Luciana Astiz (MS '82, PhD '87), NSF Program Director

Born and educated in Mexico, Luciana Astiz arrived at Caltech's Seismology Lab with personal experience with earthquakes—and a drive to better understand them. Working with many prominent professors at Caltech, Astiz conducted important research on earthquake sourcing at a time when modeling and digital stations were becoming sufficiently advanced to provide reliable data.

In her current work at the National Science Foundation, Astiz has played a key role in supporting research in earth science and seismology, and she is looking out for surprises in the future that may yield insight on the mysteries surrounding earthquake prediction.

DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Tuesday, April 12, 2022. It's my great pleasure to be here with Dr. Luciana Astiz. Luciana, it's wonderful to see you. Thank you for joining me today.

LUCIANA ASTIZ: Thank you for inviting me.

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

ASTIZ: At the moment, I work with the National Science Foundation, and I am one of the Program Directors with the program of Instrumentation and Facilities in the Earth Science division of NSF.

ZIERLER: Tell me a little bit about your work. What are some of your day-to-day responsibilities?

ASTIZ: My day-to-day responsibilities are receiving proposals and sending them out for review. All the proposals that come to our program, after they're reviewed by individual reviewers, they go to a panel of experts that we assemble, then they're discussed there. Based on that, my colleagues and in, we're three program directors in this program, allocate the yearly budget of about $20 million for this program amongst the most meritorious proposals to which we recommend awards be made. All proposals are judged on both NSF criteria: intellectual merit and broader impacts. We try to make sure that there's broad representation on the portfolio of awards, not only of topics in Earth sciences, but of the type of proposals: equipment acquisition, equipment development, or to support facilities that serve different communities in Earth science. Awards are recommended to a diverse cadre of principal investigators, whether they're early in their career, in the middle, or more senior and the type of institution. Some of them go to R1 institutions like Caltech, some of them go to state institutions, which serve other types of student populations. It's sort of a juggling act. We choose from the proposals submitted and try to make the best decisions possible because we are stewards of taxpayer money.

ZIERLER: You're so well-positioned, you see the proposals from the most cutting-edge research in the field. What's exciting? What's happening right now in earth science, geophysics, and seismology as you see it?

ASTIZ: I used to be in the geophysics program, but I've been removed from that for a couple years. But I think one thing that's exciting is the use of new instrumentation, like fiber optic cables, and I think Caltech is very well-positioned in that respect, and new methodologies to use artificial intelligence and machine learning in new ways. That's very exciting and cutting-edge in seismology. Also, there's more bridging between different communities, like the seismology community and engineering community. The objective in seismology is basically to reduce the hazard that earthquakes pose, in addition to understanding why they happen, the tectonic setting around them as well as the interior of the Earth. There are many things that can be done with seismic waves, but I think one of the motivations for many seismologists is the possibility to affect society in a positive way.

ZIERLER: Before your career at NSF, when you were closer to the science, what were some of your fields of research? What did you work on?

ASTIZ: I worked on more precise locations for earthquakes. Also, locating very small earthquakes, partly because they provide a glimpse of the background state of stress of the crust, and also help you elucidate faults that are at a depth and may not have a surface trace or topographic reflection, for instance. It's important to have accurate locations of earthquakes at high resolutions. That's one thing. The other thing I was interested in is seismic noise, which is kind of the garbage part. [Laugh] But like the saying goes, noise is noise until it becomes signal. [Laugh] It's sort of trying to figure out what the noise is telling us, and that was sort of a lesson I learned after I finished at Caltech, studying the large Mexican earthquakes. I went to study noise, and I felt like, "Oh my God, I've been really demoted." I went from the flashy big tectonic earthquakes to this seismic noise, but I actually found it to be very intriguing. It tells us things, for instance, about the ways the ocean waves move. There's what we call the micro-seismic peaks.

One of them, the secondary microseismic peak, is the interference of waves in the ocean that creates standing waves. And that's something you see all over the world, which is amazing. Even in the interior of the continents, it's sort of the pulse of the earth, in a way. The primary microseismic peak has to do with the waves pounding on the continental shelves. The frequency at which these peaks occur is different in different oceans. Part of that has to do with how long the waves get. In the Pacific Ocean, because the wind is pounding on the waves for a long time, they are much longer wavelengths than in, say, the Atlantic Ocean. You really have to look at things in context. For example, the seismic noise signature stations in Western and Eastern Japan are quite different. That's because Western Japan is closer to the Sea of Japan to its west while Eastern Japan is bordered by the much larger Pacific Ocean. In a way, the microseismic peak frequencies record the lengths and depths of the oceans. I found that was kind of interesting. Also, the amplitude of these peaks tells us about the weather because the signal storm waves are producing is seasonal. Increasing in the winters, both in the northern and southern hemispheres.

And because the winter is at different times in each hemisphere, they're going to be more noisy at the stations in the northern hemisphere, say, from December to March, while during the northern hemisphere summer, which is the austral winter in the southern hemisphere, that's when their stations located in the southern hemisphere become very noisy. Just to know what's causing this is kind of neat, but it's also very important because if you have a lot of noise, that's going to hide the signals, like small earthquakes. It's important to realize that this other mechanism is going on because if you're not aware of it, you may think, "Why are there more small earthquakes in the summer?" And maybe there are small earthquakes all year long, but they're being masked by the increase in seismic noise in the winter. Then, by figuring out how to filter seismic signals one can sort of open the curtain and uncover small earthquakes. If you were in the theater, you'd have to open the curtain to see them behind this noise. That's one topic.

Another topic I have been interested in more recently are Earth tides, which is the other end of the spectrum, i.e. very long wavelength signals that are very small in seismograms. How do you see the signals from the lunar pull and the sun's pull in seismic stations? Are they recording this? The seismographs, even though we call them broadband, they are sort of narrow-band for those wavelengths, half a day, a day, or whatever. But if you stack long time windows of double the length (at least) of the signal you are interested in, you can see it. Sometimes if you're in a very quiet location, you actually can see it in the raw data because they're quite small signals. My interest in this started because I was involved with collecting the US array seismic data. My interest initially was to try to figure out if the stations were well-calibrated. Some people have done this with other types of stations, so I thought, "Why don't I look at tides to see if the stations in the US array are well-calibrated?" That was my objective. But then, I figured out that most of the stations, when you compared them to Earth tidal models, they're right on the nose. But there are a few stations that are really off, like 20%, which is huge.

Usually in seismology, we're looking at signals that maybe are 1 or 2% different from average models. That's the type of signals you see in tomographic models and things like this. To see a discrepancy of 20%, it's like, "What did I do wrong?" [Laugh] You have to kind of go back and say, "I did everything right, and everything else works, so maybe there's something to be said about it." One of the stations in question was very close to Yellowstone National Park. Since I was working with the US array data collection, I was also working very directly with all the seismic analysts at the Array Network Facility in Scripps Institution of Oceanography. We would always see very delayed signals at this station. I think what was happening was, the tidal signal was letting us see that when we model Earth Tides with the standard model, it doesn't fit because the crust in Yellowstone is very different. The trick is to figure out how different it is and what's causing it. I think it's basically the magma chamber that other researchers have seen in the very shallow crust. But this is another way to verify that observation. It's almost in the fringes. [Laugh]

A Mexican Perspective on Earthquakes

ZIERLER: To establish some context, prior to your arrival at Caltech, your undergraduate education at Universidad Nacional Autónoma de México, obviously, earthquakes in Mexico are a big problem. Is there a specific approach to understanding earthquakes or being educated in Mexico? Did that give you a unique interest or even urgency in studying earthquakes?

ASTIZ: Sure. [Laugh] Throughout my life, I always lived in downtown Mexico City, where it really shakes when there is an earthquake in the Pacific coast. I must've felt in my life about 12 earthquakes larger than magnitude 7 before I went to Caltech. When you have felt this many earth shakes you're very aware that there are these phenomena called earthquakes and that shapes your life. In high school, I actually learned about plate tectonics from my geography teacher in Mexico City of all places. She said, "The people who study these are called geophysicists." I wrote it down because I'd forget by tomorrow, I've always been kind of forgetful. Then, when it became time to decide what to study at the university, I thought, "What was that word?" If I went into the science school, I would have to do either physics or math first, then go into geophysics. But in the engineering school, there was this new degree called engineering in geophysics, where one could learn all these geophysical techniques more applied to the oil industry, mining, and geology.

By the time I arrived to the Seismo-Lab, I had this great foundation on techniques, but not about the interior of the Earth or anything like that. In Mexico, there's a requirement that to graduate, you need to do one year of national service as part of your degree. If you had a non-paid internship, you only had to do it for six months, and if it was paid, it was a year. I got mine at the Institute of Engineering, working with Dr. Singh and Dr. Haskov, who were interested in earthquakes. They were both geophysicists, who were interested in connecting with the engineering community more closely to give them the data to keep Mexican cities safe. Thus, I had that awareness of seismology contributing to society very early on. There was this almost applied science sense. Like, "It's fun to learn about the Earth, about how it works inside, and about all these practical issues too because we're studying earthquakes because we also want to understand, say, amplifications in the Valley of Mexico, or because we want to understand how different the earthquakes are from each other." I would say I had that keen awareness. And I guess my last year at the university, two professors from Caltech visited UNAM, Dr. Hiroo Kanamori and Don Helmberger. I remember meeting them.

They went there to look at the data that the Mexican scientists had gathered for the Petatlán earthquake in 1979. They were talking about all this data, and I was an intern on the edges, listening. At the Institute of Engineering, we were seven or eight of us students(interns), and we were a very close-knit group. Two of them had gone to the United States to do a summer internship in Wisconsin. And they had brought with them all these applications for schools in the US. At this time, there was no internet, and Mexican mail was notorious, and still is, for losing things or arriving very late. They figured while they were in the States, they would gather all these applications to enter many schools. They basically handed them to us like, "I have two of these, who wants the other one?" And I got an application for Caltech. I thought, "OK, I'll apply." But honestly, I had no idea what Caltech was when I applied. [Laugh] I was like, "It's the place those two professors came from." But there was no awareness. And when I asked my professor, Krishna Singh, what Caltech was like, he said, "Oh, it's a small school in California." That was it. [Laugh] I think he didn't want me to get nervous and feel intimidated. If he had told me, "If it was the place with the most Nobel Prize-winners," I would never have applied. "I'm not that smart." [Laugh]

ZIERLER: How was your English prior to your arrival at Caltech? Had you taken it in school?

ASTIZ: Oh my, I actually went to a bilingual elementary school and junior high. By high school, I didn't have English classes, nor at the university. I think I learned most of my English probably from 7 to 14 at school and reading. I studied a little bit to brush up before the TOEFL exam, and I passed it. But at the beginning, it was hard because of the different accents. Since all the programs on TV in Mexico were dubbed, there was no opportunity to listen to English unless you went to the movies and listening to them with subtitles. And my family was not rich, we were lucky to go every few months to see a movie. It was a big treat "We're going to the movies." [Laugh] In that respect, you didn't have that much opportunity to practice it. At the beginning, at Caltech, it was hard particularly with Peter Goldreich, a professor in planetary sciences, compared to New Yorkers, his accent was softer, but for me, it was very hard to understand him.

And he had this idea of asking people to take notes. He said, "I want you to pay attention, so one of you is going to take notes. I'm going to check them, and we'll photocopy them, and give them to everyone." This was how he did it because he wanted people to be paying attention and asking questions. If you're writing, you cannot ask questions. My last name starts with A, so I was first on the list. He turned to me and said, "OK, you're going to take notes today." I said, "I can't." I think he said, "What do you mean, you can't?" I said, "I can barely understand you." [Laugh]. He started by the end of the list, by the time he got to me again I was able to understand him.

Sourcing Earthquakes at the Seismo Lab

ZIERLER: Once you got comfortable at Caltech, you understood where you were, the big issues, what were some of the major debates happening in geophysics in seismology at the Lab at that time? What were people working on, what was exciting?

ASTIZ: We were trying to define what an earthquake source was. We didn't really have the data. It was when the modeling of earthquake sources started. With techniques like moment tensors to really see the difference between earthquakes, there was a realization that the existing instrumentation of the time would only let you see a very narrow band of the signal, and the way we were measuring magnitude, any large earthquake would be a 7 or a 7.2. But it was a 9. But the way we were measuring, it would come in as a 7.2 because we were only looking at, say, 20-second waves. Here, Hiroo Kanamori proposed, " to measure amplitudes at longer periods because the 7s don't have as much content in the longer periods, say a factor of ten, like 200- or 300-second waves, which is really hard to imagine. A 300-second-long wave takes five minutes. In your head, you don't really have a concept of a wave that takes five minutes. But very large earthquakes produce those. And that's how Hiroo defined Mw. That was very exciting.

Now, we could really, truly distinguish the big earthquakes from the large ones. I was also interested in understanding how deeper earthquakes connect to shallower seismicity at subduction zones, that was part of my thesis. And that's something that people have kind of forgotten, and it's still there. Is there any connection? Is there any transfer of stress from the shallow part of the subduction zone to the deeper part of the subducted slab? I don't think that's been resolved yet after 30 years because it's hard to do. There's not as many intermediate earthquakes, and the truly large earthquakes come once in a blue moon or less. That part was exciting to me. Because that might be able to tell us, at least a little bit, of the state of stress at the subduction zone, for instance. We see an earthquake ready to happen. That's something that we don't know. Other things that were exciting, the instrumentation started to be digital. In Mexico City, we had records that were still on drums, being recorded in analog way. When I went to the field and put stations in the field, I still had to smoke the paper with a very smoky candle. You'd smoke the paper, then put it very carefully, so as not to smash it, in this cylinder, let it record, then get it out very carefully, then put a fixing solution like varnish on it. And you measured things like that.

When I started at Caltech, digital recordings started happening. All of a sudden, you could see them in your computer. You could rescale up, down, or stretch time in the seismic traces, and you could maybe see the real start of earthquakes that the previous instrumentation didn't allow you to see. The realization that, say, a magnitude 5, a magnitude 7, or a magnitude 9 all start the same way, so you don't know what the magnitude will be when they start out. And that was something that was being debated. People didn't have a model of asperities in subduction zones, proposed by Hiroo. There was a big debate between Hiroo Kanamori and Keiiti Aki about asperities and barriers. "What is the mechanism? Is it this runaway rupture of an earthquake that stops because there's a barrier? Or is it breaking these asperities, and it just runs out of steam because it uses the energy to break the asperity?" Not because I'm a student of Hiroo Kanamori, I would say the asperities model won, I think there's a little bit to be said about this barrier concept, which we see, for instance, in Southern California, when faults are not continuous, when there's a large step-over.

We have seen it, say, in the Landers earthquake, that some earthquakes are able to jump from a rupture in one fault to another and get bigger and bigger. But some of them, there's this geologic actual barrier, and they just stop. That was all being debated at the time. The techniques of moment tensor inversions were being established. Now, they're routine, and people know about them. They don't even think about it. But at the time, we were validating different methods. There was Adam Dziewonski on the East Coast, and there was Hiroo Kanamori on the West Coast. We were kind of trying to figure out who was right, which method was better and if they differed. If magnitude results were 0.1 different, why was it? How were we measuring it differently? We were trying to establish standards. There was a lot of first-order science that was done then. I didn't want to graduate, it was just so much fun. People kept on telling me, "You're ready to graduate. You have enough." I was like, "Why should I graduate? I'm having fun here." [Laugh] It was a very exciting time, actually.

ZIERLER: Was Hiroo your thesis advisor from the beginning?

ASTIZ: Not from the beginning. I kind of fell into his lap in the middle somehow. I worked with many people. I worked with Clarence Allan, Brad Hager, Don Helmberger, and Hiroo. That was the amazing thing about Caltech and the Seismo Lab. Also as someone coming from another country, I found the fact that when I arrived, they gave you this code of conduct that was one page with four or five things scribbled on it, how to behave ethically doing science, with your colleagues, with your fellow students, with your professors, and everything seemed like common sense, I signed it, and they gave me a master key to the Seismo Lab. Everybody had a master key. I actually thought this was how things were done in the US.

ZIERLER: No, that's just Caltech. [Laugh]

ASTIZ: When I went to other institutions, I was waiting, like, "Where's my master key?" [Laugh] But it took me that long to realize that this was something very special.

ZIERLER: How did you go about developing your thesis? What were the topics you were working on at that point?

ASTIZ: I had worked on different things. It was sort of developed late in my student career. I had looked at smaller earthquakes that had occurred in offshore Mexico, like the doublet in '82,. Then, I started looking at doublets all over the world, e.g., in the Solomon Islands. Why do doublets occur? Why were they common? Doublets are earthquakes that are about the same size and occur very close in time, within a day or two of each other, like twins. I was doing that and trying to figure out some other things about the energy of earthquakes. This was a period with a lot of advancement in techniques because there were not many large earthquakes. There was a lull in these exciting earthquakes. People had time to kind of dive into techniques, instrumentation, theories of what was happening, the asperities or the barriers models. Then, the 1985 earthquake happened in Michoacán, Mexico. I actually saw it coming on the drum in the morning, and I thought, "That looks like something from Mexico. That looks big."

Then, I tried to call my parents, and they didn't answer, and I thought, "Oh, maybe they're busy." Until that evening, somebody told me, "Did you know there was a big earthquake in Mexico?" I was like, "What are you talking about?" [Laugh] Then, I realized how big it was and how the city had been devastated. I actually went to Mexico City the next day. At that time, you couldn't just ask for data from a data center like we do now in seismology, but I had this master list of all the addresses of all the seismic stations in the world. I went to the post office, I bought a lot of international stamps, printed letters asking if they could please photocopy this earthquake on this day and send it to me. At the time seismograms were collected from all the seismological stations worldwide, the station operators gather seismograms each month, package it and send it to the USGS in Albuquerque.

Then, the USGS photocopied them. But with this process, nobody saw the data again for a year or two because it took time to do all these things. I short-circuited the system. By the time I got to Mexico all the letters had been sent, and over the next couple of weeks, photocopies of the seismograms started trickling in from all stations all over the world. I basically had the most precious possession, the list of addresses of the seismic stations. [Laugh] And it was very exciting to be able to look at the data so soon. All of a sudden, I had this amazing dataset that nobody else could put their eyes on, the body waves, the surface waves, everything. That became my thesis. And Hiroo was very excited. It was probably the first time we were able to document and determine directivity in a seismic source, how the rupture was moving. Because we had all these data, like it can be bidirectional or one side to the other, but we hadn't actually seen it in a record. And this was the first time it was actually documented in global records. It was kind of nice. The other topic I tackled in my thesis, stem from an earthquake that happened in Nicaragua, a magnitude 7.2, which occurred at intermediate depth, like 70 kilometers, and people don't think too much about those type of earthquakes at the time.

But they can be very damaging because they occur underneath the continent and are much closer to cities, like in Seattle. The earthquakes that have caused damage in Seattle have been intermediate-depth earthquakes. So, I started to make an inventory of all the intermediate earthquakes to try to figure out globally how they act, how they work within the theory of plate tectonics. At the time, there were all these debates about whether it mattered how old the subducted oceanic crust is, how it age relates, to the angle at which it subducts. There were all these interesting things, and I wanted to put these intermediate-depth earthquakes in the context of this whole big theory. That was part two of my thesis. And I did a lot of other stuff I didn't put in my thesis, like the work I did with Clarence Allan on the Garlock fault. I said, "What should I do for my project, my qualifier?" He said, "There's this big fault in California that nobody looks at. It's the second-largest after the San Andreas, and we have no idea what's happening with it." I said, "OK, I'll look at it." That paper was my first paper at Caltech. It's probably one of the papers I got the most citations on because nobody else worked on that for years and years, so I was the only one to cite. And that was thanks to Clarence Allan. He said, "That's what you should do, look at things that nobody's working on."

The Advent of Digital Stations

ZIERLER: What were some of the advances, both in instrumentation and perhaps even in theory that made your research at Caltech possible?

ASTIZ: In instrumentation, it was the digital stations. I had analog data that had to be photocopied then digitized manually. [Laugh] But at the same time, there were these stations that were funded by the Department of Defense, and Caltech had copies of the 9-track tapes including data recorded by long-period instrumentation which made possible to analyze surface waves of earthquakes, and that's how Hiroo developed the moment tensor techniques, using the surface waves from these digital recordings. Those were not analog recordings, but digital recordings with larger dynamic range. It was sort of at the transition from the analog world to the digital world. Also, the transition from going from what we call empirical modeling, where we built models and understood what happened to the signal if we modified parameters on models, then we'd compared with the signal and try to build a model that would replicate that signal. Then, you tried to fit a model not just one record, but maybe with 5, 6, 20, however many you had. And we did this by trial-and-error. All of a sudden, we had the capacity of saying, "We have all this data." making possible the start of inversions of data to get models. "Let's get all the data and try to give it some parameterization to invert from the data to the model without this trial and error." Those were two things that were happening that were instrumental in my thesis work at Caltech. And I'm very grateful I had the opportunity to do all this trial-and-error fitting, which I think provided me with a very keen understanding of the connection between physical properties of the Earth and the signals, per se.

ZIERLER: As a woman, as a Latina, were you made to feel welcome at the Seismo Lab?

ASTIZ: Yes.

ZIERLER: That was never an issue?

ASTIZ: No. People sometimes made a little bit of fun of how I pronounced some words, including Don Helmberger. But there was a student whose parents were Greek, but he was very good with languages, and he was blunt enough to tell me, "You're pronouncing that wrong." I was like, "How do you pronounce it then?" I had a lot of trouble with short and long vowels in English because in Spanish, we only have one size vowel. [Laugh] It's very hard to hear the difference when your native language does not have those sounds. Until I was able to hear the difference, because I would say words that, with a short vowel, have a nice meaning, and with a long vowel, may not have such a nice meaning, people would laugh at me. But I was like, "OK, fine, you know what I mean." [Laugh] But it was very collegial. I always loved that we were mixed in the offices, students from different professors, classes, and academic stages. There was a lot of mentoring and cross-pollination of fields that doesn't happen at other universities that segregate the students. This professor here, this professor there. I think it's a loss for other people.

ZIERLER: Did you feel like the Seismo Lab was an intellectual center, a place that researchers from all over would come to as a place to present their research?

ASTIZ: Not so much to present it, but to be there. There were people who would come for the summer just to be able to work with Hiroo, or Don, or whoever. There were a lot of visitors from the East Coast, but also from Japan, from Europe. And it was just exciting. The fact that the faculty were committed to coffee time–I don't know if you've heard about coffee time.

ZIERLER: I've heard about coffee time, it's very important. [Laugh]

ASTIZ: The commitment that it was a safe space just to take a half-hour break, for people to just come like, "Oh, look at what I found." It was like, "I'm having trouble with this. What do you think about this? Do you think this fits this theory?" I remember that during one of those coffee times, Tom Heaton, who had been charged to review an environmental study for a nuclear plant in Washington state. He said, "I have read this report, and everything seems very solid. They said, 'Yes, it's safe to build this plant.' But if we really consider our theories of earthquakes, seismic gaps, the earthquake cycle, maybe there are no earthquakes there because we're in a period of quiescence. Maybe this is just a seismic gap. We know that there must be a subduction zone there." Then, Hiroo, Don Anderson, Don Helmberger, Clarence Allan, the students, everybody talked and gave their opinion. It was sort of this daily ritual in which there was a lot of intellectual give-and-take, but in a very honest way with a lot of curiosity. Everything was presented as a puzzle that needs to be solved. "Let's talk it out." People felt safe in saying even crazy ideas because nobody was being judged at coffee time. Seminars were a different thing because they had to be kind of polished. But at coffee time, everybody was safe to say, "I don't follow that," or, "What do you think that is?" You learned a lot. And the faculty were committed to doing it not once but twice daily.

ZIERLER: As a result of being at Caltech, as you looked to the next stages of your career, what are some things you took with you from the way of doing research at Seismo Lab that have stayed with you in the overarching years?

ASTIZ: I think the curiosity, asking questions. Asking questions in a non-judgmental way with the purpose of finding things out. Not really with the objective of being right or wrong but trying to figure it out. I might not have all the answers, but anybody can contribute. I think that's the main thing I took with me. And that you could learn from anything if you really dig a little deeper, that you didn't have to take things at face value. You really had to think about it and say, "OK, wait a minute." Tom is responsible, really, in my mind, for the awareness that there's a Cascadia subduction zone. He put a halt to the approval of that report, because he used intellectual integrity. Shortly after that, Brian Atwater, a geologist, found this dead forest along the northwest Pacific coast. This observation fitted the model of deformation at subduction zones after a big thrust earthquake. The fact that subsidence was observed confirmed as he said, "a very large earthquake what must've happened in Cascadia."

I think the other thing I took is that everything in isolation might not make sense, but you need to look at all the little parts as part of this system. How does it fit? Does it make sense in the larger question? I don't know if that type of thinking was particular to the Seismo Lab, but I think over the years, I've realized that it is. Also, that kind of giving the master key to all the students–you could go into any professor's office at night and get a book. They just had a piece of paper to sign it out as if it were a library, then you brought it back. That, to me, was like having the keys to the kingdom in a way. Like, "This is your home, and you're welcome in any room." That was very powerful. It felt like I belonged not only in the Seismo Lab, but also doing science.

ZIERLER: I've come to appreciate that in the history of geophysics and seismology, there are things that are believed to be true, and new evidence will come along and shake that up, so to speak. From your time at the Seismo Lab, what has stood the test of time? What were some theories or observations that were known to be true and that today remain understood to be true?

ASTIZ: Asperities, the earthquake cycle. That's something that was developed around that time, the understanding that maybe there's micro-seismicity, but where is it happening? Is this illuminating a fault that we don't know about? All those things started to be understood then. I think these are the guidelines we still follow. Also, the realization that the truly large earthquakes are a different beast. I don't think there's that much awareness in terms of how they affect the built environment. Because for instance, say a San Andreas-type earthquake like the 1906 earthquake. Yeah, it's a 7.8, but in 1906, there were no high-rises. I'm actually very worried about what's going to happen if there's a 1906 with all those high-rises in the Bay Area, and with all the landfills around the Bay, all the liquefaction. I think things are well-understood, but it's just too big a problem.

ZIERLER: To flip that question around, what were some of the orthodoxies that have since changed as a result of new research?

ASTIZ: That the duration of the earthquake is important. The time it takes for an earthquake to propagate, to rupture.

ZIERLER: Meaning what? What was thought then, and what is understood now?

ASTIZ: Going back to this analogy I was talking about before about earthquakes being all the same size after 7.2 or 7.3. It was very hard to determine why some earthquakes that are 7.2 didn't seem to do anything, and some that are 7.2 are very damaging. Then, the realization that maybe it was more a matter of the measuring rod, how we were doing it. The ruler we had was faulty and couldn't measure the truly big ones. The follow-up thing to that is that, say, a magnitude 7, the ground is going to break in about 20 seconds, give or take. A magnitude 8 is going to be, like, 60 seconds. But a magnitude 9 could be up to five minutes, like the 2004 Sumatra earthquake. The fact that you're going to have this shaking, it's not only the propagation afterwards, but just from the source, there's going to be, at minimum, five minutes of shaking, plus whatever the ground resonates or amplifies. That has changed, and I don't think the engineering community quite knows what to do with that information yet because it has implications on materials. If you shake something for 20 seconds, it might not get to its breaking point, but if you shake it for a minute, or two, three, four, or five, it might. That kind of conundrum is still unresolved in my mind.

The Holy Grail of Earthquake Prediction

ZIERLER: That gets me to my next question, which is sort of the in-between question. What was poorly understood when you were a student at the Seismo Lab, and what, despite all the advances in the field, remains poorly understood today?

ASTIZ: That's a big question. Earthquake prediction.

ZIERLER: It was the holy grail then, it remains the holy grail today.

ASTIZ: Yes. And some people, may avoid saying prediction because there was such a backlash from politicians, like, "You guys can't predict earthquakes? What are you doing?" But I think the seismology community has learned to be a little bit wary of using the ‘P'-word because it carries too many expectations. The next best thing, I think, is earthquake early warning.

ZIERLER: A nomenclature question. It sounds like earthquake prediction and earthquake early warning really mean the same thing, no?

ASTIZ: No, no. Earthquake prediction is like where I get my crystal ball and tell you, "There's going to be a magnitude 7.5 in Turkey tomorrow at 3 pm." [Laugh] Earthquake early warning, there are a bunch of instruments, and it happens to be that they're in Turkey, and they start feeling the waves, and they go above a threshold that tells you it's a big earthquake. Because the communication travels faster than the seismic wave, you can provide an early warning of the incoming shaking. The communication basically is going at the speed of light, of electrons, while the seismic waves are going pretty fast–five or six kilometers per second for the slow ones, so not that slow–but much slower than the speed of communication. You can say, "In 15 seconds, we're going to start feeling shaking. Hold onto your table, or get out of harm's way." It's very different. Earthquake early warning is based on data that you have at hand. Earthquake prediction might be part of a larger system, maybe measuring stresses.

And maybe it's possible. But we haven't figured out what we need to measure to be able to do it. What we can say in terms of predictions is, we have these models where you know there's going to be a large earthquake in Cascadia. Why? Because paleo seismicity studies, and historical records, and the trees, and all these things tell you that there was a large earthquake in Cascadia in the 1700s. Investigators actually found a record from a tsunami in Japan that they haven't been able to attribute to any earthquake in the world, and they were able to attribute it to this Cascadia earthquake in, I think, January 26, 1700. We actually have the date because there are records in Japan of this tsunami. We know from the trenching that some geologists have done that there were previous earthquakes. In a sense, we're predicting, "Yes, there's going to be an earthquake there," but we don't really know when. In the San Andreas, we know there have been big earthquakes. There's the proof, you look at the rocks, and they've been displaced, there are records in the missions saying when they were damaged or when a tower fell.

We may not have seismic records, but we have other kinds of accounting to tell us there was an earthquake there, and we can do some averages. We can say, "Well, in the Southern San Andreas, we expect there to be a large earthquake every 150 years." We're close to 200 years, and we're still waiting. How regular the occurrence of earthquakes is, we still don't know. We're still grappling with it. Or in Japan, the 2011 Tohoku earthquake. There were a bunch of off-shore earthquakes, they were all around 7.5s magnitude. And people were all happy because no much damage had occurred, and nobody even thought that there could be a bigger one. Nobody else but Hiroo because he's always thought about this model for the Nankai Trough further south in Japan. Now, people are starting to think that maybe these very large, shallow earthquakes are possible in other places, like the Middle America Trench offshore Mexico. We have these regular earthquakes, 7, 7.5, 7.7 that occur every few years and that everybody in Mexico City feels. But there's a possibility of having, say, a large magnitude 9. But we don't have the records to distinguish that.

ZIERLER: Last question, looking to the future. Because, as you emphasize, earthquake prediction remains the big question mark, how do you see the work at the Seismo Lab contributing ultimately to developing the theory, the research, the instrumentation that will make viable earthquake prediction at some point the viable necessity it will be?

ASTIZ: I think they tried to get on that bandwagon when I was a student with a Russian scientist, Boris something. He was all into earthquake prediction. And the Chinese, with the frogs, the dogs, and whatever. They claimed that animals could feel the waves that we humans couldn't feel. I think that the Seismo Lab is frankly wary of it because they got burned then. I think there's room for study and further understanding. One of the observations that has been made tied to some earthquakes is changes in radon, how these gases release from the ground. But there's not enough known. What's the mechanism? Also, because it happens sometimes and not others, or maybe there were no instruments to measure this. I think going back to this systems science, trying to figure out what the other processes are that happen prior to earthquakes, to understand the whole system that might be happening that we could be happening, which might tell us about this impending earthquake. I think it's not going to come from the earthquake itself, but from something out in left field.

ZIERLER: We'll see. Luciana, this has been a lovely conversation. I'm so glad we connected and were able to do this. I'd like to thank you so much.

ASTIZ: I really appreciate it. This has been very fun.

[END]