skip to main content
Home  /  Interviews  /  Lianxing Wen

Lianxing Wen

Lianxing Wen

Professor of Geophysics, Stony Brook University

By David Zierler, Director of the Caltech Heritage Project
May 4, 2022

DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It's Wednesday, May 4, 2022. I am delighted to be here with Professor Lianxing Wen. Lianxing, it's great to be with you. Thank you so much for joining me today.

LIANXING WEN: Thank you for your invitation. It's great to be here.

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

WEN: I'm a Professor of Geophysics at Stony Brook University.

ZIERLER: How long have you been at Stony Brook?

WEN: 22 years.

ZIERLER: And how has the program in geophysics and seismology grown over the years at Stony Brook?

WEN: Actually, geophysics is more of a similar size, but we've grown a lot in planetary science in the past few years, and that was really due to a major focus on Mars and other planet explorations.

ZIERLER: Being so close to Brookhaven National Laboratory, is there any overlap? Is there any opportunity for collaboration there?

WEN: My personal research is not there, but I have a lot of colleagues working on the mineral physics side and material sciences. They have a lot of work and collaborations there, and I collaborate with some mineral physicists in that regard, but I'm not directly involved with Brookhaven National Lab.

ZIERLER: There are so many subfields in geophysics and seismology. What have been some of the main areas of research you've pursued in your career?

WEN: My research is mostly three parts. One is the earth's interior, trying to look at what's in the earth. The second part is trying to understand the dynamic systems of the earth in the longer timescale, geological timescale, and also in the modern human scale, in the scale of years and hours. The third part is looking at special seismic sources, not traditional earthquakes, like nuclear tests, induced earthquakes, and some seismic sources from hurricanes and other types of waves in the ocean, and deep earthquakes. Those are the three major parts.

ZIERLER: What part of your research is more theoretical, and what is more on the observational or experimental side?

WEN: Almost every part of my research is based on observations, like using seismic data geodetic data, like we see from GPS, gravity variation, or strainmeters. But over the course of this research, a lot of observations require the theoretical interpretation of the data. It's both observational and theoretical, but I'm more focused on the observations, what I can tell from them. Most of my method's developments go toward that goal.

ZIERLER: What role does geodynamics play in your work?

WEN: Geodynamics gives us an idea about how the internal system works, which explains some of the things we can see at the surface, like how the surface is moving, how the mountains are built, volcanic eruptions, and the chemistry associated with volcanic eruptions. Also, some of the dynamic systems might trigger an earthquake and other [seismic] sources. That's part of the role, from a dynamic point of view, trying to understand how this system fits into what we see at the surface or what we know about the interior. And we have a lot of knowledge about the earth's interior from seismology.

ZIERLER: How far back in your career did you start to take an interest in planets beyond earth?

WEN: I started when I saw there was global mapping of gravity and topography on Mars, seeing the beautiful dataset. Although, at that point, we didn't have seismology, but it gave us ideas of how the internal system worked on Mars, and I did a little bit of work on that. Also, there was the geodetic data, which is the moment of inertia of Mars, and I had previous experience working on mineral physics, and that could build into possible composition models trying to figure out what possible density profile to fit those data. Recently, we were working on the seismic data on Mars, but we haven't gotten any results yet.

ZIERLER: When in terms of your motivations and curiosities is studying other planets simply about learning what they're like, and when is it useful in refining your research for earth science, your studies here back on planet earth?

WEN: My interest is primarily, when I look at the earth, I have some other research on the seismic anomaly inside the earth, especially in the mantle side, which I think has a lot of compositional changes in the deep mantle. I think a very similar scenario applies to Mars, which is, for instance, a big volcanic eruption in Tharsis province, and that could be very similar to what we see in Africa, just that we have many things that have been destroyed [on the earth due to plate tectonics]. I see those similarities, and that's what convinced me that there must be some kind of similar process happening between different planets.

ZIERLER: Your work with seismic waves, how far down does it go? Does it go all the way down to the core? Do you focus on the mantle? Where is your area of research?

WEN: The seismic waves actually go past the center of the earth, so they can go as deep as the center, they go right through the center of the earth. My research is from the bottom to the top, and the deepest one's in the inner core. The second emphasis is on the boundaries between the mantle and the core, so on the mantle side of that core-mantle boundary. And the third part is focusing on the shallow part, crustal [structure], the top 60 or 70 kilometers beneath China.

ZIERLER: Do you see your research contributing to debates about both the structure and the movement of earth's inner core?

WEN: Yeah, well, it's one debate right now. It's a very interesting observation. I think it really points to new discoveries in the inner core. There is a story about the inner core differential rotation that was published by two [past] Caltech graduates. The idea is, the inner core is rotating faster than the rest of the earth. The primary evidence is that the seismic waves going into the core change over time. When we look at seismic waves ten years later, they appear to be slower or faster compared to ten years ago. That was interpreted as the rotation of the inner core with respect to the rest of the earth. But my work using different dataset has indicated that the earth's inner core surface is actually changing, enlarging and shrinking. Because the inner core has a different structure, it has a different speed than the outer core, so you have shrinking or enlarging of the inner core, which also changes the travel time. I'm arguing, based on what we've seen, that the inner core surface is shrinking and enlarging in human time scales, not that the inner core has a differential rotation. That has totally different implications.

If the inner core is rotating faster than the earth, there will be a completely different understanding of core dynamics than locally changing topography. Changed topography relates to energy release. If you add some inner core material, you release latent heat, and you also release some light element that will screw up the convection in the outer core, and vice versa. Those are really the driving forces of it [the outer core convection]. Other than in the mantle, you have radiogenic heating that provides energy into the system, outer core convection is mostly because of the solidification from the outer core to the inner core. I think that's a reason why it's so important, because we thought that [the inner core surface] had never changed [in that time scale]. The scale is so long that the inner core has been there that the growing and releasing was [thought to be] a very slow process. And we also thought the growth was uniform globally. [We] certainly see very different cases there.

ZIERLER: So much of your research is dedicated to understanding how the processes of both the inner and outer cores of the earth affect what's happening at or near the surface of the earth. If we could run through a few of those areas where you've looked at that relationship, let's start with vulcanism. What do we know now about volcanic processes as they relate to what's going on internally within earth?

WEN: The volcano we see is mostly related to the deeper part of the mantle, which is outside the outer core. One of the regions we see [is] in the South Atlantic Ocean [and Indian ocean], we call a hot spot of volcano regions and in Africa, and if we look at the relative motions in the past, they [the volcano hotspots] seem to be relatively fixed, so they don't move with respect to each other a lot. If you look at the geochemistry in those volcanoes, you see the geochemistry's quite unique compared to other volcanoes. [The volcanism in] this region, I think, is something coming from a chunk of very different material in the base of the mantle. Because the material is different, the formation processes are different. They came from the early earth, so they have a very different composition. That gave rise to thermochemical plumes and the chemical anomaly observed in those volcanoes. So you had a plume coming out [of the chemically different material at the base of the mantle], then you see those plumes are relatively fixed and also plumes rise with different kinds of chemistry that come to the surface. This is on a large scale. There are small-scale ones as well, like Iceland, for instance, in which we see very different kinds of seismic anomalies at the base of the mantle. That could also [be] compositionally very different and give rise to the volcanism.

ZIERLER: What about intra-plate deformation? I wonder if first you can define what intra-plate deformation is, and then explain its relations to earth's inner processes.

WEN: If you're looking at the surface, we can divide it into, say, blocks in general, which is what plate tectonics says. Within each block, there are uniform velocities and uniform rotations. All the deformations are either colliding or sliding with each other across the boundaries, which is what plate tectonics says. But that's just the first order of approximation. Beyond those boundaries between two plates, inside the plate, there are also a lot of deformations. They shrink, they expand, and they have stresses inside. That's what intra-plate stress and deformation mean, anything within the plate. And according to plate tectonic theory, it's supposed to have no deformation [inside the plates] whatsoever because they are rigid.

ZIERLER: What are the most important concepts in geochemistry that are important for your research?

WEN: What I'm interested in, there are regions, like different kinds of volcanoes, and they give different kinds of chemistry. And this is very well-known in geochemistry. What I'm interested in here is a certain group of volcanoes that give a very unique geochemistry signature and how it relates to what we see in a very different part of the earth. We do see some examples like the volcanoes in Africa and Iceland, maybe Hawaii, too, that can be tied to what we can directly see at the base of the mantle. And the other ones, that could be a different story. We have not explored but I think of those things that could be related to different mantle.

ZIERLER: What about, of course, the major topic of plate tectonics? What do we now understand about plate tectonics as a result of our enhanced knowledge of what's happening internally in the earth?

WEN: The plate tectonics theory, the first order of approximation, there were rigid plates, and that the earth surface is composed of several pieces of plates, and there was no deformation in the interior. Also, when you have different plates moving each other, they had motions also bypassing or shearing between the plates. [The physics of] that piecewise velocity was not very well-understood. But the internal deformation, I think, now is quite clear, and that's because each of the plates itself is not rigid. And the plates also have variations of densities inside. We have mountains, and mountains have roots that have light density, so we have something which is very lateral. Some regions have thinner crust, some regions have thicker crust. And that will create deformation. The higher crusts, because they have higher gravitational potential energy, they tend to spread out, then to produce stresses and also deformations. And the deeper part, the deeper circulation, the deeper materials that are moving are also coupled to what we see at the surface, and that coupling could be shearing or pushing. And that also could contribute to deformations in the intra-plate. That's the integral system. In terms of velocity, we do see velocity has a sheared component or rotated component. We cannot explain that in a model because in all the models in the past, everything's uniform rigidity [laterally]. The properties are uniform [laterally], so you have something pulling down [or pushing up], and everything is converging toward that or diverging away from that. You cannot create shearing or rotational motion. That [shearing or rotational] process is because you do have that contrast of rigidity between the continents and oceans, and you also have a layer beneath that, a couple hundred [kilometers] or so beneath that, a soft asthenosphere layer. With these combinations, you can explain why the plates move [bypassing each other] or have a rotating component other than just the converging or diverging component. I think those observations can be explained in the context of what we see and observe, also the density changes and the rheologic properties in the earth's interior, so we built a model to explain that.

ZIERLER: In light of your response on plate tectonics and your interests in relation to the earth's internal processes, what about continental drift? Where is the overlap between your ideas and work on plate tectonics, and where is it a separate discipline or separate approach to looking at continental drift?

WEN: The overlap component is if you're looking at general draft, no interactions, without details inside intra-plate, it is quite clear that you can build [the draft] into what we call a dynamic system. We know what the past subduction has been and where they're distributed. Those are the major driving forces that have created a circulation system so the continents move in the direction they move. In general, that's quite consistent with that model and could be explained in that context. In more detail, which is where the continental drift or plate tectonics part fail, there are more details, like lateral changes of densities and lateral changes of properties, and those are attributed to the secondary effects. But they are even more important effects that tell us what the dynamic system is coupled to. I think there's no contradiction, it's just that the continental drift is the first order, and there are some things we explain really well. Directions of the draft, we did not understand the rotating component, but we do now because we think it is [due to] geological unit's rheologic properties. But there are also details, internal deformations, which tell us more about the internal [dynamic process] and coupling [between] the shallow part and deep part of the earth.

ZIERLER: An overall question relating to your entire career in seismology. Where do you fall within the debates about whether earthquakes can ever be predictable? Are you more or less confident that will be something that can be realized?

WEN: I'm more confident that earthquakes could be predictable. I'll give you a reason for this. In the past, I was very skeptical that earthquakes could be predicted. The first reason is that there are a lot of things that could be different in the field than in experiments. We know that earthquakes occur when rocks move with respect to each other, and we also understand in general terms that you make things overcome that friction, and it will occur. But I think in the field, it's much more complex than that. It's not a simple block, and what's between [and] the friction coefficients are not necessarily what we understand in the lab. But I'm more [leaning] toward now that they can be predicted. One example we're looking at is induced earthquakes, and that's one thing I'm focusing on right now. Because we know the seismicity, and we know what we did to the earth and internal stress. One example we looked at is a gas field in China in the Xinjiang province. There was an oil field, and they emptied it and turned it into a gas storage field. They imported gas in the summer, and they extracted gas for domestic usage in the winter. With that process, it created perturbation of stress and generated earthquake.

We're looking at those histories, injection-extraction histories, which occur with seismic activity. And we find that the seismicity occurs when material's injected. And that's not surprising because injection will compress things. But the seismicity occurs not proportional to how much you inject, it's proportional to how fast you inject it, so we were lucky we had their operating history. In other words, what we suggest is that when you have a fault, and you push that fault to a certain extent toward failure, then all of a sudden, you change the stress, because the friction coefficient is not static and has a dependence on the stress changes you have, that creates an earthquake. Because the major earthquakes occur when you suddenly shut down the operation during the injection or increase the operation during the injection. Both ways, that can create seismicity. That pointed to a direction that if we understand the stress field, and we understand how the fault behaves with respect to the stress field, we might be able to predict earthquake.

ZIERLER: As you know, the argument against the idea that earthquakes can be predictable is that the earth itself doesn't know when an earthquake is going to happen. How might you respond to that from your perspective?

WEN: I think if we're looking at those induced earthquakes, we don't see much of the extraction period, and that's seismicity for that certain [injection] period. Certainly, the earthquakes tend to occur when a critical stress is reached. Whether we will be able to know that for sure is a different kind of argument. But I do think earthquakes occur following some of the principles we know. The earth has to be pushed to a critical state, and the trigger is much more complex. But I think if we had the field example we could understand, when we've accumulated enough field experiments, [we may know] how the friction changes because of the changing of stress and stress rate. Of course, we won't be able to know exactly where [we may be able to first predict], but I think earthquakes have some rules to follow of their own, not complete randomness, based on the induced earthquakes we see. That's my opinion.

ZIERLER: Just a technical question, how do we go about inducing an earthquake? What does that mean?

WEN: The terminology means that for places you don't have earthquakes ordinarily, you do something to change the stress field, which changes the local environment, and you make an earthquake occur. Otherwise, it won't occur if you don't have any field operations. One typical example is fracking, you inject a lot of water, then you perturb the stresses to make an earthquake on a fault that's otherwise stable. Slipping begins, and you have an earthquake. That's what my definition is. You have a change either by injection or extraction of the gas field, for example, and you create an earthquake where it otherwise would not occur under the ambient tectonic environment.

ZIERLER: Your work on deep earthquakes, what's the dividing line? How far down before it gets the title "deep earthquake"?

WEN: Different people have different definitions. Most people are referring to 400 to about 600 or 700 kilometers. Between 100 and 400 is intermediate, and above that is shallow. The reason is, it's mostly divided by the distribution of earthquakes, the frequency or occurrence, but also relating to some material changes, phase changes, or structure changes of the earth's material. We see in the shallow part, lots of earthquakes. Then, we see very few earthquakes between 100 and 400. Then, we start to pick up in 400 to 600. The earthquakes I focus on are between 400 and 600. The shallow earthquakes, even though we don't have the ability to predict, we know how they occur, the fault slips by overcoming friction. And between 100 and 400, people think they know what's occurring, and there are mechanisms suggesting that you have the subducted slab, you bring water down to greater depths. Between 100 and 400 is where the slab starts to get rid of water. When the slab gets rid of water, it increases a lot of the pressures [in the pore space] between the existing fault, and effectively lows the normal pressure on the fault and earthquakes occur [by overcoming the friction]. Now, from 400 and 600 have people running out of ideas of what occurs, because water is gone about 400. As to [that], why we have earthquakes there?

There are many physical mechanisms proposed for why we have these earthquakes. That's why I was interested in looking at these earthquakes, to see whether they were different from shallow earthquakes and to see what different or similar mechanisms occur to make that happen. Because at that depth, you're not supposed to have earthquakes. The ambient pressure is gigantic. You have the rock from the surface to 400. It's impossible to make that fault slip. You have to have gigantic stresses to make them move. That's how we look at it, we look at those rupture processes in detail from seismology, then we will build that into interpretations. We find that the bigger ones are not ruptured in one point. They look like mulitple smaller events occurring in short time interval at different places. And the rupture directions also seem to be different than shallow earthquakes, where you have more uniform slipping. We proposed that in some of the regions, we have weak zones. If you deform the weak zones, you have viscous dissipation and that will create heat.

If you create heat, they'll weaken more. And if they weaken more, it'll create more heat. At some point, you create too much heat, it'll melt the interface and let it slip. That's my research into that. The first part we looked at–because a lot of studies study the earthquake using seismology based on the assumption that everything slips at one plane, we just threw out that assumption. When we look at those, it's actually not one plane, it's just one collection of multiple smaller sub-events that rupture on various planes. Also, we don't require that to be more like a uniform mechanism. It turns out we have different mechanisms. That's the part I'm interested in. Also, that could point to other quakes on other planets, like a moon quake, and there are also deep moon quakes. They [may] have the same mechanism. That's the part I'm focused on with deep earthquakes.

ZIERLER: I'm curious about your work with hurricanes. First, is it possible that hurricanes can have seismological or geophysical effects?

WEN: Hurricanes can produce a lot of seismic noise. It's because the hurricane interacts with the ocean. It's a pressure system and a wind system, and that perturbs ocean's surface, as well as the pressure and movement of water. That [perturbation] could transport down to the deeper part of the ocean, and that [deeper part of the] ocean could interact with solid earth. That perturbing of the earth, similar to what produces earthquakes, generates "special earthquakes". That perturbation propagates out as general seismic waves that go inside the earth, then propagate out and can be picked up by seismometers. Ocean waves have been recorded [that way in seismic sensors] for a long time.

It's just noise coming from the ocean. We just didn't know what to do with it. We were looking at other things [in the recording], because the noise is the nasty stuff. You don't have something coming out [from the seismograms], it can't be recognized. You know if it's noise, it can be random and chaotic. You know the amplitude's increasing, but you don't know exactly what time it arrived, and you don't know how to locate it because it's noise. You can't pick a time or particular shape. But the hurricane system could have good record, especially for [Hurricane] Sandy. We happen to have a project called the Earthscope project, which happened to [be in the time] moving from the West Coast to the East Coast. And we picked up so much data, so much noise there. But what we saw was, between the noise, there actually is a coherent signal. If you extract the coherent signal, you can actually track where the source is. And the source happened to be the hurricane center. That [source] also is moving along the center of the hurricane. That is a part, you're using those coherent signals to track the location in real time, because seismic waves travel very quickly, just a few minutes.

And then, if you develop a model with an atmospheric system, coupled with ocean, coupled with ground, you can actually determine the strength of the hurricane based on the strength of the seismic source it creates. It's surprising, but it is doing remarkably well, tracking both the path and strength of the hurricane. I think it can be a very helpful simple method, because the current knowledge we have, the monitoring system, mostly is based on ground observations which come from farther away from the center. You can't put it in the ocean. Most of these [monitoring systems instead] are based on satellite images to see how big the hurricane is, then you quantify that and assign an intensity of that particular hurricane system. Or you could have the sample [of the hurricane system]. You have drones [flying over hurricane], and you can drop some measurement [tools] inside the hurricane system. But that's just one point of measurement. Seismology could track where it is on the ground and how strong it is.

ZIERLER: Several questions as they relate to your work simulating viscous flow. First, why is viscous flow important to your research, and how do you go about simulating it? What does that mean?

WEN: Rock is very funny stuff. If you push the rock, the rock will respond. If you squeeze the rock, the rock will respond right away. If you squeeze, they shrink. That kind of behavior is called elastic behavior. If you release that, the rock will recover. And that's what we apply this property, assuming it's elastic, to deal with seismic wave propagation. Seismic waves, you punch the rock, then the rock will shrink, expand, or shear, and the wave will go on. But if you push the rock long enough, it's just like mud. And mud will slide in response to the forces you apply. That behavior is viscous behavior. When we're looking at the same rock but on a much longer time scale, like millions of years, or even a scale of a couple of decades, the rock is actually pretty soft material, flowing. That's what we call viscous flow response. And that's responsible for carrying the material from the mid-ocean ridges to the subduction zones and then back. And that's something we see, the recycling of the surface. But internally, it has a similar situation. The materials are moving, but very slowly, like on the scale of our fingernails growing. That kind of flow, we define as a viscous flow.

ZIERLER: I'm curious if any of your work has led you into the world of policy, working with governments at all.

WEN: There are two things I'm working on, mostly in China. One thing I'm working on is trying to put together the problems for seismic hazard [mitigation in China], and of the near future on the national level, what kind of scientific projects should be pursued and what actually should be invested. That's a part. And I was the lead author for a particular report to the National Natural Science Foundation of China, Chinese Academy of Sciences, and China Earthquake Adminstration. The report was trying to identify what kind of research was needed for the seismic mitigation.

The other project I'm working on now is, we're trying to build a seismological reference model in China, and that's primarily looking at the seismic structures within China. China is a very interesting area, tectonically and geologically. It's much more interesting than the rest of the world because we have big continental collisions in China, and Tibet [region] is one of the most active research areas in the world. And it also has a lot of subduction and a lot of diverse geological backgrounds in China.

But the other aspect to consider is that, right now, different groups of people are working on different parts of research areas in China, and there's a lack of communication of data and a lack of collaboration between different research groups. My interest is, established by the National Natural Science Foundation of China, trying to build this model to bring different groups of people together. You have data, you have tools, you can contribute to those. And that project, we're hoping to build a model associated with the data we have and the secondary product between the data and the model, and make it as a basis of our future collaborations of different groups. That's the goal, and it's not just [dealing with] a science problem, but it's [also] a cultural problem and a communication problem, but we are making good progress.

ZIERLER: While we're on the subject of China, let's go all the way back to when you were an undergraduate. First, in the Chinese system, did you declare an interest in geophysics and seismology right from the beginning? Or that happened later on?

WEN: I declared my major at the beginning when I was applying for the University of Science and Technology of China, although I didn't know what it was. But I did [declare]. [Laugh]

ZIERLER: What was your interest? Where did that come from?

WEN: At that point, the department was called earth and space science, and it just felt cool. Earth and space science. I didn't know what that was, but the name was already attractive to me, so that was what I got into in college. Space science and earth science just sounded cool, both things together.

ZIERLER: Tell me about graduate school. First, the master's program in China. Did you move institutions?

WEN: Yeah, I moved to the Chinese Academy of Sciences. University of Science and Technology of China was in Hefei, Anhui province, in the middle of the country. I went for graduate studies in Beijing. At that point, it was called the Institute of Geophysics. Then, it was combined with the Institute of Geology, so now it's the Institute of Geology and Geophysics in the Chinese Academy of Sciences. I spent three years as a master's student, and two years, I think, I worked as staff there. I'm not sure what my title was. It was just natural when I graduated that I continued to work there.

ZIERLER: Tell me about the opportunity to travel abroad for your studies. First, who suggested to apply to Caltech?

WEN: My advisor in China, the master's advisor. He was a visiting scholar at the Caltech Seismo Lab in 1981 or '82. He was among, I think, the first group of people sent to the US by the Chinese government at that point. He knew the Seismo Lab very well. He had a great impression at that point that has continued to develop really well. It was just natural that he would recommend me to take a look at Caltech. Also, I think his recommendation was really important for me to get into Caltech as well.

ZIERLER: Given that you had already done graduate school at this point, how well-defined were your interests when you got to the Seismo Lab? Did you know what you wanted to work on, and did you know who you wanted to work with?

WEN: I did some seismology work, mostly on the theoretical part, in my master's, but I wasn't quite open-minded about what I wanted to do, so I didn't know what I wanted to do when I got to Caltech. I know I wanted to do some seismology, and I wanted to find a job after graduate at a point. But I didn't know what earth science was, to be honest. It's just seismology, but at that point, it was just all the equations and programs. And this was coding, and this was how to collect seismograms or calculate synthetics. But in terms of the big picture, why are you doing that? why are you picking up these particular seismic data? what kind of things do you want to say about the science other than wave propagation? I didn't know until I got to Caltech and really started to get it pieced together [at Caltech].

ZIERLER: How was your English before you arrived in the United States?

WEN: That's a really good story. My English was really, really poor. I was okay with writing and reading, but my spoken English and listening were terrible. I couldn't understand and I don't think anybody could understand me either. I don't know how my advisors survived in the first year or two. But it was really when I started to take the courses at Caltech that my English really started to improve. And when I started to work with Don Anderson for the first paper, my English started to improve. My first paper with Don Anderson went through, I think, 40 or 60 iterations, and that helped a lot. And at Caltech, the first year of courses is really, really heavy, so it was a challenging time for me, but it was also when I started to learn English. I remember the first course I took was Ge101, Intro to Geology. Man, that was really painful. First of all, it's a lot of rocks, a lot of minerals, a lot of things to memorize. But there's also a lot of terminology in geology. I started to go through the vocabulary and comprehension. Thanks to the courses at Caltech and the interaction with Don, which helped me with my writing skills.

ZIERLER: Did Don ultimately become your thesis advisor?

WEN: I had two thesis advisors, Don Anderson and Don Helmberger. But I started my writing, the publishing, with Don Anderson first. My first few papers were with Don Anderson.

ZIERLER: What was Don Anderson working on at that point?

WEN: He was working on integrating all the results in seismology, geodynamics, and geochemistry. He was trying to build them into general models of earth's internal processes, trying to explain all the data. At that point, he was focused on that. That really helped me open my eyes to various topics, also, his ability to put things together.

ZIERLER: When did you start with Don Helmberger? How did that relationship develop?

WEN: Don Helmberger, I actually also started with [him] when I first got there because his research was mostly [close to] what I did for my master's thesis, which was developing methods for calculating seismic wave propagation. I started with him [with a project], and it was actually one of the two pre-oral projects. But that one is on the shallow part, to simulate the basin responses. I sort of took a break and worked with Don Anderson, then I came back when I was about to graduate with Don Helmberger. Either way, we were happy both working [out]. But I spent more time with Don Anderson in the middle.

ZIERLER: Once you got more comfortable in the Seismo Lab and saw what was going on, in the mid-1990s, what were the big debates? What were people interested in at that point?

WEN: At that point, there was a revolution in seismic tomography and geodynamic systems with viscous flow. Seismic tomography, at that point, they had seismic data, and they were trying to build a three-dimensional picture inside earth. It was like a CT scan of the entire earth using seismic waves. You see the large scale [structure] popping up, you see high velocity and low velocity. The debate, at the time–well, there was a minor debate about who had the better resolution, but the debate was really what that velocity anomaly meant. If you see low velocity, the easy way to interpret it is because it is relatively hotter. Same material, relatively hotter, so the seismic speed's relatively slower. Or you could have something compositionally different. For instance, you have more iron there, and that appears to be relatively slower. I don't think that came to people's minds yet, that interpretation. But a lot of debates assumed those temperatures were there. What did that tell us about the convection system inside? Either that the convection system is uniformly one layer from the top to the bottom, or the convection is actually two separate layers, and one's separated from, like, 670 kilometers, with the upper layer and the lower mantle separatly convect on their own. That was the major debate in chemistry and also in geophysics, and everybody was trying to get arguments based on the tomography we were seeing.

The geodynamic part was that we knew the earth was viscously deforming. That was also a famous geodynamic framework developed by Caltech professor Brad Hager and his student Mark Richards. And we knew it was circulating and also could explain the gravity data that became available in global scale from remote sensing. But there were internal inconsistencies. The induced topography we had was inconsistent with what we saw at the surface and the model was also inconsistent in explaining the plate motions, especially the shear component, the rotation component. That was, I think, the major debate.

Later on, we saw the low velocity that had sharp edges, I started questioning, that is really hotter or something. Because if it's hotter, it has very different consequences. If it's hotter, it's going to be lighter. Lighter is going to rise. But it could also have more iron, just more iron, they're also slower, but they're heavier. And heavier means they're very comfortable there. They just sit there. That's also the part I was trying to get into there because I thought, "Is it going to be forever?"

Either you had low velocity that's–well, heavier or lighter. After graduate, I was focusing on the edges to see how they changed, to see whether not just the velocity is slower or faster, but the transition from that particular chunk of material to outside material. If you have very smooth transition, it means that the temperature could explain it. But if you have very sharp transition, the temperature cannot explain it. Because everything sitting there has to be a long period of time, tens of millions of years. And the temperature is going to diffuse, and can't maintain the sharp transition. Looking at those features, it really could bring us a step forward. Then, I'll start to look at the data and methods trying to constrain that feature. And that feature is also difficult to constrain because you have the wave propagation that you will have the entire earth to digitize, and computation is expensive, but more importantly, the computational ability cannot give you that kind of frequency to match the data, or you cannot use those kinds of details to constrain your high-precision structure.

ZIERLER: Of course, you were there in the mid-1990s, which is exactly halfway between the use of computers in their earliest forms at the Seismo Lab and where we are today. In the early 1960s, 30 years, this is the very first primitive computers that the Seismo Lab used, then it's 30 years in the future with the modern computers we have right now. With that in mind, what aspects of the computers that were in use in the Seismo Lab at that point felt modern while you were there, and what seems very antiquated now, looking back?

WEN: It certainly was less powerful, but at that point, it was actually quite good. For instance, I wanted to focus on the region in the deeper part. I just applied my method there. The programs certainly were standardized, the C programs. They're very similar to what we have today, although we didn't have other high-level languages at that point. But C and Fortran were really standardized. And the programs were able to calculate the regional [wave propagation]. A couple hundred kilometers was no problem in two dimensions, with as high frequency as we had. It wasn't as powerful as what we have now, you build the entire earth, and you can simulate the waves for the entire earth. But at that time, maybe 1 100,000s of the earth, but if you were able to develop a method just focusing on that part with a numerical one and another part with an analytical one, you could do that. That's what we called the hybrid method. A numerical method you just put into the places you're interested. The others, you just figure out analytically. You just bring the wave down there or bring the wave back to the surface.

ZIERLER: In the 1990s, just as the internet was starting to be adopted, I'm curious if there was still data at the Seismo Lab that could not be shared over internet lines, if people still needed to come physically to the Seismo Lab to look at data that might not have been available elsewhere.

WEN: Yeah, there was some. At that point, new data was mostly available from the internet. The older data, the analog data, was still not shared very widely. Also, people had to come to the paperwork. And the Seismo Lab was one of the places that stored this paperwork. A lot of the data used in my thesis was actually from analog data from the basement in the Seismo Lab. You had to take the paper, copy it, and digitize it.

ZIERLER: Was there a field work component to your thesis research?

WEN: No, I didn't do any field work component.

ZIERLER: Where did the data come from? What was most relevant as you were developing your thesis?

WEN: One category was a lot of paperwork, the data stored at Caltech. Another set of data is available from the internet, but I only used the part on the LA Basin. At that point, the TerraScope was operated by Caltech, and that was also available. But there were other data available at that time, I just did not spend much time doing other stuff. But after I graduated, I used more data from the internet, which was [distributed] by the Incorporated Research Institutions for Seismology, which has been really great at archiving and distributing data. I feel it's a really lucky generation [for me] because the previous generation had [created] this great organization that put data together and shared it, which made everything easy and advanced science significantly. This is what I wanted to do in China because it's really a great model. It not only advances science but saves a lot of money.

ZIERLER: What were some of the main research questions for your thesis?

WEN: My main research question, the first part was trying to understand the circulation system relating to the inconsistence between the gravity and the topography at the surface. Because you have a system, and you're circulating, and you generate deformation. At the same time, you have topography changes, and you have gravity variations, those could be observed at the surface. At that point, those two datasets were not consistent with the model. It turned out that if you divide the circulation into two layers, you solve that problem. That's what my thesis was on. The second one was trying to understand why we had this rotation component of the [plate] motion. At that point, it was not able to explain it because the existent models produced no rotational motion and had inconsistency as well. I developed a method to calculate viscous flow with lateral changes of rheologic properties and identified the continent-ocean [contrast] and the weak zone beneath that. When you had that kind of model, you could explain the rotation component.

For the seismology part with Don Helmberger, the early part was working on the LA Basin, the [basin] amplification [of ground motion]. But most parts later on were trying to work on the deeper part of the mantle. The goal was to get a high-resolution mapping, not just the velocity, but also the structural feature, the geometry. And that's what I think could help identify [new things] or resolve the debate about temperature origin or other stuff. I focused on the core-mantle boundary, trying to explain not just how much travel time is delayed but how much the waveform has been distorted when it [seismic wave] goes through this structure. That part, you wrote [a computer code] with a developed method, hybrid method. You just put that numerical simulation right in the spot you're interested in, and you put analytical solutions everywhere. We looked at the different part of the core-mantle boundary, we located one region in the Pacific, and we saw a very low-velocity structure, [similar to] ultra-low-velocity that was discovered earlier by one of Don's other students, Ed Garnero.

But we put it [the effort] into localized geometry. We had a set of the data sampled the base of the mantle, we tried the existent model, and it was not able to explain the data. And that was because lots of this kind of anomalous data could be explained by a one-dimensional model, which is that you have analytical solutions, they are easy to do. But those sets of data are very difficult to explain in one dimension. We used a different model, [different from the model] that goes to infinity, only changing vertically, you're not allowed to change it laterally, horizontally. When the method became available, it began to be very easy to explain the data with a dome-shaped low-velocity anomaly that we attributed as the root of the Iceland volcanoes at the base [of the mantle]. That's a brief summary of most of my thesis topics.

ZIERLER: Did Helmberger and Anderson collaborate at all? Were you a point of connection in their own work?

WEN: They didn't collaborate, and Don Anderson used a lot of results from Don Helmberger. Don Anderson was a big-picture guy. He integrated almost everything. He was always looking at the big picture. Don Helmberger was looking at the wave shapes. He always thought [about] distortion of the waveform and the way the wave goes, and he modeled, then moved on. They didn't have much communication. I was surprised that–I think I was the only student to work with them, but I was surprised there weren't more. I think it was great looking at both sides because you could have big picture and the details.

ZIERLER: What were the major conclusions of your thesis?

WEN: My major conclusions for the convection system, the two-layer, stratified layered system, is consistent with what we see, the topography and gravity at the surface. And the continent and ocean difference, the geological property differences, and the existence of a soft asthenosphere layer beneath that, are responsible for plate-like velocities, [and] why they have the shear motions as what we see at the surface. The core-mantle boundary, from what we see from this geometry, the very low-velocity anomalies were not just [due to] temperature differences, they had a geometry, had sharp transition, and that supported the partial melt idea at that point, and that is also compositionally distinct. But [among] those regions, we just started to pick up one [of great interest], that's Iceland. There's always debate on those we call hot spots, volcano regions in the middle of the ocean. And people always speculated where they come from. Of course, Don Anderson was of the school of thought that it came from shallow parts, but a lot of people argued they were coming from deeper parts. What we have mapping at the base in Iceland suggests that at least there is a possibility connecting what's near the surface we see in Iceland to the base of the mantle in Iceland. Because if it's a very low velocity [and has a geometry], it requires something different other than temperature. We attribute that as the root of Iceland.

ZIERLER: A personal question, when you got to Caltech, were you hoping to make a life and career for yourself in the United States, or originally, you thought you'd go back to China?

WEN: I was quite certain I wanted to stay at that point. I wasn't quite sure I could stay in academia, but I wanted to stay in the US. The reason I wasn't quite sure about academia was, statistically, at that point, there were no Chinese being hired as professors. Back in earlier years, there were. But for my generation, and just a few years before me, there hadn't been successful hires within the Chinese group. I was prepared that I might not find a job in academia, but I wanted to stay.

ZIERLER: Besides Helmberger and Anderson, who else was on your thesis committee?

WEN: David Stevenson. Michael Gurnis. I forgot the fifth one. [Laugh]

ZIERLER: As a graduate student, were you looking at Mars at all? Was anything happening at JPL relevant for your work?

WEN: No, I was not looking at it at all. I didn't know what Mars was like. [Laugh]

ZIERLER: What opportunities did you have after you defended? What was available to you?

WEN: I applied two places for post-doc. One was UC San Diego in Scripps and the other was the Carnegie Institution of Science. Carnegie was a fellowship, and UC San Diego was a part of a grant support, not fellowship. I chose the Carnegie Institution in Washington.

ZIERLER: What was the main mission of the Carnegie Institution at that point?

WEN: The Carnegie Institution, one of the major missions is always looking at astronomy and planets beyond our planet. That's a very strong component. Another strong component is in seismology and geodetic observations, especially the strainmeter, which is putting a very long instrument buried underground that records tiny deformations near the surface. I was attracted there by two things. One, there was a fellowship, and they told me I could do whatever I wanted to do. I'm not sure if that was 100% correct. The second thing was, the Carnegie was doing a lot of field seismic experiments. And there happened to be one in Africa, and that happened to be the point that the data began to be available. I was attracted to that data. When I was a graduate student, I knew there were two big low-velocity anomalies in the base of the mantle, and people always attributed it to the temperature, being hotter, a rising super plume. But I wanted to look at the transition features. That data mostly was collected looking at the shallow part in Africa. I thought it would be very complementary to look at the deeper part of the earth. And I went to Carnegie.

ZIERLER: Tell me about the opportunity at Stony Brook. How did that come available for you?

WEN: When I had just graduated, I also applied to faculty positions. I didn't get any interviews the first year. But my first year as a post-doc, I started getting interviews. I think that year, I had about nine interviews for faculty positions. I got two, and I decided to choose Stony Brook.

ZIERLER: I'm curious if you knew Bob Liebermann before you got to Stony Brook, if he was connected to the Seismo Lab at all.

WEN: I did not know him before I interviewed. Actually, some time before the interview. When I applied, I looked at the research webpage of the department, I didn't pay much attention to who graduated from where because I also wasn't quite sure what my chances were at that time. But when I got an interview call from Stony Brook, I started to look in detail at the faculty, but I also talked to Don Anderson, sought his advice, and Bob Liebermann's name popped up. I think Don Anderson knew Bob quite well. At that point, I started to know of Bob and heard a lot of stories about him, stories about college football and other things. But that was when I learned about him.

ZIERLER: When did you start to develop collaborations in China after you got to Stony Brook? How did that process begin?

WEN: I actually started quite earlier. When I got here, I did some work for the waveform modeling of the African anomaly, I started mapping out the changes there. And people started thinking, "This is a compositionally different material." There were ideas about how the compositionally dense material fed into the circulation system. People in China, mostly at the same school I graduated from, noticed that. They started to have the budget to do some of the dense observations at the time. That was about 2000, 2001, that period. It wasn't much of a budget, but people had the budget. There were data there, but people were frustrated because they didn't have the expertise to interpret this data or come out with any possible results associated with this data. They contacted me, my previous teachers and people I knew, and they wanted to send a young person to Stony Brook. At that point, they didn't have much of a budget, so just one person. They were trying to say, "We used exactly your method with your data." I told them, "For one person, the data application won't necessarily be copied because it depends on what data you're recording, what anomalous waveforms you have. Why don't I just come back so I can do a group of people? I can spend some time in the summer, and we'll start there." At that point, I think two or three years into my assistant professorship, I started to go back to China for very short periods of time because I couldn't afford to go for too long. It turned out there were three post-docs at a time, and one was interested in the deeper part of interpreting data, but the other two didn't have much of an idea, coming from a different background. They were coming from an exploration background. At that point, one started to look at data in a deeper part, another was looking at the inversion process, and the other was using the method from exploration seismology and applied to dense natural earthquake data. That was the starting point. Later, about ten years ago, I started to work with USTC, the University of Science and Technology of China, which is where I did my undergrad, and I started to realize that China has much more data there. At that point, I started to think about educating the young generation, but also [was] trying to think about how to put various kinds of groups together in China.

ZIERLER: Tell me about your work with the AGU. How has it been helpful in your research?

WEN: The AGU has been really helpful. The publication, getting a lot of the policy they have just right–because it's a community publication, so it always maintains high quality. Serving as an associate editor, for instance, is a very rewarding experience. You just feel like you're helping people. AGU publications aren't like other publications that are money-oriented, they're mostly community-oriented. Also, I view the AGU awards very highly. I think that nomination process is just amazing. There are a lot of things to learn about how great the community is.

ZIERLER: Just to bring our conversation up to the present, what are you working on right now? What's most interesting to you in the field?

WEN: The first thing is, I'm trying to finish the China Seismological Reference Model, for not just putting the science together, but also trying to unite and integrate the community in China with the US and around the world. The second thing I'm focusing on is the shallow part of the physical mechanisms, which I don't think we know. That's based on putting all the possible forces acting on the earth, which is the time we are on the position of getting those. Also, the strainmeter recordings of tiny deformations. Most of these are in the US, but not many people are looking at the data in the US. The reason is that to understand those physical mechanisms is really important for understanding the earth's responses and triggering of other processes. The shallow part is very different than what we think. One example is, you have water, and the traditional view is that water pushing the surface down. But water could also go into the pore space, and that will push the surface up, and water could flow. And all this comes from the strainmeter, and that's something we've never seen before. That's a part I think I'll spend significant time on. Later, maybe I'll spend some time looking at the Mars data, trying to get a sense of the Mars quake and structure. The dataset is released and about completed, I think.

ZIERLER: For the last part of our talk, I'd like to ask a few retrospective questions about your career, and then we'll end looking to the future. First, I wonder if you can reflect on some of the important opportunities that the Seismo Lab provided to you, both on a personal and a professional level.

WEN: I think the first one, on a personal level, the Seismo Lab was always supportive. I didn't have to worry about any other financial problems when I was at the Seismo Lab, I was always well supported. My personal and/or professional life is really attributable to the environment here at the Seismo Lab. The Seismo Lab has a distinct faculty, and you could work with different faculty on different aspects of the field. But the faculty is also very good educator. They allow students to do whatever they want, at least in my case. That, for me, was extremely important. It wasn't just supportive, but also built my confidence in letting me do whatever I wanted. And I knew at the time it was a really good environment. But when I became a faculty member and worked with different scientists in the field, I started to appreciate much more how lucky I was. It's not just that you have distinguished scientists at the Seismo Lab. The collaboration or the environment created depends not only on scholarship but personality and culture. You could have two distinguished scientists who have never worked together, and sometimes try to destroy each other. But the Caltech Seismo Lab really provides that [collaborative] environment. They're so supportive of the grad students, and it's such a collaborative environment that makes the student feel comfortable working with different professors who always trust the ideas the students bring. At least in my case, I brought different ideas like a convection stratified at a thousand kilometers [deep]. And Don Anderson was happy to say, "That's your idea, and I'm trying to help you understand the logic. Go ahead, you can do it." That's the environment that I really think made a lot of grad students successful. And that's what I appreciate about.

ZIERLER: What did you learn at the Seismo Lab in terms of how to do the science, how to analyze the data, how to work in collaborative teams, that's stayed with you ever since?

WEN: This really comes from my advisors. I was so lucky to have these two advisors. They had very different styles. Don Anderson was a big picture guy, and that helped me understand the big picture, but also helped me understand the kinds of important problems you should pursue and where the important problems were. But he wasn't a detail-oriented guy, and he used to also take everything and integrate it together. I'm more skeptical whether results represent what they look like. That came from Don Helmberger. He was so detail-oriented and had the insight into waveform wiggles that gives you some of the detailed structures here. If I wanted to say something skeptical of Don Anderson's particular structure, I could always try to find the data, I wanted to say whether I could resolve those ambiguities before I integrated things together. But for the Don Helmberger side, he had broader modeling interest than I wanted to pursue. I did not have his passion for every waveform wiggle. When I had doubts about whether I should pursue something, seeing something in the data and whether I should pursue that, I was trying to think about what kinds of things we could learn and whether they were important or not. And that's from Don Helmberger's side, this is anomalous waveform wiggle, put it into the context of "that structure, it's important for us to understand?". Or from Don Anderson's side, this is important, well, "try to look at what kind of supporting evidence we have or what kind of other data, not just the seismic data, that can help resolve this issue". I think that combination, really…

ZIERLER: It's a good combination.

WEN: It's a good combination. [Laugh] I don't know, but it's a good combination.

ZIERLER: Finally, last question, looking to the future, what's most important to you? What are you interested in that you haven't yet focused on?

WEN: That's a really good question. I think what I would really like to focus on, if possible, is looking at the strange sources from the seismic data. I think there are a lot of things we don't know. First of all, we have recorded ocean waves and volcano eruptions undersea, and we do see these observations. We just don't know how to interpret them. I think a lot of this has implications about what's going on [in the oceans]. That's part. Let's spend more time on that later on.

ZIERLER: Lianxing, it's been wonderful spending this time with you. I'm so glad we were able to do this and capture your recollections. Thank you so much.

WEN: Thank you. Thank you for the opportunity.