Allen Husker, Seismologist

Allen Husker splits his time at the forefront of both applications and fundamental research in his seismology research at Caltech. As manager of the Southern California Seismic Network and Southern California Earthquake Data Center, Husker is responsible for the effective management and monitoring of earthquake activity in the California, which is run as a joint partnership between Caltech and the U.S. Geological Survey. As research professor of geophysics at the Seismo Lab, Husker pursues a wide range of basic science research relating to subduction zones, transform plate boundaries in Southern California, understanding the phenomena of slow earthquakes, and seismic network data analysis.

Following his graduate work at UCLA, Husker became professor at the Institute of Geophysics in the National Autonomous University of Mexico (UNAM) where he chaired the Seismology Department. As with Caltech, Husker was dual-hatted in Mexico. In addition to his research at UNAM, he also worked as Station Manager for the Comprehensive Nuclear Test Ban Treaty Organization seismic stations in Mexico. In that role he visited the United Nations in Vienna, where he represented Mexico as a foreign national.

In the discussion below, Husker conveys the challenges and pleasures of juggling his many roles, and he explains the dual value in his involvement in seismological applications related to earthquake preparation and mitigation, and fundamental research that investigates the nature of tremors and subduction zones - topics of inquiry that have animated the field for decades but which remain ripe for further discovery.

DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It is Friday, September 16, 2022. I'm very happy to be here with Professor Allen Husker. Allen, it's great to be with you. Thank you so much for joining me today.

ALLEN HUSKER: Thank you for having me.

ZIERLER: Allen, to start, would you please tell me your title and affiliation here at Caltech?

HUSKER: My name is Allen Husker, and I'm a Research Professor in the Seismological Laboratory. I'm also the manager of the Southern California Seismic Network and the Southern California Earthquake Data Center.

ZIERLER: The Network and the Data Center, are those Caltech projects, or are those bigger affiliations?

HUSKER: A little bit of both. The Data Center is purely Caltech. The Network is a joint project with the USGS. The USGS has an office in Pasadena. It's just right across the street from Caltech.

ZIERLER: These dual affiliations, did you come to Caltech with all of them in place? Was it sort of part of a package deal? Or you took on some responsibilities later on?

HUSKER: No, I actually only joined Caltech in the last year and a half. I was replacing Egill Hauksson who was the previous manager of the Network and the Data Center.

ZIERLER: Tell me a little bit about your overall interests, the kinds of things you've worked on in seismology.

HUSKER: I'm probably one of the seismologists who is all over the place more than anybody else in the Department. In fact, Mike Gurnis was asking me the other day what I do. Because it's kind of everything. It's observational seismology, so at least that narrows it down. I guess the easiest way to describe it is to explain where I started. My research—this isn't where it started. The starting point for everything I'm about to say is looking at tremor and slow slip. In the early 2000s is when they first found out there was something called slow earthquakes—when the scientists found this out, first discovered them. My interest spawned shortly after that. The slow earthquakes have tremors associated with them. A slow earthquake could be just like a regular earthquake, but instead a magnitude 7 earthquake, say, normally lasts for about 20 seconds—the duration of the rupture, like how long it takes the rupture to occur. In the case of it being slow, it could be months for that same rupture. So, it's really slow. It takes that rupture a really long time to propagate across that full fault trace. So, if it's many months, it's slow. It doesn't trigger the normal waves. The seismic waves don't come out of it, so it's not as dangerous. My interest started there and the tremor associated with it. As it's moving, it kind of vibrates as it's moving very slowly and creates this low vibration that's called tremor. You can see tremor on seismometers, and the slower rupture, you can see it with the same techniques we use to study tectonic movement. So, it's really slow. I started there, and then I started looking at the tremors. The tremors are really interesting, and there's a lot you need to have developed different techniques for detecting them. Some of those techniques are called—I don't know how specific you want to get into the science.

ZIERLER: No, this is what we're here for. As deep and technical as you want to go so you can convey the point.

HUSKER: They're called cross-correlation techniques. Basically, you see the wave form moving, so if you have two different parts of the waveform that look similar to another part of the waveform, you can compare them. It's just comparing the waves from different parts of the waveform to see if you can find similarities. It's that similarity you can use to locate it. They found little signals within the tremor called low-frequency earthquakes. Basically, it's tiny earthquakes that are occurring inside the tremor. Then we realized that these tiny earthquakes, if you add them all up, it creates the whole tremor episode. You can find these point sources and use that to locate where it's coming from. Otherwise it's difficult to find. So, there are these techniques that we started off looking at for these small positions of where it's located. It turns out we can use these to find lots of things. Another professor in the Department at Caltech here had a paper where, using the same technique, he found a billion new earthquakes in Southern California that nobody had noticed before. We had done the same thing previously, looking for the tremor sources with some other colleagues. This is where I spawned the interest. This technique is really powerful. You can use it in different places.

From there, I started exploring other places to use this, detecting earthquakes in general, and also I had always had an interest in planetary seismology. One of the ideas at the time was the Europa mission. They were going to send a mission to Europa and try to land a seismometer on there that has since been scrapped. In order to develop the techniques that might be feasible for detecting icequakes on Europa, we looked to floating ice in the Arctic. Norwegians had put out arrays on the floating ice in the Arctic many years ago, around 2010, a whole series of them around that time. The data is free on the web, so most seismologists will, after they're done with an experiment, put this data in a repository that everybody will have access to. Nobody had looked at the data in a long time, so we downloaded it and tried to run these techniques on it. It looked like it was really promising. Then machine learning came along. We said, "Oh, this is perfect." We then applied machine learning to it and found that you could detect the icequakes really well, so that was an interesting technique. I started working with this post-doc who had come from NASA and was also using the same machine learning code for moonquakes, so I branched out and worked with him because he was using the same code for moonquakes, for icequakes. It turns out ocean-bottom seismometers have a lot of noise as well. So, you have all these noisy sources, so how do we detect the quakes in it? This technique was also really useful for detecting earthquakes on the ocean-bottom seismometers. It's the same detection technique across all these different sources. So, that was one area of my research.

Another area: like I said, I started with slow earthquakes. In order to measure the slow earthquakes, I mentioned there's two different things: there's the tremor and the slow earthquake, and they kind of happen at the same time. The tremor is measured with seismometers, because it's a vibration that's happening, but the earthquake movement is so slow that it requires a GPS to measure it. So, we have these networks of GPS sensors that normally tectonic movement, and a slow earthquake is a little bit faster than tectonic movement. Tectonic movement is about the speed at which your fingernails grow, but this is a little bit faster than that. We can still use these to measure the slow earthquakes. The problem is—in Mexico, I used to work in Mexico. They have the largest slow earthquakes in the world. There are about ninety-two 7 or 7.5 earthquakes that happen every four years, or every three to four years, roughly. The question is, was that taking up a lot of the slip that would normally happen at the interface? So, instead of having a really large, regular earthquake that causes damage, do you have these slow earthquakes happening? It seems to be that's really the case, that these slow earthquakes are happening. A couple different things there: one is slow earthquakes themselves, how do you measure them? Because it could happen over months. Then, you get close to the cycle of weather. It's a really common problem that can happen in data analysis. You might have another source of noise in the data, and you have to look at something on the order of the same frequency. Weather has an annual cycle, or you get a six-month cycle because of rains and things. In Mexico, they have a really strong rainy season, and it turns out the rainy season pushes, actually it deflects the ground up and down. That signal is getting mixed in with the slow earthquake cycle, so that was an interesting thing to look at. Then I started looking at, "Oh look, you can measure the rain almost with these same techniques." So, it's an interesting spawning of different ideas all from this initial looking at slow earthquakes.

The other thing that came out was interest in studying—well, two other things that came out of it. One is the statistics. A lot of seismologists study the earthquake cycle. That's like you have a buildup of earthquake, of stress over time, where two plates are pushing against each other, and it releases in an earthquake. There's this idea. Many years ago, some different scientists came up with something they call the gap theory in the seismic cycle. The idea is that if you haven't had an earthquake in a very long time in one place, that should be the most probable place for a future earthquake. You're building up energy and whatnot. It turns out, somebody in the 1970s, around 1973, somebody wrote a paper where they basically said, "Around the Pacific Rim, what are the different places? Where are the gaps?" They said, "These are the most likely places to have an earthquake in the future." Then ten, and then twenty years later, somebody went back and analyzed that to see if that was the case, and it turns out it didn't work at all. It means we can't predict earthquakes. It wasn't that surprising, in some sense. Some people weren't surprised, and other people were very surprised. But nonetheless, this gap theory persisted. In Tokyo, they have something called a Tokai gap, and they said this is most likely the place for an earthquake, and this is where the big one's gonna happen. It turns out it didn't happen there at all. It was the mega-earthquake that happened in Tohoku that was nowhere near this. So, this is a showing that it didn't work, trying to predict earthquakes in this way. The same thing happened in Mexico, so it's a different repetition of the same thing. Then there's something called the Guerrero gap where they said it was the most probable place for an earthquake in the future. It turns out, the slow earthquakes are happening there every three to four years, so part of my research was trying to understand why is it happening there? What's going on? What's the particular reason that these slow earthquakes are happening there? It seems they're eating up all the slip. So, instead of regular earthquakes, slow earthquakes are occurring. Then I got interested in the geology: Why is this happening? What's the structure that's making it so slow earthquakes happen instead of these regular faster earthquakes? I worked with several colleagues to do a magnetotelluric survey. So, now this is branching out of seismology all together. [laughs]

P Waves and S Waves

ZIERLER: Allen, what would you call the new field when you say you're branching out? Where were you going?

HUSKER: It's all geophysics, but it's just using other techniques to get at the answer. There are different techniques to measure structure with seismology, and typically it's done with something called tomography. It's one of the most common techniques. This is getting really detailed now. [laughs]

ZIERLER: Please.

HUSKER: There are two different seismic waves. There's the P wave and the S wave. They're kind of the most common. They go through the Earth. They're affected by the fluid content of the Earth, so the more fluids there are, the more the S wave will slow down. We look at the velocities of the two waves and compare the difference. If the S wave is really slow compared to the P wave, we'd say there are probably fluids there. We look at a lot of different waves passing through a certain area from different angles, then we can invert those and look at the zones where these slow S regions are compared to the P wave. Then we can infer that there are probably fluids there. There are a number of things that could potentially make these differences in the two different wave rates. Another way to look at whether or not there are fluids is with the magnetotelluric technique I was mentioning. It looks and sees if the Earth is conductive in certain areas or not. High conductivity means there's probably fluids is the idea. With this technique, we can see zones of high conductivity in the Earth. Sure enough, we can see at the interface in the Guerrero gap there's high conductivity, which means there's probably fluids there at the interface that's greasing, it's lubricating where it's interfacing and makes it so there is less friction. So, that looks like that's the answer as to why there are slow earthquakes there instead of these fast earthquakes.

ZIERLER: Allen, if I could ask, maybe it's a very basic question, but the idea that there was a new category of earthquakes—slow earthquakes developed relatively recently, within the past 20 years—is that to say that these earthquakes simply went undetected? Or that the seismological community simply categorized them in a different way?

HUSKER: They had been undetected beforehand. It wasn't until we had really strong GPS networks all over the Earth that we could see these things were happening. In fact, the first time I saw them, I was a grad student at UCLA, and I thought it was an error. As I mentioned, it's really easy to mix things up. It was about an 11-month cycle they were first seen in Cascadia, which is a zone near Seattle. It kind of runs from Oregon through Seattle and into Canada.

ZIERLER: Was there a theoretical basis that they might exist, and that instrumentation was improved and deployed in order to find it? Or was the discovery more happenstance than that?

HUSKER: The discovery was more happenstance. There wasn't a theoretical basis for it, it was rather that it was observational first, and then the theory came afterward. When it was first seen, it was about an 11-month cycle, and that's close enough to a year. Like I said, when I saw it, I was a grad student, but I thought, "This has got to be, something that has to do with a seasonal change. It's almost a year. There's gotta be something to it." It turns out I was wrong; it actually was a real signal. Now it's being seen in all over the Earth in different areas.

ZIERLER: How does this contribute more broadly to the debate or the general mainstream opinion that earthquakes are fundamentally not cyclical? Because it sounds from what you're saying is that slow earthquakes do seem to have a cycle to them.

HUSKER: Yeah, they definitely do have a cycle to them. In fact, it got to the point where they now have dubbed the name "Episodic Tremor and Slip" in Cascadia because of their cycle that they have. They do what we hoped regular earthquakes do, that they have this distinct—I mean, it's not 100%. It varies a little bit, but it is much more cyclical, I would say, and less variable than regular earthquakes. A regular earthquake, it might be you have 20 years, or 2 years, or 100 years between them and on the same fault. So, it can vary quite a bit.

ZIERLER: Let me just think this out, because I have now been steeped in all these discussions about earthquake prediction being fundamentally impossible, because it's not about our observational or theoretical limitations, it's about the underlying chaos of the system, whereby the Earth itself doesn't know. So, the obvious question, the thought process that I'm going through here is: clearly there must be some connection between slow earthquakes and large earthquakes. So, if slow earthquakes, as you're saying, are cyclical, does that get us a little closer? Does it open the door just a little bit to the idea that there might be some cyclicity with large earthquakes and we simply haven't gotten there yet?

HUSKER: I would say, yeah, maybe? That hasn't been connected yet. That was certainly a hope, but it doesn't seem to be the case. I should say, if you look at a single fault or a single zone, the more you narrow it down to the space you're looking at, it becomes more difficult to predict. If, say, you want the average rate of earthquakes in California, that's actually doable. You can say we've had so many earthquakes in the past within this zone, so in the future we're probably going to have a similar amount of earthquakes in the future. So, in that sense, it's actually more predictable, but if you try to say in a specific fault that you're going to have this many earthquakes, or when the next earthquake's going to be, then it's going to be a lot more difficult.

ZIERLER: So would the onus be on the research? Is the starting point that you're assuming that slow earthquakes are not connected to large earthquakes and that you have to demonstrate it as such? Or is the deeper assumption that it's all connected, it's all happening deep in the earth, and we just have to figure out what those connections are?

HUSKER: They're definitely connected, but for whatever reason—and I'm an observationalist. You probably going to talk to, or already have talked to Nadia Lapusta who does rock mechanics. She's the proper person to ask about what's the connection there. But just observationally? They are separate in the sense that they, at least in the subduction zones, that you do have a separation from where they occur, for the most part. One type of slow earthquake can invade a little bit into where the regular earthquake zone is, and the regular earthquakes can get into a little bit of the slow earthquake zone, but they are separate physically in space. So, yeah, that's an excellent question. I don't really have a good answer. I don't know that it has been solved yet. That's something we're thinking about, I guess, but I don't quite have an answer yet.

ZIERLER: There's always something where you always have to figure out what is it, where it's simply an expression or a statement of our observational limitations, and when is it about recognizing fundamental chaos. It's just fun to see the outer reaches of knowledge in seismology at any given point.

HUSKER: Yeah.

ZIERLER: Allen, you alluded to it, but it just bears a little more explanation. What were some of the technological and computational advances that allowed for this happenstance discovery circa 2001? What was happening at that point?

HUSKER: What really happened is the GPS networks were good enough and well enough distributed over an area to measure this, and there had been enough recording for enough time. Because, if you have something such as an 11-year cycle, if you only see it once, then on your signal, basically you see a line, because the plates will be converging, and you watch that rate over time. Then suddenly, they go backwards, because it would be the equivalent of an—it is an earthquake, but a slow earthquake. So, that line regresses. It goes back and then it keeps going again. So, as it goes back, if it only happens once, you question, "What in the world was that? There's something weird. Give me my data." You need to have that happen a couple of times. It happened over many cycles, and then they were starting to believe it, basically. You need to have at least three or four years of data, or perhaps more, of these GPS stations weren't working. So, a full network that's completely working and having continuous data recording was what brought this about.

Slow Earthquakes are Important Earthquakes

ZIERLER: Allen, I've heard Tom Heaton talk about gentle earthquakes. Is that an entirely different category? Or are we just in nomenclature territory here?

HUSKER: That might be nomenclature. I haven't heard him speak about that specifically, so I don't know if he was referring to low magnitude earthquakes or the slow earthquakes. They're sometimes called silent earthquakes, slow earthquakes, or slow slip events. Those are the three names I've heard. I haven't heard the gentle one before. It might be the same one. I'm not sure if he was referring to that.

ZIERLER: Allen, obviously, large earthquakes from a funding perspective, from a societal benefit perspective, they're sexy. People care about them. They affect people's lives. What are some of the inherent challenges—from funding, from getting your colleagues interested in this—what are some of the challenges in terms of focusing on a topic that nobody even knew about 20 years before and don't really affect us obviously as far as we can tell?

HUSKER: Let me take one step back, because you asked if they were completely separate, and I guess I gave the impression that they are completely separate. There is a question if slow earthquakes can provoke a regular earthquake. It certainly happened. The last many cycles, now, of a slow earthquake occurring in Mexico because they last for months, and during that time regular earthquakes occurred. Like a very large, regular earthquake. A magnitude 7 has occurred now, I think at least four—it's been about four times, four or five times—that a large slow slip event or slow earthquake is happening. Then an earthquake will happen. So, it looks like there's some causality, like these slow earthquakes are perhaps provoking these regular earthquakes. The thing is—

ZIERLER: Would that fall under the category of earthquake clustering? Or this is a larger timescale than that?

HUSKER: It's a larger timescale than that, I think. But it also doesn't happen all the time. That's the problem. Most of the time you have a slow earthquake and there's no regular earthquake. It just happens. Then there's the question: does it just raise the bar a little bit? We might have: that earthquake was ready to go anyway, but then you have a slow earthquake that comes along and nudges it a little bit more to make it happen. So, that's a little unclear, but it looks like there is some causality there, and certainly the other way around, too. That you can have a large earthquake that then causes a slow earthquake to occur has also been observed. So, there's both sides of that coin. There is at least that interaction that seems to be happening between the two. But back to your question, I'm sorry…

ZIERLER: The question about the inherent excitement or sexiness of large earthquakes and the challenges of dealing with an earthquake that's hard to detect and perhaps even get people excited about.

HUSKER: Right. At first it was really exciting because it was brand new. There were a lot of questions. Specifically this question: Can a slow earthquake provoke a large, regular earthquake or vice versa? It was an unknown. It seems to becoming a little bit clearer now, but that certainly wasn't the case at first. So, there was that question. Then suddenly a tremor was also detected and it was whether or not it was completely silent. In fact, I think the first paper about this was along the lines of "The Noise of Silent Slip" or something to that effect. It was some kind of clever name. Then there are different flavors of tremor that have come out where they see these—I told you about these low-frequency earthquakes, which are these tiny earthquakes within the signal. Then they have very low-frequency earthquakes. There are a number of things that were observed at first.

Then there were a thousand papers coming out—not a thousand, but a very large volume or number of papers that were suddenly spawned or coming out of this new field once they realized this was happening. People started looking any way they could to figure out what was going on. Then it was very quickly observed that the hyperfluid pressure—I know that I had mentioned fluids, basically, were observed in association with these. There were a lot of different observations that came out. Scientists—there's a little bit of excitement, there's competition. I think competition drives a lot of people. When you see your colleague suddenly writing five papers about something, you get interested and "I want to write five papers, too." So, there's the competition aspect of it. There's also a little bit of publish or perish. If you know you can write a paper about something and get that published, there's a little bit of that going on. So, I think there are a lot of motivations. It's not just "How do we detect large earthquakes?" I guess beyond hazard, beyond seismic hazard, there are plenty of other reasons, is what I'm trying to get at as for motivations for people to study this. The fact that it was something new, I think, made it the big thing. After this, I would say, in general, one of the big things that came along afterward was the induced seismicity, like what was seen in Oklahoma, was very exciting. It's almost like a fad in seismology. The first one was these slow slip events, slow earthquakes, and tremor was a thing that lasted for a few years. Then the stuff in Oklahoma came out, and that was really exciting. That caused a lot of interest. Now, there's machine learning, and that's really exciting. Now, there's DAS. Actually, I was going to mention that too, that's another thing. I haven't reached that yet, but now I'm starting to work a little bit on these, now as being the manager of the Seismic Network, but those are the big things. I would almost see them as waves; fad maybe is too strong a term, but interest. This is new and exciting, so a lot of papers are written, people get excited about it, and then it gets incorporated into the main body of seismology. Then it's something that's part of it, as opposed to the exciting new thing. Certainly, slow slip and tremor has passed. It's not the latest thing anymore, but it certainly was in the early 2000s.

ZIERLER: Allen, just a nomenclature question: slow slip and slow earthquakes, what's the relation? What's the overlap? Where's the distinction?

HUSKER: They're exactly the same. It's just that one is used more with lay people, and the other one is used more between scientists. Slow slip is used more between scientists, and slow earthquakes we use to describe it to the rest of the world.

ZIERLER: For your own career, it's always exciting, obviously, when there's a discovery that happens during graduate school, during a formative period of one's research career. Over the last 20 years, have your research interests and opportunities more or less tracked with where slow slip research has taken the field?

HUSKER: It did for a long time. There's a part of me that once I feel like I figured something out for the most part in my head, I'm like, "Okay I did that. I want to move on now." That was a lot of the reason why I started looking at these other things, like, "How do I detect something that's in ice quakes? That looks interesting, let me go play with that now." So, part of me feels like a jack of all trades, master of none at times, because I really excited about something and then I move on to the next thing. So, I'm happy to move and try new things. Some of my other colleagues, I think, do the deep diving to really get into a field and explore it really deeply, whereas I don't. I feel like I touch on a lot of things, just what excites me, and I want to move toward that. If a student also is interested in something, it's easier to push in that direction.

ZIERLER: Allen, I'm curious, doing so much research at UCLA, how did the Seismo Lab loom in your mind in terms of its status in the field, the kinds of people that were there, the kinds of research that the Seismo Lab was known for? Where did that fit in with your overall interests in graduate school?

HUSKER: Caltech is the pinnacle. I mean, I still can't believe I have a job here. [both laugh]

ZIERLER: Me too. We're in the same boat.

HUSKER: One of my post-docs said it best, "It's like seismological royalty." You had Gutenberg and Richter. The pioneers in the field were working here. You see their pictures hanging up down the hall, and you think, "I'm not worthy," right? These are the guys that started the field. So, yeah, it's the biggest names, and I think the most exciting research that's happening seems to be coming out of Caltech, certainly. I mean, not everything. Like I said, there's this discovery in Cascadia that had nothing to do with Caltech, but nonetheless, you can see their influence on so many different things.

ZIERLER: I'm always interested in the concept of—I don't even know if it's a word, but—extrapolability. The idea that you're studying something locally, but you want to see if it has global ramifications, if this is something that's true across the planet. So, in that vein, what have we learned about slow earthquakes? If you're studying at the local level and trying to figure out how they work on a planetary basis?

HUSKER: That's 100% important, I would say. We want to have a universal law, like why is this working where it is working? When I was looking in Mexico and trying to figure out "Why is this slow earthquake only happening in this one spot? Why does it repeat here every single time?" Whereas with regular earthquakes, there's, like I said, a jumble. If you look at one part of a fault, you don't see the repeatability quite the same way on quite the same scale. So, why is that happening? The answer seems to be that there are fluids or something that are trapped in a certain area that really allow for this to happen. That seems to be a great explanation, and you want to see that across the globe. If you see it in one place, is that true? You need to be able to repeat it in some sense. For the most part, that seems to work, but there are certain situations where you don't see this sort of relationship with fluids, so you wonder "What in the world is going on?" It's not 100%, so it's interesting. Why does that occur? But you need to have universal laws.

Another issue that comes up is just because you see a coincidence, is that causality, or is that a coincidence? As I mentioned, these slow earthquakes are happening, at least in Mexico, for many months, then a large earthquake comes along and occurs at the same time. So, did one cause the other? Or is it just random that an earthquake happened? So, you need to see this happen over many cycles in order to get an idea that yes, there seems to be a higher rate of these large earthquakes happening whenever a slow slip is happening or a slow earthquake is happening. Then you want to think about this universally. Is the same thing happening in Japan? Is the same thing happening in Cascadia, or wherever? I think with everything, not just seismology, but studying the Earth and geophysics, you want to be able to extrapolate and see if this is happening across the entire globe or if you can say something more than just the area you're looking at.

ZIERLER: Allen, tell me about your work in transform plate boundaries. I know plate boundaries, of course. What does the term "transform" mean that precedes "plate boundaries"?

HUSKER: There are a few different types of plate boundaries. There's convergent. Convergent is what I was working in Mexico. That's where they're converging and pushing against each other. Transform plate boundary would be more like here in California, the San Andreas, where they're moving side by side against each other, but they're not pushing against each other. I actually haven't worked a lot here in California. I went to grad school. I did write a paper for my master's here in California looking at side effects a long time ago. Then I went off to Mexico, and now I've come back. So, now I'm just kind of back to the transform region. I've been away for a while. My research right now is getting into that a little bit. We're managing the network, but I haven't done a lot of research in that area.

ZIERLER: Allen, how did you get involved in nuclear verification and nuclear security issues?

HUSKER: Oh, okay. [laughs] That was one of the cool things about the job in Mexico. When I was working in Mexico, I was a grad student. I went to Mexico for my research, because we went and installed seismometers there for their seismic network. Then I ended up meeting my wife, falling in love, and getting a job there. [laughs] So, I stayed there for quite some time after grad school. Well, you know, I was a professor. UNAM, the university there, is the National Autonomous University of Mexico, is a national university. In the US where we typically have state universities, the national university is with the federal government. Mexico is a much smaller economy than the US, and therefore doesn't have all the same parts of the government that the US does, or have the same money. So, it works with the federal university to do a lot of things, so the National Seismological Service is actually run out of the federal university, as opposed to here in the US, we have the USGS that really runs on the national level. We have state universities that have regional networks, but there's the USGS to hold it all together. Whereas UNAM does that in Mexico. Because of that, they're the ones that have the contract with the CTBTO. The CTBTO is the Center for the Nuclear Test Ban Treaty Organization. The Commission, I'm sorry. The Commission for the Nuclear Test Ban Treaty Organization. They have a contract with them to run the stations that they have with the CTBTO and give their data to them. I had experience running local temporary network in grad school, and they were looking for somebody to do it, and they said, "Could you do it, Allen?" I said, "Sure, let's do it. That sounds great." [laughs]

ZIERLER: You did this as a foreign national, I assume.

HUSKER: Yeah, exactly.

ZIERLER: Were there any sensitivities? This is sensitive information. How did you deal with that?

HUSKER: It was a little weird. The information is sensitive, and it's not. It's seismic data; it's open to the world, because you can install a seismometer, and the seismic network are already open to the data. So, there's data that's available for many of these stations anyway, because they're shared typically between the local country and En Piena. In that sense, that was not a problem. It's not sensitive in that sense, because it's earthquakes. Earthquakes aren't a national secret. What's weird though, is when I would go to Vienna, and I'm sitting around the table with everybody else and have Mexico sitting in front of me on my name tag. People would say, "Why in the world?" I mean, I'm at the United Nations representing Mexico in some sense. That part was a little weird.

The other thing is the CTBTO is a huge bureaucracy, so they have to answer to all the countries that give money to the United Nations. So, they're not actually allowed to even say if there was a nuclear bomb threat or if a nuclear bomb occurred, or a nuclear explosion. What they have to do is, they can show, "Here's the data. Look at this, here." But then countries have to individually say that there was a nuclear explosion, or confirm it. The CTBTO can only provide the information; they're legally not allowed to say it because of different countries not wanting them to. It's kind of an ironic situation to be in.

ZIERLER: Yeah, I mean, there's so many different unique vantage points that you're getting. First of all, regardless of your citizenship in representing Mexico, what did you learn more broadly about the world of nuclear verification and nuclear security in this experience?

HUSKER: What's interesting is, the countries that have the most don't want to be verified.

ZIERLER: Of course. [laughs]

HUSKER: So, Russia, the US, China, I think Israel is one of them, I can't remember for sure. But those, India, again, none of those have signed the treaty. So, the treaty won't go into force until all the countries sign, and nobody foresees that it will be signed. Nonetheless, the US contributes millions of dollars to have this global network monitoring to see if anybody's testing nuclear weapons. So, in that sense, it's kind of ironic, again, this balance of what all the different superpowers want. It was interesting to see how these negotiations occur. Nonetheless, since it's gone into force, the number of nuclear arms testing has just diminished completely. It's almost nothing. In the ‘80s it was happening constantly. Since this has gone into place, it has completely diminished, the amount nuclear testing.

ZIERLER: What role does the end of the Cold War have in all of this, do you think?

HUSKER: It's all tied together, right? The end of the Cold War was a huge part of it, and that's what led to this organization. My role in it was very tiny in general, other than just being a piece that kept at least the Mexican station moving forward.

Seismic Delineation in Mexico

ZIERLER: Is there a field of seismology that's specifically focused on delineating tremors from a nuclear explosion versus tremors that are occurring naturally?

HUSKER: Yes. A nuclear explosion doesn't look like tremor. It looks more like an earthquake. In fact, you can put them on a similar magnitude scale, because it's just how much energy was released in that moment, or how much seismic energy you can measure. So, you can use, for instance, the Richter scale to measure the size of the explosion. But yeah, they do look quite different. I mentioned there were P waves and S waves. A P wave is a compressional wave. It moves in and out as it is expanding away, whereas an S wave is a shear wave, so it goes back and forth like this. A bomb doesn't cause shear. Typically, a bomb really has a very strong P wave and almost no S wave, whereas from an earthquake from a fault will have a smaller P wave and a much larger S wave. So, that's the easiest way to distinguish between the two. But they're almost impossible to disguise. If anybody looks at a seismogram of the two and is a seismologist, it's very easy to distinguish between the two. But I'm sure there's somebody doing machine learning algorithms right now to classify them automatically, as opposed to having somebody to need to look at them.

ZIERLER: Allen, did you know Spanish prior to going to Mexico? Or was that some crash course that you had to take to pursue this opportunity?

HUSKER: No, I didn't know Spanish. Even when I arrived, I would say I didn't really know Spanish. I knew some words, but definitely not fluent at all. I didn't speak fluently at all. So, I just learned it by the seat of my pants kind of thing. I was out in the field having to find places to install seismometers, and I had some Mexican colleagues that would go out with me. But they were hired by UNAM, and I was the one directing, "Let's go over here and check this out. Let's try this. Let's try this." So, I was kind of fumbling through Spanish and as much as I could and just building it up over time.

ZIERLER: Allen, you mentioned the Richter scale before. I'm so fascinated by the relevance of keeping his name attached to the modern magnitude system where maybe it really should be called the Kanamori scale or something else. We just call it the magnitude scale at this point. Is the Richter scale, since obviously Richter himself would not have known about slow earthquakes, are the magnitudes that the field of seismology has traditionally dealt with, are they relevant for slow earthquakes? Or is it simply too small, you need a different parameter set?

HUSKER: Okay, let me explain. Just to take a step back, actually I just learned recently, the reason it's called the Richter scale. Gutenberg and Richter, who were both professors at Caltech, actually came up with the scale. So, it actually should be called the Gutenberg Richter scale, right?

ZIERLER: Yes indeed, yes indeed.

HUSKER: But I only recently found out, maybe you already know, is that the reason it's called the Richter scale is because the local news media wanted to interview scientists about earthquakes, so the person they would always interview was Richter, because he had no accent. He was American, whereas Gutenberg wasn't American. So, it got the name from the news media, the Richter scale. I only recently learned that, so that's interesting. Beyond that, the scale itself is measuring—it's a relative thing—how big is the seismic wave at a point, at a certain distance away from a fault, or from where the source was. So, people got really used to this scale. So, all scales that have come since then are associated with this scale. So, even if you have a magnitude 5 on the Richter scale, and you said the Kanamori scale, his is on the MW scale, you also have a magnitude 5, then you know that they're really similar. That's done on purpose. Everything was related back to the Richter scale because it was the first one. They're all very similar. That way it gives people an idea of what's going on.

Back to the slow earthquakes. Modern earthquake scales are based on the rupture size of the earthquake and basically the dimensions of the earthquake. So, it's not about how much you feel it, it's the dimensions of the earthquake itself. If you have a really large rupture that's, say, maybe 50 kilometers or 100 kilometers in width or length, and something like 50 kilometers or 20 kilometers in width, whatever it is, it means it's a very large area. If that area is really large, then the magnitude is going to be large. It's all associated with what that area is and how much slip occurred, so it's the actual dimensions of the fault and how much slip was in the fault. That scale then works whether or not it's a slow earthquake or a regular earthquake, because in either case, all you're doing is measuring the dimensions of the fault.

ZIERLER: Your time in Mexico living among earthquakes, just on a personal level, did that influence your science in any way? Just experiencing earthquakes in a society that has too much experience with earthquakes?

HUSKER: Yeah, it got me really interested in people's perceptions of earthquakes. Definitely. So, toward the end, like I said, I'm a jack of all trades, master of none, so I got really interested in earthquake early warning. I'm heavily involved now with earthquake early warning from the management side of things, and operations with the Network, but at the time in Mexico before I left, I started getting really interested in how do people experience it? What do they do? How do they experience it? What do you do if you hear an alert that's occurring? These kinds of things. So, I started to work with a post-doc who was interested in these kinds of questions, the societal questions. So, now we've written a couple papers together. [laughs] Then I left Mexico, so she's still there working on things. So, that was a whole other aspect. Definitely being in the middle of these large earthquakes, hearing the alarms go off, and seeing people's reactions made me think about that part of it, definitely.

ZIERLER: As prelude to your involvement with the Southern California Seismic Network, Southern California Earthquake Data Center, how much involvement did you have prior to these appointments in just the general field of network monitoring? What was new to you, and what did you come in with specific experience?

HUSKER: You mean here in California?

ZIERLER: Correct.

HUSKER: Okay. I had a lot of experience. I had done a lot of the fieldwork to install. I didn't have to do that as a manager, but I had done it personally. Pouring concrete base and putting a seismometer down on it, all that kind of stuff in the field. Hiring people to help out. The installation of a seismic station in general. Also, part of my research in grad school was actually to install telemetry. So, we had radio sites on all the stations I installed to radio the data back to some central hot site. So, I had that, again, all the way from the field up to actually running code and logging into the different sites and making sure data was flowing. So, I was doing some kind of network monitoring personally. I had very much from the ground up kind of experience. I was managing a team of people to go out and install equipment or to do whatever. My advisor completely left that up to me. I just took over and started managing people right away to make the network run, because it was 100 stations and it had to run. When things would go down, I was the one in charge of making sure it was running. I would say immediately I was doing this with my own—just because it had to happen. I don't even know if I was assigned to it, it just had to happen, and I was the one doing it. I naturally just fell into the role in some ways. Then the CTBTO I was managing but at a much higher level. I was looking at finances and asking the CTBTO for the next round of funding and things like this, and then also managing people that were going out into the field. They had a station on an island, so we had to get permission from the military who controls the island to go and service the site, and then worry about weather. We had a satellite connection, so that was a new form—it wasn't new, but it was new for me—but they had been working with us to work on satellite telemetry. So yeah, I got all levels of that in my experience there.

ZIERLER: Allen, you mentioned of course Mexico is a much smaller economy. It's a poorer country than the United States. What are some of the structural limitations from a budgetary perspective, working with Mexican authorities to make sure that the network is healthy, it's doing what it's supposed to do?

HUSKER: To clarify, I was never the head of the entire seismological service, I only had a very limited piece of it, which was the station that we were working with the CTBTO. From that perspective, the CTBTO was actually giving us plenty of money, so there was never an issue with funding those specific stations. There was only four of them, so we actually had more than enough budget. We were doing fine. But in general, I would just say for my research project it was almost impossible to get money for research projects. A lot of what we'd do is work with foreign countries that bring in money, so we'd go out with Japan or France that helped us install equipment for the CTBTO. So yeah, you look for further funding sources when you don't have the government to help fund things.

ZIERLER: Are there any analogues in Mexico to the USGS, the NSF, the traditional places that American researchers would go for funding support?

HUSKER: Yeah, the equivalent of the NSF is called the CONACYT. It's very similar, the same idea, just nowhere near the same level of funding. For a while that was okay, but it was near impossible by the time I left to get any funding out of them. Each president in Mexico has a very strong control over the budget, so the most recent president has just reduced the funding to it quite a bit.

ZIERLER: In maintaining these networks, to make sure that they're working well, that they're healthy if that's the right word, what are the metrics that you use? How do you know that they're doing what they were designed to do?

HUSKER: Again, good question. There are a few different ways. One would be station up time, just looking at the percentage of time the stations are up and running. There's also how noisy stations are. This might change over time. There's a station here, for instance, that's in Rancho Cucamonga. Rancho Cucamonga has grown, so now they're building more warehouses and things in the area, and it's causing more noise. So, we look at the quality of the station and see that it's getting noisier. Now people are thinking we're going to have to move it, because the noise level has just gotten so bad. Those are the two main ways, I'd say, just looking at the quality of the signal coming in, and the other one is how long is it actually working for? What's the up time?

ZIERLER: When you were in Mexico, I wonder if you thought more broadly about the humanitarian situation in Central America, the issue of refugees and drug violence and political instability. Did you tend to look at those things through a, forgive me, a seismological prism? Was there that aspect? Or just being a scientist around these things, it was just inescapable for you to think about them?

HUSKER: More the latter. I was out in the field, I did a lot of fieldwork where I was just in places where you see things, and it was impossible to escape it. As much as you hear things on the news, here, there to actually be in the place where you see it is different. You see what people are going through on a day to day basis in some of these areas. I was at a school where you could see bullet pockmarks on the side of an elementary school. It's hard to avoid, you know?

ZIERLER: I'm curious if you've ever thought about neocolonialism, things like that: tensions between developed countries and underdeveloped countries, again through the prism of seismology. I'm thinking for example the controversy over the 30 meter telescope in Hawaii and its connection to native peoples of Hawaii who are upset with how the astronomy community has dealt with them over the decades. Are there analogues to that as it relates to seismology and research stations?

HUSKER: Not in a negative way. In general, people want to know more about earthquakes. In fact, it almost made life easier for us, because we could drive into narco-traffic, drug lord territory and be okay, because they are people too, they want to know about earthquakes as well. They would see it as a way to find out about hazard as well, I think. If anything, it was more of a positive. People are interested in earthquakes and want to know what's going on, so people were very interested. In general, for seismology itself, I'd say that wasn't the case. A bigger overall picture, seeing these kinds of things happening, outside of seismology I would say, yeah. Americans obviously buy tons of drugs, and it's why these kinds of different things are happening in these areas. It was more like being an American in these regions I would say more than the seismologist side of things was what I saw.

ZIERLER: The fields of earthquake engineering and earthquake mitigation are obviously adjacent to your area of expertise, but being in Mexico, being in a land of unreinforced concrete and fragile infrastructure, how might that have changed your perception of these issues relative to had you been in say, Switzerland, for example?

HUSKER: I'm probably going to answer in a way that's odd. If anything, I was surprised how often things survived. These kind of earthquakes, there's magnitude sevens on the coast that happen often enough, and you see these tiny hotels, a few stories, maybe one, two, or three stories tall, they survive the earthquake impacts and they're basically sitting right on top of it. It always blew my mind how well they did for the number of earthquakes that were happening there. Typically, it wasn't always the case, obviously. There were situations that were a lot worse. But it did surprise me in some of these smaller communities. I guess I saw that the more complex the infrastructure, the more the potential problems. Mexico City, for example, had a lot more damage in previous earthquakes, but it's a much more complex situation than what's going on. It's not fair just to say it's the buildings and whatnot, because Mexico City itself is on a basin, and there's a huge amplification that happens anytime a wave enters the basin. But certainly, the larger the city, the more complex it is, the more things going on, and the more people, the much larger the vulnerability, and so the much greater potential for disaster. The other things I saw is that people don't follow building codes. Actually, Mexico City has really good building codes. The problem is people wouldn't always follow those building codes. That could be because of money; it could be because of just ignorance. There's different reasons. There was one building that collapsed, because, the store keeper on the bottom floor, for example, wanted to have two stalls next to each other. Then broke the wall that was between the two of them so they could expand the size of the store. The building collapsed during the earthquake, because it no longer had that support wall. The person that did this obviously didn't fully realize the importance of that wall that was there while they were trying to expand their business. So, those are the kinds of situations that are really sad. I guess education is key in a lot of these situations.

Thinking About Past and Future Earthquakes

ZIERLER: Allen, a few questions related to chronology, both backward looking and forward looking. How do you assess earthquakes that happened before the advent of seismic networks a century or more ago? How do you think about the location and meaning of earthquakes before we had effective data to understand them?

HUSKER: I tend to be on the skeptical side of that. There are people who have done a lot of work trying to figure that out. The tip of it is going back to older records and newspapers to try to figure out more about what the magnitude actually was at the time, where it happened, and the location. I read some of them—at least I should say the Mexican newspapers—and some of them are quite confusing. Even time of day will switch from one account to the next when it gets to be older. So, as a scientist who's really trying to narrow something down, it's really difficult to fully count on these records. Even here in the US, they have the New Madrid earthquake that happened a couple hundred years ago, and it's supposed to be a very large earthquake. There's been very little seismicity since then in the same area, and it makes you wonder. There are people, again, who've done very excellent work to look at this and try to figure out what really happened, looking back at those account, and they certainly have not done that. But I at least start to get a bit more skeptical, the older the earthquake, as to what the location might have been and the actual true magnitude. We'll know that it's big, but never narrowing down that number I think is difficult. There's going to be more variation of what that might actually be.

ZIERLER: To flip the direction of that time scale for the question, thinking about the future, in what ways do you consider the moment that we're in now as a future historical capsule? How can we preserve this data so that future generations will make the best use of it? That it'll simply be around, that it'll be accessible 100 years from now in terms of whatever their computers look like.

HUSKER: Yeah, I think a lot of it is the metadata that goes into it. The metadata is all the description behind the data, so that they can piece together and find and use whatever new techniques might exist. One other interesting thing is we have an explosion of data right now, which is different. So, when we're looking back and trying to piece together what happened 100, 200 years ago, there's just very little to go on. If anything, we have just such an explosion of data right now, we're trying to come up with new techniques to handle this mass of data—analysis techniques like machine learning and whatnot. We're even getting more now with new types of instrumentation. The challenge almost becomes: how do we even preserve so much information without it getting thrown away? I know that a lot of different temporary arrays that have been installed in the past. The tapes get thrown away. It gets old information, and how do we really preserve that is a really fundamental question that's going to be hard. Moving forward, I'm seeing how much data isn't multiplying. The storage we have needs to change to accommodate that, and even the techniques to analyze it. So, that's all moving forward, and it's exciting, but it is a real challenge. I don't have a good answer, I guess. I'm still sort of grappling with this as well. But I think just documentation is what we have to go with right now.

ZIERLER: Is there still a role for paper and analogue records? We never have to worry about what computer system can access a book, or a printed map, or things like that. Is there still that role? Or is there the concern that we're entirely digital now, and now it's just a matter of leapfrogging this data from one generation of systems to the next?

HUSKER: I'm of the mind that we need to go completely digital. The reason is usability. All of our tools are moving to electronic tools, so the more digital you can make it, the easier it is to use. If we have it on paper, it's just less usable. There's not many people that are coming back to paper records at this point. If anything, we need to scan it and make it digital so it's usable. If we don't have it digital, it's just going to fall by the wayside is what I see at this point.

ZIERLER: I want to ask about your sense of the history, first with the Southern California Seismic Network. What are the institutional partnerships that allowed this network to come into being?

HUSKER: It has a long history. I don't know if you're interviewing Egill, but he's certainly the better one to know the past, the history, than I am. It's gone through a number of phases. California itself has invested quite a bit. The office has gone through a couple of name changes, but currently it's Cal OES. The California Office of Emergency Services has invested quite a bit. They did that originally through the California Integrated Seismic Network—is what it's called. That's a combination of Caltech, the USGS, and Berkeley. The local USGS, not the national level, but the local office in Pasadena and Menlo Park are the two that—

There's that part of it, and then there's the USGS; the federal government also gives fund. So, those are the two main ones for a long time. Then earthquake early warning came along. Caltech was actually heavily involved in the effort to lobby Congress, actually, for the earthquake early warning system. That was a lot of work from Tom Heaton and others from Caltech and also Berkeley, to move this forward. Now, the entire west coast has earthquake early warning. It now accounts for the majority of our funding comes from early warning.

ZIERLER: What is the institutional relationship? Obviously the Seismo Lab is part of Caltech, but is the Seismo Lab the main point of contact between the USGS as far as the SCSN is concerned? In other words, is there an independent relationship between Caltech and the Network independent of the Seismo Lab, or it all runs through the Seismo Lab?

HUSKER: The Seismo Lab is the umbrella above the network, so we're definitely within the Seismo Lab. The SCSN is purely within the Seismo Lab as well as the Data Center, and then we work with the USGS. But yeah, we're all underneath the umbrella of the Seismo Lab.

ZIERLER: So, the reporting structure: you would report to Mike Gurnis, who reports to the division chair, as you would with anything else?

HUSKER: Yeah.

ZIERLER: What is the overlap? And where is the distinction between SCEC, for example?

HUSKER: SCEC originally helped fund a lot more. I guess they still do, but I think it's been flat funding for a while. They were certainly very involved early on in helping the Data Center get started, and had a much more vital role there. As other funding has come in that's a lot bigger, that role is diminished a little bit. It's still there, and it's still important, but it's just not quite the same. I went actually to SCEC this last week, the last few days, and I just got back yesterday. What's interesting is that everybody uses the data, and it's a fundamental role for all the Southern California earthquakes. It's actually now called the Statewide Southern California Earthquake Center. There's not a lot of communication as to what they want, so it's me more as a scientist trying to get a feel for what the community wants. A lot of it is driven from the people at Caltech in the Seismo Lab, but certainly I'm as a manager now trying to see the greater community, what are the needs, and what do people want. That level of communication maybe isn't there, and it's probably something I want to work on a little bit more.

ZIERLER: When you hear on Twitter or the evening news there was a 7.7 something. Where does the Network play in first the detection and then the dissemination of the information?

HUSKER: It's kind of at all levels. All of the data comes through us. There are immediate algorithms that start to detect. Those first few seconds are when earthquake early warning will detect and make a determination of magnitude and start to send out alerts. Those algorithms were developed at Caltech by different groups. One of the most important ones was developed by Tom Heaton and his group, then were implemented by computer scientists or computer engineers within SCSN, within the Network. A lot of that actual programming was done here as well. Berkeley has their version, too, that they've done a lot of the programming, so it's been kind of combined between a lot of what Berkeley and Caltech have done. So, there's that early warning piece, and then beyond that there's, depending upon the size of the earthquake, you have to wait some time to be able to determine the magnitude. The reason for that, I mentioned earlier, if it's a magnitude 7 it can take 20 seconds just for the rupture to propagate. So, that's at the fault itself, but then the waves have to propagate out from there and reach seismic stations, and you have to be able to measure at enough seismic stations to get an idea of what the actual earthquake was. So, the larger the earthquake, the more time you have to wait, basically. If it's say a magnitude 8, you might have to wait 20 minutes to be able to know everything about the size of the earthquake. So, it really depends on that. But then all that data comes through us. If it's above a magnitude 6, the USGS basically takes over. So, our agreement is we do everything below a magnitude 6, locally, or regionally here in Southern California. But if it's a larger earthquake, that's when the USGS takes over and they run the show. But they're getting all of their data from us.

ZIERLER: As an educational institution, what are the boundary lines where Caltech has to be careful that it's not dispensing with policy advice and public safety information? Those are the kinds of things that are best left to the government. How are those delineations made in real time, particularly when an earthquake is happening?

HUSKER: That's a difficult one, because the public wants to hear it from an expert. They want to hear it from a scientist, so they want to know what we have to say. In fact, I think we drive a lot of it. So, Cal OES will ask us for information immediately as well. They want within 20 minutes some piece of information—the size of the earthquake and whatnot—so, there's automatic assistance they can get from the USGS when there's a rupture, who's it's affecting, these kind of things. We give out that information and the public wants to know, so I think a lot of it is our own idea of what's appropriate. There are so many scientists in Southern California that are working on this, that the USGS on their side, and with us, it's almost an organic thing, somebody will start a Teams group and immediately add everybody, and people will start throwing in information. It's this interesting way to see how science is done on the fly. There is suddenly all the aftershocks are thrown on a map so people can look at that and analyze that. The earthquake, the magnitude of the main shock itself, and all the information for that is thrown in, and talking points are developed, and all this kind of comes together. It's actually a really cool thing to watch. [laughs] I haven't been through a very large earthquake, but there was a magnitude 4.4 some time ago that I was here for, and I watched it all come together. So, it's interesting that there's so many minds working on this that the talking points generate themselves. People seem to adhere to it for the most part. We don't have anybody that goes off and says something crazy. [laughs]

ZIERLER: As manager, do you directly liaise with the mayor, the governor, the policy organizations? Or does that all happen through the Survey?

HUSKER: That happens through the Survey. We don't have that kind of direct level of communication. So, yes and no. There's an interesting relationship between Cal OES and the federal government. Cal OES a lot of times prefers to work with us or with Berkeley. They don't want to deal with the USGS, because there's some tension between government agencies sometimes. [laughs] So, we'll tell them the information, but they could easily get it from the USGS as well.

ZIERLER: Just so I understand the nuances, the Southern California Earthquake Data Center, that's simply the archive of the network. Does it have a different or separate function?

HUSKER: That's the primary function. Typically these are Caltech professors, but it could be somebody outside of Caltech could analyze that data, and then we put that analysis on the web, also. For example, we have the archive of the data of the original wave forms from all the stations that are coming and we save that. We also locate earthquakes to the magnitude and make a catalog, and we save that catalog as well, which is a product of the data. But other people, like Zach Ross, I mentioned had made his own catalog, so we archive that as well and make that available to the public as well, that he used with the machine learning paper. We also use DAS as a new technology I mentioned a couple times, but maybe we can get into that some other time. That's John Nguyen's field. He and I are certainly interested in working on making this part of the Network as well, which is distributed acoustic fiber, but using basically the telecom fibers that are laid turning those into seismometers. That's a whole other large amount of data. We have now, in the data center, some of this fiber, but it's coming in from one of John Nguyen's experiments. But now we're hosting it on the website. I would say the majority of it comes from the Data Center. It is the archive for the network, but we do have specific data from other areas that we put on there too.

ZIERLER: Allen, the last topic I want to touch on for today's talk is just your balance of responsibilities at Caltech with all of the things that you're doing. On a given week, between managing the Network in the archive and being a research professor, is it all sort of jumbled together? Is it Monday, Wednesdays, and Fridays? How does that work for you?

HUSKER: It is jumbled together, but I would say the majority of Mondays and Tuesdays are probably dedicated more to the Network, and toward the end of the week I go more toward the research. But yeah, it's a jumble. I was doing field work in Southern Mexico some months ago in March, and while I was there, I was on my cell phone. I was out in the field digging, and I would stop and look at my cell phone and answer a question to my administrative assistant about something in the budget, put my cell phone back in my pocket and keep digging.

ZIERLER: [laughs] That's great.

HUSKER: So, yeah, it's kind of constant, everything.

Managing the Network

ZIERLER: As manager, do you have direct reports? Is this essentially a one man show?

HUSKER: Yeah, there are 20-something people in the network there. So, yeah. It's not a one man show, we all make it happen. I have three, no, four people who I work with more directly. They have their own teams beneath them, but I'm still the one that has to make the decisions on staffing. We have a few positions open for instance right now, so I'm the one who drafts the description of the position so we can put that online and look for job applicants and whatnot.

ZIERLER: Is it a two-way street? In other words, are there aspects of your administrative duties that inform the research? And are there aspects of the research that are helpful in your administrative duties?

HUSKER: Yes, 100% both of those. Yes, both ways. [laughs]

ZIERLER: What's an example that illustrates the point?

HUSKER: Going back, it's hard to say which way this goes. The DAS that I mentioned, it's a word I'll use again as an example because it works really well. The Caltech engine was always looking for the next thing, like what is the newest technology? What is the latest that we can apply here? Try to always move the envelope. I think it's part of the idea of Caltech's philosophy in general, and we certainly try to do that in the Data Center. That's what excites me. That's where I want to be. What can we do that's new? The DAS is the newest thing. There's nothing newer. It's this new technology. There are all these different questions: How useful is it? Will it really be revolutionary? Is it just adding another headache? Is it actually something new? But that's what makes it exciting. It's the new thing. But it's also a ton of data. It's difficult to even deal with, but we can see it's the new thing. From a network perspective, I look at this and say, "Oh, we need to start working with that and figure out how, because in the future, we need to incorporate this in the Network." Looking at it on the research side, I see the volumes of data coming in. How is this going to be useful to a network? What can this additional information provide? So, I flip back and forth on this same one. I feel like I don't sit on either side of it. John Nguyen and I are certainly both working on it from both perspectives constantly. He's certainly more on the research side, but I'm moving in that direction, too. That's the easiest one, that we're just both.

ZIERLER: Besides not having teaching responsibilities as research professors, is your day to day more or less the same as other faculty in the Seismo Lab, in terms of the meetings you're going to, the things that you're involved in?

HUSKER: I think I'm actually quite a bit different. I'm not on any committees or anything else. The Network takes up all my time. A lot of times I'm at meetings with the Network and I'm not able to go to, say, talks or something that the rest of the faculty might be able to go to. So, it takes quite a bit of time. I would say somebody like Mark Simons might be the exception to what I just said, because he's also in a number of meetings with everything he's doing with NASA, for example. He just doesn't take on these other responsibilities. Certainly before, when I was just a professor—well I wasn't just a professor—I oversaw, and ran the department, and did other things, so I was in a number of meetings, but it was a lot more of just faculty meetings. I don't have any faculty meetings as a research professor.

ZIERLER: Where might the funding in terms of your various jobs—the research professor, the manager of the Network and things like that—what funding is coming from Caltech, from the Seismo Lab? And what's coming from the Survey?

HUSKER: Everything for the Network comes for the Survey and from Cal OES, that's pretty much everything. A little bit from SCEC, too. All my research funds—I've gotten support from the Seismo Lab and also from NSF for all my personal research. Those are the only two. No, also, I haven't even mentioned there's this whole geothermal well research that we're doing for tremors in a geothermal well, and that funding has all come from a project with Jean-Philippe. So, collaborations, the Seismo Lab, and NSF, I would say from the research professor side.

ZIERLER: Finally, do you have opportunity to interact or supervise post-docs or graduate students?

HUSKER: Yes I do. I have two post-docs right now. I still, since I only moved from Mexico a year and a half ago, have a couple of grad students in Mexico, and I'm hoping there might be another grad student down at Caltech that's joining my group.

ZIERLER: Is that, as research professor with all of your other duties, is that an expectation, beside what you personally want to do? Is building up a research group, having your own graduate students and post-docs, is that very much within your expectations and aspirations at Caltech?

HUSKER: It's very much within my expectations. Within my aspirations? I think in general I'm not expected to do as much as the academic professors just because I have the responsibility of the Network, and I think that the position was really focused on the network previously. But I certainly have the freedom to do that as long as I have the time to do it.

ZIERLER: In terms of your interest and your expertise and your bandwidth, if a graduate student or a post-doc would want to work with you, would it be just as likely because of your academic expertise, the things you've done research on, or is it more because of your involvement in the Network and the archive?

HUSKER: If it's a student, I would say probably more on the research side. If it's a post-doc, it might be more the Network. The reason for that is I have funding for post-docs on the Network, so there's the natural connection there, a way to move things forward with a post-doc that I don't have with students. But the students seem to be more excited about the other research, so far. I've only been here a year and a half, so we'll see what happens in the future. [laughs]

ZIERLER: We'll see what happens. This has been a terrific conversation. We've covered a tremendous amount of ground in under 90 minute. Our next one, we'll go all the way back to the beginning, learn about your family background, and we'll take this story right up to the present.

[End of Recording]

ZIERLER: Okay, this is David Zierler, Director of the Caltech Heritage Project. It is Wednesday, October 12, 2022. It's great to be back with Dr. Allen Husker. Allen, great to be with you again. Thanks again for joining me.

HUSKER: Thank you. Thank you for having me.

ZIERLER: Allen, in our first discussion, we did a great tour of all of your research endeavors and professional responsibilities. Today, let's go all the way back to the beginning, establish some personal narrative for you. Let's start with your parents. Tell me a little bit about them and where they're from.

HUSKER: They're both schoolteachers. Or they were, they're retired now. My dad is from Wisconsin, and my mom was born in Arkansas but lived most of her life as a kid in Texas. Both ended up in Washington state, which is where I was born.

ZIERLER: Did you grow up in Washington state?

HUSKER: Yeah. I was actually born in Seattle, and then my parents moved to the suburbs when I was a toddler.

ZIERLER: Washington state, of course, is very seismically active. Did that register for you as a kid? Were you interested in earthquakes and things like that?

HUSKER: No, my interest was always space. I always wanted to be an astronaut when I grew up, so that was more of what I was interested in. I got into earthquakes because I wanted to study planets, and Earth is a planet, so that was my interest there.

ZIERLER: What kind of schools did you go to growing up?

HUSKER: Just the local public school. We only moved once when I was a kid, but it was the local public school, whatever it was.

ZIERLER: Did you always gravitate more toward math and science in school?

HUSKER: No, I did both. In fact, I think we were all overachievers at Caltech, but I was one of these people that my senior year, I think I won more of the awards at the final assembly at the end of the year, and it was just across the board in math and science and everything. I mean math and science and English, history, kind of across the board. Then when I went to undergrad—I actually have two degrees. I think I might be skipping ahead in what you're going to be asking, but [laughs] one was Comparative History of Ideas, and the other was in Physics, and that was specifically because I like everything and I still like everything, so I couldn't decide between the two originally.

ZIERLER: What kinds of schools did you apply to for college?

HUSKER: This is going to sound weird: I only applied to University of Washington. That was just the local university. It was the local big university, and it was because I wanted to be close to family, and my parents didn't want to see me go and that kind of thing.

ZIERLER: Pretty lucky that it's an amazing institution, too.

HUSKER: Yeah, it worked out okay. [both laugh]

ZIERLER: Did you live on campus, or did you commute?

HUSKER: No, I commuted for the first two years, and then I moved in with some roommates and had an apartment just off campus for the last couple years.

ZIERLER: When during undergraduate did you move away from the liberal arts and more toward the science? Did that happen early on?

HUSKER: Yeah, it was kind of happening the whole time, more and more. The liberal arts felt like there were a lot of fuzzy answers, and I wanted more fixed answers. But I've always enjoyed history. I still enjoy history. It had already started, I think, from the beginning. When I was growing up, part of it—a lot of this stems from when I was growing up, just around the dinner table, my dad would ask questions about if I existed or not. It was going back to this kind of philosophical argument of—he would say, "How do you perceive yourself? Is that the same as how I perceive you? Is that the same way you're perceived by your peers? Is that the same way you're perceived by the school principal or whoever? If everybody sees you a different way, who is the real you? If none of them are the real you, is there a you that actually exists?" This has happened since I was like, ten. [both laugh]

These philosophical kinds of questions of "What is reality?" were running through my head, so I guess that's where I got the balance of wanting to explore philosophically what that meant, and also in the sciences.

ZIERLER: What kind of physics did you want to do as an undergraduate?

HUSKER: Astrophysics, cosmology—that was what I was excited about. Then I realized it got to be a little too esoteric for me at a certain point, and maybe my rebellion from my father was becoming less esoteric and more fixed on something that really would have worked. [laughs] Earthquakes and things like that are actually something that people can feel and actually affects people in their lives.

ZIERLER: When did seismology and geophysics enter the picture? Did you get advice from a professor, or how did that work?

HUSKER: As an undergrad, we have to do different projects, so I had done a few different—I was kind of trying out a few different projects. We had a series of doing the labs from the previous 50 years of Nobel Prizes, or the previous century of Nobel Prizes. So, it was a bunch of different lab work that was looking at different—I'm trying to think of any specific one right now off the top of my head. There was the photoelectric effect, which was what Einstein got his Nobel Prize in. So many years after wrote about it, he actually found the photoelectric effect. That was one of the labs we did. We had a whole series of those kind of labs. I ended up working a little bit with the professor, and what he was doing was material physics. It was looking at skaters will skate across ice, you get an edge effect, basically, on the ice where it will turn into water. So, it changes phases into water, and there's a solid underneath, and you have this right at the edge as the skaters are skating across this. So, I was looking at ice, and then I went from ice to looking at ice crystals in the atmosphere. Then I thought more about the atmosphere. Then I ended up suddenly at the Earth. It was kind of like this segue, I guess I want to say, going from looking at these different basic kind of physics laboratory experiments to ending up looking at the Earth. [laughs]

ZIERLER: How did all of that translate to postgraduate opportunities? Graduate schools, work opportunities—what were your choices?

HUSKER: I had a TA, who had just gotten his PhD, who ended up working for a company. I went to work for the company, so I was an engineer for a couple years where I just kind of followed him. He had sent this email opportunity of they're looking for somebody to work at the company, so I emailed and ended up working there. They hired me, and I ended up working there for a couple of years. That was much more hands on, looking at hardware, and what can you actually measure with real equipment, so I guess that's a lot of this data loggers that we use now. Mine stemmed from my experience there for a couple years. You have to learn how to measure things at a sampling frequency and rate—what can the hardware actually do? Now that you say that, I realize how much that informed the rest of my career, those two years, because the hardware got to be really interesting for me. When I moved onto grad school, that informed a lot of what I did, was actually working with the hardware and working in the field and doing a lot of the hands on part of it.

ZIERLER: What schools did you apply to for graduate school? Were they all geophysics programs?

HUSKER: This is going to sound hilarious as well: I only applied to UCLA. [laughs]

ZIERLER: Well, you've got a good track record: two for two.

HUSKER: [laughs] I don't know how much I should get into this stuff. There was a problem with a girl. We broke up. My life was kind of in a flux. I had visited UCLA, so I ended up applying there and no other place and was trying to figure out what to do with my life, and then I got accepted and ended up going for it after applying to only one place. [laughs]

ZIERLER: It was the geophysics program at UCLA?

HUSKER: Yeah, it was the geophysics program. In this job I was working at, I traveled around a lot and went to different companies to help them with what their applications were, so I was specifically an applications engineer. When I got to LA to help with—there's a brake manufacturer that manufactures brakes for airplanes. They have to test the airplane brakes, they have to see how the brake is spinning, and they have to measure the heat and other properties of it; and have it clamp at the right rate, so there's a force-feedback loop that presses in how much the brake is going to press down. The company I was working for gave them the hardware to have it do this force-feedback loop, which measures how much force it needs to apply to brake so the airplane will stop when it needs to stop.

I was at this manufacturing company down near the airport and took some day to go to UCLA and meet people. I actually had done the same thing at Berkeley when I was visiting there, but I wasn't too excited about Berkeley when I met people there. The response I got from UCLA was a lot better, so I ended up applying to UCLA.

ZIERLER: What was the vibe at Berkeley that didn't sit with you well?

HUSKER: I sent an email, and said I wanted to visit, and said some different things I was interested in—tried to meet with people. There was a geomorphologist that was available, and that was it. Maybe I just visited on a bad day, but it just seemed like nobody was interested in meeting potential students. I don't know; that was kind of the vibe I got.

ZIERLER: And you didn't get that at UCLA, obviously.

HUSKER: No. There were a few different people that wanted to. They were like, "Oh, sure! Come by. It will be nice to meet somebody and see what they're up to." It was much more friendly and much more open, I guess, was the vibe I got.

ZIERLER: If I could reverse engineer a question going back to undergraduate. Given your interest in facility and instrumentation, does that mean that when you were pursuing astrophysics and cosmology, it was more on the experimental side, not the theoretical side?

HUSKER: No, that was more on, I think—I had this idea before I went in to study anything. It was like, "I'm going to discover something great, like Einstein" kind of thing. Once I actually got into it and was doing these different lab experiments, I found that I really enjoyed the hands on part. That was really, actually: How do I fine-tune the equipment to get it to work and do what I want it to? I really enjoyed that, so in the process of all the different laboratory work, I realized where I wanted to go with it.

ZIERLER: UCLA of course is a big place; the geophysics department is a big place. First of all, how much catch-up did you have to do not coming from a geophysics undergraduate background?

HUSKER: A good amount; it wasn't 100%. I knew the math and the physics and that kind of part of it, so that wasn't a problem. But the geology was something I definitely needed to catch up on. It took a little more time. I had to spend off-hours reading books and things to catch up. Even in the courses I took, to some degree, there was a little bit of that, but since it was a geophysics course, they expected that I already knew the geology. There was catch-up time, but it was in my own time and off-hours for the most part.

ZIERLER: Did the job and your interest in instrumentation, did that influence the kinds of professors you wanted to work with, the kinds of research projects you wanted to join?

HUSKER: Yeah. A big reason I guess I ended up choosing UCLA—when I had gone ahead of time to meet people, one of the people I met was the person who was going to be my advisor, Paul Davis. He immediately started talking to me about different projects and travel. I realized right away it was going to be more hands-on. He really believed that students should have their own unique dataset to work with, so he really wanted people to be able to run an experiment and to have that unique dataset then to work with instead of something that a million people have looked at before.

ZIERLER: What was Paul working on at that time?

HUSKER: At that time, the main project he had was actually in Santa Monica. They had had the Northridge earthquake sometime before that and looked at focusing from the basin effects. As the seismic wave traveled underneath the Earth, it comes back up to the surface, it can focus like it would through a lens, like through glasses, and increase the amplification just in a certain area. He found that underneath the Santa Monica mountains, basically there's a route there that comes down that focused all of the waves from the Northridge earthquake to make it so the damage was worse in a certain part of Santa Monica. That was one of the things he was working on, but specifically with me—in the past he had also done international experiments. When I first came in, he said he was thinking of doing an experiment in Nyiragongo in a volcano, and that just sounded like the greatest thing ever for me. [laughs]

It ended up not working out to Rwanda, it's on the border with Congo, so it's kind of a dangerous place in the world. So, I ended up working with him, and when we were planning out this experiment, we contacted the University of Rwanda—no, we didn't actually contact them. The week we were going to contact them, the State Department issued a warning that said not to go there, because aid workers were kidnapped and beaten. So, we decided not to do that. [laughs]

ZIERLER: What was the science objective there? What would have been found out as a result?

HUSKER: Nyiragongo is part of the African Rift, if I remember the whole thing. Africa has very slowly—it's been sitting in the same place for millions of years. It's heated up, basically. Underneath it, it's built this hot place that's pushing the continent apart, slowly. There's this rift zone, and there's a line of volcanoes and volcanic activity that's starting to separate the continent up in the northeast part of it. So, that would be in that line. The idea was to try to map that out using seismic techniques.

ZIERLER: What did you do as a result? What was the pivot from there.

HUSKER: The pivot was even odder. [laughs] We had another professor in the department—well, in geology; he was also in Earth and Space Sciences. He was working in Iran and suggested that we should work with him in Iran. I thought, "Oh, this is really great," because it was another opportunity, and it was a similar kind of a thing—there was mountain building and volcanoes, and we were going to put a line of seismometers across it. This was in 2000 when we talked to him about this—yeah, the year 2000. We were getting ready to go and talk to them as well, then a plane hit a building in September 11th, and that whole thing went down the drain. Then, Rob Clayton here at Caltech, who is also friends with Paul, my advisor, suggested Mexico, so we ended up going to Mexico instead. It was a complete pivot to Mexico, which was also volcanoes and looking at them, but it was a different field.

Mexico and the Caltech Connection

ZIERLER: Was that initial connection to Rob Clayton the first time that you thought about Caltech and the Seismo Lab and what they were doing?

HUSKER: Yeah, to a large degree, that was when it came into—I knew about Caltech. Everybody knows about Caltech, but that was the first time that—yeah sure, that was my first real connection, I think, to the Seismo Lab.

ZIERLER: Tell me about the project in Mexico. What was the funding source? Who was supporting it?

HUSKER: There were two main funding sources. Part was UCLA. With Paul, we had a funding source with, it was called the Center for Embedded Network Sensing, or CENS was the acronym. That was computer science—NSF funded computer science center that was trying to take all these different scientists and use computer science with embedded sensors all over the place to find something new. Seismometers was one idea. Another one was aquatic robots. Another was in the forest, but having sensors in all of these different places and using computer science to try to bring back the data to some center and analyze the data, it was before the internet of things existed, maybe a precursor to that. So, that was the idea at the time, but we were just one wing of it. That was one part was from UCLA, and the other part was the Moore Foundation here at Caltech. The Moore Foundation, each of them paid for half of the seismometers, and then the Moore Foundation covered all the field work.

ZIERLER: Allen, last time we talked about the idea of extrapolability—studying locally, but drawing more generalized conclusions. On that basis, what was really specific to the research project in Mexico? And what might it tell us about things on a planetary basis?

HUSKER: Specifically, Mexico has a flat subduction zone. What that means is when the two tectonic plates are pushing into each other, when an oceanic plate pushes into a continental plate, the oceanic plate always goes underneath the continental plate.

ZIERLER: Why is that so?

HUSKER: The oceanic plate is denser, colder. It's denser. The continental plate floats on top, and the oceanic plate is denser, so it sinks underneath it.

ZIERLER: Is it denser because of the pressure from the water? What makes it denser?

HUSKER: No, it's a separate material. They're two different mineralogies, basically.

ZIERLER: Okay.

HUSKER: There are subduction zones in the world in general. The oceanic plate will go down underneath the continental plate, and when it sinks underneath, it goes down into the mantel. Typically, it goes at kind of a steep angle of subduction; it goes just kind of straight down into the mantel. But under Mexico, it goes flat. It goes underneath the continent and just kind of floats underneath it almost before it turns and finally goes down. So, that's how it's unusual. There are a few different places in the world where there are this flat subduction, where it goes down and goes flat and then continues sinking. There are only a few places like that in the world, and one of them is Mexico. So, that was both the unusual and also the reason to study it globally, was to understand what makes this kind of thing happen.

ZIERLER: What were some of the key findings of the project?

HUSKER: The key finding was just how flat the slab really is. It really is just incredibly flat, and it seems to go right underneath the crust, which is also unusual. They found a very thin layer of material between the crust and the slab underneath it makes it so the two don't push against each other. Those were the main findings. It's also unusual, because it's flat for such a huge distance—it's over 100 kilometers where it's just really, really flat. Usually, there's at least some angle as it's going down, even if it's mostly flat, it's not entirely. But this seems to be just entirely flat. So, it made it really, just unusual.

ZIERLER: What was the instrumentation that led to these findings?

HUSKER: It's a line of seismometers that run every five kilometers, that ran from Acapulco through Mexico City and to the Gulf Coast. I was in charge of half of those sensors.

ZIERLER: Did this project lead into what became your thesis research? Your dissertation?

HUSKER: Yeah. I did a seismic tomography of it, which is similar to a tomography of the human body when you do a CT scan. Instead of having sensors—on a CT scan, your body goes into a circular things, and you have sources on one side that go off and then pass through a human body, and then on the other side you measure that so you can see what's inside the human body. Similarly, we do the same thing with the Earth, but instead, we can't say where our sources are going to be on one side of the Earth. They are actually just earthquakes—earthquakes and the seismic waves that go through the body of the Earth. We have sensors on the other side that measure that. That's where we get our tomography of the Earth, from that system.

ZIERLER: I remember hearing from Mike Gurnis, that the turn of the century is really when high powered computation really begins to make its mark in seismology and geophysics. Did you feel that in real time? Were there computational advances that were relevant to this research?

HUSKER: It's kind of happening all the time. For this specific research…

ZIERLER: In other words, could this same research have been done 10 or 20 years in the past?

HUSKER: Not nearly as easily, no. It's happening continuously all the time. At the time, this was—I think there were clusters of things that were starting to happen, but instead of having a real cluster, what I did was we had a computer lab where we had five or six computers, and half the time they weren't being used. So, I ran simulations on all of them continuously. Remotely, I would log into them all and run things, so it was kind of a precursor to a real cluster, I guess. Even then, at this point the single computer was good enough to be able to do this kind of computation on a single computer. It didn't require a real cluster. It was just to run multiple instances of the same code over and over again. That was a precursor to a real cluster. People certainly had clusters at the time. It's just that we didn't have one in specific that we were using. Certainly the years before it would have been hard to do the same thing just because the computational power wasn't there. Since then it's just exploded. Now there's even more computational power, and it's just at a whole other level. But yeah, it's continuous.

Even simple programs. There's something called seismic analysis code, SAC, which is still used even though it's kind of a line—you type in line commands and whatever. But some of the original ways that would show a seismogram, it would do a reduced version of it, because the original computers weren't good enough to show the full version of a seismogram, and now it's just really simple to show a full version of a seismogram. It's not a big deal. Even at the time it was, and so you had to change slightly the output so it shows the real version of it. Even now, we're doing terabytes of data with DAS. I don't have the words for this, but it's been continuously getting more and more computational power and more and more data throughout my career as we go. I'm kind of going in a long and a roundabout way of saying yes, definitely you're right. I couldn't have done what I did at the time with the computational power of ten years before, but even looking back now that it's been like 20 years, you could do the same thing so simply on my laptop. It feels like 20 years ago, right? [laughs] It feels like almost nothing now at this point. Any tomography at that scale is so simple now. Things are so much more complicated and there's so much more data now.

ZIERLER: Allen, in setting up the seismic network for this project in Mexico, was the original game plan for it to be temporary or a permanent network?

HUSKER: No, it was always going to be temporary. It's a common thing in seismology to set up temporary networks, because the Earth doesn't change, at least on our human scales. If you're looking at changes over millions of years, if you just put out a temporary network you can take a snapshot of it and take it away and you have that snapshot. You don't need to leave it there for a long period of time.

ZIERLER: Is that more about how costly seismometers are? In other words, why not just leave them where they are? What's the motivation to take them out?

HUSKER: That's a good question. It's a little bit the cost of the seismometers, the cost of running the network itself. Running any network takes a lot of money. Having a hundred seismometers that you're trying to maintain in a foreign country for multiple years just takes a lot of money. It's the seismometers themselves, but also just the effort involved for managing the data, collecting the data, all the fieldwork involved.

ZIERLER: Did you establish connections with the National Autonomous University of Mexico during this project? And is that what led to your faculty appointment there?

HUSKER: Yeah, largely. I was in an interesting position, because I was a grad student, but they sent me down as the head of the project. Not the head of the project, but at least managing the project. They gave me a professor's office who was on sabbatical in UNAM which was a little odd, because I was running a project, they were like, "Oh, you need to have this office so you can run the project. You get this office here." I think the other grad students at UNAM who were grad students there didn't realize I was a grad student. [both laugh] So, I worked with the professors to run the project, so I was already kind of an elevated status compared to the other grad students at that point. Once the project ended and I had my PhD, I worked for a short time in Peru as a research—doing a similar thing, just installing equipment and whatnot. Then eventually I got a postdoc and almost immediately after the postdoc ran for just a few months before they offered me the permanent position.

ZIERLER: The postdoc, was that UNAM as well?

HUSKER: Yeah, that was UNAM.

ZIERLER: How was your Spanish before you got to Mexico originally?

HUSKER: It was not very good. [laughs] I couldn't speak it very fluently, and it was just very stilted.

ZIERLER: And you picked it up just in the course of living there and doing research?

HUSKER: Yeah. It was funny, because most of what I picked up was just what I needed on a daily basis or what I used. So, I could ask for directions really easily, because I had to figure out directions from point A to point B to install something. So within a very quick time, I could do directions, ask for permission to install a seismometer, explain what that was about, explain things about subduction zones and whatnot. But I had problems, like say when I first met my girlfriend who ended up becoming my wife, flirting was a difficult thing.

ZIERLER: [laughs] Because you could only talk seismometer. [laughs]

HUSKER: Yeah, exactly. It was a different language. [both laugh]

ZIERLER: Besides the work in Mexico, were there any other projects that informed your dissertation?

HUSKER: No, that was it. For the dissertation, it was all my work in Mexico.

ZIERLER: Did you get a chance to work with Rob Clayton on this at all, directly?

HUSKER: Yeah, no. He was on my committee for my PhD, yeah.

ZIERLER: What do you see as some of your contributions with the thesis? What did you add to the body of knowledge?

HUSKER: I did the tomography. I showed, with the tomography, exactly where the slab was going down into the mantel, and also it was cut off at the bottom. There was a huge slab called the Farallon slab, which was part of the Pacific Ocean. It has gone underneath, all the way underneath the North American continent, so it's underneath us right now. [laughs] But it went down all the way into Mexico, but it's actually this huge weight that's been kind of ripped off, so it's ripped off the bottom of the slab where it is now in Mexico.

ZIERLER: The postdoc in Peru, that was for the Peruvian government? Were you doing that from UCLA?

HUSKER: It was kind of the next part of this project. After they finished with those seismometers in Mexico, they sent half of them to Peru. So, for the half in Peru, I went down and helped with that for a few months before that project ended. Actually it was just getting started. I stopped with the project and got the postdoc in UNAM.

ZIERLER: Are there ongoing institutional collaborations between UNAM and UCLA and Caltech? Do we see an afterlife to this research?

HUSKER: Not officially, but yeah, now that this has all happened. I think me, also, having gone back between the institutions, keeps up the relationship quite a bit. There's other people also. So, Xyoli Pérez-Campos is here right now, and she's doing a sabbatical from UNAM. Not in an official way, but just the fact that we're all scientists and we're colleagues now, so we interact.

ZIERLER: We're on a pattern. Let's see: you only applied to University of Washington, then you only apply to UCLA. Did you only apply to UNAM? Did they recruit you? How did that play out?

HUSKER: [laughs] It's an interesting pattern you're noticing. In some ways, yes. My wife—and there's also one woman involved in a couple of these decisions. One was a break-up; the other was much better, I was getting together with my wife. We were looking for places to be together, and she already had a job at UNAM, so I got the postdoc there. Then they happened to open up the position when I was there for the professor, and they said, "Hey, do you want it?" I said, "Yes, please." It kind of just fell in my lap, being in the right place at the right time.

ZIERLER: At this point, you're probably fluent, like teaching and being conversant in Spanish is not going to be an issue at this point.

HUSKER: Yeah, exactly. I taught in Spanish. I had interviews on TV in Spanish.

ZIERLER: Tell me about geophysics at UNAM. How far back does it go?

HUSKER: They had a seismometer there since 1906, I want to say. I can't remember the year, but right at the turn of the century. They had already had one of the original seismometers, and it's a two ton seismometer. The thing is just huge. It's enormous. You could push it with your hand. It's made to move and really measure the movements of the Earth. But they had that—yeah, so it goes back a long time, I guess. They have a really strong history of earthquakes. Actually, they have a lot more earthquakes than the US does. It's part of their history, so now it's part of their culture. They actually had that dictator at the time, at the turn of the century. His name was Porfirio Diaz. He was the one that instituted the Seismological Service. That's been going on for a really long time. Actually, the Institute of Geophysics is only what—I can't remember how many years old it is anymore, but it started after the Seismological Service did.

ZIERLER: What kinds of courses did you teach there? Would it have been basically the same as if you were at an American university?

HUSKER: Yeah, it was more or less the same. I taught for the grad school mainly. I did Seismology 101 for entering first year students that were taking seismology. I also ended up towards the end moving toward inversion theory.

Mexico and the Subduction Zone

ZIERLER: The research that you did for graduate school and postdoc, did that feed into sort of larger questions about why Mexico is so seismically active?

HUSKER: No. The reason Mexico is so seismically active I think is well known. It's because it's a subduction zone. The two plates push into each other. Generally around the world, the most seismically active places are subduction zones, whereas here, we have the San Andreas, but it's not one plate pushing underneath the other, it's two plates rubbing against each other. Just that difference can account for the difference in why one area is more seismically active than the other. The biggest thing I actually was looking at was as a postdoc and then actually as a professor. Most of my time was looking at slow earthquakes. The largest slow earthquakes that have been recorded to date in the world are in Mexico, so that was the interest for me: Why are these happening here? What's happening? What's going on?

ZIERLER: The uniqueness that you explained to me earlier about the flatness, is that related to the phenomenon of slow earthquakes and slow slip?

HUSKER: Yes and no. It's not unique to just flat zones. They're all over the world in different subduction zones, but certainly in Mexico it was kind of easier to study them in some ways, because you have this flat zone. So, we could put seismometers on the land, and the subduction zone is underneath us, and exactly where the slow slip was happening or the slow earthquakes were happening was under the land. In other places, like in Japan, it's actually offshore. A lot of it's offshore, so it's harder to study there. You need OBSs, which are ocean bottom seismometers a lot of times. There are onshore portions, but it becomes more of a challenge, whereas in Mexico it pretty much was all right underneath you, so you could put seismometers on land to measure it.

ZIERLER: Allen, in our last discussion you explained how, of course, the Mexican government does not have the kinds of resources that the US government does to support this kind of work. Does that mean that you were supported by the NSF? That you were still sort of hooked into the US funding structure?

HUSKER: No. It was either collaborations with external partners—so there were a lot of collaborations with France and Japan, and the previous collaborations had either been done with Caltech and UCLA, or make do with what you got, kind of.

ZIERLER: Yeah. [laughs]

HUSKER: It was fewer seismometers, so we got really good at coming up with algorithms that would work on a single seismometer.

ZIERLER: Tell me about how you got involved in the UN issues, the international policy. How did that start?

HUSKER: A lot of it was, basically, I already had experience with the networks, like installing a temporary network and running that. They have national pride, so they didn't really want to have—they have their own network, the National Seismological Service, but in large part, they want to have Mexicans in charge of it or Latin Americans in charge of it. So, my opportunity might not be as strong, but they have these other four stations that were run by the United Nations. Xyoli, who I mentioned earlier, had been in charge of it, but she wanted to hand it off, so I said I'd be willing to do it. They were like, "Sure, go for it." That was that.

ZIERLER: Did you interface at all with the Mexican government? How did that work? You're a university professor, it's an international environment, but you're representing Mexico, so how do all of these things play together?

HUSKER: [laughs] It's a great question. It's an international organization, but they still have to have a contract to pay somebody to run these stations. They have the contract that paid UNAM and the Seismological Service, which was housed at UNAM, to run this, so they needed a PI on the project, basically—the equivalent of a PI. I was the equivalent of the PI on the project. In the terms of the UN, what they would call it is the station manager, or the station operator. I was the station manager for those stations and just ran the contract. But because it was with the United Nations, it meant I had to go to training at the United Nations, and when I'd be there, they'd put the little tag that said Mexico in front of me for the training or whatever I had to do there.

ZIERLER: Maybe it's a silly question, but Mexico does not have nuclear weapons—maybe Los Alamos, Livermore, that's where we're thinking. What's the relevance or the importance of having a Mexican seismic network keyed into nuclear test ban treaties and things like that?

HUSKER: I think it's just part of getting every country on board. Actually, one of the Nuclear Test Ban Treaty was signed in Mexico. I can't remember the year now, and I should remember it, but I don't remember it off the top of my head. They were within this international realm of nuclear safety. Mexico kind of played a role in that. So, that's just in general trying to make the world a safer place—putting seismometers in every country to measure any sort of nuclear tests that were going on is part of the idea. And certainly the ones in Mexico, what are they going to measure? The ones in the US, right? [laughs] The US already shares its seismic data openly with the world, so if there was ever a nuclear test in the US, it would be very obvious to the rest of the world that it happened.

ZIERLER: That would register in—for example, a nuclear test in the Nevada desert—that would register in a Mexican seismic network?

HUSKER: Yeah, easily. With a nuclear explosion, it's hard to make one that would be—there would basically be a magnitude 6 or above. Magnitude 6ish—the equivalent of a magnitude 6 earthquake is the seismic waves that you'd see coming out of it.

ZIERLER: That would be for what, a Hiroshima/Nagasaki level A-bomb?

HUSKER: Yeah, or even something smaller, like the ones North Korea was testing was about that size. So, it's actually kind of, from what I understand, harder to make them too small, because you have this explosive, potentially. You have to make it a certain size, or it has to be a certain size before it's even possible to become nuclear. So, it's got to be at least kind of that big. So, once it's at least that big, you're going to measure it from quite far away.

ZIERLER: Were there directives coming from the Mexican government? Were you taking orders form the authorities? How independent was your position, essentially?

HUSKER: No, it was very independent. It was the equivalent of running a project. It was a contract that was the UN had paid UNAM to run this contract. In fact, they had a specific position that Mexico—so, my position was to run the contract, really, but they also had positions that were kind of more within the UN that came from Mexico, so I was not allowed to be in that kind of a position because I was not a Mexican citizen. I was very specifically running this contract, basically is what my position was.

ZIERLER: What kinds of things were discussed at meetings in Vienna? What did that look like?

HUSKER: I honestly only went one time, but it was the meetings—I mean, I only went to Vienna one time, but we had other meetings like video kind of meetings, video chat kind of meetings, emails, whatever. But it was largely trainings and a very international community, so when I was there, there was somebody from Bermuda, I think it was Nigeria, Kenya, Russia. You have all these different people in the same room to talk about the different aspects of the Network. It was very much running the Network. It wasn't international politics. It was like, how do we get whatever we need installed to make sure we have enough money to keep this station upkeep, and how many times a year do we need to visit the station, and questions along those lines. It was very technical. This part of the UN was all very technical oriented questions.

The one political thing that was just really odd: the UN is not allowed to say if there is a nuclear blast or not. The CTBTO actually is not allowed to say it, because the different countries don't want to give them that authority. So, ironically, they have to produce all this information and data and whatnot, and then they're not allowed to say if there's actually a nuclear blast. They can kind of point at it and say, "Hey, look at this." But then every individual country is supposed to say whether or not there actually was a nuclear blast. So, there was a little bit of politics that was there behind the whole thing, but people just kind of did their jobs. This was their contracting of the seismic network in your country, so go do it, kind of was the task.

ZIERLER: I very much have nuclear issues on my mind right now given all this craziness with Russia and Ukraine. Based on what you've learned, both in terms of assessing if, heaven forbid, Russia is thinking about a nuclear attack in Ukraine, and if they actually do that, what role internationally will seismologists and geophysicists play both in predicting it and then also assessing it?

HUSKER: There's no predicting, there's just if it happens. It's a political question as to whether or not—or a war question as to what's going to happen. Once it happens, the seismic waves—it emanates seismic waves as soon as a nuclear blast goes off, and within seconds or minutes, it'll be very evident that it occurred. I don't think Russia will try to hide it, but it will be very obvious that it happened. Any bomb blast, actually, you can pick up. It doesn't need to only be nuclear blast. It's just that nuclear blasts are so much bigger, you can measure them much farther away.

ZIERLER: Would seismologists be the first—whether or not Russia tries to conceal it or whatever—would seismologists be among the first to be able to confirm that it was a nuclear blast, just because of what a nuclear blast is seismically?

HUSKER: Yeah, it will be obvious almost immediately that it was a nuclear blast. Yeah.

ZIERLER: Back to the science, a much happier topic. At UNAM, did you have a traditional research group? Did you have graduate students and postdocs, or was it more of a teaching university?

HUSKER: It is a teaching university, but at the institute I was working in, it was more research, so yeah, we were very research focused. I had students, and the postdocs were harder to come by just because the money. I had a postdoc towards the end that I was working with, and just in general in the group, there were different postdocs that we could work with, but not nearly as much so like here at Caltech.

ZIERLER: How much were you collaborating in the United States? Were you keeping current with the literature in the US? Were you working with American seismologists?

HUSKER: I was working a little with Americans, but more tacitly. Mainly I was working with Japanese and French colleagues.

ZIERLER: At what point, if at all, did you think about returning to the United States? Or did you get an offer? How did that happen?

HUSKER: I'd been thinking about it for a little while, just off and on, if there was an opportunity that showed up, then I might apply to it, but it was never that strong a thing. I have a son who has a disease, and the medicine was only available in the US. That was kind of a big issue to get the medicine. We were flying back and forth to get the medicine and it was a big headache. Once that happened, I started applying a lot more to try to find a position here.

ZIERLER: Besides Rob, did you have any other contacts in the Seismo Lab? Have you worked with anybody else here?

HUSKER: Indirectly with Mike Gurnis, but very indirectly. He was on the paper or papers that came out, some of the papers from that research project for my PhD.

ZIERLER: What was the initial job at Caltech? What was the offer?

HUSKER: I'm still doing it. I've only been here a year and a half. It's to run the Seismic Network, pretty much. The Southern California Seismic Network is the crux of the position. And then if I have time to be able to do research…

ZIERLER: And this is a great opportunity of course. You jumped at it.

HUSKER: Yeah. Yeah, yeah, yeah. No, I was really excited and kind of shocked [laughs] that I was given the opportunity. So yeah, it's been great.

From UNAM to Pasadena

ZIERLER: Tell me about your initial impressions when you got to Pasadena and you got to see the Seismo Lab from the inside.

HUSKER: I had seen it before, years ago, when I was at UCLA. Actually, when I was a grad student at UCLA, because I was running the Network in Mexico, Rob had actually set me up with a P-card. I don't know if you're familiar with them.

ZIERLER: All too well. I wish I didn't, but I know all too well about the P-card.

HUSKER: The Caltech credit card. I had one of those, just to be able to run and pay for things and do all the fieldwork I needed to do. I had already come over here and gotten involved with a little bit of the bureaucracy [laughs] from even beforehand. So, I was already familiar with Caltech from those days. It was really odd when I first arrived as a professor, because it was right in the middle of COVID, so nobody was here. My first day or days, if I came into the office, I never saw anybody. It was just empty. It felt like a ghost town. I opened up the door and took a picture of myself—you know, a selfie—just in front of the door and said, "This is really happening." [both laugh] That was pretty much it.

ZIERLER: How much of your experience running the Network in Mexico was transportable for Caltech, and how much was brand-new for you?

HUSKER: A lot more was transportable than I actually would have expected. Mexico has a lot of bureaucracy. I thought that Caltech being a private institution, there would be a lot more ways to cut through the bureaucracy, and it turns out that hasn't been the case. [laughs] There was a lot more than I expected.

ZIERLER: What did this mean, the move in terms of your research agenda on slow earthquakes? As you explained earlier, being in Mexico, thinking about Mexico was really fundamental to the research. How did that change or not when you came to California?

HUSKER: That's a great question. I had already been thinking about moving away from that a little bit. So much of my career was based on these slow earthquakes and understanding that, but it was getting to the point where it was like, "Okay, this is interesting, but what else is there to do?" I had already been exploring other things and trying to move away from that just a little bit. One of the algorithms—I think I mentioned this last time—that we had come up with was this technique called a match filter, which repeatedly looks for smaller signals within the tremor that was occurring. This same technique can be used for a lot different things, so I thought it would be cool to go back to my roots of when I wanted to be an astronaut and explore on other worlds if you could use these kinds of techniques. The machine learning was also coming along at the time, so I got interested in trying these techniques maybe on Europa and then thinking about ice, so just completely moving away from anything related to Mexico. It was useful coming to Caltech to be able to explore and move away from, not just Mexico, but to more global and even extra-global activities.

ZIERLER: Is that to say that JPL is something that you can definitely take advantage of?

HUSKER: Yeah, 100%. In fact, even before I left Mexico, I was already starting to work with Mark Panning who's a scientist at JPL. He and I were already doing some research together, and when I moved here it just made it easier.

ZIERLER: Institutionally, running the Seismic Network here, who are your peers? Or, how does that work beyond Caltech?

HUSKER: Another great question. There are a few different levels to that. There's the California Integrated Seismic Network. That includes us. Then right across the street is USGS in Pasadena, then there's USGS in Menlo Park and Berkeley. Also the California Geological Survey. The five of us make up the California Integrated Seismic Network, so it's all the networks across California. Then there's Shake Alert, which is all the networks up and down the west coast. We have different meetings with all the different networks to try to organize the different algorithms and different things we're doing. Beyond that, there's the ANSS, the American National Seismic System? I might have that acronym wrong. But anyway, that's the American—the National Seismic Network. It's all the different regional networks. We collaborate then with Golden Colorado, which is the seat of the USGS network for all the different products that we have.

ZIERLER: Allen, we talked a little bit about this last time. Is most of your research life running the Network thinking about applications—early warning, mitigation, engineering—and most of the fundamental research is the other stuff you do? Is that a fair way to understand the divide in your work week?

HUSKER: Yes, 100%. That's a great way to explain it. Yeah. [laughs] One is more applied, and the other one is like you said, more fundamental or more science-y.

ZIERLER: In terms of expectations, in terms of what you're capable of doing, what does the best week look like for you in terms of dividing those responsibilities? What's the percentage breakdown?

HUSKER: This is dangerous; I shouldn't say. [both laugh]

ZIERLER: I could ask like this: Do you want to be doing more fundamental research than there are hours in the day?

HUSKER: Yes, I would love to do more fundamental research than there are hours in the day. No, it's a really fun week. Let's put it this way: I enjoy the research side of doing a lot of the stuff at the Network. I also enjoy being able to see the Network progress and do new things with it. I get frustrated when it comes to limited decision making with bureaucracy and whatnot, or I have to spend a lot of time making financial decisions. I'm not an economist; that wasn't what I was excited to do with my life, but it's a necessary part of the job. When I get bogged down in that, then I might be a little bit more frustrated during the week, but if they can be a little bit more science-y, like the earthquake early warning part or whatever for the Network, then that's a lot more fun. Sometimes some of these other decisions about personnel or whatever it might be are things that are necessary portions of the job that can take up a lot of time.

ZIERLER: As you were explaining earlier, there are a lot of reasons to make seismic networks temporary; on a human scale, we can get the snapshot in time, and that works for us. Why, then—obvious question—the permanence of our Network here in California? Is that really just about resources? Or is there something else going on, seismically or administratively, that allows us to understand the permanence of this?

HUSKER: Yes, all that. There are a couple different answers. If you have the resources, then sure, a permanent network is useful. Part of that is to see patterns of seismicity over time, especially going far back. When we get to any larger earthquakes, like magnitude 5, 6, and 7s, the scale on which they occur is much longer than different events. Certainly the smaller ones are happening all the time, but these larger ones just take longer between each one. If you want to understand over a long period of time what's happening, you need to have a permanent network there that's recording it in such a manner that you trust it. You look at this recording from 30 years ago, is it the same as the one I have tomorrow? You want to make sure those are the same thing, so certainly having a permanent network that you can trust over a very period of time is very useful. To be able to see these long-term trends that really are beyond a human lifetime. What we're doing now I think is useful for science that's going to occur 100 years from now if we can keep these records, so they can see what are the earthquakes that occurred now. There's this longer-term trend, so that part of it's important. On the daily scale is the earthquake early warning and Shake Map and these kinds of products that help decision makers on a much smaller timescale as to what they can expect in case of an earthquake. Then there's just general outreach. If an earthquake ruptures our water intake—all the pipes that come in and bring 80% of the water into LA across the San Andreas Fault—if we break those, and then you break also roads that go in and out of LA, and start fires, you're going to have these multiple hazards in the city. Thinking about that on a very short-term scale, and how do we inform decision makers about these kinds of things.

ZIERLER: Is there anywhere else in the world where there is what we can consider an analogue to the Network we have here in Southern California? For example, Japan. Is there anywhere that has this level of resources and this amount of saturation of sensors that we do here?

HUSKER: Yeah, I would say Japan is very similar. In some ways, I think we have a denser system, but they have more bore holes. The bore hole sensors, they actually drill these bore holes that go down into the Earth, and they are much quieter, because they are below humans. They're just much quieter, so they actually have a very, very good system in Japan. I would say it's very comparable. They also have a lot of off-shore sensors in Japan, and that's because, like I mentioned, the subduction zone there.

ZIERLER: Are there opportunities to collaborate? Is it useful to know what your peer is doing in Japan as they might be for you?

HUSKER: Yeah, definitely. In fact, Japan had their early warning system, or has had an early warning system quite a bit longer than the US, and there have already been a number of—as they were setting up the system here—a number of collaborations for the early warning part of things with Japan. So yes, definitely.

ZIERLER: You've only been here a year and a half. I've been here a year and a half as well. Neither of us have experienced any earthquakes yet. Heaven forbid the big one happens tomorrow. Walk me through what your day looks like. Let's say it's 9:30; you're already at work. What happens?

HUSKER: There's the media thing: drop and cover. Everybody goes through the earthquake. Then there's the impact after that initial wave passes—

ZIERLER: I'm sorry to interject, but in terms of getting the message out that people should—you know, early warning, even if it's only 30 seconds—are you involved in the system where we would get an alert on our phones? Where does the Network play in that?

HUSKER: That's all automatic. It's all happening all the time. Humans are too slow to have that system—to be involved at all. It just is all automatic. [snaps fingers] A computer will basically decide—the algorithms we already have running will decide if it's an earthquake or not, and it'll send out those alerts to phones. I'm involved in different committees that are ongoing all the time, but are higher level for these things. The Network itself is providing all the data from Southern California for it. The computer servers right across from my office are running those algorithms right now, so we're definitely involved. But, in terms of any decision making, decisions have already been made once the earthquake starts. The automatic systems all run, so all the work that's been done ahead of time hopefully pays off at that moment when the earthquake actually happens. That's been going on for years before I ever got here, too, so I can't take credit for all of that work. The instant the earthquake happens, all that information goes out, and then there's a number of systems that happen after that. If it's large enough, if it's a big one on the San Andreas, Golden Colorado actually takes over, because then it becomes a national crisis. We have to color coordinate with Golden right away, so there are phone calls that are going to be happening with them, where we say, "Hey, are you seeing the same things we're seeing?" Things like Shake Map and a bunch of other products that will come out, then immediately there's going to be a million aftershocks—well, not a million, a lot of aftershocks. [laughs] Thousands of aftershocks. We'll have a team of people looking at the aftershocks and have to support them. Then our data analysts will be looking at the aftershocks. Then we have to look and see what's broken in the system, like has anything gone offline? Have we lost power to any sites? Are any sites down, damaged from the earthquake itself? Do we need to send out teams to fix those sites? There's going to be a whole response of all the different seismologists, both in Southern California and around the country, that will want to help out in some way, so we have to coordinate that with the USGS. So, that's a lot of phone calls with them. There's also, they use Teams—Microsoft Teams. Immediately there's a Teams thing that will go up that people can log in to and it all happens organically. We have to work with the media within minutes to an hour will probably be here and want to have an answer as to what's going on, so we have to coordinate with the rest of the Seismo Lab to see who's going to be giving updates to the media and work that out.

ZIERLER: What is the protocol for media appearances? Just on a general level, when is it most appropriate for somebody from the USGS to speak on a given topic, and when somebody from the Seismo Lab? How does that work?

HUSKER: That's a good question. I don't know that it's that thought out. It might be more: Who does the media know to contact? They just start reaching out to the previous people they've been in contact with in the past. They reach out to the Caltech team, certainly, and Lucy Jones gets called in almost immediately, so at this point, she probably still would be… [laughs] I don't know if you know this, but she's his wife.

ZIERLER: Yes, yes.

HUSKER: Okay, so we would contact her, and she would be happy to come over. We already talked. So, she would probably would be here within minutes as well and be helping and taking over a lot of that burden. [laughs]

ZIERLER: Allen, do you have a stock answer ready when inevitably the reporter asks why we couldn't have predicted this?

HUSKER: I've answered it so many times, and I guess the answer is kind of always the same. Here's my problem with it: seismologists kind of have different reasons as to why this is. I guess the answer that reporters seem to like the most—well there's two things. One, there's going to be chaos. There are so many different systems happening at the same time. It's hard to predict what's going on, because it's a chaotic system. So, it leads to uncertainty that you just can't get around. There are too many unknowns, so it's just impossible to predict. So, that's one easy answer. The other one is on a geological timescale, we can predict it quite easily. With geological timescales, we're talking about millions of years, so if we can say within this 50-year time frame we're probably going to have an earthquake, we're actually very exact. But that doesn't help with human timescales. But within geological timescales, we're actually really exact. So, [laughs] in the age of the Earth, we're good.

ZIERLER: It's pretty good, all things considered, is what you're saying?

HUSKER: Yeah. [laughs]

ZIERLER: Allen, now that we've worked right up to the present, for the last part of our talk, just a few retrospective questions, then we'll look to the future. Do you see—even though you said you're consciously moving away from it, what might be new and interesting in the field of slow slip and slow earthquakes that might pull you back in?

HUSKER: The last little bits of research that I was doing with—I have a collaborator at MIT, William Frank. He and I have been working for a number of years on this. One of the things that we're seeing is that slow slip seems to self-organize. How do I explain this better? It seems to kind of clump together or almost quantize. For example, in slow slip, it will occur for just one week, or roughly a week, give or take a few days; and then it won't happen again; and then it'll happen again for about a week, give or take a few days, etc.; but it's always that kind of week timescale. Or it might form together to go to a six-month timescale, or it might come together for an hour timescale. But there's nothing in between those timescales, so it quantizes as those specific units. The question is: Why those specific units? How come it's always that specific unit and not something else in between? So, I guess that was where I would go to explore, is why is it self-organizing just in those specific clumps? Why is there no in-between?

ZIERLER: So, there's definitely big, unanswered questions in this field? You haven't moved away from anything that's necessarily settled?

HUSKER: Yeah. No, no, no. Not to place this back on my physics, but in the atom, the atom only has certain energy quantizations that it can go to. It's only at the lowest base level, or you can jump up to the next shell or the next shell, but there's nothing in between. It felt similar to that way to me; like, why is it only these fixed specific units that you get the slow earthquakes?

ZIERLER: In what ways has your research, so far, contributed to where the field is headed, whether or not you remain involved?

HUSKER: Certainly everything that's describing the phenomena in Mexico, but also tying together tremor and slow slip. Part of this quantization that I was talking about was what we've seen, first in Mexico, and then William, my collaborator, pointed it out in Cascadia and Japan as well. I guess those are the big contributions I've given.

ZIERLER: You've had such a unique career trajectory up to this point. Very few American-born professors have done something like what you've done in Mexico. Just culturally, what do you think will stay with you? Your world perspective, your time in Mexico—what do you think is of ongoing relevance to your research and the way you see the world?

HUSKER: That's an excellent question. Just the way I see the world: I already feel a little bit like a foreigner in the US sometimes. Having come back, it feels—just an example: if I go to Mexico and I see any of my colleagues, the first thing we do is all give each other hugs, and if it's a woman you kiss on the cheek because it's a part of the culture. The secretary of our department where I was working in seismology, if I saw her, she'd want to run up and give me a big hug and kiss on the cheek and be like, "Hi, how you doing?" Just very warm and embracing, but it's part of the culture; that's how people greet each other. If I did the same thing here at Caltech, I feel like I'd probably be fired.

ZIERLER: [laughs]

HUSKER: So, it's such a different culture in that sense. So yeah, just kind of getting used to the difference, I guess. [laughs] Again.

ZIERLER: Between the institutional and personal relationships you've built, your language abilities, is maintaining a research agenda in Latin America important to you? Is that something that you want to maintain?

HUSKER: Yeah, definitely. I think so, but maybe moving beyond to not just Mexico, but just thinking further. In fact, I've already been invited to be part of SZ4D, which is a subduction zone initiative, and they're going to be running experiments in Chile. Because I have experience already in Latin America and speak Spanish, they're like, "Hey, you'd be a great candidate to help out with this." It's certainly something I'd be interest in. The problem I have is just balancing my time. I'm running the Network, and I already have a research project in Mexico, now maybe Chile, and also thinking about the Moon. So, how do I balance everything? It becomes a bit difficult. I like to say yes too much to everything, because there are so many exciting and cool things to do, but I just have to make sure to limit myself to be realistic.

ZIERLER: Again, only a year and half into it, what might be the trajectory where increased resources, just getting more comfortable in the job—where might you be able to pull back a bit on your responsibilities running the Network so that you do have more bandwidth for the fundamental research, all the other things that you're interested in? What would that look like?

HUSKER: [sighs] That's a good question. I don't know if I can too much. That's the other part: I really enjoy the Network to some degree. I want to try to get more funding for offshore equipment and also, even in Northern Mexico. The area that we actually cover, that we report earthquakes in extends into Northern Mexico, because earthquakes cross borders. Their network is much sparser there, so I've already talked to them about donating equipment and helping to increase the size of their network in that way. As we take old equipment out of the ground—it might be a bit older, like 10 or 20 years old; maybe the equipment still works—we can give it to them, and they can still keep using it for the next however many years. It's a way to maintain their network or make it a little bit bigger possibly. That helps us in that we are locating earthquakes better, and also extending Shake Alert. They're interested in extending Shake Alert into Northern Mexico, so they're going to be talking to the governor of Mexico. That was something I started when I talked to them, and said, "Hey, would this be interesting?" So, that's another thing we're working on: to try to extend Shake Alert also into Mexico, and putting sensors offshore. So yeah, I'm excited about working here, too, with these kinds of things—

ZIERLER: It sounds like you want two of you. You want one of you doing the Network, and another one of you doing the research. [both laugh]

HUSKER: Yeah, exactly. It's hard for me to pull back, I guess. I don't know how that would look, I guess.

ZIERLER: Allen, it's such a long running network. There have been so many advances over the years. Where do you want to make your mark? However long you're doing this, what are the real opportunities to improve the Network in whatever timescale is reasonable to you?

HUSKER: I guess that's where I see it now: what I just mentioned. Making my mark would be to try to improve those problem areas. The problem areas for us are the offshore and across the border. Those are the limits of our network. That's where whenever there's an earthquake that seems to have an error in location or magnitude or whatever, it's coming from one of those two areas, so that would definitely help. It also brings more data for researchers. The offshore I think is a lot of fun, because you have data that can be used for the earthquakes themselves and understanding the Earth, but maybe you can hear things like whales or something you can get that goes beyond seismology. So, there might be other opportunities that are interesting, so I think extending in those two regions are fun. Also, the types of data. I mentioned, I think last time, DAS. The data—it's like fiber-optic cables, so being able to run those offshore and we can get access to those cables, that would be really exciting to find a new data source, if we can find ways to bring that into the Network. So yeah, any ways we can increase the scope, but also new data from DAS I think is exciting. It's just a whole other ball of wax. You mentioned at the beginning of the interview about how I was doing this research with the computer that would have been difficult to have done before. I have in the back of my head DAS, which has terabytes of data, and it's just a whole other level. We're at the limit, basically, with the hardware we have right now being able to use that, and I can imagine years down the road, people will look back and be like, "Oh, it's so easy now." But we're at that stage now, certainly with the DAS data, where it's just too much data.

ZIERLER: You knew coming in all the opportunities there would be to improve the Network? That was sort of baked into your interest in taking on this challenge?

HUSKER: I actually didn't know, honestly, what all the different opportunities might be. I kind of had some idea that the Network would just run itself, was kind of what I was told, and that's really not the case. [laughs] As I've been here, I've seen these different opportunities, and if I see an interesting way to go or a fun way to go, "Hey, let's see if we can do this. Hey, let's see if we can do this."

ZIERLER: Finally Allen, last question looking to the future five years out, ten years out, all of the fundamental research that you want to focus on here on Earth, the Moon, other planets, other satellites, what might that look like? Given all the things to work on, what's most important to you, and where do you see your contributions?

HUSKER: Oh jeez, that's a big question. [laughs]

ZIERLER: That's why I save it for the end.

HUSKER: [laughs] There's understanding the subduction zones better; that's part of it. Certainly in Latin America, but beyond. That's what I've been doing previously with fundamental research. I definitely need to switch more, since I have a network here sitting under my feet at my fingertips, to do some more research with the network that I actually have here. I'm slowly moving in that direction, but that's one of my goals: to be able to do more with the Network here. But certainly I'm working with some great scientists and postdocs that are doing a lot with the Network already. The other thing I guess is the DAS, the combination of using DAS with everything, both in terms of incorporating it into the Network and hopefully running it offshore, but also it would be exciting to have DAS on the Moon. If there's any way that could work, I think John and I are both excited about this, and people at JPL and a postdoc I have here would all be excited about that opportunity as a new data source.

ZIERLER: What are the mechanics? How does this get to the Moon? Artemis II?

HUSKER: Yeah, right? [laughs] If Artemis could do it, that would be great. Another potential might be just doing some sort of satellite. It's a fiber-optic cable, so if there could be some way to unroll, unravel the cable away from the lander as it comes down and it just has something that unrolls it away from the lander, that would certainly be an interesting way to go. But if it needs to be the astronauts that put it down, that's another way to go.

ZIERLER: I'm sure would be thrilled to do it himself.

HUSKER: Yeah, wouldn't we all? [both laugh]

ZIERLER: Allen, this has been a great series of conversations. I'm so glad we were able to do this. Thank you so much.

HUSKER: Thank you.

[END]