Erik Ivins

An Oral History with Erik Ivins

Senior Research Scientist, Jet Propulsion Laboratory (Ret.)

By David Zierler, Director of the Caltech Heritage Project

June 15, 2023

DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It is Thursday, June 15th, 2023. I am delighted to be here with Dr. Erik Ivins. Erik, it is so nice to be with you. Thank you for joining me today.

ERIK IVINS: Thank you, David.

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

IVINS: I formally retired from JPL on February 5th of 2022, so it has been a little bit more than a year that I have been retired. Although I'm not a fully retired employee, I retired as a Senior Research Scientist, which allows me to stay on as a Research Associate, which means that I don't get salary but I get all the full privileges of a person with a badge and all the access to the library, access to VPN and all the electronic services that JPL can provide, and I continue to do research. I continue to, most importantly, advise younger scientists at JPL. I'm still traveling for JPL. I'll be going to Berlin to the IUGG meeting in July. I'm still an active scientist reviewing papers and so on. But I don't have anyone who gives me deadlines, so to speak.

ZIERLER: What is the kind of research that has compelled you to remain active? What is interesting to you in the field?

IVINS: I would say that one of the interesting things is the reconstruction of the late Quaternary, which means basically looking at how both life forms and topography and the water system has evolved over the last million years or so. I think that's one of the things that—I just happened to be reviewing a paper where that was the discussion, so that brings me to that particular topic. I would also say that I've always been interested in mantle geophysics, in all of its aspects. Since I'm at an institution that does planetary science, that would include why planetary orbits become circularized through dissipative interaction in the mantle itself or within water systems. That brings me to probably the topic that I'm most active in recently, and that is the constitutive nature of solid Earth and how creep phenomena occur in ceramic materials, and how the experiments that are in that field can be translated into geodynamic models of data that we observe in geodesy.

ZIERLER: Given that you are serving now in a mentor capacity at the senior stage of your career, when you look at some of the things the upcoming generation are interested in, what kind of window does that provide into where the research is headed?

IVINS: I would say that in the modeling realm, the fact that we're looking at computational capabilities that we had not conceived of 25 years ago. I'd say 25 years ago in my career, if we were talking about doing a hundred simulations, that seemed like sufficient to come to conclusions about what model parameters were, and so on. I think today we see ourselves moving to computing millions of model simulations and then looking at the nature in which the parameters vary and how the datasets that we use are capable of being able to say anything that advances the field of study. It's sort of interesting—this brings up to this direction of artificial intelligence, and you can almost see this field kind of breaking open and having its growing pains. Because you can see some young scientists have gone a bit too far with this, and not really being able to grapple with the processes that are involved in each of the links that a simulation code will give you, versus the data constraints that you might have. If you have any of these things slightly mismatched, you will wind up with a solution that has used a hundred million solutions but may be the wrong solution. So, I think we have to understand it. We really have to scrutinize what we get out of that kind of science.

ZIERLER: With all of this computational power, with what AI can do, beyond what it can't do, what are some of the topline items that we now know that weren't possible 20 or 30 years ago?

IVINS: If you look at remote sensing from space, we're now capable of being able to look at images of ice sheets, or even smaller ice systems, and we can see how they flow in time. We can do this by looking at speckle tracking. That kind of dataset simply was not available at the dawn of the era of remote ice sheet sensing, which really goes back into the 1980s and was led by people who are now long retired.

ZIERLER: To get a sense of your research expertise from the broadest to the specific levels, what is the umbrella discipline that would cover everything that you do? And, when you get right down to the panels that you would sit on or the journals you would publish in, what would you call yourself in that context?

IVINS: That's kind of a difficult thing to say, because I'm pretty interdisciplinarian. I would say the specific topic that I would always be brought to the table on, in terms of forming small working groups and so on, is the interaction of the cryosphere and the solid Earth. This can come about in many ways. For example, one of the ways it can come about is the formation of landforms during ice advancement and subsequent retreat in the most recent of the Late Pleistocene glaciations, and even from the Little Ice Age. For example, as the Last Quaternary ice age came to a close, you get outwashes from the melt, and that can change the topography in a dramatic way. Foer example, you get ridges that formed where ice was grounded on the sea floor, great accumulations of sand where catastrophic flooding has occurred on land, and so on. Many of these features can also act as recorders of relative time and rates of change. That's not so much what I'm interested in; it's more the fact that the stress at the surface of the Earth changes dramatically at a level that can create large earthquakes. It's that stress level that induces flow in the mantle and even changes in the shape of the Earth. Since the glacial age is a global phenomenon that involves 130 meters of sea level change, it affects all aspects of the solid Earth, including the core. Because the earth's mantle is viscous, these past dramatic changes in water on the surface of the Earth are recorded as slow motions of the mantle and crust at present-day. As a consequence, it's a global phenomenon that can be recorded today in changes of the Earth's gravity field, and also of changes in the long-term rate at which the pole moves through the crust of the Earth.

The Last Glacial age ended about 11-9 thousand years ago, or at least this is the period when the last major transfers of ice mass from the continents to the oceans occurred. In the past millennia we have not seen mass transfers to or from the oceans in any form or scale such as this. However, there is a the link to the study of smaller glacial systems like Patagonia, where the same phenomenon occurs on a much smaller scale. Because there are many such ice systems experiencing change today there is a connection to a very large scale. Here there is a connection to global gravity variations. At some level, one has to have some expertise in satellite gravity to conduct the detailed research into mass change and stress-induced motions of the solid Earth. That's kind of the niche that I'm in. I am fortunate to have had the educational background to have developed an expertise in geodesy such that I can make this link between ongoing sea level rise for example, and the gravity field of the Earth. If you combine that with the fact that I'm an expert in how the mantle also participates in this process and produces a gravity field, then that's my unique niche in the science world.

ZIERLER: When scientists started to ring the alarm bells about global warming in the late 1980s and early 1990s, did that affect your research agenda, and how did that affect cryospheric research in general?

IVINS: Yeah. In December of 1984, I read a paper by Mark Meier who was a PhD from Caltech, 1957 I believe. He studied the mass balance of glaciers using both field measurements and airborne data. It was Mark's paper that made an inference from the negative mass balance he could measure from smaller glaciers to generalize about global glacier mass, and make the connection to global rise in sea-level which was recorded by tide gauges since the late 19th Century. He further made the connection between glaciers, sea-level rise and increased carbon emissions and other Greenhouse gases. I modeled the numbers that Meier reported for the decrease of glacial mass and the rise of sea level as a change in the gravity coefficients of the Earth's field. The lowest degree in the Earth's gravity field expansion in spherical harmonics is degree two, essentially representing the gravitational ellipticity of the Earth. Because we had two satellites in orbit, one a NASA satellite called "LAGEOS" and the other "Starlette" launched by CNES, the French equivalent of NASA, we could study the rate of change Earth ellipticity since the orbits were tracked starting in the mid 1970's.

My computations revealed that the global effects of glacier loss would have an influence on the orbital changes of the satellites. At that time, we thought both Antarctica and Greenland were stable, so the glaciers became sources, the known sources, from which we referenced global change in gravity to climate change. There weren't too many glaciologists back in the 1950s and 1960s, but Mark Meier was one of the people who organized field campaigns to do observations of glacial systems. He also founded the glaciology office at the US Geological Survey. It was his paper published in Science, that was was the trigger point for my involvement in what we might call climate science. So, it was fairly early on. I have kind of honed that area of expertise.

ZIERLER: Understanding that climate change is extraordinarily complex, and that there are so many scientific disciplines that are required to understand these trendlines, where does someone with your expertise add to what we really understand, to minimize the uncertainties, to minimize the error bars?

IVINS: If you look at what is required to move the mantle around, it requires very long time scales, because the viscosity of the mantle is so high. We're talking about climate changes that have a span of nearly five million years. We actually need a five-million-year simulation to be able to predict the correct direction and magnitude of the secular position of the pole as it changes in time, because the viscosity of the Earth is controlling the redistribution of the moments of inertia of the planet. This viscous time scale forces us to try to understand what the changes were in the hydrology of the Earth going back five million years. We then begin to understand how rapidly sea level can change, for example. When we're confronted with someone who is a bit skeptical about the importance of present-day observations of change it is useful to have knowledge of the past changes and their causes, so that we can put present-day changes into context. For example, I sometimes get e-mails from climate skeptics who ask "Well, how do we know that sea level doesn't always change as rapidly as we're currently seeing it?" The question is a relevant one. It is important to know, for example, that we now know from geological observations globally that over the last 5,000 years, we've not seen any sea level changes that are even a hundredth of the rate that we're seeing sea-levels rise in the 21st Century, which is now about 4 mm/yr. That's a pretty profound, very quantitative fact that we're knowledgeable of. It may not necessarily be something where we have a bell-shaped curve that we can move around and say, "Here's what the statistics are," but we know in an order of magnitude sense what we are facing right now with ongoing climate change.

ZIERLER: The road not traveled, do you think that if you had taken an academic path, if you had become a professor, would you more or less have studied the same things that you've pursued at JPL? How central is the mission of JPL in driving the things that you've worked on?

IVINS: Oh, I think it's absolutely essential. Because I think in academia, one of the things—and I could see myself going down the academic path and I think I'd be perfectly happy there—that when you're in an academic institution, you really are trying to adjust the research that you have by the opportunities that you have to receive research funds. Those opportunities may or may not include space sciences. The resources that you are provided with by an academic institution include, hopefully, energetic, high quality graduate students. As a consequence, you may be going from working on a particular area—say, working in the mantle or working in geomorphology—that's fundable at the time. But the academic environment as a whole involves education, and if the students generally have learning resources provided by that institution, it may tailor the focus of your research. Whereas working within NASA and JPL, you have the extraterrestrial environment that is your experimental playground -- you have what the engineers call "metal in the sky", and that metal in the sky, and the associated capabilities of those instruments on those spacecraft, will determine both the direction and, ultimately, the full reach of your research. The time span over which you dedicate a segment your research career is also determined by the space missions.

I think scientists at JPL tend to have research continuity because of the time span over which there is the conception of, the concept of a space project, and its ultimate conclusion. The conclusion being the answering of science questions that motivated the mission in the first place. This is about a 20-year cycle. In academia, you don't tend to have those kinds of very long time-span cycles. I know this because I have worked very recently with a group of scientists, all very esteemed scientists, who generally get their funds from the National Science Foundation. We worked on a field project for Patagonia in which we integrated seismology, a field called limnology for studying proglacial lakes, and also other environmental factors that surrounded the Patagonian Ice Fields. This was an eight-year project. This was huge. It involved multiple millions of dollars. It is a significant anomaly for the National Science Foundation to fund something like this. Whereas, when we have a satellite program, we're talking about $750 million for the project, and then, oh, by the way, there's peanuts involved for the scientists who analyze the data. But those peanuts tend to be relatively large amounts of funding, relative to what you'll get out of an individual National Science Foundation grant. There are advantages in being in the National Science Foundation world, and that is that, with a good graduate student or postdoc, you can produce a lot of results in a relatively short period of time. But you tend to lack this continuity that you get by having a space project that is producing data over a time span of two decades, which has essentially been the case with our GRACE gravity program that has been so successful at JPL.

ZIERLER: Has your research always been tied to a major mission at JPL? Have you ever pursued something on a smaller scale just based on your own interests?

IVINS: Actually, when I became involved in the geodynamics program at JPL, this was a program in which a lot of technology was going to be developed that would, in theory, help us understand plate tectonics and fault mechanics, and ultimately help us understand more about the how earthquakes cycle, and even how earthquakes are generated. In the early stages of this, the technology was just barely presenting itself in quite small bits and pieces because of the sparse sampling of geodetic measurements in the early 1980's. Scientists were blessed with a period of time where we could pick out those areas where we thought geodesy was going to become successful in the scope of the space-based measurement strategies. If the science goals of the technology were to be successful, we were pressed to use both our knowledge of models and of geological facts to test and validate the initial results of the measurements. This would then give us the confidence to pursue research focused on either smaller-scale problems (like motions of individual faults or geological structures), or to contemplate "big picture" problems that could be solved by the emerging measurement capabilities. In other words, to address scientific questions that weren't necessarily tied to the current dataset being provided by the NASA Geodynamics program. That allowed me to be able to take some time to be able to understand California tectonics and the tectonics of the Western United States. I found that to be very rewarding, in the sense that I could really be looking at datasets that were generated by the U.S. Geological Survey, for example.

So, yes, we do have periods of time in which we can use our own scientific intuition to know where to look for smaller-scale phenomenon that eventually will become something that NASA will use. Patagonia for me is a good example. I have been working with colleagues on global problems of glacial isostatic adjustment and ongoing ice mass balance, and how these affected solid Earth phenomenon and the gravity field. I realized that turning back to smaller glacial systems, that if we had GPS receivers that were placed very close to these glacial systems that we might learn some interesting things about the rheology of the Earth over hundred kilometer scales should the entire ice load be shrinking. We might be able to detect a vertical signal of the ongoing isostatic rebound in response to the change in load. If the load loss had a history that could be traced back into the last century, or in other words to the end of the Little Ice Age, then we could use uplift measurements to tap into properties of the viscous behavior of the mantle. The point I am perhaps trying to make here is that if I didn't have the opportunity within the science calls for the Geodynamics program, I wouldn't have been able to spend the time to write the computer codes that then allow you to make that a justifiable prediction of this phenomenon, and then support a field program to go out and measure it. It is space technology that allows you to be able to do this. In fact, we have a whole Telecommunications Division at JPL whose calling is to develop this kind of technology for precise point positioning. We have engineers at JPL that are fully aware that we want to make these kinds of measurements.

ZIERLER: Over the course of your career at JPL, I'm curious if Caltech has been an asset for you. Have you had opportunity to collaborate with Caltech professors or interact with Caltech students?

IVINS: I would say not so much with students, but we have had good collaborations periodically with Caltech professors. Now, I specifically say students because that may change in the future, and in fact we even have an ongoing presidential fund in which we're actively working with the students. But historically, looking over the past 50 years, I would say generally the Caltech professor tends to be relatively protective of the graduate student. Only in relatively rare cases will the graduate student find new work at JPL that will translate directly into the PhD thesis with the JPL collaborator playing a dominant role in the interpretation of the scientific data set generated by a NASA mission. A good example of that would be a guy who currently is a senior scientist at JPL—I won't name who it is, but you might have run into him at some time—became involved in interpreting the lunar gravity field. At the time, JPL was having great difficulty finding people that were interested in doing research that would use the detailed mapping of the lunar gravity field performed by JPL, ask the important science questions and then provide a path to answer them. This graduate student just sunk their teeth right into it, and it became the standard model for the gravity field of the Moon. I don't know who the Caltech professor was that was the advisor, but I think the advisor at Caltech just decided, "Well, the guy is so into it, just let him go!" But that's a relatively rare case. Often the presidential funds for collaborations with Caltech professors will have a broad range of active collaboration, say anywhere from 5 to 95 %. Usually there have been good things that have come out of that, no matter how the actual ‘collaboration' is measured quantitatively.

ZIERLER: I'm curious, having spent so long at JPL, if you have ever gotten involved or have lent your expertise to non-terrestrial science, looking at planetary science beyond Earth?

IVINS: Oh, yes. When I was hired at JPL, the funds were coming to analyze lunar heat flow that was measured during one of the Apollo Missions. When hired we didn't pay much attention to these things, because we just figure the money is around. I was young and naive. But in historical perspective, it's kind of interesting that I was hired to interpret lunar data from the Apollo program, and specifically related to the heat flow experiments. I became involved in building a computational capability to be able to do mantle convection for both Mars and the Moon. So, I did gain a great deal of expertise about that, and particularly for Mars. I had always been kind of tempted to put my hat back into the ring of being an investigator for planetary science. In 1981 I won a PI-ship for a small amount of funds for planetary convection simulation. But a problem for me entering planetary science at that time was that the Geophysics Program was being severely cutback, and geology promoted. There was a need to find a long-term project that was in its initiation phases and there were not very many around, if any. It was known at JPL as the era of the "Purple Pigeon", a term coined in Bruce Murray's office.

The desire at JPL was to find projects that would capture the imagination of the public, and of congress, that were comparable to the Apollo Program. A tall order! One aspect of the scramble to find the Purple Pigeon was that one might spend an awful lot of energy chasing a ghost of a mission concept, and this wasn't for me. There are some scientists at JPL who have worked in planetary science and they have been pretty much able to completely fund themselves through the calls for science investigations. The numbers of those scientists has been very small over the years, and I'd dare say is below 10%. One example of one who is very successful, who was my first officemates at JPL: Bruce Banerdt.

ZIERLER: Oh, wow.

IVINS: He is the PI of the InSight instrument. It's interesting—we also shared our PhD advisor, me much later than he. I am interested in the topics that he worked on for his PhD thesis, still, today, and I could study those things within the context of terrestrial science and what's broadly called the Geodynamics program of NASA. Bruce, in parallel, became involved in the development of the InSight instrument, the seismic instrument that was placed upon Mars. It may not be appreciated by many people, but Bruce understood exactly how that instrument should be built, and in fact is responsible for the high quality of that instrument capability. So he had to make a decision. He was a very good, well respected scientist within the Mars community in particular, and had done very nice modeling, but he had to make a decision about whether he was going to dedicate himself to making sure that this planetary exploration endeavor, really great science, was going to be successful. That absorbed most of his career. So, you have to make a decision about whether that is the kind of scientist you want to be, or you want to be a scientist that is constantly writing proposals that will keep you fully funded to do scientific research. I decided the latter.

ZIERLER: Erik, you have a tenure where you have seen so many directors of JPL come and go. Are there any directors that have been specifically impactful for your career, that have emphasized the kind of research that you do?

IVINS: I'm a good friend of Mike Watkins, so I would certainly say that I'm somewhat in parallel with Mike's research, he was studying time-variable gravity from the time of his PhD work at the University of Texas, Austin. I have a great deal of respect for Mike's broad skills, and his especially important role in making the GRACE and GRACE-FO missions quite a success. Actually, I have a great deal of respect for all of the directors at JPL that I have experienced. There isn't a director of JPL in my entire career at JPL that I didn't like, in one way or another. I remember when Lew Allen was brought to JPL, we were afraid that we were going to become part of Ronald Reagan's Star Wars effort or something like that. The Caltech faculty actually objected to the idea that JPL change its contract under which 90% of JPL's work is for NASA. I think that was a turning point for JPL, and we credit the Caltech faculty for sticking up for the contract that is with Caltech and JPL and NASA. Lew Allen, he was very much in line with that, actually. He was from the Air Force, and it turned out that he was actually a very reasonable scientist. He had his own interests in balloon experiments at very high altitude for atmospheric science, actually. I would say at every turn, I found that the qualities of the scientists that have become lab directors are very admirable. I'd also say, characterizing science at JPL, we also have had good leadership in terms of those that have held the position of Chief Scientist. Those are very important people, as they ensure that the highest standards are perused.

ZIERLER: You mentioned NASA. Have you had opportunity for direct interface? Has there been cause for you to go to Washington and meet your peers at NASA?

IVINS: Yes, certainly. I have sat on review panels, and I have also been invited to some panels in which we met, for example, Soviet scientists when the Cold War essentially became a thing of the past in 1990. There often are missions that are collaborations with Goddard Space Flight Center (or GSFC) as it is known. We have many science meetings in Maryland where it is located. It is an official NASA center, and a hour or so drive from NASA Headquarters in D.C. We have come to know our colleagues at GSFC very, very well. Sometimes we know some of our colleagues at Goddard better than we know some of our companions at JPL within the Science Division.

ZIERLER: This has been a great overview of your career. Let's go back and now establish some personal history. Let's start with your parents. Tell me a little bit about them.

IVINS: My parents—I guess because I'm a Southern Californian, maybe they are normal, maybe they are not so normal; my father's family history involves Mormon settlement of the western US. My great grandfather was asked by the Mormon Church in the late 1890s to go to aid in the establishment of a series of Mormon colonies in Mexico, in the states of Sonora and Chihuahua. These colonies were set up to avoid U.S. law against polygamy. My great grandfather was not a polygamist, and my grandfather did not grow up in a polygamist family, but he grew up in a polygamous environment. It was also the time of the Mexican Revolution, so my great grandfather—not only met Pancho Villa; he was a friend of Pancho Villa. When my grandfather's family moved back to the United States, to Southern Utah, this history kind of characterized the way that my father, my grandfather, and of course my great grandfather viewed the world. They were multiculturalists. My grandfather was sent as a young man to be a Mormon missionary in Tokyo. There he studied the Japanese language for translation of the Book of Mormon into Japanese. He became familiar with Buddhism and Shinto. I think he discovered that he was Shinto at heart. My grandfather was very influential in my life. He and my grandmother joined the Unitarian Church sometime in the 1940's. He was an avid horseman and fisherman, and especially a lover of the land. Very Shinto!

Among all of the grandchildren of my great grandfather, my father most thoroughly aligned himself with this kind of multicultural world view. And it was a world view; it was not a regional, Southern Utah view. My father grew up on a dry farm in Southern Utah, where you were really one with nature, to the point of which he was much more comfortable riding a horse than he was cranking up a Model T. He kept that attitude all of his life. Although he became kind of a man of the world—he fought in the Second World War in North Africa, and Sicily; and he studied art in Geneva, Switzerland; and he married a Norwegian—he always kept this very rural U.S. kind of cowboy approach to the world. Only he despised the Hollywood stereotypical cowboy.

That's kind of the background that my father had. He was very interested in science, much in the vein of the Enlightenment. My grandfather actually taught animal husbandry, which is part of animal science, at Brigham Young University. This is, perhaps the segue into my story of growing up in northern Claremont and having a father dedicated to a belief in nature, and pursuing an understanding nature. That's a part of my history that actually goes way back to these experiences in Southern Utah and even going all the way back to the colonies in Mexico.

My mother also can be credited for science and my science education, but in different ways. She grew up during difficult times in Norway, but her family were relatively well off. Her father was a sea captain in Stavanger, Norway. To become a sea captain you basically have to be the kid in the class that scores the highest on your math tests, and then, and only then, do they send you off to do navigation and become a sea captain. That was life in coastal Norway at the turn of the 20th Century. He became the captain of a ship and he was gone at sea for most of the time, and the household was basically run by women—the mother of my mother, and my aunt. They had an appreciation for science, mathematics in particular. They always believed that you should study mathematics as well as you can.

The Germans took over Stavanger in September of 1939, and the relations, you might say, were rather hostile between the Germans and the Norwegians, although the Germans will never tell you that. [laughs] My grandfather's ship was taken over, and he basically lost his ship. It became part of the Third Reich's military operations, as did the school grounds where my mother would have entered the equivalent of the 7th grade in Stavanger. The family decided that they would leave Stavanger and go back to the traditional fjord lands where their extended families still lived, where they didn't really run into the Nazis. They kind of escaped from them. So my mother, as a teenager, grew up in a very rural environment in which they were using boats in the fjords. So, she also had been someone who was very close to nature. Not only very close to nature but believed in a strong science education. So, when I grew up in Southern California, she would always introduce me to some science program that was for elementary school kids, and I would always wind up in that. While other kids were going off and playing baseball or playing tennis or something like that, I was going off to a little science camp. That was very influential. My parents actually are very unique in the sense that they were very close to nature, in very, very intimate, strong ways. As I look back, this was generally a contrast to my classmates in primary school.

ZIERLER: You mentioned Claremont. Is that where you grew up?

IVINS: Yes.

ZIERLER: Was that more suburban, more rural, when you were a kid?

IVINS: Very rural. We lived north of Baseline Road, which during that time was all orange groves or chaparral. We were on a piece of property that was surrounded on several sides by orange groves, but to the south of us was all chaparral and set within a mile or so of the great San Gabriel mountain fault structures that promote the uplift of Mt. Baldy. So we were always out with the snakes and the lizards and the rabbits and the chaparral environment, and when the season permitted, in flowing waters of the arroyos.

ZIERLER: When did that area really start to get built up? Was it during your childhood?

IVINS: Oh, yeah, during my childhood. About 1960 is when the rumors of land developers who wanted to build houses started to come to those who owned the orange groves. Some of my friends whose fathers owned orange groves—they were actually lemon groves—started to get proposals from companies who wanted to buy their land for development. Between 1960 and 1967, which is when I went off to college, suddenly we had these housing developments popping up in lands that had previously been orchards. And so, suburbia had come.

ZIERLER: When it was time to think about college, were you already focused on science? Was that the career track you wanted to pursue?

IVINS: Not necessarily. I was more passionate about history and even economics, psychology and biology. So, I had a potpourri of things that I was interested in pursuing. But one of the things that I—maybe it was partly my personality; I'm rather a shy person—when it comes to delving in psychology, it tends to invite people that are good at verbalizing themselves. What I found in both biology and physics was that you relied upon the ability to understand quantitative—relationships, I'd say—and I found that to be internally quite satisfying. In high school, I had always been good at physics, so physics was a natural avenue to follow.

ZIERLER: Was the draft something you needed to deal with? Was the prospect of going to Vietnam a reality for you?

IVINS: Yes, it was, so being in school was a good thing. They developed this numbering system, I think it was when Nixon was president. They drew lots for a numbering system based upon your birthday, and I got a number that was like 307, so I was basically not going to be drafted when I finished college, which was a great benefit. I had roommates who all got low numbers, and they had to do all kinds of various things to escape the draft. I had one roommate who was probably the best physics student we had at Cal Poly; he had to escape, he had to leave the country. He went to Canada. He was not a U.S. citizen, either. He was a citizen of Argentina.

ZIERLER: Tell me about your experiences at Cal Polytechnic, Pomona.

IVINS: Good education, I would say. Since I went off to graduate school at UCLA, UCLA was hyper-focused on research and not so much teaching, whereas at Cal Poly, you got very good teachers, especially in physics. I think the time was very unique for that, because we had one professor who went on to write a great undergraduate physics textbook named Douglas Giancoli. He went later to UC Berkeley and started a physics education program there. He was a high energy physicist who basically burned out on high energy physics research. But he knew all of modern physics. When we took his courses, he was like Aristotle to us. He was just fantastic. I probably wouldn't have a career where I could have applied to UCLA if it wasn't for him.

ZIERLER: Being a college student in the late 1960s and early 1970s, was the counterculture relevant on campus at all? Did you see any of that?

IVINS: Oh, of course. Even at an engineering and agricultural school like Cal Poly, it was definitely there. I remember in my freshman year, we had the assassination of Martin Luther King, followed by the assassination of Robert Kennedy. These were tumultuous times. The Black student movement, and the Chicano movement were all getting going really strong. So, yeah, these were tumultuous times. When you think about it, I had all of these interests—history, and so on; to kind of settle on physics was one way of seeing the world in a less chaotic state. In fact, much later in my career, when chaos theory became a little bit en vogue, there was somebody who was a secretary who once asked me, "Oh, Erik, aren't you going to take the courses in chaos theory?" I said, "Look, my life and my research is all about stability theory."

ZIERLER: [laughs] Erik, as an undergraduate, did you have exposure to geophysics or Earth science at all?

IVINS: As an undergraduate, no, not really, not in any meaningful way. I took a geology course which was called Historical Geology, where all we were supposed to do is figure out what a sedimentary history was, based upon faulting and reorganization. It just wasn't meaningful at all. The most meaningful thing that I took as an undergraduate that was applied to geophysics was in the Math Department. I studied Legendre polynomials and Bessel functions, and then differential equations, and then high-pressure statistical mechanics and thermodynamics. Those became essential to studying geophysics.

ZIERLER: Without so much of a background, how did that lead you to the Geophysics program at UCLA?

IVINS: The Geophysics Program at UCLA at that time didn't require any understanding whatsoever, of geology. In fact, the way the department was run, if you were a geologist, there was no way you were going to make it very far. You had to be a physicist. In fact, we were required to take physics courses in the physics department.

ZIERLER: Did you start at UCLA directly after college?

IVINS: Yes, I did. I went directly to UCLA, so I entered as a 21 year old.

ZIERLER: Was this a terminal master's program, or could you have stayed on for the PhD?

IVINS: I could have stayed on for the PhD, but I started working at JPL, and JPL was so fascinating to work at. I think at one time I was told, "The research that you're doing with Roger Phillips is in direct opposition to some of the research that is being done by UCLA professors, so you better watch out!"

ZIERLER: [laughs] What was that initial opportunity? What was the point of contact that got you to JPL?

IVINS: Actually, one of the guys who was in one of the battery of courses that we were told to take in our first year of graduate school was a JPL radar scientist. He also happened to have a lot of money to do planetary radar sciences, and so he hired—I think he hired three of us. We all got to go out there and interview at JPL. I think that during the interview process, they said, "You really ought to go and talk to this guy Roger Phillips, because he really does mantle stuff, and you're more interested in mantle stuff." So I went down and I talked to Roger Phillips, and that's how I got a job. So, it was really this individual at JPL who didn't have a PhD but wanted to go back and get a PhD, who kind of instilled this excitement in many of us about research at JPL. One of the people that he also hired was Charles Elachi. So, in those very early years, at lunch, we were with not only Walt Brown, who was the person who had hired us, but Charles Elachi, who later became the director of JPL.

ZIERLER: When you accepted that offer to join JPL, did you have at the back of your mind that one day you would return to graduate school to finish the PhD?

IVINS: Yes.

ZIERLER: Tell me about your initial work at JPL. What were you involved in from day one?

IVINS: There had been a proposal that had been written by someone at JPL who was quite a gifted scientist, and I would say he was from the old school of analytical approaches to geophysics problems. His advisor was actually Barclay Kamb at Caltech, and Barclay was a very similar kind of a guy, extraordinarily good at both observations and analytical theory. This particular individual's proposal had been rejected by the NASA panel, and he was very upset about this fact. The reason that this proposal had been rejected was because it did not have the fluid transport of heat owing to mantle convection. So, I was given the prospect of writing mantle convection code. That was my first project at JPL.

ZIERLER: Where did you sit administratively? What division or section was it?

IVINS: There was a reorganization that had occurred. There was a science division, but the sections had been reorganized. I think we were called the Planetary Science Section. So, while there wasn't a Planetary Science Section when I was hired, that was the section that we became. We became the Planetary Science Section. It did remote sensing and it did Earth gravity fields and geophysics.

ZIERLER: What were the major missions happening at JPL in the early and mid 1970s that were important for you?

IVINS: There was Viking. There was Viking 1 and there was Viking 2. After that, there wasn't much. Because we had had such a flurry of data on the Mercury missions such that most of what we were relying on was data analysis. So, it was lunar data analysis and Martian data analysis. There was a prospect of a mission to Venus, but actually what had happened, NASA had gotten so much money during the 1960s to basically compete to get to the Moon and Mars that there was a dry spell that occurred when Bruce Murray became the director of JPL. In fact, what we called it was the "purple pigeon era" of JPL that I mentioned before. This was an era in which JPL was struggling to find a new flagship mission that could be the focus of a lot of engineering activity. There was a fear that JPL had lost its direction in terms of having a new, post Viking-era mission. Of course the mission that did replace it, the flagship mission, was the Voyager mission. There was a tremendous amount of effort between, say, 1978 and 1982, to get this Voyager mission basically off the ground, going from just talk around a table to actually agreeing on being able to cut metal and start the project. That was a very difficult era of JPL.

ZIERLER: Did you ever interact with Ed Stone? Did you see him in action during these early years?

IVINS: No, I did not, because I had made this decision in 1982 to go into geodynamics, because I saw terrestrial sciences as being a place where not only did my understanding of mantle convection lend itself to the current datasets that were being produced—and those datasets involved seismology—when the IRIS global seismological system was put together, in I think about 1975 to 1977, the datasets that were produced dramatically changed our understanding of the Earth. I'll give you an example of how dramatically they changed our understanding of the Earth. In 1981, Adam Dziewoński and Don Anderson at Caltech produced a paper called "Preliminary Earth Model." It later became known as PREM. That preliminary Earth model is so preliminary that we're still using it. [laughs] That's how profoundly the seismological community had changed how we understand the Earth. Now, this put me in kind of a situation where we had geodesy coming, and we had this profound new understanding of the interior of the Earth, and how do we put these things together. There were a small number of geodetic scientists who understood the importance of this, and I have to say without the existence of those people, I would have been a lone island.

John Wahr is one scientist who particularly stands out in the memory of the 1980s who had produced a PhD thesis at University of Colorado that just profoundly changed our understanding of Earth tides. Earth tides are of course related to the elasticity that comes out of this preliminary Earth model. So, Earth science had suddenly exploded, right in this period where we were struggling to get new space data for scientists for new projects. I recognized this opportunity to work in Earth science, and in particular to work on mantle geophysics, which had greatly expanded its capabilities by having this global seismic network.

ZIERLER: You alluded to it, but it bears going into a little more detail—in the early 1980s when you made this switch, what were some of the key technologies that made this field compelling and exciting for you?

IVINS: One of them was VLBI, very long-baseline interferometry. The idea of very long-baseline interferometry is that you understand quasars very, very well. And so, these antennas are huge. They're the size of many basketball courts put together in radius. These are capable of being able to not only detect quasars, but they're able to understand the waves that are emanated by quasars very well. They are used to study astrophysics, and it turns out that basically because of some scientists that worked at the Center for Astrophysics at Harvard, and at MIT—Irwin Shapiro in particular—the concept of being able to use these to study plate tectonics was put forward in the mid 1970s. A lot of people were very skeptical if this would be feasible, partly because they didn't quite understand what the noise level would be around a VLBI facility. Goldstone is an example. The Goldstone antennas are an example of what a VLBI center could be.

But the capability was such that one could understand the phase of the signals that were coming out of the quasars very, very well. If you locked onto the same phase with a global system of VLBI antennas and understood how the phases at the different VLBI centers across the globe shifted over time, you could make a time series that would tell you where the plates were going. This actually turned out to work, which surprised a large part of the community. A large part of the community thought, "This is just a theoretical idea. People don't understand what the full noise systems are." But over the course of about a decade, it was shown to actually work.

ZIERLER: In the early 1980s, what aspects of plate tectonics felt like settled science, and what were some of the big debates at that point?

IVINS: I think that one of the things that those of us who were required to have an in-depth understanding of the science, that went into the possibility of doing space measurements, of which VLBI was one, was what we already know. What we already knew about plate tectonics. We knew how to map the magnetic variations of the Earth's core in newly formed basalts that form at mid-ocean ridges, and the sequences of field reversals recorded a pattern of extension at all the Earth's oceanic spreading centers. By also understanding the geometry of earthquakes that are associated with both the region where ocean spreading is occurring, and subduction zones, we get a very good idea of both the rate at which plates are moving, the direction in which plates are moving, and therefore scientists can put together a system by which basically the geometry and kinematics of plate tectonics, and even the rates at which they occur can be assembled. It was actually research that was done at Caltech by Dan McKenzie that was one of the forerunning papers on this in the 1960s. Bill Menard, also at one time a Caltech graduate student, was also instrumental in this type of high impact research on plate tectonics. Those of us who were well aware of this type of science, we already knew pretty much where the plates were going.

The question was, how much better could we do if we had these space measurements? It turns out we could do a lot better. Because we know this data from the ocean floors, which is 67% of the planet, so we have a very good idea of where the plates are moving, and actually a lot about how the plate system works, because we have heat flow experiments that tell us how the heat flow drops away from the mid-ocean ridges and the topography. But we didn't have a quantitative sense for how to use that space data to do a little bit better. It turns out by doing a little bit better, we suddenly understand continental tectonics a lot better. So, the focus during the late 1980s was to try to be able to develop a system by which we could understand the Western United States and its tectonics. Naturally, tectonics is occurring everywhere, but since we, NASA, live in the United States, targeting the Western United States was a very opportune thing for the Geodynamics program to focus on. That turned out to be successful because GPS became successful as a geodetic technique.

ZIERLER: Tell me about some of your work on thermal convection during this time.

IVINS: At that time, the codes that were being written were almost exclusively finite element programs. For example, there would be classical finite element programs that could be written for spheres, because in mechanical engineering, you look at the deformation of the interior of spheres. You can also do fluid dynamics for the interior of spheres. But one of the things that characterizes the kind of convection that we want to simulate for a planetary interior is that the differential equations that govern the free thermal convection are nonlinear and they're higher order. They're sixth order partial differential equations. So, they're rather challenging mathematically. The technique that I developed was a spectral technique, and we took on the problem of temperature-dependent viscosity, which makes the equations even much more nonlinear. That was a major effort, to try to develop a new computer program that would do the simulation of temperature-dependent viscosity, planetary heat transport. What I did was I tried to develop enough simulations so that we could get a scaling law between an amplitude measure of the interior convective rate to the heat transport that would come out of the planet. This goes back to the original Apollo funding where this proposal had been rejected, because they don't have this fluid transport. I successfully developed a program that would do that, and that became part of one of the first publications that I did at JPL.

ZIERLER: Was this associated with, or did it rely on, data from a specific JPL mission?

IVINS: No. Except that we were asked to match the plate motions, or lack of plate motion on Mars, and also the heat flow data on the Moon. However, what I found in working with the fluid dynamical equations was that it was much more important to get these scaling relationships, because if you had the scaling relationships, then you could apply it to Mercury and Venus, and you could apply it to all the icy satellites. That was my focus, basically, was a more general approach to being able to use the computer program to be able to generate generalizations that could be applied to all terrestrial planets.

ZIERLER: Moving into the mid 1980s, just a nomenclature question—interplate shear zone, what does that mean?

IVINS: What had happened by I would say the mid to late 1970s was that the U.S. Geological Survey had done enough geodetic experiments such that they understood that the interaction between the North American plate was spread out over a zone. They didn't know, and it was in great debate, how far this zone was. Was it from the Sierra Nevada's to someplace, some offshore fault, like the San Clemente fault? Earthquake data at the time was helpful in defining those zones of earthquake related slip, but we didn't understand enough about earthquake mechanics and silent slip to be able to extrapolate from the fact that there was an earthquake to what it meant about regional tectonics on the scale of the entire western United States. The question was, we knew the shear zone existed; we didn't know the details. We didn't know how wide it was, and we didn't know how intense it was. We didn't know its temporal structure, either. In fact, we're still refining our understanding of that.

ZIERLER: When did you start to get involved with cryospheric research?

IVINS: I started getting involved in cryospheric research when, through this paper that was published—it actually appeared in the literature in 1984, I believe, by Mark Meier. This is global. This relates to another space program that JPL had, in collaboration with the University of Texas and Goddard, and also with the French. There were two spacecraft that were at very high orbit. One was at one Earth radii and one was at about half an Earth radii. These are up high enough so they can be in very drag-free environments. They were spheres covered by mirrors, and there was a laser system of terrestrial stations that tracked them. Originally it was proposed that this laser system could help us understand plate tectonics. The motions of these freely orbiting, drag free, objects in space were entirely driven by the Earth's gravity field. So, we could determine its orbit extremely accurately by placing a set of mirrors on that sphere—it was about the size of a basketball—and determine its orbit so well that we could understand how the Earth's gravity changes as a function of time.

Part of the change was being caused by ongoing postglacial rebound. This concept of being able to measure time-dependent gravity from these types of observations was actually suggested by a Japanese scientist in the early 1960s. This idea had actually been around among people that knew satellite geodesy fairly well. So they launched these—it was rather simple. You have a sphere, you launch it into space, then you do laser tracking of it. This was a project that actually had gotten started in 1976, so they already had a lot of data. This is how we made the link to sea level rise and glacier mass; it actually turned out to be a sphere that was basketball sized, with the mirrors on it. Meier's paper was important, because the glaciological details therein mattered, a great deal. Hence, my pursuit of these details leads to my interest in cryospheric sciences.

ZIERLER: I got to know Charlie Sammis during my time documenting the history of the Seismo Lab. What were some of your collaborations with Charlie? How did you get to know him?

IVINS: I got to know Charlie Sammis through Bruce Banerdt. I knew what Charlie did, and I realized that there was an alignment between what Charlie did and what his interests were—in fact, rheological sciences is one of his passions—and understanding rheological sciences from a fundamental physics standpoint. Charlie also has a very artistic way of thinking about these things. I found the interaction with Charlie to be really, really fun.

ZIERLER: That gets me to my next question. How does rheology work on a planetary scale? I understand rheology in a laboratory environment, but how do we look at plate tectonics, geodynamics, from a rheological perspective?

IVINS: I might point you to the most extreme example, and that would be the planet Io. I call Io "the hot ball bearing of the Jovian satellite system." Because it's the place where all of the tidal dissipation is kind of forced into this one planetary moon. So it winds up with a very high temperature, and so it winds up with a system that is constantly belching out these plumes of material, because it basically has these rock-water ice volcanoes in it. So, it naturally has to have a planetary environment that is very fluid. The question becomes, then, how fluid is fluid? Are you talking about fluid that would be a material I would find in a cold jar of honey in your refrigerator, or is it comparable to anything we find in solid Earth material, on Earth? The fact is for Io to operate the way Io does in its interior, it has to be much lower in viscosity. If you look at another example, which would be, say, Venus—what's the viscosity of the interior of Venus—we don't know the answer to this question within many orders of magnitude, so it becomes a question of what clever things can we come up with that will tell us what this time scale is.

In trying to understand rheology on a planetary scale, it always comes down to, how can we measure what change that occurs over a period of time? Because basically what viscosity is, it's a coefficient that stands between the rate of deformation and stress. So, if we know something about stress and we know something about the rate of deformation, then we can infer something about the viscosity. If we're looking at a planet like Venus, then we've got to make this connection between time scale and stress, and we've got to know something about stress, and we've got to make some kind of observation about the rate of deformation. That is one of the key things that we want to get at in the future, with planetary science studies, is to make better inferences of what the viscosity is.

ZIERLER: As you were explaining earlier in our conversation, the rate of sea level rise right now is beyond anything that we have seen in geological history. When you started to think about these—

IVINS: Not in geological history, but within the last 5,000 years.

ZIERLER: Thank you for clarifying. When you started to think about these things in the 1990s, and you were thinking about sea level rise and its relationship to Earth deformation and gravity and cryospheric research, was it then that it became obvious that the human-caused aspects of sea level change really separated it from these other natural processes?

IVINS: That's a very difficult question, because we have had this phenomenon called the Little Ice Age, that occurred sometime between 1500 and 1900. Depending upon which region you're in, this could be very variable, but we believe that in North America and Europe, that the Little Ice Age, its maximum occurred sometime around 1870. Climate emerged out of the Little Ice Age, and the question is, how much of an impact does emerging out of that Little Ice Age on change that we so accurately measure since about 1980, and that is well recorded through the 20th Century. In other words where is the boundary between human related warming and a natural cycle having the same phase? Where does the end of the Little Ice Age intersect with what are human-caused sea level rise? One of the things that we can understand is that the Little Ice Age is a phenomenon that produces glacier changes that are actually small in rate compared to the changes we see occurring after about 1990. Even if the Little Ice Age is the largest of what are called global Neoglacial events that occur during this 5,000 year period, what's called the late Holocene in cryospheric and Quaternary sciences, we are quite sure that we never had any kind of emergence out of one of these Neoglacial periods that is as ferocious as what we now have. The GRACE satellite missions have been important in establishing the profound rates at we now see ice sheets in demise.

But finding that kind of sweet spot that defines the difference between anthropogenic induced changes and how we're evolving out of the end of the Little Ice Age, which terminated at about 1870, and possibly as late as 1930 in some places, is a difficult question, and is a question of ongoing research. I currently have a paper in review where I try to couch the ongoing uplift that we see, very rapid uplift that we see around glacial systems in the solid Earth, and connect that to what we see from post-seismic events. I realized that we have to sort of move our paradigm to new kind of theoretical expressions for how we define rheology in these simulations. I'd like to think that 1945 is the time at which we can start to quantify what our human impacts are on the environment. Post-war industrialization occurs at such a rapid rate, and there's a defining horizon within the geological horizon—it's called the bomb-horizon, and that's—when the first nuclear tests were done, they put out so many short-lived nuclides that it's very easy to define. In fact, what it does is it kind of messes up radio carbon dating, so people that do radio carbon dating are very familiar with this particular sedimentary horizon. You can find it globally; you can find it in the ice cores. That's what I define as kind of the beginning of the Anthropocene. People have different definitions but that's kind of a nice one where there's both a rationale for it, because of this postwar industrialization has so many other quantifiable changes that are generated.

ZIERLER: Tell me about your decision to return to graduate school at USC to complete your PhD.

IVINS: There was a friend of mine, a very close friend at JPL, who is Slovakian. He was in radar sciences, and he was actually one of the Cassini radar scientists, basically Charles's right-hand man on the Cassini radar experiments. He sat me down sometime in the late 1980s and he says, "Erik, with all the research that you're doing, you're going to be a research scientist for the rest of your life. You just don't have the union card." Here's a guy who, his father was taken off to Auschwitz. He had to watch the Soviets come in, literally seeing the tanks come on the battlefield, when he was a young man. He had escaped the Iron Curtain in the 1960s out of Czechoslovakia. So, I knew this was a person who really understood the world. He sat me down and he says, "Erik"—this is just he and I at lunch one day—"you have to get a PhD. Because if you don't get a PhD, you're going to find yourself without a union card, and you're going to too old to be competing with people who have the PhD, who are young, new scientists." I took him at his word. We have a few people that do not have PhDs in my field. Some of them are quite famous. But I didn't see my trajectory of my career going in that very fortunate path.

ZIERLER: Given all of your experience, was the PhD somewhat fast-tracked for you?

IVINS: Yeah. I'd say I enjoyed it so much that I probably took more time than I should have, but it was what it was.

ZIERLER: Did you take a leave from JPL or you did it part-time?

IVINS: No, there was a program in which if your section approved, they would pay your tuition, so JPL paid my tuition. I was still required to write proposals at JPL as principal investigator, so this was a dual thing. I had to both be a student and be a competitive research scientist at JPL. It was a lot of work!

ZIERLER: Was the thesis directly responsive to what you were doing at JPL?

IVINS: Yes, it certainly was. The PhD thesis, at the time we had beautiful geodetic observations of the post-seismic mantle flow that was occurring after the Landers earthquake in 1994. This was a big topic not only at the Southern California Earthquake Center but also within NASA as well. One of the things they had to understand was, why was everything flowing so rapidly, and over such a large area, and for so long? The rheological model that Charlie and I conceptualized and put together and quantified was something that could have explained this. In fact, it did explain the time scales.

ZIERLER: By the mid 1990s, was your sense that JPL was fully invested in Earth science, generally, and climate change research specifically?

IVINS: No, I did not believe that. I became a little bit more closely in tune with what the ocean sciences group was doing at JPL. They also had this sense that climate change was not really something that we were supposed to talk about. We had Al Gore, who was a great proponent of studying climate change, but it wasn't clear that NASA wanted to take the step of saying, "We are going to involve ourselves in the study of climate change." Because it was more of a question of quantifying things like the rise of carbon in the atmosphere through space-based atmospheric sounding. This is an observation. This is not the study of climate science. So I did not feel as though NASA necessarily embraced the idea of studying climate change. However, in ocean sciences, we had this rise of sea level that was being observed by the late 1990s since the TOPEX/Poseidon altimetry satellite began taking data in about 1992.

We also had remote sensing observations of the cryosphere that demonstrated the rapid nature of changes that were occurring in the cryosphere. So you'd go to a cryosphere meeting and everyone would be talking about the disappearance mountain glaciers. However, what was known about the ice sheets was still a great debate in the 1990's. Some NASA scientists favored growth of the ice sheets, and some scientists favored their shrinkage. The numbers were all very small, since understanding change required subtracting one big number (say input) from another big number (sat outflux) and have confidence in the accuracy of that difference year-after-year. We didn't think that ice sheets were really contributing to sea level rise very much in 1990. That all changed in the 2000s.

ZIERLER: To return to this earlier point, it seems in retrospect the connections would be obvious between cryospheric change and sea level rise. But where does Earth deformation and gravity play a role here?

IVINS: It plays a role in corrupting the data signal that we obtain from the gravity signal. In other words, if we have a spacecraft that's only measuring the gravity change on the Earth, we have two signals that we have to worry about. Well, we have more signals than that, but two long-term competing signals are the fact that the surface mass is changing, shifting, and moving around—it's not disappearing, but it's just taking mass that is on continents and moving it into the oceans. We also have places that are very interesting like the Amazon, where not only do you have huge seasonal changes, but you also have decadal-scale changes. When these phenomenon occur, you have the solid Earth, which encompasses a very long-wavelength gravity field of the Earth that is also changing simultaneously, in part in response to the changes in the surface load, which cause a change in the stress and therefore cause creep in the mantle, but we also have this longer-term phenomenon that comes from the Last glacial age, some thousands of years ago. I think you mentioned earlier that in the geological record, we don't necessarily have rapid sea level changes, but actually at about 18,000 years ago and about 14,000 years ago, and again 12,000 years ago, we have huge changes that occur in the system, and these cause deep responses, deep in the mantle, hundreds or thousands of kilometers down. Even the shape of the core-mantle boundary, that also causes slow changes in the gravity field. Those have to be filtered out, and they're one of the more difficult things to filter out. This sort of corruption of the water change signal in gravity data has to be modeled, and removed. This is something that is absolutely required to run a gravity mission that will get you detailed information about how the present-day ice sheets are changing.

ZIERLER: You mentioned this debate, this scientific disagreement about ice sheet loss versus growth. Your work in the late 1990s in Patagonia, was your focus on land uplift and you took it at face value that the glaciers would shrink? Or were you actually part of that argument that we would see glacier shrinkage?

IVINS: No, actually there was historical data that was remote sensing data that occurred right after 1945. This was because, in war times—it was called the Army Air Force; the Air Force was part of the Army—my father was a pilot in the Second World War, but he was always part of the Army, Army Air Force—they took a special camera and they went throughout the Andes, and they photographed the entire mountain range system. With these photographs, it had three cameras that basically allowed you to be able to do geometrical rectification so that you knew the geometry of what you were looking at. These were classified defense photography that was designed for any future conflict that we might have with any country, Asia or Russia or whatever. These were picked up by glaciological scientists, many of whom were working in Japan. So, the Japanese scientists interpreted theses photographs as a baseline observation of the geometry of the glaciers. They were able to compare them to photography that was taken later in the 1950s, 60s and 70s, and they were able to actually develop a crude time series so they knew the glaciers were shrinking. Starting in the 1990s, there were altimetry measurements that were made over the Patagonian ice system, so we already had some knowledge that we could quantify how the glaciers were changing. It was nice that the glaciers were geographically isolated —so we didn't have to worry about unmeasured areas corrupting the inferences about shrinkage. In turn, we were able to conduct a study of what this shrinkage might do to the mantle. It was a place where we already knew the glaciers were shrinking, we had quantitative estimates of what those rates were, so it was a perfect little laboratory, terrestrial laboratory, to study mantle geophysics.

ZIERLER: You mentioned it was only in the 2000s that the scientific establishment coalesced around ice sheet shrinkage. What were the developments that led to that?

IVINS: You have to understand how crude things were in the 1980s. Although our ability to do technology in the 1980s was quite significant, the vastness of the Antarctic interior meant that glaciologists had to go out with stakes that were pounded into the snow. They actually put stakes in the ice covered ground, came back a year later, and saw how much snow accumulated. They were putting these stakes all in very coastal regions, and so naturally, they were seeing a lot of accumulation, because this is where all the precipitation occurs, but they had no ability to be able to understand what the flow characteristics of all of the outlet glaciers were and only crudely understood calving rates, because you just had to make too many measurements in the field. So the bias was that the ice sheets are stable. And because there's so much accumulation that's occurring, possibly are in growth. These were really quite good scientists that were making these estimations that the ice sheets generally are growing, at the current time. It was because of the bias that was created by the fact that you didn't have space observations that could provide comprehensive surveying and thus provide quantification of secular mass changes.

Sometime in the 1980s, and I believe Barclay Kamb was involved in this, we began to understand that using remote sensing, we could understand what the Antarctic ice sheet was doing in terms of its flow systems, where they were, but we didn't have a measure of how quickly the ice was moving out of Antarctica. Using radar interferometry, it was Eric Rignot at JPL who was one of the first to determine the change in the grounding line in Greenland and Antarctica for a handful of outlet glaciers and also to measure in the ice velocity at the terminus at outlet glaciers using repeat radar surveys. So, we got a new sense for how we could quantify the outflux. Along comes radar altimetry, which starts to occur in the early 1990s, and this radar altimetry also gives us a sense for how height changes. The problem was that the signal is corrupted by the various calibrations that have to be done, so there was a lot of controversy as to how to interpret this radar signal that was coming from both Greenland and Antarctica in terms of the height changes. There was a raging debate that was occurring within about 1998 to about 2002 over, are the ice sheets gaining mass or are they shrinking in mass.

The thing that kind of became crucial was that we developed ways to be able to measure the outflux velocity using remote sensing very accurately. We also developed the computational capability to use atmospheric data and reanalyze that atmospheric data and make very good models of the amount of snow that was falling on Greenland and Antarctica to within plus or minus 5%. Knowing the outflux and knowing the influx meant that we were able to build models of the rate at which the changes in these processes were occurring, and they suggested negative mass balance. Along comes gravity, and we have an overwhelming signal that tells us Greenland is losing mass. Overwhelming signal. So gravity is really—it's kind of like you've got the golf ball on the green, and you just want one putt to put the ball in the hole, and it's a long putt; well, that long putt wound up being space gravity.

ZIERLER: Many questions on the GRACE mission, but first some background. What were the prior missions upon which GRACE was built? How did it advance previous missions?

IVINS: I would say the whole NASA gravity program including the Lunar gravity program contributed to this. I'll give you some historical context. When I was a child, the big event that occurred in space sciences was Sputnik. Sputnik was kind of a joke satellite in some ways. It was just a technological demonstration that the Soviet Union could put something into orbit, have it come around, and they would know where it is. Actually, they didn't know where it was; the U.S. knew where it was, because the U.S. did the tracking of the Sputnik space satellite because we had much better optical tracking technology. One of the things that came about using the tracking of the Sputnik satellite, was that we determined that the Earth had a pear shape. The pear shape was caused by the gravity field that affected the Sputnik satellite orbit. What this pear shape is, is basically a first-order determination conducted by satellite geodesy. Determining orbits is intimately connected to how you measure the Earth's gravity field. That's how this concept that putting these balls way up in space, these basketball sized objects, and putting reflectors on them, occurs. At the time, one of the problems is to try to be able to determine not only the time-dependence of the change of the oblateness of the Earth, which is basically the potato shape of the Earth as opposed to the pear shape, but we had a great deal of difficulty in being able to determine the time rate of change of the pear shape of the Earth, which would have been very important for understanding the mass balance of Antarctica.

We already knew, from the two space missions launched in 1976 that continued to collect data into the 1990s, how were we going quantify the lower order time-dependence of the Earth's gravity field. Such measurements were related to mass balance of the ice sheets, but not definative. There had been another program that had been thought about by space scientists that did space gravity earlier on, and that is to put a pair of satellites that would track one another, and as they flew over a region that had more mass, the trailing satellite would accelerate towards the mass, and even the lead satellite would be attracted back to the mass.

So, using the range measurements between the two satellites, you could tell where there was mass, and where a mass or a mass deficiency was. And, if you had these two satellites tracked one another for years and years and years, then you would be able to determine what the rate of change of the mass is below. People understood that this technology could be investigated. The problem was that we didn't know enough about how to accurately track these two satellites and measure all the other forces on the individual satellites. As the technology developed so that we understood both using GPS and understanding on-board satellite technology—and all aspects; thermal and all kinds of control that we have on what's affecting the distance between these two satellites—then we were able to get the GRACE program off the mat. But the concept had been around maybe for 20 years before that.

ZIERLER: Were you involved from the beginning on conceptualizing and launching the GRACE program?

IVINS: No. I would say that the champion of that was Byron Tapley and Mike Watkins. Because we didn't understand what kind of breakthroughs there could be, in that technology. It was really Mike and a lot of very good engineers at JPL that spearheaded that. We began to understand—I personally began to understand—how that was coming fruition by about the year 2000. I was completely on board with the mission concepts and the mapping capabilities. In fact, I remember running into Mike in the cafeteria shortly before it was going to be launched, and I said, "Do you really think we can get so accurate to be able to measure Greenland's ice mass balance?" He said, "Well, maybe? Maybe not." This is one of the surprising things, I think, in all of space exploration, is that you actually get there, you actually get the data, and it turns out the data is just overwhelming. It's just—it's one of the fun things about space sciences. And it was. Just stunning data.

ZIERLER: You've alluded to it, but just to go into a little more detail, what were the primary scientific objectives of GRACE, and what were the engineering breakthroughs required to get there?

IVINS: The big driver of GRACE turns out not to be a scientific objective. This is really weird, because there was so much skepticism over the success of GRACE that it was flown as an experimental mission, meaning that, "We are going to see how well you guys can do your technology, and we expect that there's going to be a three-year mission, and we expect that we might be able to find water changes that might occur across the entire United States over that three-year period, or the Amazon." The Amazon was one of the big targets. It was one of the big advertisements: "We can determine the gravity changes of the Amazon Basin." It turned out, we could do a lot more.

ZIERLER: When launch day happened, what were your interests, and what were your hopes? What did you hope that it would accomplish?

IVINS: Since it was launched in Northern Russia, I was hoping it just got up there and went through the first phases of its calibration. The data wasn't released very quickly, because there's a lot of post-processing that occurs at a very basic level with the GRACE program. Both the Center for Space Research at the University of Texas and JPL had extraordinarily competent engineers that worked on this part of the data analysis, to take the range and range-rate distance between these two spacecraft and make good science out of it. They did, and that took about a year from the time that it was launched to the time that it was released. I knew what was coming down the pipeline, so as soon as the first data were released, I was ready to make gravity maps. I think people were a little bit surprised, because I wasn't on the GRACE advisory science team, that had been doing a lot of the pre-estimates of these things. Suddenly people at the University of Texas were saying, "Who is this guy, Ivins?" One of the things about the GRACE science team is that they're extraordinarily happy to have users. This isn't necessarily the case in planetary science, where the discovery—so much investment has been made by the scientists that have been promoting and overseeing the engineering that they're not so happy to be sharing the data as soon as it becomes available. In the GRACE science team, quite the opposite was the case.

ZIERLER: The collaboration with Texas, did that run through Mike Watkins? Was he the connecting point there?

IVINS: What had happened in the 1970s is that Byron Tapley taught courses on satellite gravity. There were not very many competitors, and so he had a staff of people at Center for Space Research that had intimate knowledge of orbital mechanics and the data reduction processes. He basically had built a graduate school that had a large number of people that were very good at understanding the software and the engineering. Mike was certainly one of those people. He came out of that group.

ZIERLER: When the data started to come in, what was it telling you? What were some of the new things we could know?

IVINS: The first data that we got was basically to build a gravity map of the Earth. We had never had gravity fields for the Arctic Ocean, and we had never had gravity fields for Antarctica. Since I had expertise in understanding Antarctica, understanding its morphology, tectonic history and its basins, basically where the ice was and where the rock was, I first looked at the gravity field over Antarctica. And you could immediately see the basins; you could see the geological structure just immediately. I actually reported this I think at the first meeting in Germany when GRACE data had become available. We were not supposed to talk about GRACE; we were supposed to talk about a German gravity mission, CHAMP. But [laughs] I used both the German data and the new GRACE data. I remember the heads of the German component of the GRACE project and Byron Tapley, the head of the American effort; they had somewhat nervous looks on their faces when suddenly I showed a GRACE map of Antarctica that I had been able to reduce from the spherical harmonic coefficients. But it was all just beautiful. You could see all of the Wilkes Basin in East Antarctica. You can see all of this geological structure. So I said, "Hey, you guys, you have science. You've already done science right here. Nobody has seen this gravity field before." They were quite pleased by that.

ZIERLER: What was the critical mass of data and even some of the theoretical advances where in the early 2010s it became obvious that you could make some pretty definite conclusions about ice sheet and glacial mass and things like that?

IVINS: Part of it was the fact that we had a good idea of how to correct for glacial isostatic adjustment of the solid Earth gravity field. That became a very crucial factor in being able to determine the mass balance of Antarctica. Not so much for Greenland, because Greenland was losing mass at such a tremendous rate that the gravity field change signal that would be caused by postglacial rebound there turned to be at the 15% or 20% level. Whereas for Antarctica, it was actually at the 100% level. So we had to really reduce the error bars on these glacial isostatic adjustments models, for Antarctica. I was a participant in that modeling. That was my primary modeling effort.

ZIERLER: In all of this work, did this increase your interface with NASA, given NASA's overall interest in this area of research?

IVINS: Yes, it did. I became the lead of the Ice Mass Balance Inter-comparison Exercise for NASA, which was a combined effort with ESA.

ZIERLER: Tell me about working with ESA and some of their different perspectives on the research.

IVINS: It's interesting; we had a discussion in the cryosphere group about this two days ago—I would say that ESA is an interesting entity in and of itself, in that there are the British, and then there are the continental people. The British area of cryospheric sciences, and therefore determining mass balance, there are too many scientists and too little funding, so they are sometimes creating controversies among themselves that don't even necessarily have to be there. The Europeans are a little bit more cold and objective: they just want the measurement, and they're willing to live with the measurement. They don't invest their careers so much in formulating what the result should be. The Europeans themselves actually form a good mixture of straightforward science, I would say. The Americans were—and it may just depend upon the individuals that were there—are much more contentious, and have much more of a tendency to fall into a particular position on mass assessment. And then it's very difficult to move them out of that position. So it was a lot of personal dynamics that you had to contend with when bringing together about 27 scientists who could all sit down at the same table and say, "We are going to do inter-comparisons of gravity, altimetry and the mass flux method with the following set of rules." We were able to do that. Between Andy Shepherd and myself, we were able to pound out what those rules were and make sure that everyone followed them. John Wahr was an invaluable leader of the gravity component.

ZIERLER: Later on, in the 2010s, as you started to think about the relationship between glacial melt and sea level rise, I'm curious if you thought about this as a uniform phenomenon or if different coastal areas would experience more flooding, would experience more land loss than other areas of the planet.

IVINS: We certainly recognized that the ice sheets themselves contain a gravity field, and as that gravity field changes, the rates at which inundation will occur in some locations will be different than it will in other locations. One of the things that we did understand from research that had been done by Jerry Mitrovica's group was that the more intense sea level rise that would occur from the changes in the gravity field of Greenland and Antarctica could possibly be on both the western and eastern coasts of the United States. And also Japan. You're talking about some of the most important industrialized part of the world being affected much more by sea level change than the rest of the world. A preliminary estimate that was based upon rather crude models of Greenland and Antarctica but these heightened the interest in determining where the change in the environment was going to be the most. Of course, heat content of the Earth's oceans is also an important factor in this, and circulation.

ZIERLER: This obviously has international political ramifications for people who live on the coasts and planning and things like that. Did you get involved in those political-style conversations?

IVINS: I would say that I did not become so involved in the political conversations. I don't think that there have been many political conversations. I think the way this has worked out is that the NASA Sea Level Change Team tries to inform people as well as we can what is in store for them as the climate changes, and what communities might be most vulnerable, so that they can make decisions about what they should do. For example, if you go to the Eastern Seaboard of the United States where you have the Chesapeake Bay and communities that are very vulnerable to storm phenomenon, you find that county to county, there are different policies about what to do with sea level rise and coastal degradation. For example, there are some cities where storm surges actually enter the sewer system, so their remedies depend upon a very local infrastructure. Becoming involved in the actual politics of what you should do or the decisions that you should make, that's something that we really want to stay away from. But we want people to understand that they have a resource, that they can look at NASA datasets that can be presented in a way that local decision makers can make prudent and well-informed decisions. This is the most important thing that we can do. We can have nothing to do with the local politics, because local politics may have to do much with how much do people want to invest in infrastructural development as opposed to abandoning those coastal sites that they know are going to be inundated. It is their cost-benefit analysis to be considered, not NASA's.

ZIERLER: Of the many ways that your research has been honored and awarded, one I want to specifically ask you about is the NASA Exceptional Service Medal. What was it like when you got that, and what research were you being recognized for?

IVINS: I was being recognized I think for the work that I had done going back to the early 1990s involving making predictions of what the geodetic responses would be of glacier and ice sheet retreat, and that many of these predictions had actually turned out to be proven by geodetic observations. I think that's what the citation was for. And it was for the sustained research in that area.

ZIERLER: Moving in more recent years, as you reflected on the way that your research became somewhat even more interdisciplinary, did you see this as opportunity to meld the more recent cryospheric research with your earlier terrestrial geophysics research?

IVINS: Yes, I did. Going back to the original convection code that I was asked to write at JPL that involved temperature-dependent viscosity, because the mantle properties are strongly affected by temperature, and we know that a convective system has very large lateral variations in temperature, this means that the solid state properties of the minerals are going to be very different depending upon these changes in temperature. For example, Iceland is very near a mid-ocean ridge, and we know—or we infer—that the temperature beneath Iceland is very much greater that they are at an equivalent depth beneath Scandinavia. This means that the mantle uplift response time that we're going to have for glacial changes in Iceland is going to be very short in comparison to that od Sweden's coastline, for example. The viscous response that we get from the mantle is going to have very large deformations over relatively short time spans. Therefore, our design of geodetic observational systems should be sensitive to the kind of changes that Iceland has experienced since the end of the Little Ice Age. In fact, geodetic observations and models reveal this ‘fast' deformation (10s of cm/year uplift).

ZIERLER: I asked earlier about the road not taken, not being a professor. More recently, you did have the opportunity to work with a few postdoctoral scholars. Tell me about that work and what you were able to accomplish.

IVINS: I think it came at a very opportune time. Surendra Adhikari, who is trained as an engineer in Nepal and became a glaciologist with some of the most eminent glacial modelers in the world both in Belgium and in Canada, came to JPL to do glacial modeling. At the time, Eric Larour and I were interested in being able to do a more global model of sea level. One of the things that we were interested in is being able to use the changes in the solid Earth as they might change the environment in which the ice was changing itself. In other words, to try to look at changes in the slope, for example, of the bottom topography beneath an ice sheet, as the ice sheet changed. Because this has a feedback, and that feedback can be positive, or it can be negative, meaning that the changes in the solid Earth can slow down, or they can speed up, the rate at which mass exits the continent and goes into the oceans. It turned out that Surendra Adhikari, this particular postdoc, was extraordinarily gifted and unintimidated by any higher-order mathematics.

I realized at that point that he was probably capable of doing global geophysics. In various discussions—because we would have long discussions—I mentioned to him that once he had written a computer code that did self-gravitation of the sea level in response to ice change, that he would be able to predict Earth rotation parameters. He was very surprised. He said, "How would I do that?" I said, "There's this thing called Euler's equations." Euler's equations basically operate as a way of being able to take changes in the Earth's inertial tensor and then be able to drive the changes in the rotational properties of the Earth, the rate of change in position of the pole, or the rate at which the planet is spinning. I said, "Do you know that there's actually an extraordinarily good dataset for the changes in the pole position of the Earth?" He put two and two together and he said, "Euler's equations, those don't intimidate me." Even if they're just a set of nonlinear ordinary differential equations, and very low order. So, he set about both getting the dataset and using GRACE data, putting together a change in inertia tensor model for the Earth, and lo and behold, it matched the data. It matched the rotational data. The greatness of postdocs.

ZIERLER: [laughs] Moving our story right up to the present, in what ways did COVID and the lockdown affect your research, and did it also influence the timing of your decision to retire?

IVINS: It did not influence the timing of my retirement. The timing of my retirement is basically tied to the ages and joy I have of my grandchildren. That's really what motivated me to retire. But one of the stories I tell about retirement is that I worked with Hélène Seroussi at JPL, who was an ice modeler. When we'd get together—and my oldest granddaughter was getting older—and she was still looking very young, I was thinking, "When a young person I work with at JPL, who has a PhD, starts looking like my granddaughter, then I know it's time to retire."

ZIERLER: [laughs]

IVINS: That became part of that story. The COVID lockdown was interesting, because it put me in a position where—I could do Zoom meetings, but Zoom meetings just aren't very much fun, and you don't really get in long conversations with people at Zoom meetings—it put me in a situation where every day, I could get up and I could continue doing mathematical analysis of advanced geodetic problems, and that's what I've done.

ZIERLER: Now that worked right up to the present, 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, what value do you see having this duality in your research where you started more in terrestrial geophysics and then you moved more into cryospheric research? In what ways have both aspects of your research agenda positively influenced the other?

IVINS: I would say it is because at this unique ice-solid Earth-ocean interface occurs at the surface of the Earth. You have a changing ice load stress that is forcing the deep interior of the Earth to flow, a phenomenon for which we have the means to measure using geodesy and we have records of past ice sheet change, a forty year history of space observations, and ice sheet physics models that can predict the future changes. I'd say we are in exciting times. I just came back from a workshop that we had at the University of Minnesota, where we had all the people that were involved in the NSF funded Patagonian field project and interpretation—one of the things we want to try to understand is the ice history that will explain the ongoing uplift of the region, a problem that is complicated by large ongoing response to mass change of the present-day.

We had someone, Lars Hansen, who came from another part of the Geosciences Department there who does experimental rock physics and creep. He said, "This is just the perfect field rheological laboratory that you have here. You have great knowledge of how the ice is changing, and you have this unknown that is how the Earth is going to respond." And he says, "I can't see a better laboratory that you would have." So that's how it affects solid Earth sciences, it allows very different fields to become intertwined and inter-dependent, but with solid Earth sciences being a big beneficiary if we can solve that rheological question. But I would say on the side of cryospheric sciences, that you now have a grand Earth system science question in front of us all. If you're able to use a 3-D thermomechanical ice sheet model coupled to climate to predict the gravity field of the Earth as the ice changes and the associated solid Earth deformation, then we have a great way to quantify the feedback to the ice sheet. The feedback can either slow down or speedup the rate at which ice is irreversibly lost to the oceans.. So you're able to make contributions to cryospheric sciences and also to projections of future changes in the environment that we are going to experience as the phenomenon of global warming goes on.

ZIERLER: Of all the scientific research you've been involved in, what has been most satisfying, and really making a discovery and changing the way we understand these processes?

IVINS: I would say in terms of discovery, I think it was the knowledge that the GRACE science dataset was so full of new information. I would say that it reinvigorated my career as I passed my 50th birthday. I felt as though the research I did complimented a revolution in hydrological and cryospheric sciences. I had the sense that my research was helping to shape this new scientific frontier. I'd say the dataset offered a new vitality for me in science.

ZIERLER: Because JPL contributes more in analysis and observation, and not in climate mitigation, where do you see other government agencies, other non-governmental organizations—how do they take all of this research actually to make us adapt and mitigate climate change?

IVINS: I didn't know this interview was going to go so long, because we have a scientist that I worked with at JPL who's having a lunch together—I'm perfectly happy to continue this interview and miss the lunch, which started at 11:30—but her name is Nicole Schlegel, and she is leaving JPL, leaving the cryospheric sciences group, and she is going to NOAA, National Oceanographic and Atmospheric Administration. She there is going to put together a new ice physics group that is going to try to help make projections of sea level change that are related to the cryosphere. Because everyone now understands that to be able to be a viable community-serving organization, which is what they are, what their calling is, they need someone who is capable of being able to do this kind of data-driven ice sheet simulation. That's the relationship, I think, between NASA science and the need for agencies like NOAA to help the society protect itself from future changes in sea level. In the case of Nicole, we see the direct link of expertise honed at JPL that will now be put to good use, creating newly emergent applications at NOAA.

ZIERLER: Last question, looking to the future—as you explained, in the 1990s JPL was not fully invested in Earth science and climate change research. Now, obviously it is. What are you most excited about? What accounts to this transition to JPL's investment and where can it lead in the future?

IVINS: Oddly enough, I see the capability to finally be able to perform full Earth system science in the modeling sense using NASA and other space data sets. Ecosystem sciences may also emerge as major components since these can be monitored much more effectively, so that we can start to understand what kind of trends there are, possibly crucial for forecasting our ability to sustain a food supply for a growing population. NASA is always involved in new technologies, and as long as we remain NASA's most capable laboratory for this, we will see Earth sciences from new perspectives with science activity centered at JPL. The new perspectives may come in unpredictable ways, such as what we have witnessed with ocean topography, GPS and gravity field determinations. The data sets are now becoming long enough that viable data-informed fully coupled Earth system modeling is sitting just around the corner. These will be capable of making statistical models of future change. Inevitably, JPL will be involved in that.

ZIERLER: Erik, I want to thank you for spending this time with me. It has been a wonderful, wide-ranging conversation. I'm so appreciative of your time. Thank you so much.

[END]