Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
Please contact [email protected] with any feedback.
Credit: LSU Dept. of Physics and Astronomy
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of Gabriela Gonzalez by David Zierler on March 22, 2021,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
Interview with Gabriela Gonzalez, Louisiana State University Boyd Professor in the Department of Physics and Astronomy. Gonzalez explains how the pandemic has slowed down data analysis for LIGO, and she recounts her childhood in Cordoba, Argentina. She describes her early interests in science and her physics education as an undergraduate in Cordoba. Gonzalez explains the circumstances that led to her graduate studies at Syracuse University where she studied relativity under the direction of Peter Saulson, and where she first became involved with LIGO. She discusses her postdoctoral appointment at MIT to work in Rai Weiss’s group, and she explains LIGO’s dual goals of detecting gravitational waves and building precision instruments toward that end. Gonzalez explains her decision to join the faculty at Penn State and she describes the site selection that led to the detection facility in Livingston, Louisiana. She describes the necessary redundancy of the LIGO detectors at Livingston and Hanford, Washington, and the importance of “locking” the mirrors on the detectors. Gonzalez describes the overall scene at LIGO in the months up to the detection and the theoretical guidance that improved the likelihood of success. She describes the intensive communication and data analysis to confirm the detection prior to the announcement, and she explains how she felt honored as part of the overall Nobel Prize award and subsequent celebration. Gonzalez describes LIGO’s work in the current post-detection period, and her own focus on diagnostics of the data, and she explains why this work, and the constant concern in missing something important, can be stressful. At the end of the interview, Gonzalez surveys what mysteries LIGO can, and cannot, solve, and she conveys optimism for LIGO’s long-term prospects to continue to push fundamental discovery.
Okay. This is David Zierler, oral historian for the American Institute of Physics. It is March 22nd, 2021. I'm delighted to be here with Professor Gabriela Gonzalez. Gaby, great to see you. Thank you for joining me.
Thanks for thinking of interviewing me.
Okay, so to start, would you please tell me your current title and institutional affiliation?
I'm Gabriela Gonzalez, I'm a Louisiana State University Boyd Professor in the department of physics and astronomy.
Now, the name Boyd for the chair that you're honored by, who is or was Boyd and is there any connection to your research?
It's an honorific name given to a small number of professors at LSU, named in honor of brothers David F. Boyd and Thomas D. Boyd, presidents of LSU in its formative years.
Well, Gaby, I'd like to start with a question that we're all dealing with right now. How has your science been affected in the pandemic, by not being able to see people, by working remotely, by being farther away from the experiments, perhaps, that you'd like to be? How have you fared during this past year?
Well, everything has changed and not all for the better. We are all looking forward to going back to normal times. There are things that we have learned, like we can communicate even without meeting in person. We can make progress. We can analyze data. But we cannot make progress as fast as we want. We cannot communicate as well as we like to. So progress has been a lot slower.
Where would you be now in terms of LIGO and the detection of gravitational waves, absent the pandemic, do you think?
We certainly would have taken a month of more data than we took. We were taking data in what we called our third Observing Run, that started in April of 2019. And we were going to finish at the end of April 2020, but we finished before the end of March because of the pandemic, because we weren't sure that it was safe to have people at the observatory taking data. In those early times, very little was known. We had taken enough data, and we are still analyzing those data. We can analyze data remotely and we have made already a number of detections, even though we haven't finished analyzing all the data.
At the observatories, things were stopped for a while, but now they are safely meeting to make progress with the installation of new hardware, with the installation of new systems. I and my students would have been part of that, not so much because we are a critical part of that (we are not), but because we like having several people, not just the critical people needed for the job, to help with the diagnosing, with the communication. And that is not happening. Only people who are needed and with many safety precautions, are at the observatory. My individual research has to do with diagnostics of the data, and I'm still doing that with the data we have taken in the past. I'm still doing this with the instruments as they are being changed, as we speak. But not in communication, not in real communication, with the people doing the job. So again, it's a lot less efficient this way.
Well, Gaby, let's take it all the way back to the beginning. Let's go back to Argentina, and let's start first with your parents. Tell me a little bit about them and where they're from.
Well, they are from Argentina. [laugh] They were born in the city where I was born, in Cordoba, Argentina. My dad’s family has several Argentinian generations, which was strange for my generation. I remember when I went to school, we all would ask: where are your grandparents from? On my mother's side, my grandfather was Polish and my grandmother was Spanish, they immigrated to Argentina in the early 20th century. On my father's side, my grandmother, whom I didn't know, was an indigenous person. My grandfather's parents lived in the countryside.
Tell me about Cordoba, where you grew up. Is it a large city?
It is a very large city. It's one of the largest cities I lived in, or the largest depending on how you count [laugh] Boston or Cambridge. Cordoba has a million and a half people, closer to two million now, but a million and a half when I lived there. My mom was a high school math teacher, my dad was an accountant. They studied in Cordoba, and they graduated there. I have a brother three years younger than I am. I went to school and to college in Cordoba.
Gaby, when did you start to get interested in science? Was it early on?
At the time I went to school, there were seven years of elementary school and five years of high school. In elementary school, I liked everything. Especially math, but I liked everything. I didn't have very good memory, though, so in geography, history, things that I had to memorize, I didn't like so much. My parents looked for a high school for me that had emphasis in science and technology. And I went to school that had an orientation in chemistry. I loved chemistry.
Again, I liked math, but I liked math applications, and I loved chemistry, but then I learned about atoms and molecules, and I loved the particles and the atoms and the quantum physics of the particles. And I thought that particles and physics could explain everything. [laugh] When I was about in junior in high school, I wanted to study physics. Not to be a physicist, I didn't know any physicist. I didn't think about science as a career. I just wanted to know the explanations to everything. And that's what I went to college to learn.
Gaby, as a girl and as a young woman, in -- if you would consider it somewhat of a conservative society, were you ever made to feel like science was not an appropriate path for you? Were you ever discouraged or not?
Well, certainly not by my family. Although long after [laugh] I migrated to the US, my mom and my dad told me that when I wanted to study physics in college, they thought, “Oh, what is she going to do with that degree?” But they never discouraged me for what I was going to college to do. There wasn't a question about us going to college, my brother and I. But what we wanted to study was up to us. In college, and even graduate school, I did hear from people saying women are not good in science, and women are not good for physics. Especially for physics and they said that in front of me. In front of other women. And I think with the intention of discouraging us. Of course, these were older physics professors. But somehow, that didn't discourage me, although of course I understand how it discourages many people. It's terrible, terrible. And somehow it made me more, I don't know...
[laugh] More resolved, yes, because I said to myself: I'm going to show them. I lost respect for those people, I have to say. I was brought up to respect senior people or people who were older than I was, because they knew better. But then I realized that sometimes they don't know better.
And Gaby, did you know when you started college in Cordoba that physics was what you wanted to pursue, or that came later on?
In Argentina, you have to know your major when you start college, because it's not easy to switch majors. If you switch majors, for the most part, you have to start from scratch again.
It's like the European system.
So I went to college to study physics, yes. We started together with a class of students studying physics, math, and astronomy, but by the second year, we were all taking different classes. But like I said, I just wanted to learn explanations for things. I wasn't thinking about physics as a career. I wasn't thinking about a career in general, if I have to be honest. [laugh] But it was in college from the beginning that I learned that the physics professors were also investigators. They were all men. Well, I had one female professor in the third year. But everybody else was a male professor. But they were all investigators. They were all scientists, so that's when I learned that not all the answers are known, but not even all the questions are posed. And I loved that.
I'm learning so much these days about physics education in Argentina. In January, I interviewed Juan Maldacena. In February, I interviewed Marcela Carena. And now I have you for March. So this is, I'm on a roll, this is great.
[laugh] And we probably all told you the same. You have to know what you want when you start college in Argentina.
Gaby, during your undergraduate, what kind of physics did you enjoy the most? What branch or more generally, did you know if you were more interested in theory or experimentation?
I was interested in theory. In Argentina, college physics is a five-year degree, and the last year, you work on a thesis, that it usually leads to a publication, my thesis was in the theory of general relativity, which I learned to love. [laugh] We have to choose a specialty to take advanced courses in the fourth year, so that's when I chose to work on general relativity. And I liked theory. I don't think I was better at theory or experiment. We didn't have a lot of experimental classes. There were experimental research groups, of course, more than half the groups were experimental groups. But somehow, the experiments that we did in college were more, I wouldn't say boring, but they didn't have unsolved questions. We all knew what the answer should be, and we worked to get that answer. While in theory, even in some of the homework sets, there were different ways of getting answers and it was like more open, and that's what I liked. Until I came to the States. [laugh]
Gaby, what kind of advice did you get when you were thinking about graduate programs? Were you specifically encouraged to go abroad for your thesis degree? To come to the United States specifically?
No, no. I have to say that, of course this was a long time ago, but in Argentina, in the university where I was going, and there was a PhD program and it was very good. It's not very, even now, structured like here in the US. There are not, for example, required courses. It is a bit more structured now than it was then, but at that time, it was very, very unstructured. There were few people who followed the PhD. You could get a job at the university as a teaching assistant without getting a PhD, so graduate school is not something that I really thought about.
But I actually fell in love with [laugh]-- this is an anecdote I like to tell. I have a poster behind me about it, but you cannot see it, with Einstein saying that his theory of gravity should not be blamed for people falling in love. But I met my husband [laugh] studying relativity, studying Einstein's theory. His name is Jorge Pullin. He's a physicist too. He's from Argentina, but he was studying in a different university in the south of Argentina, in the Instituto Balseiro. It's the best physics school in Argentina. But they didn't have general relativity for a PhD thesis, and he wanted to do his PhD in that, and there was a relativity group in my university, in the National University of Cordoba.
So he came to do his PhD research, to Cordoba, and that's how we met, [laugh] because I was studying in that group too. He was finishing his PhD at the time I was finishing my college degree, and he got a postdoctoral fellowship in the US. I had registered already for the PhD program in Cordoba, and I thought, well, we'll go two years to the US, I'll take graduate courses, which would be good for me, and then we would go back. But we never went back. [both laugh] I think my whole career has been kind of zigzagging into things that happened and then I loved and then I make choices.
Where was your husband's postdoc?
It was in Syracuse University.
Okay. That answers my next question, why you went to Syracuse for your PhD.
That's right. My husband had this postdoctoral fellowship at Syracuse University that had one of the strongest groups in theory of relativity at the time. Even now. At the time, it was very well recognized, with many young people and professors like Josh Goldberg, Abhay Ashtekar, Lee Smolin, Raphael Sorkin. A year after we got there, for several reasons, we decided to stay in the US, and I registered for graduate school at Syracuse University with Abhay Ashtekar as my PhD advisor for a thesis in general relativity.
Gaby, how was your English? How was your husband's English when you got to Syracuse?
[laugh] Well, his English was very good, because his father was from a British family, his grandparents were British and they spoke English at home with many relatives. So his English was very, very good. I had studied in English in school. At school one takes English, but that's not very much. My parents sent me to private academies, to learn English. Because of that, I thought my English was very good. But it was not. [laugh] I realized that when getting to the US. If people didn't speak slowly to me, it was so difficult to understand and for them to understand me. I still have a very, very thick accent. It was even worse, much, much worse then. But I managed. [laugh]
Gaby, how did you develop your relationship with Peter Saulson?
I arrived to Syracuse University in '89, in the middle of a winter storm, which was the first time I saw snow falling from the sky. That first year, was a cultural, weather, intellectual shock. [laugh] In all senses. Peter Saulson joined Syracuse University in '91, after I had already been there for a year or so. I had to take a lab course in graduate school in which one had to choose an experimental group to work for a semester, to be familiar with laboratory techniques. I chose his lab, which was not set up yet, because I was doing my thesis in relativity, and he was talking about this measurement of space time. And I loved it.
We had a small, what looked like a small experiment that was going to be done in the semester. I changed advisors and thesis topic and that experiment ended up taking several years, and this was my PhD thesis. I loved the idea of experimental precision and the idea that you had to build this instrument that was so precise to measure waves that were coming form so far away, that actually had been up to that point concepts. Just intellectual concepts. But here, they could be measured. And the instrument, the precision that was needed and how you didn't know how to get that precision, and you needed to work on prototypes and you had to work on this and that. So it wasn't just one big thing. It was a lot of small things that had to be done right. That just fascinated me.
What were some of Saulson's areas of expertise?
His area of expertise was for most of his career, apart from LIGO in general and detection of gravitational waves, was thermal noise. And that was what my PhD was about, and how Brownian motion or thermal noise makes object vibrate whenever they are at the non-zero temperature, which is always. How that vibration can cause noise in an instrument that measures changes of distances. So the vibrations of the mirrors that we used as test masses to measure gravitational waves vibrate because they are at room temperature, and how to predict how much of that vibration happens at different frequencies was using a theorem from statistical physics called the fluctuation dissipation theorem that had connected the microscopic and the microscopic. [laugh] The use of this theorem it hadn't been proven in experiments, in macroscopic experiments, as much as it needed to be for predicting the noise in LIGO.
Were you aware of LIGO? Were you aware of the collaboration between MIT and Caltech while you were a graduate student?
I was. Even before Peter Saulson got to Syracuse, I remember Kip Thorne visiting. Because, again, the group at Syracuse was a big group. And I remember Kip Thorne saying that he was pushing and it was almost approved—this project to build an interferometer—and the project was called LIGO (Laser Interferometer Gravitational-wave Observatory). And I remember him saying, "I don't know if we're going to call this lee-go or lie-go, but this is the acronym." I learned from Peter a lot more, because he had worked especially with Rai Weiss in MIT, before spending a year in Colorado. Having been at MIT with Rai Weiss for several years, he knew all about not just LIGO at MIT and Caltech, but also some of the rivalries between Caltech and MIT. And some of the personalities involved.
Gaby, did you see your thesis research as being specifically relevant to the LIGO collaboration? Before-- I mean, as you were considering postdocs.
Yes, yes, definitely I saw my research relevant to LIGO. In fact, as a graduate student, I attended conferences and I talked about my research in these conferences. Some of these conferences were specific to gravitational waves and detection of gravitational waves. These were early 90s. [laugh] The field was not that large, but all these people were very interested because thermal noise was actually a very hot topic at the time. My advisor, Peter Saulson, was the one who said that the predictions that had been used early in the design of LIGO could actually be wrong, if the source for the dissipation, in the fluctuation dissipation theorem, came not from viscous dissipation, which is what's usually assumed, but from what's called structural damping, which is due to the loss of energy in the material. The material of the mirror, the material of the fibers, and the prediction from that mechanism was very different. So that's why my experiment was important, because it showed that, yes, if you had this kind of dissipation, and the prediction of thermal noise was different than if you assumed viscous dissipation.
What were your impressions when you got to MIT and you joined the collaboration?
When I joined MIT, it wasn't a collaboration, it was a project led by Caltech and MIT. The collaboration was funded later in 1997. There were several groups working on this. There were conferences where people attended from different groups. For example, the Colorado group, where Peter Saulson had been, but also Stanford, LSU. LSU had a group on detection of gravitational waves with bar detectors that they were interested in this. They had proposed the Livingston site to build one of the LIGO Observatories. But when I went to MIT, when I went to interview at MIT, I was just, oh, so...
I had never been in a lab like that, and it wasn't all shiny. On the contrary, the gravitational waves group was in building 20, which is now demolished, but it's historical. It was built in the Second World War [laugh] for radio research. So it is that old. And it was built as a temporary building in the 40s. So, it was not a modern building at all. [laugh] But so many things were happening. There were two experiments, two prototypes that were going on. There were students, there were postdocs, there was Rai. There was Peter Fritschel and David Shoemaker, who had been in Europe. And we're there and it was all about creating this instrument, creating it for real. It was not designing but building for real this interferometer.
At MIT, I worked on a highly precise instrument, not LIGO, but a prototype to test a technique called “power recycling” with smaller suspended masses. There was another prototype to test ways to align the systems. I didn't work on thermal noise at MIT, I worked on other things. But there were so many things going on, and everybody was so enthusiastic about it. It was transformational.
Now was Rai officially your postdoc advisor?
I was hired as a research scientist, and Rai was the group leader. But I never asked who my supervisor was. [laugh] I was not very good at, in general, at paying attention. And I'm still not paying attention to paperwork. In fact, I didn't get my retirement benefits, that I didn't realize until a year late. [both laugh] Rai Weiss and Peter Fritschel were the people I interacted the most with.
Yeah, yeah. And what, as you say, there was so much going on. Everybody was so enthusiastic. What did you slot into? What project did you contribute to when you got to MIT?
Well, I was mostly working on a prototype that was testing power recycling with a suspended interferometer. For detecting gravitational waves, we essentially use a Michelson interferometer. A laser is split in two with a beam-splitter, the beams travel in perpendicular direction, and we measure differences in distances between the center and the ends. Power recycling was Ron Drever's idea at Caltech, although I think he had this idea before going to Caltech. The idea was that if one operates the output the of the Michelson interferometer at what we call the dark port, meaning that the two beams coming out of the detector have destructive interference, then the beams that go back to the laser have constructive interference. So the Michelson detector is acting like a mirror. And if one considers the Michelson interferometer as a mirror, because it's sending the light back, and one puts in front of that a partially-transmissive mirror, one can make the light go back and forth and that increases the circulating power, the number of photons that are used to detect the gravitational waves. And that is called power recycling. Recycling the power of the laser.
And so that technique had been tested totally on prototypes, but on optical tables with mirrors that were bolted to optical tables. But for LIGO, the test masses, the mirrors, need to be isolated from the ground, so they are hanging in pendulums and that introduces all sorts of technical complications, apart from other noises like the seismic noise moving the mirrors differentially and thermal noise. And so the need to test this powered recycling technique, when the mirrors were hanging in pendulums, had not been done, and that's what we did at MIT, with again Peter Fritschel, who was the main scientist leading that experiment. I was a postdoc. There was another postdoc that was working in this other experiment I was talking about, testing the drastic alignment. The student there was Nergis Mavalvala, who's now a dean at MIT.
In our power recycling experiment, I worked first with a graduate student, Partha Saha, using a big laser that was an Argon-ion laser, because that was the laser that was going to be used in LIGO. At the time, the solid state lasers were just being developed, and LIGO then decided that using infrared lasers, solid state infrared lasers, was going to be a lot more promising, which was the right decision, but then it meant we had to test this prototype again with a different laser. [laugh] Because all these techniques depend a lot of reflectivity of mirrors and wavelengths, so we had to do it all over again, so I did the experiment twice with another student, another graduate student, Brain Lantz. And so there were two PhD thesis on this.
So Gaby, this was not a postdoc. You were a full-time employee? There was no end time to your employment at MIT?
There wasn't, no. At the time-- actually, when they first interviewed me, it was for a postdoc position, but then they had a scientist position and they offered that to me. At the time I went to MIT, that was in '95, my husband was a professor in Penn State, and so we were looking for something permanent together. [laugh] So I do know that I never thought about MIT as permanent, we were looking for things together, so for us neither one nor the other were necessarily permanent. In fact, we spent six years in different places.
The two-body problem. Difficult.
Yeah. And we have another anecdote. [laugh] Jorge, when he finished his postdoc at Syracuse, he went for another postdoc in Utah, in Salt Lake City, at the University of Utah, and I was a graduate student in Syracuse, so that was a very long distance. That was very hard. We had to fly, that was expensive. Then in 1993, he got a position as an assistant professor in Penn State, so he was in Penn State, I was in Syracuse. We could drive back and forth to see each other almost every weekend. But then when I moved to MIT, that was too far away to drive for a weekend. So we actually bought a camper and parked it halfway [laugh] in a campground, and then we both would drive about four or five hours to meet there almost every weekend.
Gaby, I'd like to ask your perspective while you were at MIT, and to bookend your experience... well, to centralize your experience with the bookend of the amazing story about how Rai started this project decades earlier, and the dramatic -- not the conclusion, but a crescendo of the detection, you know, almost 20 years after you had left, right? During your time at MIT, did it feel closer to 2015 or did it feel closer to 1975? Looking back.
[laugh] Oh, I think it was certainly closer to 2015. We were building, we were not planning. We were building LIGO.
And how well-formulated was the goal of the project? In other words, was everybody working toward this singular goal of actually detecting a gravitational wave, or was the research more broadly conceived than that?
Of course, the long-term goal was detection of gravitational waves. But the promised goal was achieving the precision of these instruments. It was building an instrument that could measure strain of ten to the minus 21. That was almost the magic number we had in mind. And I think that we didn't know that it was going to take so many years. We began realizing that when working on prototypes. There was a 40-meter prototype at Caltech, and experiments that might have taken six months, that looked like a PhD thesis, would instead take a couple of years. So maybe we should have extrapolated from that, but we didn't. We were going to build this detector. That was going to have this precision. We didn't know it was going to take that long, but we knew we were going to get that precision, and the goal was to get that precision. Of course, we wanted to get that precision to detect gravitational waves. But if we go that precision and better, and we didn't detect gravitational waves? Well, it would have been a very big disappointment. But we would have reached our precision goal.
How did the opportunity come together for you at Penn State?
Well, of course [laugh] people at Penn State knew that Jorge had a wife that was a physicist. But they were also interested in gravitational waves. They have a physics department and an astronomy department, and they have people in the physics department that were doing other things in gravity theory. LIGO was still a very new field. There were very, very few groups in the US, especially experimental groups. There were beginning to be more groups doing gravitational wave theory. Some those groups began doing data analysis, but very few new groups that did experiment. The groups had already been established: Stanford, Colorado, LSU. But Pen State was interested in gravitational waves, and they opened a position that was general. It was a bit more general than gravitational waves, and I applied. They interviewed several people, and they made me an offer. [laugh]
Gaby, had you done any teaching at the university level up to this point, or would that be a new adventure for you?
It was a new adventure. I had been a TA as a graduate student in Syracuse. I like teaching. I always liked teaching. Even when I was in high school and then in college, I did a lot of tutoring. Even as a girl, I played being a teacher with friends in the neighborhood. They didn't like it as much as I did... [laugh] But it was a new experience, especially with US students, because I didn't know much about their high school preparation or even the format, in the college format was all different than the college I had gone to. But I loved it.
Was the intention for you to stay active, current with your former colleagues at MIT?
Yes, yes, definitely. I was going to set up a lab in Penn State, because that was something that they were very keen on, that I procured a grant to set up an experiment that would continue to LIGO, but it would be an experiment in Penn State, and that was a lab where I set up prototypes—where I was going to set up the prototypes of multiple suspensions, double suspensions, that we were going to use in a second generation of LIGO. But at the same time, and I asked before I accepted the position at Penn State, is that I also wanted to travel to the Hanford Observatory and to the Livingston Observatory once every other month or so.
So [laugh] I did travel a lot to help install the new detectors that were being set up at the time. I did keep in touch, because people at MIT were doing the same thing. And they were going to the observatories to set up these instruments. We kept meeting and then in '97, there was a meeting at LSU to start the collaboration, the LIGO scientific collaboration. This was something Barry Barish had insisted was needed to make the field grow in collaborative fashion. In '97, there was a meeting at LSU where different groups proposed associating to the LIGO projects as members of the LIGO scientific collaboration. I did that from Penn State.
Now, were you involved at the site selection? Do you have a sense of why LSU was chosen?
It wasn't LSU. It was Livingston.
Right, that's what I mean.
I was not part of the site selection at all. That happened probably before I even came to the States. As I understand, it happened when NSF decided to invest in this. And since it was a significant investment, they didn't want to just accept the choices that people had made that weren't very convenient, or the most convenient thing. They were good sites, but more convenient for people at Caltech and MIT. But NSF decided to make a national search for sites, so they accepted bids or studies from different sites, and it was people at LSU who studied the site at Livingston who proposed this site with the help of the state of Louisiana. But that was one of I think around 20 sites. I've never seen the list, the complete or even incomplete list.
And then there was a group of scientists that, as I understand again, it was Rai Weiss and Kip Thorne and others that looked for good pairs of sites, because you wanted two sites. You wanted them as far as possible, but not aligned with Europe (there was a detector being built in Europe, so you could triangulate). There were several things, apart from noise and big facilities, internet, well internet was not as important, but [laugh] data transfer and things like that. And then there were several pairs that were considered good, and then it was NSF looking for Congress support that made the final selection. That's the story I've been told, but I was not part of that search.
Administratively, what exactly was the relationship between Livingston and LSU? How would they work together?
The proposal was that the state of Louisiana would buy the land needed, which is not a square four by four kilometers, but it's just an L-shaped land from a logging. The state would buy this land and then give it to LSU, which is a state university, to lease it to Caltech. The people at Livingston Observatory are all employees of Caltech. They don't have an institutional relationship with LSU. However, at LSU, being so close of course we adapt our research to things that many things can be done, or best done, at the observatory, so we contribute more than other groups, and of course we are more useful to LIGO than people from other universities, in that sense. But it's not an institutional relationship. Although I should mention that the observatory head of the LIGO Livingston Observatory is Joe Giaime, who is a physics faculty member at LSU. So in that sense, there is a relationship.
And what is your sense of the relationship between Hanford and Livingston? Is it designed to be redundant or are there actually different things that happen at each site?
They were designed to be not redundant, but as identical as possible. [laugh]
Yeah, I don't mean redundant like one is not necessary, but if they're both detecting the same thing, they can provide confirmation of a detection.
That's right, that's right. And that's the way they were designed, installing identical technologies at two places, so they have the same laser technology, the same suspensions, the same everything. Also, they are L shaped and they are oriented as parallel as possible -- of course, they are not parallel because they are on a curved Earth, and they are far enough apart that they are not parallel, but they are oriented so that they can be as parallel as possible, and that is not necessarily best for science.
Gravitational waves have two different polarizations. In one polarization, for an L-shaped detector, it would make one arm shorter and the other arm longer. The other polarization would do the same, but for an interferometer rotated 45 degrees. So the two polarizations are 45 degrees apart. If you have two detectors that are parallel to each other, with the arms parallel to each other, then they detect the same polarization, which means you cannot resolve polarization, [laugh] you cannot tell the polarization of the system. You get more information if you have them oriented differently. But what was needed was solid confirmation, because gravitational waves have never been discovered, and ten to the minus 21 is a very small number. It could be produced by many different things. So you needed to have as similar a wave form as possible in the two places. That’s why the LIGO detectors have the same orientation.
Gaby, it begs the question, if two sites are better than one, would three sites be better than two?
Oh yes, yes, definitely. And that's why we work together with the Virgo detector in Europe, because that's how you can triangulate. When you detect the same thing with two detectors, you don't detect it at exactly the same time. There is some time difference because the gravitational wave travels at the speed of light, so depending on where it's coming from, there will be some time difference. But if you only have the time difference between two detectors, then your localization is a ring in the sky. It's a thick ring, because it depends on how much noise you have. So that's not very good localization, even if you use some extra information, it's a good fraction of a ring.
But, if you have three detectors, now you have three time-differences and you can triangulate the source. And that's why we are working with Virgo together and like I said, because there are two polarizations, even more than three detectors is even better, because then you have more information than just on localization, but on polarization. And that's what you need to do astronomy.
Now, when you got to Penn State, did you know how focused you would be on Louisiana? Or this is a responsibility that sort of grew over time?
No, I didn't know. I went to both places, to Hanford and Livingston equally, so I wasn't focused on Louisiana when I was in Penn State. It was only in 2001 when they had an ongoing search for a chair in theoretical physics at LSU, and that's what they considered my husband for, and of course it was very close to Livingston, so in the beginning, when we were considering this, it wasn't that I was going to be a faculty. It was not necessary that I looked at a faculty position at LSU. I was thinking about maybe being a full-time scientist at the LIGO Livingston Observatory, a research scientist, because I was—even though I loved teaching, and I still love teaching—between teaching and the lab at Penn State, and the travel to the sites, I was not having a good time balancing all these things.
Gaby, what was the experience like, building that lab at Penn State?
Like everything, it took a lot longer than I thought it was going to take. [both laugh] But it was a great experience. I mean, I had not built a vacuum system myself, and I did. I had not supervised a postdoc myself, I had just been a postdoc myself, even though I was a scientist, a postdoc is not very different. So, it was a great experience. Especially when I began recruiting graduate students that were also very interested in working with me. And I'm very proud of them, even though I moved from Penn State before they graduated. Two of the students that I recruited then are still working in LIGO. They finished their PhDs at Penn State.
Did you take graduate students with you to LSU?
I didn't. They were already set up on their studies, and although I offered the possibility of course, they preferred to stay. But we kept in touch. We are still friends with both of them. And both of them are women, too. [laugh]
Now, when you got to LSU, was the plan, as you said, at Penn State, you were equally focused on Hanford and Livingston. Just by virtue of being so close to Livingston, were you more focused on Livingston when you got to LSU?
Oh definitely. Definitely. Once I got to LSU, I could drive to Livingston, so I flew to Hanford a lot less often. I focused on Livingston. And Livingston had more problems, too. Because even though this wasn't known at the time [laugh] at the time that the sites were chosen, the seismic noise in the band of gravitational waves frequency is actually smaller at Livingston than at Hanford. But in order to keep all the systems operating, one needs to push mirrors against large motions at low frequencies. At much, much lower frequencies. The ground moves and moves most with periods of several seconds. And that's called a micro-seismic noise. And at Livingston, because the soil is softer and it's closer to the coast, it moves a lot more, so one needs to push a lot harder, which introduces more noise. So, the interferometer at Livingston took a lot longer to get operational than the two at Hanford. In the initial LIGO project there were two detectors at Hanford.
Now this is sort of a broad administrative question, but you have Caltech that is running Hanford and Livingston, and you have MIT. How is everybody working together? How is the sort of overall direction of LIGO since it is all geared towards detecting gravitational waves? How is it all sort of working together to make sure that everybody is doing what they're supposed to do, where they are supposed to do it?
Well, you should ask people at Caltech and MIT now. [laugh] I keep in touch for them, but I'm not sure I can talk on their behalf.
But I think it has been the merit of very good executive directors. People who have defined the responsibilities of everybody in the project very well. So it's not that the more important things are done in one site or another. There are four LIGO laboratory sites. Caltech, MIT, Hanford, and Livingston. And there are different responsibilities for each site, but all of those are equally needed and equally critical. And they know that. Everybody knows that. So, I think that's how it works well.
Of course, we always-- I mean, I say 'we' although I'm not in any of these groups now. I'm now a more regular member of the collaboration. But we take advantage of certain competitiveness between Hanford and Livingston, between the institutions that are not Caltech and MIT, and Caltech and MIT, but that is only healthy competition. We all know that it's working together when we get the research done. And we have proven that.
Now, when you got to LSU, just like at Penn State, did you build up a lab? Or just by virtue of having a proximity to Livingston, essentially that became your lab?
Well, I should say both. Actually, I was given lab space and I moved equipment I had from Penn State to LSU. But I didn't set it up, because at the time, I thought that working at the Observatory was a lot more important. And LSU actually allowed me to do that, which was not very, I think, it was not very typical from universities. But LSU not only allowed me but encouraged me to be at the site, at the observatory, really more of a critical part of the operations. [laugh] The attempt to make it work. And that's what I put the priority in, and then I never went back to the lab. So that lab space is used by my colleague, Thomas Corbitt, who has a lab for quantum technologies applied to gravitational wave detectors now. But I never used that lab space.
And Gaby, at what point did you feel like your both administrative and your scientific interests were changing as a result of this new arrangement at LSU?
It wasn't, no, it wasn't. I said that when we talked about moving to Louisiana, I was thinking about a scientist position at the site. But LSU actually was interested in having a faculty, another faculty, because they already had Joe Giaime, who was actively working on seismic isolation, which is another of the topics needed for LIGO. There was already one faculty apart from the two other more senior faculty who had proposed the site, who were working with this other technology, a bar detector called ALLEGRO. I would be a fourth faculty, but the second one in gravitational waves—and they actually asked me to apply for a position in astronomy, but I had never thought of applying because I’m not an astronomer, but at LSU, the department is a department of physics and astronomy. And because these are related to gravitational wave astronomy, again, LSU was incredibly supportive because at the time, nobody believed the gravitational waves as an astronomical field.
But they had been involved with proposing this LIGO site at Livingston, so they asked me to apply to that, and I did. I interviewed, and they made me an offer, and I had to think hard about it, because like I said, I had had some trouble balancing working at the site and teaching. But because they said, “Well, it's okay if you don't work on the lab at this time,” then I accepted, and I'm very, very glad I did. I did very well.
What changed for you as a result of accepting this opportunity?
Well, I recruited students at LSU. Of course, there are students from Caltech and MIT that go to do their thesis research for a year or two to the sites, but I could recruit new students myself at LSU that would also work at sites, but with more independence. So I was more independent, they were more independent. But we all worked together. I made very good friends. Still, I'm very good friends of all the scientists on-site, but I'm also collaborating with many other scientists outside the observatory. So being a faculty member at LSU, rather than a scientist at the observatory, I think gave me a broader perspective.
You gave a really good overview of some of the technical challenges while you were at MIT. At this phase in your career, what were some of the key technical challenges you were facing?
At this moment, in 2021?
No, no, no, back when you started with this new appointment.
Oh, oh, well at the time, it was actually making LIGO Livingston work. I had been to LIGO Hanford and while in Hanford, the two detectors there were operating in the sense that the interferometers, the masses in the interferometer could be controlled with power recycling. At Livingston, they couldn't because the seismic noise was too large. So I was involved in solving this problem Joe Giaime was involved in installing some more advanced techniques for seismic isolation, and I was involved in how to have more reliable ways of aligning the systems, aligning the mirrors at the end of the 4km arms and at the corner so that they would be facing each other.
And all of these—most of these activities—had to be done at night. The large low-frequency ground motion happens night and day, but during the day, there's also more noise due to human activity, even though this is far from cities, but there's still traffic. Nights were quieter, so we worked nights. It was all work during the night. And during the day I was teaching, so it was very hard. I remember very patently at night, it was more than one night, but it was one night where Rai Weiss and I and several others were in the control room at Livingston, looking at the images made by cameras and looking at the signals in the computer on how the mirrors were oscillating, thinking about how to make them stop. How to make the detector “lock,” as we call it.
And why is that so important, to make them lock?
Well, what we call "locking" is pushing the mirrors, cancelling the noise, especially the seismic noise, to make the light that goes back and forth between mirrors have constructive interference. We have two mirrors that are four kilometers apart in each arm, and we make the light go back and forth between those, which is like making the interferometer longer. That also gives us more sensitivity, but that happens only if the mirrors that are at the right distance and they're only at the right distance if we push them. We also need to keep the beams that come out of the interferometer in destructive interference so we can use this power recycling. We need to have this power recycling mirror at the right place so that the light can go back and forth between the Michelson mirror and the power recycling mirror with constructive interference.
So, everything needs to be pushed and all the mirrors need to be facing each other the right way... but they want to move and tilt. So, the control system of the detector is very, very complicated. And that's what needs to be tuned and it takes years to not just tune it, to make it operational, but also to not introduce excess noise. You know, to push, but not introduce noise.
Gaby, when did the site become viable, scientifically viable so that you were actually spending more of your time not being involved in getting it up and running and operational, but that it's ready to do what it was built to do and now it's just a matter of detecting the gravitational wave? When did that happen chronologically? How long, and then when did you feel like it's reached that point of maturity or operational viability?
If I have to be honest, I don't think we're there yet. [both laugh]
What was the Nobel Prize given for, then?
[laugh] Yeah, but it's not like a detector that you can leave it running and then it's doing what you want. We don't have the sensitivity yet that we wanted. From the beginning, there were these two phases, Initial LIGO and Advanced LIGO. Like I said, when I was in Penn State, I was already thinking about these double suspensions we were going to use in Advanced LIGO. So we were always thinking about these two phases. But in the initial LIGO detector, the goal was reaching that ten to the minus 21 sensitivity, which was actually ten to the minus 21 in the noise. We wouldn't be able to detect the 10 to the minus 21 signal. At least not clearly, but reaching that sensitivity was important because that was what in principle what we had promised and thought we could get with the hardware that we have installed, with mirrors in hanging in single pendulums and passive seismic isolation.
Technology that seemed relatively simple but was so complicated because of all these complexity in keeping the mirrors in the right places. And we reached that in 2005. Installation was finished around the year 2000. It was only 2005 when we achieved the sensitivity that we wanted. We took data for a year. And we analyzed that data. And only then, the Advanced LIGO budget was approved. Because that's what they wanted as proof.
Of course, we were hoping to be lucky and detect gravitational waves, but we knew that was very unlikely, and in fact, during all that time, more and more was known about the frequency at which neutron stars coalesce. Very little was known about coalescence of black holes until we detected them, but about coalescence of neutron stars, more and more was known, and it looked like the new predictions were not more optimistic than they had been. If anything, they were to be more pessimistic. So, it was clear that we would need this second generation. It was only approved in 2008. And it began installation in 2010. [laugh] In 2014 was when we locked the advanced detectors. In 2015, we had already planned this first observing run, but we did not have sensitivity, enough sensitivity to detect gravitational waves from neutron stars. We were not expecting any detection, although I'm on record saying, “We don't know about black holes.” [both laugh] That's what happened. But again--
Gaby, can you talk a little more about the creation of Advanced LIGO as an administratively separate endeavor from LIGO? Or do you not see it like that?
I can’t see it like that. No, it is the same people. It was a different budget, different technology. So administratively, and this is something that of course people at Caltech would know best. The initial LIGO was a cooperative agreement to build LIGO and take data. And that agreement was renewed every five years. Advanced LIGO project was a certain amount of money to buy and install the hardware for Advanced LIGO, that was approved in 2008. But then once that was done, then Advanced LIGO is done. Now there's only one NSF cooperative agreement with Caltech that is renewed every five years to operate the observatories.
We have not reached the sensitivity that we want with the hardware that has been installed in Advanced LIGO, but already there has been a different budget approved for some other improvements that we call A+. So, as I said, this is all a process of improving sensitivity of the detectors. We have been taking data, we have had three observing runs since 2015. And we have not yet the sensitivity we want, but we have plans to even install new hardware to improve sensitivity and I think it's going to be a cycle like that. It's not going to be, we are done, we are detecting gravitational waves. I don't think we will ever be done.
[laugh] Gaby, it's such an interesting perspective, as you say in all honesty, we're not there yet in terms of the operational viability of the sites. So, I know you're much closer physically and professionally to Livingston, but would you say that that's true for both sites? Is that sort of a universal statement about where each site is?
Yes, yes. Yes, it is. Especially now in Advanced LIGO, the detectors, because of the experience in initial LIGO, in Advanced LIGO they are a lot more similar now. There's only one detector in each observatory. In initial LIGO, there were two detectors in Hanford, one in Livingston, and that made life a lot more difficult for Hanford. In Livingston, the problem was seismic noise, in Hanford it was that there were two detectors. [laugh] So you always have to take care of two-- [laugh] of twice the number of problems. But now there's only one detector, so it's a lot more similar at both observatories. And the challenges-- and the detectors are not identical, and people of course are not the same, so the priorities sometimes, and what is the problem that they are working on, are not the same, because you want to work with problems in parallel. Not in series. So the challenges are a bit different, but they are at the same stage, yes.
So Gaby, to set the stage, we're roughly in 2010, 2011. To work up to the gravitation four or five years down the line, right? If what you're saying, which is so interesting, that even today, we're really not there, is that to say that in the years, months, weeks, leading up to the detection, that it came as a surprise that you weren't particularly expecting it because as far as you were concerned, the labs were not where they needed to be? I wonder if you can sort of narrate how that felt in real time.
Yes, yes. Although maybe I can start with saying that this collaboration that was founded in 1997 was first led by Rai Weiss as spokesperson. The spokesperson is the scientific leader of the collaboration, which is different than the executive director of the LIGO laboratory. That is a person that organizes and pays the salaries of everybody working at Caltech, MIT, Hanford, and Livingston. But the collaboration is a collaboration, international collaboration of people who agree to, not to pool financial resources together, but to pool work together. So that was proposed by Barry Barish in 1997. Rai was the first spokesperson, then Peter Saulson in 2003, then Dave Reitze in 2007. I was elected spokesperson in 2011, right after we finished with initial LIGO. In 2011, we had just finished taking data with initial LIGO, and we were beginning to install Advanced LIGO.
We called that time the “dark years” because we were not going to have data to analyze. However, we took a long time to analyze the data we had taken. But we were also preparing for Advanced LIGO. In 2011, there was also a need for reorganizing the large collaboration in different ways, with Barry Barish led that process too. But by 2013, we began preparing for the Advanced LIGO era. And we said, while we are going to detect gravitational waves at some point, we don't know when, it's probably going to take several years, but we are going to detect gravitational waves, so we need to prepare for that.
Gaby, if I could just interject. Why the confidence that sooner or later you will? If the labs aren't where you need them to be, and all we have are gravitational waves as a theoretical proposition, going back all the way to Einstein, how do we know that it's not going to be like Blas Cabrera and the magnetic monopole that never happened, right? Why the confidence that it's going to-- that the detection is actually going to happen sooner or later?
Well, the fact the gravitational waves existed had been proven already. They had not been detected, but their existence had been proven because the prediction is that two objects that are orbiting around each other, like two stars, would emit gravitational waves, lose energy and get closer together. And that had been observed with two neutron stars, and that was a 2003 Nobel prize. [laugh] So gravitational waves existed. I mean, some people said that that was a prediction of Einstein that hadn't been proven, but we didn't believe that. We believed fully in the existence of gravitational waves.
What we didn't know was how often they happen and with what amplitude, and for that, we counted on the knowledge about the binary neutron star systems, like the one that was used for that 2003 Nobel prize. Those systems are known to exist in the galaxy. They are going to merge in hundreds of millions of years, we are not going to wait for those. But from that number of binary systems and the number of neutron stars known in the galaxy, one can extrapolate how many binary neutron star systems there are in each galaxy, how often they take and coalesce, so the prediction is that there is one coalescence like that every 10,000 years or so in each galaxy. If you can measure the noise in your detector, you know what's the amplitude of gravitational waves you can detect larger than the noise. It’s a relatively simple calculation to translate detected amplitude of gravitational waves into the distance at which they were emitted.
So if you know you can see up to this distance, then you know how often you're going to see gravitational waves. This prediction for the sensitivity of initial LIGO said that detection would happen once every 50 years or so. So, we didn't detect anything, but we would have had to be very lucky. With Advanced LIGO, we also knew it was going to take a while to get the sensitivity we needed to expect one detection a year of merging neutron stars. Which is about the sensitivity we have now. But not the one we had in 2015.
If I can revise the question, if there was the strong theoretical basis plus the 2003 Nobel prize, if there was that strong basis that gravitational waves are out there, where is the confidence that you have that LIGO has built the right tools in the right way to detect that?
Well, it was the best tool. [laugh] It is still the best tool for the price, of course, I mean you could have made it longer, and then they would have been more sensitive. But that would have been a lot more expensive. So at the time, it was a trade-off between the cost and the sensitivity that you needed to expect detections, even if not many detections. And I think that was a problem, astronomers especially didn't think that this was an astronomy project. They thought this was a physics project. That it was about the detection of gravitational waves, not detection of systems emitting gravitational waves in general. Now of course they have changed their mind, [laugh] especially because we have detected so many black holes. But not only black holes, we also detect the coalescence of neutron stars already.
So Gaby, let's bring it right up to 2015. Set the stage for me. Is there advancing optimism that LIGO was getting closer and closer, or did that detection, was it a surprise, and this was a really dramatic, out-of-the-blue event? What was your experience as these things were unfolding?
Well, like I said, we were preparing for a detection sometime in the future. And we actually had begun seriously preparing for that, having what we called a detection plan. What would we do? What confidence we needed to have? What statistical confidence we needed to have? We were going to have a red team, a detection committee, that was going to try to poke holes, to look at all the angles. [laugh] If we had a detection, we actually talked about what journal we would publish in. We talked about waiting for peer review before making any announcement. So we had been talking about all these things, but with the expectation that this was for years in the future. We needed to do it and I was really insistent that we needed to do it before the first observing run, because I was afraid that if we didn't do it, then we were going to just set it aside.
So luckily, we actually decided on the journal and the wait for peer review and all that, we approved that detection plan in September of 2015, [laugh] before we were going to start the first observing run at the end of September of 2015. The detection appeared on September 14, before we started taking data 24/7. It was a huge, huge surprise. And not only because it was a detection, but because it was a strong detection. It was a very large amplitude. Even now, after having published 50 detections, this one has the record of largest amplitude.
The first one.
The first one has the detection of largest amplitude, yes.
What besides serendipity do you think explains that?
Nothing. [laugh] Nothing.
And Gaby, just so I can understand, so are gravitational waves always out there to be detected, and it's just a matter of being at the right place at the right time? Or is it more like, a comet that you only see every once in a while? How does that work in terms of the regularity of being able to actually detect gravitational waves?
I like to compare gravitational waves to neutrinos, [laugh] rather than supernovae or comets. We know that neutrinos are all the time traveling through us, traveling through the air. It's just so difficult to detect them, because they don't interact. The same is true with gravitational waves. They are going through all the time. There are black holes and neutrons stars and white dwarves and all kinds of stars and exploding and binary systems coalescing, producing gravitational waves. The waves are in general just too small to detect.
Now when you say they don't-- like neutrinos don't interact, is that to say gravitational waves don't interact either? Or that we don't understand how they interact because we've not yet figured out how to integrate gravity into the Standard Model?
No, no, no, we know how they interact. We know how neutrinos interact too. They just interact very weakly, I should have said. They do interact, because if they didn't then we wouldn't know about them.
And they interact very, very, very weakly. And the same is true with gravitational waves. The effect they have, this stretching of distances, is so small that what you need is not to be there at the right time, but have enough sensitivity. If you have an instrument with the right sensitivity, then you will see gravitational waves. How many gravitational waves depends on your sensitivity.
All right, so let's pretend we're doing a movie, and it's the moment. How much drama is there? Is it the person sitting in the control room and nothing's going on and it's just sleepy, and then all of a sudden, “Peak amplitude! There it is! Get everybody on the phone.” Or is there a build-up of drama in the days, hours, minutes, where there's a sense that it's coming and then there it is? What's happening in the moments that the detection happens?
Well, it's different for the first detection and the ones that followed, in the second and third observing run. That first detection appeared on a computer. We were not routinely taking data yet. We were preparing to take data, but we were doing diagnostics and calibrations of the detectors. Sometimes, especially at night, detectors were left alone, and they were operating. Meaning mirrors were in the right place, all optical cavities were resonant, detectors were locked. And if they were locked, both at Hanford and Livingston, then the algorithms to detect gravitational waves were being tested on that data. We call that an engineering run. And one of the algorithms being tested on that September 14, and in the previous days before then was an algorithm to look not for coalescence of binary systems, but to look for unmodelled transient signals. And every time one of these algorithms worked, they produced web pages that have a catalogue of candidates.
And of course, most candidates are insignificant, statistically insignificant. But there was this candidate that was found by this algorithm, and then people in Germany and in Florida--they woke up very, very early in Florida, and people around noon in Germany looked at this thing. Note at the time that it happened, the signal had been already passed two or three hours earlier. They looked at this candidate and they said, “This is just really big.” It seemed to be very large. I mean, you could see the filtering and in a spectrogram, you could see it with your eyes. It looked incredible, and they thought that it was a test, because we had been injecting gravitational waves, pushing the mirrors with simulations, because we do that all the time to test the system.
So they thought it was that. They called the observatories to ask if that was the case. They said, no, it wasn't. But we also had been planning to do something we called blind injections, where we charge a small team of people to test us injecting simulations without telling anybody what they injected, or whether they injected anything. We call that blind injections. And we had done that in the past in the last two data taking runs in Initial LIGO. So, emails began floating. I got text messages that woke me up saying—
So you were at home, you were sleeping at this point?
Yeah. Yeah. I mean it happened at 5am and this was discovered at 5am local time, and it was—emails began floating like an hour or a couple of hours later, and I woke up to this. I received these text messages like, “Who put this there?” Everybody thought it was a blind injection. It took about a day to realize that at least it wasn't a planned injection. It could be a hacked injection, and we had to worry about that. But by the next day, we knew that it wasn't a test. It wasn't a drill. [laugh] But it was so big, and from the frequency we could also tell roughly that these were not neutron stars, but black holes. And big black holes that were not known of that size. So, this was all incredible. The size, the fact that...
So, these are discoveries within discoveries that are happening right now?
That's right. It was a gravitational wave of large amplitude. It was a binary system of black holes. No black hole binary system had been known before. People thought they existed, but they had not been seen, because they're black, they don’t emit light. And they were black holes that were very large. 30 solar masses, the largest one known was 20 solar masses. So this was too good to be true, too incredible. We first had to make sure that it wasn't a glitch in the instrument that looked like this and happened to be appear at both detectors the same time. So everything froze at both detectors.
Everything was left in the place and we took data for a few days to make sure that this was not happening all the time. But then, when we realized that that wasn't the case, we decided that we needed more background. We needed more time, more data, to get the statistics that we needed. We had not taken data at any time in the past with these detectors. We didn't know what the noise looked like. We didn't know whether things like this happened by chance, once every month... we couldn't know. So we decided that we had to take at least a month of data to get 15 days of real coincidence. We did, and it was by the end of October that we analyzed that month of data, and by this time, yes, the tension had been raising.
There had been rumors all over the place that we had detected the gravitational waves. But we weren't sure. And we had to be sure. I mean, there had been false positives about gravitational waves twice in the past. I mean in the ‘60s, Joe Webber had claimed seeing gravitational waves, he did it later too, and that was not reproduced. And then with the BICEP-2 experiment in 2013, they had claimed that they had seen signatures of the gravitational waves in the microwave background, and that that was not confirmed, was later understood to be signature of dust. So we just couldn't be wrong. [both laugh]
And we couldn't be sure, until we really had all the statistics, and then after that I think we really knew that those 30 solar masses were the right number, because that was all so difficult to believe. It was end of October when we finally convinced ourselves that there was statistical confidence. But it took another two or three months until we had the parameters, the analysis for getting the parameters, all confirmed. In January, we sent the paper for review, and we received it shortly after. And we made the announcement on February 11th, 2016.
Gaby, to go back to the administrative question about the overall direction. You have the two sites, you have Caltech, you have MIT. In this moment of intensive communication, all of the emails and the phone calls going back and forth with this statistical analysis, what was the basic division of labor among the sites and the universities in terms of analyzing this to confirm that what you saw was actually what you saw?
The two biggest computer clusters where the analysis was done were at Caltech and in Germany. The groups that did the analysis were in many different groups around the world, most in Europe and the US. But everybody was involved in asking questions. Like I said, we had set up a detection committee that was acting as a devil's advocate. So there was that going on, and there were people from many different institutions doing that. Stan Whitcomb from Caltech was the chair of that committee. We were also working with Virgo together in this, this was the LIGO collaboration and the Virgo collaboration working together. It was Fulcio Ricci, spokesperson of Virgo and myself making all these difficult decisions. We actually appointed what we called the paper team. This had been part of how our detection plan-- a paper team of six people that would be in charge of drafting the detection paper. There were also lots of other papers that were drafted with details.
Each of those papers has like 20 people writing the paper and doing analysis. But this paper team were six people who were not the experts on the analysis itself. We actually made a decision of having people who would write a very clear paper, and giving the right balance between describing the complexity of the instrument and the importance of the discovery and the astrophysics. There was a committee chair from Virgo and a chair from LIGO. The chair from LIGO was Peter Fritschel, and from Virgo was Pia Astone, from Italy. And they did a tremendous job. The number of opinions—they set up a computing system, a git hub system as we call it—to receive comments on the paper, and they had hundreds and hundreds and hundreds of comments on every single word, on every single paragraph, on every figure. So, everybody was involved in this. At the moment in the LIGO collaboration, we had more than a thousand members, but we probably had about 3,000 opinions on everything.
[laugh] Gaby, in terms of all of the statistical analysis leading toward confirmation, was there anything in the analysis that gave anyone pause? Like maybe we didn't see this? Or was it all about confirming this from different angles, different perspectives?
It was a bit of both, but we were actually really, really hoping that there would be another detection, and that wasn’t true in the analysis we did in October. Because you know, the natural thought is that if this happened without us even trying, then that means that there are a lot, so maybe we should have seen another. [laugh] Like you said, we were terrified of the Blas Cabrera scenario in which this is a one-off. That was very clear that it didn't happen again. And in the analysis that we did, there were hints of another signal, but it didn't have a lot of statistical significance, although later on, we could confirm it. But at the time it was not very strong. We were a bit worried about that. That gave us pause.
But in December, while we were writing the paper and confirming all the parameters of this first detection, the data showed another detection. And that was actually very nice, because it was December 26th. It was 26th UTC, but it was 25th, Christmas day, still in the US. In principle, we were going to stop the first observing run before the holidays. But because of that first detection, we decided to extend it a few weeks until mid-January, and it was December 26th when we saw that second detection, with a high SNR, high signal to noise ratio. It wasn't as strong as the first one, but it was very, very statistically very significant, so we didn't have any doubt. We couldn't announce it on February 11th, because we hadn't finished all the analysis for that one, but we had more confidence in the first detection because of that additional detection.
Gaby, was there debate within the collaboration about whether or not to announce if there wasn't a second detection?
No. No, there wasn't. No. By the time we did the statistical analysis of the first detection, we knew that we had to announce that. Not only because there were rumors all around, but because it was statistically significant. So it was there. It was-- [laugh] we couldn't hide it, we didn't want to hide it. So we didn't doubt about making that announcement. We were, like I said, we were a bit more... we were a bit scared before then, getting that second...which in the end, was the third detection, not the second, because we had another one in October.
Gaby, of course, everybody knows this is going to be a major, major announcement, that everybody has to get correct. What were some of the considerations in terms of confidentiality, protocols to make sure that yes there are rumors, but everybody needs to keep a tight lid on this until it's official? What kinds of measures were taken, and were you involved at all in those discussions?
We were in very, very frequent communication with the collaboration. We sent lots of emails updating everybody of what was happening, what was the analysis, how long it would take to get the statistical confidence. How long it would take to get the parameters right, that we needed to get these things right before we announced everything. That we didn't want to give numbers. We didn't want to say 9.5 SNR. We didn't want to say [laugh] ten to the minus 21. We didn't want to say 30 solar masses. Because those numbers could change, and we didn't want different numbers floating around. That was the main argument. And I think for the most part people just respected that. We didn't strictly enforce everything.
We knew-- I knew and several other people knew, of people who had talked. Members of the collaboration who talked to colleagues who were not in the collaboration about some of these things. But we decided that this had been with enough of a context without leading to more rumors. So we decided not to punish anybody, because we also knew that even though we knew about a few people, there were probably many other people we didn't know about.
What were the various options for making the announcement? Would it be one person in front of a camera, would it be a written statement? What were some of the possible options or weren't there any because there was one clear way to do this?
We had thought from the beginning in this detection plan, about a press conference, we talked that the collaboration leaders would be there. We had backs and forth on who would be on the stage, and where was the press conference, the press release had lots and lots of drafts, and what it said and how it said it, and what institutions were named. There was quite a bit of back and forth on that. There were some people unhappy, you couldn't make everybody happy. That's for sure. But there wasn't a lot...happiness was the main, prevalent feeling the day of the announcement.
Who was decided to make the announcement? How did that decision come about?
Well, of course the proposal came from Caltech and MIT. But the conversations were between the LIGO executive director, David Reitze, the European gravitational director, Federico Ferrini, The Virgo collaboration leader, Fulvio Ricci, and myself. Consulting with other people, but those were the main people, the four people making the decisions.
Where were you on the day of announcement?
On stage. [both laugh] We actually had... we decided on two parallel press conferences, one in Europe, one in the US. Most people looked at the US one, but there were especially people in Europe, of course the European press, who paid attention to the European one. There were LIGO collaboration representatives, and LIGO laboratory representatives in the European one and conversely, there were European Virgo representatives in the US one, although they weren't on stage. On stage, we were four people, me representing the LIGO scientific collaboration, David Reitze representing the LIGO laboratory, and Kip Thorne and Rai Weiss, because they were the pioneers. I mean, they started this in the ‘70s together. [laugh] I was so glad they could both be there. France Córdova from the National Science Foundation was presiding the press conference.
Gaby, when did the Nobel prize buzz start? When did that become something that was unavoidable to think about?
Oh, I think from the beginning. I think it was in January that some people realized that there was a deadline, I didn't know and I really didn't follow these things, but there is a deadline for nominations for the Nobel prize. Only some people are invited to nominate, to make nominations. It's not open nominations. So, some people are invited to make nominations for the Nobel prize, and there is a deadline in February, and I think it was February 10th or something like that. It was before February 11th. So some people said that we were going to miss the 2016 Nobel prize, because of that, and I made a decision, a conscious decision, not to participate in Nobel prize talks, because I know it was a difficult topic, and I didn't want to be part of that.
I know MIT and Caltech were very strong institutions, and that they felt they had good candidates for this. And as I said, I thought that Rai Weiss and MIT had to be there. Had to be there. In 2016, however, I was still the spokesperson of the collaboration, and in October, we were not well-prepared, but we had made some preliminary preparations in case the 2016 Nobel prize was awarded to LIGO. And I have to say, we were lucky it wasn't, because like I said [laugh] we were not very well prepared. By 2017, I was not the spokesperson, when in October when the announcement was made, but by then everything was very well-prepared because we knew it was very, very likely to be the case.
Gaby, of course the Nobel prize goes back to the 19th century when there was no such thing as a scientific collaboration that had thousands of people on it. And so that raises the obvious problem: how do you pick out the maximum number of three from a collaboration in which so many people contributed to the accomplishment? How well do you think Barry Barish and Kip Thorne and Rai Weiss did their best with the limitations that they were working in, to make sure that the world knew that it wasn't just about the three of them? That the detection, the collaboration required so much more expertise from so many more people? In other words, did you feel part of the award? Did you feel part of the honor of receiving the Nobel prize?
I did. I did, yes. In several different ways. First, actually, it was the Nobel Foundation itself. Because in the press release of the Nobel Foundation, if you look at it, they actually have the three names of Barry Barish, Rai Weiss, and Kip Thorne, but the affiliations are not Caltech and MIT. They say LIGO-Virgo collaboration. That is very unusual. I think that had not been done before, even though there had been other prizes given to leaders of collaborations, like for the neutrino Nobel prize or the supernova teams. Those were large teams, they were collaborations. The collaborations were mentioned in the press releases, but the affiliations of the scientists of the Nobel prize winners were institutions. Here, it said LIGO-Virgo collaboration. I always put that in a slide with that when I talk about the 2017 Nobel prize.
I can't imagine that the Nobel committee is easily swayed in changing the way it normally does things. Do you have a sense of who pushed for this unique designation?
No, I don't. I don't. Like I said, especially from Rai, Rai has been somebody who has -- well, of course he was my mentor at MIT, so I am a lot closer to him than to Kip Thorne, but I know them both very well, and I know Barry Barish also very well too, but Rai's… [laugh] he's my mentor. I always mention two mentors, Peter Saulson and Rai Weiss. And he was actually very, very careful to give credit, especially to the people who had worked on the instrument, on the experiment. If anything, I think that some people doing the analysis, which are also younger people, felt they weren't given the right amount of attention. But that has happened later on.
Kip Thorne and Rai Weiss have been wonderful about this. The three of them made invitations to a lot of people for a private party in Stockholm. There were very, very few people attending the Nobel ceremony itself, I was not one of those. But they organized a private party that night and the night before with them to celebrate the Nobel prize. And there were lots of us in there, and they paid for everything. Rai and Kip also gave the money to endow an annual prize of the American Physical Society, the Richard Isaacson Award, for people who have been instrumental in advancing gravitational waves. I know they have nominated people for that award, not just donated the money.
Gaby, the story as you tell it so far makes it so clear that there's a popular misconception. That when LIGO was awarded the Nobel prize, [brushes hands] job's over, we can wrap up and we can go home. But of course, in many ways this is just the beginning. I wonder if you could talk administratively and scientifically, how LIGO changed, Advanced LIGO changed, post-Nobel?
Well, I don't think it was post-Nobel. It was post-detection. [laugh] It wasn't the Nobel prize, it was actually the detection, especially getting from those three detections in the first observing run. For the first time, we could calculate the rate at which black holes coalesce, even though three is a small number, but still, we knew it wasn't zero. And we had three, so we could estimate at least the order of magnitude so we could plan better for the future. And on top of that, we had the number from neutron stars, which actually have a lot more physics. Black holes are very exciting, and you know, there's a lot of strong gravity and space time curvature near the coalescence time, and all of that. But there's a lot more physics and astrophysics with neutron stars, but because they have smaller masses, the waves are of smaller amplitude unless they are very close, so it was known that we had to work really, really, really hard to get better sensitivity to begin to see those.
And that happened in 2017, and again that was completely unexpected, because we didn't have enough sensitivity to expect that. And it happened so close [laugh] that it was really, well, I can't describe it other than lucky, because it happened in the few weeks we were taking data with Virgo, so we could triangulate it. We could point astronomers to look for light from the same origin. Astronomers saw electromagnetic signals from there. In the year, or almost a year, that we took data in 2019 and 2020, we saw at least two other, and we are still confirming, other coalescences of neutron stars, and neutrons stars and black holes, which are yet a different one, but we have not seen any as close as in 2017. These new events have all been farther away, and there haven't been electromagnetic counterparts. So again, the first was the best so far.
So far. And if you had to guess, I know we're not in the prediction game, but is it only a matter of time until there is a detection that matches the peak amplitude of the first one?
Oh yes, yes. Because we are increasing sensitivity all the time.
Is it possible theoretically that there will be a detection even more powerful than the first one? Is there a theoretical ceiling, I guess is my question?
Well, to get larger amplitudes, they would have to be closer. The system would have to be closer or would have to have larger masses. Now closer is not as likely, because the closer you are the smaller the rate, the fewer binary black hole systems that are around. So they will happen, but they will take many more years in between. The frequency of those will be much lower. However now, now that we are taking more data, we have been seeing larger and larger black holes. And larger black holes have larger signals. So, we don't need to see one closer. We just need to see one at the same distance, but of larger masses. That's going to happen any time.
Gaby, as you say, so it's really the detection that changes the collaboration. Is that to say that once you have the detection, there are now new questions that can be focused into the collaboration?
Well, yes. We have changed a lot of the planning because now we know better. Before, we could take all the time we wanted to publish. Although we shouldn't have taken that long, but nobody was expecting our results with anxiety. Now they are. And it's not just the publications, it's not just the results, it's also the data. The data, everybody wants the data to be public and for other people to analyze the results, for astronomers to use it, and there's a lot more pressure. So that has changed. The planning of the collaboration on the publication side. And also looking at the future.
Now, it's not like we can stop for many years, like we did between initial and Advanced LIGO to improve the detectors. Now, if we want to take time to improve the detectors, we have to be careful about what people expect, and what's happening, and who's following the signals, about the electromagnetic counterparts. Now there are telescopes and satellites that are planning their observing with taking into account when we are taking data. So everything is different, but of course more exciting. More important. We are part of the astronomical field now.
So Gaby, just to bring our conversation right up to the present. We talked a little bit about, at the beginning of our talk, about some of the difficulties of doing all these things in the pandemic. But remote work aside, what are some of the major questions and research areas that you're involved with these days?
Well, for the last several years, I have been very involved in the diagnostics of the data. We not only not have the sensitivity that we want to have, because we have excess noise from the ways we push, especially but also from light that goes that gets scattered. But also the data that we take has lots and lots of transients that are not astrophysical. And we know they are not because they are very large, and they don't happen in coincidence in the other detector.
But of course, they can obscure real astrophysical signals. In fact, the 2017 collision of neutron stars was almost hidden behind a very large glitch in the Livingston detector. So we almost missed that one. I have been studying these transients to see ways of diagnosing where they come from. Sometimes we find out where they come from, and then people can come up with solutions. Usually it's to steer the light going in different places or to tune the control system. I'm just a member of a very large group that's called Detector Characterization Group within the collaboration doing this. But that's what I have been mostly doing myself.
Well Gaby, now that we've worked all the way up to the present, for the last part of our talk, I'd like to ask a broadly retrospective question about your career and your path, and then we'll end looking to the future. So the first one is, going all the way back to your days in college when you first decided you were interested in general relativity, and then you fast forward all the way to today, where you're working and have been working at the cutting edge of the field with LIGO. What have been across that journey some of the most intellectually satisfying moments for you as you've always pursued these interests in general relativity?
Ah, that's a difficult question. Intellectual satisfactions? I think my first publication. [laugh] That was about the research I did in college that actually received a referee report pointing out a mistake. But then correcting it, and having a published work was satisfying. My PhD thesis took a lot of work and a lot of changes in the way we thought about it, with Peter. I took learning to overcome different challenges, both on the analysis and on the way we analyze the data and in the way we built the experiment. But getting it all together, writing that thesis, that was very, very satisfying. Seeing the initial detectors in 2005 when we got the sensitivity that we had been working for, taking data and being able to analyze that data. Actually seeing that noise spectral density that was as good as we wanted, and then improving it. Because we didn't stop there, we improved it. Realizing that we got there, but we could go further in 2005, it was great. And then of course the first detection, although being the spokesperson for that first detection made that time so stressful. So stressful that it didn't really produce satisfaction until a few days after the announcement.
Gaby, if I may ask, what was so stressful? Just the pressure of getting right and doing it the right way?
Making sure we didn't make mistakes. Yes, making sure that this wasn't going away, that this was not a mistake, that we looked at everything that could be done. I mean, people looked at what could have-- could have been something in the environment? It turns out that there was a lightning strike that was one of the largest ever recorded within seconds of the detected gravitational wave, and that's limited by the precision of the recording of the lightning. So could that have produced it? It was in Burkina Faso, which is very far away, but we had to convince ourselves that it didn't, that the lightning couldn't have produced the signals that we saw. And we went that way over so many things. So many things, so I couldn't sleep thinking that we had forgotten something. [laugh]
Like I said, it was satisfying only after making the announcement and seeing that, well, not just the referee reports of the paper, but the reaction of the scientists in general. They were not skeptics. They actually looked at the data, because we published the paper and the data that same day and they downloaded it and they saw it too, and they were convinced then. Later on, there were some groups that said that they didn't believe it, and published some negative things, but it was a small group and many other people took care of refuting that.
Gaby, last question –
And then that was not the last intellectual satisfaction. I have to say that [laugh] when we published the first catalogue of detections, with ten different signals. Seeing that we had actually ten different signals from black holes, and one from neutron stars, that was the latest one, yes.
Well that's perfect. For my last question, Gaby, looking to the future, all of these satisfactions that you've felt, that you've contributed to. To understanding how the universe works. What are some of the big que-- I mean, there's many big questions that remain out there. Dark energy, dark matter, cosmological constant. What are the big mysteries outstanding in the universe that you feel like your area of expertise, the field that you work in, the collaborators that you work so closely with. What are you most optimistic about in terms of contributing to solving or better understanding some of those major outstanding mysteries?
I have to say that I think that those mysteries will not be solved with gravitational wave detectors. The big mystery that I would like to get an answer to, but it's probably not going to come from Advanced LIGO, but from third generation detectors, is the history of black holes. Were there primordial black holes? Do all black holes start from star explosions? Or were there black holes from the beginning? We will know that when we can see the complete history of black holes. When we can see black holes with a very, very large red shift. And that will take third generation detectors. That we are already working to push, right? I mean Rai and Kip started in the ‘70s, and they got the first detection in 2015. And Advanced LIGO was approved in 2008, so it takes decades. So we are planning already now for the future. I hope I can see that.
What's the timeline for the third generation?
We hope that those will be built in the 2030s. But it is a very large investment, and there's no committed funding yet, but we know that funding agencies are interested in this field.
A lot to be excited about.
There is, yes.
Gaby, thank you so much for doing this interview with me. It was wonderful to hear your personal perspective and your institutional knowledge, and I'm so glad we were able to do this. So, thank you so much.
No, thank you. This was a great interview.