interest, and then picked an advisor. At Caltech, like a lot of places, people applied to work with someone. But without having a geophysics background, with not having a clue, I just showed up, and then just drifted into this. Then Don Anderson, I talked to him about maybe doing something. But it wasn't any sort of rational decision. In fact, when I applied to MIT and I applied to Caltech, I got accepted first at Caltech. I immediately accepted it because I thought that's great, California and surfing. I'd been on the East Coast my whole life. About a week later, I was accepted at MIT. I wrote back the same day to say I'd already accepted the Caltech offer, and I wasn't interested. [laugh] I got a letter back immediately saying that they'd never been rejected so fast, and maybe it would interest me to know that Frank Press was leaving Caltech and coming to MIT. I think that was supposed to influence my decision but I didn't know who Frank Press was and I didn't know [laugh] anything about geophysics. The answer is no, I didn't pick it based on an advisor. I just found something interesting when I showed up, and I worked on it.

ZIERLER: What was interesting to you? How did you come to that topic?

SAMMIS: I guess I tried to think what I could do best with my physics background. That led to problems in the Earth's interior. There are problems of high-temperature creep, and what is the Earth made out of? At that point, all we had was seismic velocities and densities, so we had to turn seismic information into material properties. I did a whole thesis on the lattice mechanics of minerals. It was a bit of a dead-end, but it interested me at the time.

ZIERLER: Now, did Anderson become eventually your thesis advisor?

SAMMIS: Yeah.

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

SAMMIS: It was the composition of the Earth's interior from seismic data. He was doing it based on empirical relations between rock composition and seismic velocity, and I thought, well, with my physics background, maybe I can do it a little bit better, and look at the mineral structures, and figure out what the elastic velocity should be based the structure. It was fun but it didn't go anywhere.

ZIERLER: What were computers like at the Seismo Lab when you were a grad student?

SAMMIS: We had a computer lab, Booth Lab. I think it's still there, but it was just one big mainframe, and it was all punch cards. You had to go over and sit in the basement. They had little tables, and punch-card machines. You'd sit and type punch-cards, put them in a box, and give them to an operator upstairs. He'd run it, and you'd go and have a cup of coffee, and then come back an hour later, and see if your program had run or not. It was not very efficient. [laugh] It took forever. If you made a mistake, you'd wait two hours to find out you had done some stupid thing. One day they put it in an enormous flat-bed printer that had a huge roll of paper, like 100 feet of really fancy graph paper. I wrote a program that unspooled about 50 feet of it before they could stop it [laugh] because I had some glitch in my program, and they hadn't worked out safety protocols. I was singularly responsible for motivating a safety protocol to make sure no one could unspool $1,000 worth of graph paper in a few seconds. That was my achievement at this computer center.

ZIERLER: Charlie, tell me how your thesis came about.

SAMMIS: As I was saying, it was a problem that Anderson was interested in. I also had a colleague—Hartmut Spetzler—who was doing ultrasonic measurements of velocities in minerals. I just thought, well, I could help interpret that by putting together models of the minerals, little simple force and mass models that would predict how structure affected velocity. I interpreted Hartmut's data and wrote a paper with him. Then I also wrote some things on the Earth's interior. But it turned out to be sort of a naïve endeavor, since it was much more complicated. [laugh] I don't know if I made any contribution but it was interesting.

ZIERLER: Is that to say you realized later that it was more complicated?

SAMMIS: Well, I thought maybe I wasn't getting there because nobody in the Seismo Lab knew about lattice mechanics. So I got a NATO postdoc at the University of Bristol, and worked for a year with a man who did lattice dynamics. Then I realized it was hopeless. [laugh] It was a hopeless undertaking. I had hoped that maybe I wasn't getting anywhere because I didn't have the right advisor. It turned out not to be true. It was just a stupid thing to do.

ZIERLER: What else did you do, or was that the entirety of your thesis?

SAMMIS: That was it. I looked at a bunch of different structures, and I talked about how you could calculate elastic properties from lattices.

ZIERLER: What was Anderson's style like as a graduate mentor? Did you work closely with him? Did you check in with him infrequently?

SAMMIS: [laugh] Well, we were all in that small, little Seismo Lab up in the San Rafael Hills, so I saw him every day. We'd all have coffee together. But he was pretty much a hands-off advisor. I was doing my thing, and then he just let me. He supported me, and he let me do my thing, which was great. At that time, there was probably a lot more money in geophysics, certainly at Caltech. I think he could afford to do that, and I appreciated it. I had a chance to develop my own style [laugh].

ZIERLER: How did you see your thesis research contributing to some of those bigger debates that we were talking about?

SAMMIS: I don't think I made any contribution. My thesis was a disappointment to me. I think all my achievements came after. I certainly took a lot out of Caltech in terms of how to do science. But I think I didn't really get on the right track until I had a [laugh]—I hate to say this—till I had my first sabbatical at USC. Then I went to Cambridge and worked with a Professor Mike Ashby, who was my real inspiration, because he thought about problems the way I did. You have mind melds with certain people, and he was very good at what he did. He was a Royal Society professor, and he was in material science, but he was interested in geophysics problems. That's when I think my career really took off, and that was like '83. [laugh] I hate to say it. I graduated in '71. That was like '83. [laugh] It was about 12 years of casting about, not doing any harm but not doing much of anything.

ZIERLER: Charlie, when you defended, what opportunities were available to you?

SAMMIS: I didn't even look around for jobs. I decided that I was going to do a postdoc. I said, well, this NATO postdoc looks like a good thing. Go to England. I was not very career-oriented at the time. I thought, well, a lot of people at Caltech do a dynamite thesis, and then the next step is, obviously, you go to some top-notch place, and do that. My thing was I felt fortunate to graduate. I wanted to go to England to see if I could make something out of the thesis that I thought I was really unhappy with. I did that. Before I left, I'd gotten a job offer from Penn State, and so I had a job to come back to. Penn State's a good university, but I never made much headway there either. It's a funny place. I don't know if you know Penn State. [laugh] It's kind of off in the middle of nowhere, and it was kind of a strange place. I spent five years there. They weren't happy years. Then I got an opportunity to go to USC. I had a couple of my colleagues from Caltech that hired me at USC, and then things were good from then on.

ZIERLER: Now, what did the NATO postdoc at the University of Bristol allow you to do?

SAMMIS: Like I say, I set that up with a man who was an expert in this lattice mechanics. How do you calculate material properties using the structure of the lattice. That was his speciality, so I went and worked—I pursued that for a year, and I didn't really get very far with it. I wrote up a paper when I got done, but its one of my least-quoted papers.

ZIERLER: Charlie, was the program at Penn State, was it a big program in geophysics?

SAMMIS: Yeah. It was a big program in geophysics, but it was very industry-oriented. We put out a lot of master's students in the oil industry. A few were good PhDs but not very many. It was a fairly large faculty, but a lot of it was exploration, geophysics, and monitoring mining spills in Pennsylvania. Shelton Alexander, who was the head of the program then, had a program with the Air Force, and I kind of got to know the Air Force people then. But then once I had done this work at Cambridge on fracture mechanics, then they came to me and said, "Can you help us with these problems?" But it wasn't until I actually did some real work at Cambridge that I got plugged in.

ZIERLER: Now, Cambridge, was that in the early '80s when you were there?

SAMMIS: Yeah, I think it was '82/'83 or '83/'84, something like that.

ZIERLER: What was happening in fracture mechanics at that point at Cambridge?

SAMMIS: At that point, we were trying to understand nucleation and propagation of fractures and how they led to macroscopic failure. I made up a model of nucleation, growth of fractures, and their interactions. Mike Ashby then turned it into what we called damage mechanics or micro-mechanically based damage mechanics, where damage is the fractures in the material. You could use that to predict failure? But for the underground explosion, they wanted to know if you set off an explosion, and you crack a lot of the rock around an explosion, how does that influence how the energy from the explosion gets converted into seismic waves. What do you see, monitoring the explosion? How does it depend on the fracturing that goes on? I could make a contribution to that.

ZIERLER: Charlie, at what point in your career did you become involved administratively with the AGU?

SAMMIS: I didn't. [laugh] I served as Associate Editor for a while, but I did that just for a few years. But I didn't really do any administrative work at AGU. I was a member all the time but I never really contributed to it.

ZIERLER: What about NASA and their planetary science work? Did you do any advisory research for them?

SAMMIS: Well, in a sense, I was on the program review panel for a few years. I went down to Houston, and helped them evaluate proposals, and then put out the call for new proposals, and so on. I got involved in that for a few years. But it wasn't really at the upper program level. It was more putting out the call for proposals, and evaluating proposals, and sort of influencing how the program went that way.

ZIERLER: Charlie, tell me about the decision to move to USC, and join the faculty there.

SAMMIS: Well, as I said, that was because my wife and I were very unhappy at that stage. It was kind of funny and isolated. People describe it as a little bit of Alabama between Pittsburgh and Philadelphia.

ZIERLER: I've heard that, yeah. [laugh]

SAMMIS: —[laugh] It really is very rural, and we're city people, so it didn't really suit our lifestyle. We went to USC mostly to get back to the big city, and also my career wasn't going anywhere, so it was an easy decision to make. When the opportunity came up, we took it.

ZIERLER: Was part of that that just Penn State was not a good research environment for what you were interested in?

SAMMIS: It should have been perfect. It was just funny. They had a material science lab. I never really fit in there very well. It's all sort of strange stuff. It's not very interesting. One of the people there that I was supposed to work with had a student that he thought geophysics should've hired. They hired me instead, so he didn't want to collaborate very much. It wasn't a happy place for me to get integrated into. I don't know. I don't want to say too much about it. It was a long time ago and I think it's changed a lot since we were there.

ZIERLER: Charlie, when you got to USC, was SCEC up and running at this point yet?

SAMMIS: No, it wasn't. That took a while. When I was at USC, there were only three of us in geophysics at the time. There was Tom Henyey and Leon Teng and me. We were all Caltech graduates from the Seismo Lab. They were a little bit older, and they hired me. After that they we hired a few people, SCEC got going when we hired Kei Aki from MIT. That was mostly Tom Henyey, who had a vision of kind of trying to pull together all the various people who work on earthquakes into some sort of collaborative group. Tom was very good at organizing things. He was a good administrator. Kei set it up so that it was very inclusive, which I thought was a key to its success. Everybody could participate, and everybody could get a small grant to pursue some idea, to participate in workshops, and go to the meetings. But there was no big funding. The big funding was all group funding. If there was some big project that everybody wanted to do, we would do it. Kei used to say, "If you depend on SCEC for your individual funding, you're pathetic." [laugh] If you have good ideas, you go to NSF or someplace, and get it funded. SCEC was not the place to fund your program = it's the place to try out ideas, and then leverage them into NSF projects. If we had to put an instrument in someplace or put out an array or something, then that was the kind of thing that SCEC could do because they had funding for that. It was good. Mostly, Tom Henyey and Kei Aki were a very good combination in getting that going.

ZIERLER: Charlie, how did that affect your research agenda, moving to USC?

SAMMIS: By the time that started, I participated in it. It affected it because I think we had a lot of very good meetings and a lot of very stimulating sessions. It affected it in the sense that it identified problems to work on, and how to contribute to the general community effort. That was good. Not a lot of people came through USC - it's not on their bucket list of places to visit, particularly [laugh], for the beauty. Having SCEC there meant we got a lot more people coming through that we could interact with, and that was good.

ZIERLER: Being closer to Caltech, did that allow you to have some collaborations or connections at the Seismo Lab?

SAMMIS: It did, in the beginning. I lived in South Pasadena. It's about a five- or six-minute drive away. At Caltech, I made my strongest connections with the material science department, and I spent a year at Caltech working with Ares Rosakis in materials. Then I published probably five or six papers with Ares, and none with anybody at Seismo Lab. [laugh] It was great. Caltech was a great place, but I didn't really participate in the Seismo Lab that much.

Seismology and Materials Science

ZIERLER: Now, you mentioned material science earlier. What was your point of entrée? What were you interested in, in material science?

SAMMIS: Everything. I just liked this idea. It started even with my thesis of the lattice dynamics approach to material properties, sort of looking at the micro-mechanics of things. The same thing with ductile flow. I wrote a big paper on flow mechanisms, and how they scale with pressure temperature based on little atomic models. I always liked that sort of micro-modeling. Same thing with fracture mechanics, this thing—damage mechanics. It's called micro-mechanical and damage mechanics. You model the growth and interaction of micro-fractures, and how that leads to large-scale failure. It's a pervasive theme in my research, I guess, kind of small to big. [laugh]

ZIERLER: Charlie, tell me about your visiting professorship in Paris in 1987.

SAMMIS: I went there to work with Jean-Paul Poirier. He was an expert in high-temperature creep applications in geophysics. That was before I got into the brittle work. I worked with him on faulting at the brittle-ductile transition. Not a lot came out of that, although it was a wonderful cultural experience. But I can't point to any big breakthroughs I made there. I learned a lot from him because that was his field.

ZIERLER: Then 12 years later, in 1999, you were a visiting professor in Mexico City. Tell me about that.

SAMMIS: That was with Cinna Lomnitz who had a program there. I went down there to work with him, basically, on again on the nonlinear mechanics for earthquakes. We did some work there and, again, that was only six months, and I didn't actually publish anything with him. Nothing big came out of it. We did a lot of work together but it didn't have a big result. It was a tough time in that the university was having a strike. The students had blocked everyone out. In order to get on campus, you had to go past the student barricades, sign in, and so it was a little bit of a funny situation. Also, Mexico City was—it may still be—it was very scary at the time. [laugh] We lived in a hotel, but you couldn't walk around town. If you wanted to go visit someplace, you had to hire a taxi for the day, and the taxi had to pay people—to watch the taxi when it was parked. Mexico City was a mess. It may be better now. But, at that time, it was really a tough place to work.

ZIERLER: Charlie, when did you start collaborating with Yehuda Ben-Zion?

SAMMIS: As soon as we hired him. Yehuda collaborates with everybody. He got interested in stuff I'd done on fault zones. He wanted to write a paper together. He'd say things like "Let's write a paper together on this or that." OK. He was right next door, so we talked a lot. That was just a matter of his ability to find things that he found interesting that people were doing, and to collaborate. He's good at that.

ZIERLER: How did that collaboration turn into not one but two book projects with Yehuda?

SAMMIS: That was Yehuda who pushed that. I would never start a book. Yehuda organized that, and I just made my contributions. But I thought about writing books at various times, but I'm not organized enough or disciplined enough to write a book, I found that [laugh] Yehuda had the discipline to do that, so we did.

ZIERLER: Had the program at USC grown significantly over the course of your tenure there?

SAMMIS: Yeah. It was not much when I started. It was remarkable. It's one of the few things I really found rewarding. It's not a matter of pride in because I don't know that I had a lot to do with it but, really, it started out with just the three of us. They hired me, and people couldn't believe that I'd left Penn State to go to USC. The university itself, it was a place for undergraduates. If you couldn't get into a good UC, and your parents had money, you went to USC. Since then, USC has come up a lot. The real turning point was when they could actually afford to turn down applicants. As soon as they could start to turn some applicants down, then it became a desirable place to go. [laugh] That's all been all uphill. Then the department's been good in the sense that we got a lot of support from the university, and managed to keep a high research profile. When I showed up, geophysics had more funded research than any other department at USC, including like physics and chemistry and everything. We had a lot more research money, and that sort of, I think, propelled us up in the eyes of the administration, and they were supportive, let us hire people, and build the program.

ZIERLER: Charlie, did you retain your Air Force contract when you switched to USC?

SAMMIS: I didn't have it till I got to USC.

ZIERLER: Oh, I see. I got it.

SAMMIS: I got the Air Force contract after my sabbatical at Cambridge. I published all these papers. The scientific director at the Air Force Research Lab at that time, also a Caltech graduate, Bob Blandford saw this stuff, and he thought, gosh, there are a lot of things they could see in the seismic records of nuclear explosions that they didn't understand. Maybe it was this fracturing in the explosions that was producing these strange seismic signals. He talked to me at AGU after I gave this talk when I came back from sabbatical in Cambridge. He said, "How would you like a few hundred thousand dollars year to work on this?" [laugh] I said, "OK, I'm game." That's been a very good collaboration. It supported my career and a lot of students over time.

ZIERLER: Now, was the work for the Air Force related to any of the arms control agreements that were going on in the 1980s?

SAMMIS: Yes, it was directly related to that. It was a question of how big is the explosion, and how do you measure that size using seismic waves? It depends on how much of the energy goes into seismic waves, and how that works. Could the Russians set off an explosion in some weird environment, and we'd not see how big it was? There were all these scenarios. What could you do to hide the size of the explosion? It was also a great mystery. There's also an interesting scientific mystery and that is you think when you have an explosion, all the energy goes out as compressional waves. Yet, there was a lot of shear-wave energy [laugh], sort of sideways motion. Where did that come from? How did that work? I had a large part in solving where that shear energy came from, and why explosions look more like earthquakes—you would think explosions would be easy to identify because they wouldn't look anything like earthquakes. It turns out they look a lot like earthquakes [laugh].. We made contributions as to why that was and how it worked.

ZIERLER: How did you come to that conclusion? What were some of your findings?

SAMMIS: Well, it turns out that if you look at a pressure wave goes out the material cracks, and that actually introduces a shear component into the motion. It's in a micro-mechanical model and, actually, if you break spherical symmetry and integrate over the whole explosion, it produces coherent shear waves, which you wouldn't expect otherwise.

ZIERLER: Charlie, some questions about your teaching and research career. First of all, at USC, were undergraduates getting interested in geophysics?

SAMMIS: No, geophysics undergraduates were very rare. That's true almost everywhere. Geophysics is not a very useful undergraduate major. If you want a geophysics career, you are better to major in math or physics. If you want to go to work for the oil companies, you really need a master's in geophysics. The undergraduates at USC, of course, all wanted to become doctors, so they all went into the biological sciences, or they wanted to go into computer-related fields, in order to design games and make a million dollars. None of them [laugh] thought they wanted to go figure out what the Earth's made out of or anything. That was not on their to-do list. We even had trouble with some graduate students in my early years at USC. They would come over from China, particularly, and other places, and they'd get in the geophysics program, and then they'd take all computer science courses because they didn't really want to do geophysics. They wanted to get the computer science degree so they could go back and get a good job or stay here and get a good job in a silicon valley start-up.

ZIERLER: Charlie, tell me about some of your graduate students. First, what are the kinds of interests that would lead to you becoming their advisor?

SAMMIS: They probably didn't know what they wanted to do. I picked up a lot of students that were a little bit disenchanted with their original advisor. I was kind of the person that would have some interesting sounding projects they could work on.

ZIERLER: The clean-up guy?

SAMMIS: Yeah, they were looking for something different to do. [laugh] They would get interested in what I was doing. I picked up a lot. I recruited a couple students, but mostly I would just pick up strays on the rebound.

ZIERLER: What have been some of the kinds of careers that your graduate students have gone into?

SAMMIS: Well, my most high-profile student, Bruce Banerdt, is now at JPL, and has been since he graduated. He's now the lead scientist on the Mars program. I've had students that have gone to work in consulting companies, even though you wouldn't think because I didn't really train them for that, but they found good jobs there. My last student [laugh] when he finally graduated. I didn't have very many of them work on the Air Force because they couldn't see where that led. I had this last student who wanted to be my student. I said, "OK, let's work on the Air Force program." Now, he is working [laugh] in the Monterey area making flying cars. [laugh] They wanted him because of this material science background, even though he did a geophysics thesis on seismic coupling of explosions in anisotropic rock. He went in and sorted out all their material evaluation programs. Now, he's a star at the flying car company. They do different things. I'm trying to think of some of these other students. I actually sent one student back to the oil patch. I had a student that came to work with me specifically. His father, as it turns out, was the head of exploration at Exxon. He came and did a thesis with me for some reason. He did a seismology thesis, and he did it in three years. He was the only student I ever had that got through it less than four, and probably most of them take five. He got through it in three years. Put his head down and did it. Now, he directs a crew of about 30 people at exploration geophysics for Exxon. [laugh] That's my one success in that industry. But I'm trying to think what all these other students did. A lot of them went into administration. One of my best students did his work on earthquake prediction, and he got a job at Fullerton because his family was there. In the end he went into administration there. Now, he's Dean of Sciences at Eastern Washington University, so he's gone the administrative route. One student did the best thesis ever was the inspector for the patent office. He came and did his thesis [laugh], and it was a great thesis. It was so good, he got a postdoc at Lamont working with Chris Scholz, who's a bright light in his field. Another student Sandry Steacy went the administrative route to Colrain in Northern Ireland and ultimately as dean of natural sciences in Adelaide, Australia.

I've also mentored some very good post-doctoral fellows. David Scott spent a year working with me after he received his PhD at the Seismo Lab. He then took a Professorship at University College London. He now works as a "quant" in the London financial district. Isabelle Managheti spent a year with me after she received her Ph.D. from IPGP Paris. She took a faculty position at Grenoble and currently edits the Journal of Geophysical Research - our premier journal. Finally, I worked with Harsha Bhat following his graduation from Harvard where he did his PhD with Jim Rice. He is currently a professor at the prestigious ENS in Paris.

ZIERLER: Charlie, tell me about becoming at AGU fellow in 2007?

SAMMIS: That was nice. I got all kinds of awards from USC. The two highest honors they give are the Associates Teaching Award, and the Associates Research Award. I'm the only person at USC that's won both of them. That was good. The AGU award meant a lot to me because it signified some outside recognition beyond USC.

ZIERLER: [laugh]

Long Term Partnership with the Air Force

ZIERLER: Charlie, in the years up to your retirement, what were you working on at that point?

SAMMIS: The same thing, just the Air Force research, and working on earthquake mechanics, and so on. I had phased retirement for four years, so I was half-time, so I got paid half-time. That's when I set up my appointment here in Vancouver. I established connections here in Vancouver, and I began teaching and research.

ZIERLER: Now, have you been able to continue the Air Force work in British Columbia?

SAMMIS: No, I've quit all that. I probably could if I wanted to but I've done that for 20 years. I don't have quite the drive. I was more interested in working with the guys here. Plus, the move itself took a lot of energy. [laugh]

ZIERLER: Sure.

SAMMIS: While I was setting up my move to UBC we had a condo here in Vancouver, and the house in South Pasadena. Then we had to sell that. But if I wanted to keep social security, if I wanted to keep my pension, I had to have a US residence, so we rent a small apartment in Hayward. Then we finally bought a house here. A lot of our effort has been just plain orchestrating this complete life change. [laugh]

ZIERLER: [laugh]

SAMMIS: But that's all settled down. But I'm not going to stop working at UBC. I would take up painting or violin playing or something if I could do that, but I don't have any talents. [laugh]

ZIERLER: This is what you do. It's what you know how to do.

SAMMIS: Yes. As long as it gives me pleasure and it's fun.

ZIERLER: Well, Charlie, now that we've worked up to the present, for the last part of our talk, a few retrospective questions, and then we'll end looking to the future. First, with the Seismo Lab, I wonder if there's any approach to the science that you learned while you were there that stayed close with you ever since?

SAMMIS: Yeah. I think I learned an empirical approach to understanding data. When I first showed up with this physics background, I thought you could really model everything in detail. At first I didn't appreciate a lot of the stuff that Don Anderson did in terms of making empirical diagrams of this versus that. My other mentor, Mike Ashby, calls it enlightened empiricism. [laugh] You use empiricism, but you sort of know where you're going with it. I learned that from Don. Don was very good at that, but I didn't appreciate it till I left. But I do all my work that way now. I've given up on being an expert in anything. [laugh]

ZIERLER: Charlie, in the way that when you were at the Seismo Lab, data was not nearly as freely shared. There wasn't the internet, of course. How has the Seismo Lab changed as a result of the democratization, if you will, of data?

SAMMIS: I think everywhere has changed that way. I don't know if the Seismo Lab is particularly unique in that. We all do that now. I'm sure the Seismo Lab excels in massive data analysis. The Seismo Lab has always been known for clever approaches to problems, and for this early warning system, and getting that kind of thing going. They recently hired Zack Ross, one of our graduate students from USC to do major data mining on the seismic catalogs. He's very good. Caltech's certainly been in the lead on that aspect of data mining or big data sets. I was glad to see him succeed. He is doing well.

ZIERLER: Finally, Charlie, for you, looking to the future, given that this is what you know how to do, and you'll keep doing it, what do you want to focus on next? What might be the next project for you?

SAMMIS: I'm going to stick with the analysis of tremor and slow earthquake phenomena in the Pacific Northwest. People have some ideas, but they haven't really figured it out yet. The reason is that I've got very good data sources for that. This fellow that I work with, Michael Bostock, is really at the top of that field, and knows exactly where the observations are at. My hope is that he'll get early key observations, and then I can work with him to understand what's going on. That's what I'm now devoting my energy to.

ZIERLER: Well, Charlie, it's been great spending this time with you. I hope good things happen for you in the future. I'd like to thank you so much.

SAMMIS: Thank you.

[END]