LIGO’s Next Steps

The 11 February announcement of the first detection of gravitational waves, by members of the Laser Interferometer Gravitational-wave Observatory (LIGO) Scientific Collaboration, dazzled the physics world and the general public alike. What’s next for the search? To find out, OPN talked with LIGO Chief Scientist and OSA Fellow Stanley Whitcomb a week after the discovery was announced. Here’s a transcript of that interview.

A 35-year journey

Let’s talk first about some of the changes you’ve seen to the LIGO project. You’ve had a particularly long association with the project—I guess at least a couple of decades now?

Whitcomb: More than a couple of decades actually. I first started with gravitational wave detection at Caltech in 1980; I arrived basically just after getting my PhD as an assistant professor, when Caltech was first starting its activities in this area. It was an exciting thing; it seemed really challenging and sounded really important to me. I think I was a bit naïve. I had some unrealistic goals—I thought we'd probably have this detector built and up and operating in something on the order of a five-to-seven-year time horizon. So I missed that one by nearly an order of magnitude. But at some level, I think it's great to come in with high goals and high expectations.

After that period, I left and went into industry, and spent six years working in the aerospace industry—still in optics. In that regard, I actually learned what it takes to do big projects. So when I first arrived at Caltech, I was like your average new Ph.D. or grad student, very naïve about what it takes to do big things. But working on a space-based product in the aerospace industry, I learned about project management, how to manage budgets, how to do scheduling and so on. I learned what it takes in terms of an engineering team to put things together.

But eventually you came back to Caltech—why?

Whitcomb: When I returned to Caltech, which was in 1991, it was because I was still in love with the challenge of gravitational-wave detection—the idea of making such a precise measurement, of doing something that nobody else had ever done before. That still called to me, it really resonated for me as something I wanted to do. And I thought that I'd matured in some ways, and that I brought something back to this venture when I returned in 1991. I've been with it ever since.

An international partnership

How has the program evolved since then—particularly internationally? Certainly there's a pretty strong sense that this is an international partnership.

“We now have close to 900 people in the LIGO Scientific Collaboration, I believe, and about half of those come from the U.S. and about half of them from other countries.”

Whitcomb: In the very early days, we were doing prototypes in the laboratory. You were never doing something big enough that you couldn't sort of reach over and make an adjustment, or walk to the other end of an optical table and tweak something. Now it's grown to a scale where it's hard to even keep track of what everybody is doing, let alone be able to participate and do everything yourself. We now have close to 900 people in the LIGO Scientific Collaboration, I believe, and about half of those come from the U.S. and about half of them from other countries.

Part of that is that this is a unique facility, and there are people around the world who want to get involved in it. But part of it is also that the nature of this research allows smaller groups to get involved and to participate in a meaningful way. And that's true for groups both inside the U.S. and for outside the U.S.

So we have these large activities at Caltech and M.I.T., where we've got tens of scientists and engineers that work on this full-time. But we also have groups at other universities that might consist of a professor and two part-time undergraduates who work on this. And they're able to make a significant contribution and participate. And that kind of small-scale participation is also available to people outside the U.S.; so if you happen to live in someplace that can't afford to build a gravitational-wave detector on its own, you can still participate in this and make a meaningful contribution that will be appreciated by the bigger collaboration.

And something else that’s changed since 1991, when you got reinvolved at Caltech and with the project, is that now we have the World Wide Web—those communications advances must have made a huge difference in enlisting other partners, too.

Whitcomb: Oh, yeah, absolutely. Having the ability to set up quick online meetings to share visual material or graphs quickly at the same time that you're talking about them makes it very, very possible to do this, and it would be extremely difficult without that.

LIGO’s mirrors: The “crown jewels”

Let’s turn to the LIGO instrument itself, and discovery of gravitational waves announced in February. LIGO is a rather amazing piece of equipment—is there anything that, to you, particularly sums up this technology?

Whitcomb: Well, in some sense the most important thing we have in our interferometer are the mirrors that form what we call the test masses, which reside at each end of our optical interferometer. I consider these mirrors the crown jewels in this whole thing. They are 35 cm diameter and 15 cm thick, so they're quite sizeable mirrors; they weigh 40 kg. They're ultra-high-purity fused silica—just one grade higher uniformity than you can order from a catalog—so it's a special order in terms of the uniformity.

“Just from the specs alone, you can see [these mirrors] are really, really special items.”

They're polished to a figure—there’s a small curvature to the surface of the optic, and the spec on the accuracy of the surface is an RMS of λ/1,000. And then they're coated with very-low-loss, ion-beam-sputtered multilayer coatings; the losses in those coatings, fom absorption and scatter, is of order of few parts per million. And the uniformity of those coatings in each layer is about a part per thousand of the thickness of the layer. So to first order, those coatings are uniform in thickness at a plus-or-minus one-atom level.

Just from the specs alone, you can see they are really, really special items. And I think we sometimes, in all of the other things we talk about, don't really talk about how very, very special those things are. To me, that is the fundamental enabling technology—the mirrors and learning how to get these mirrors.

Learning how to get them?

Whitcomb: I'll tell you a little story. Back in 1989, when we wrote our proposal, they did a calculation and came to the conclusion that mirrors that were a λ/4 would be fine. They said, “Our specification should be λ/4; that's what we need, but just in case we'll make them λ/10 because that's easy to do. We hadn't actually done the additional calculations you needed to really understand that they needed to be much, much better than anything we had imagined at the time of the proposal.

And I truly believe that if we had actually known how good the mirrors had to be when we wrote that proposal in 1989, we never would have actually submitted the proposal, because nobody would have believed you could make mirrors that good—I don’t think the proposal would have been accepted. It took a lot of work with industry…. Part of what we had to do was actually help them develop the metrology so that they could know what they were producing, and would actually then be able to institute quality control and actually sign up for the contracts that we were doing.

So the way we worked with industry on developing those optics is a story that doesn't get quite as much emphasis as it ought to.

Next step: Boost the sensitivity

It is just a mindboggling, amazing setup and facility. Let's talk a little bit about what's next for LIGO, and for gravitational-wave detection generally. There was a flood of interest when you announced the first detection. Has the instrument yielded up additional observations since then since then?

Whitcomb: I can't say anything more about what we've seen. We did see this first event last September, very early on in the run, and continued to take data until we had enough data to fully characterize the instruments—to give us enough data to provide a background estimate and a no-false-alarm probability—you know, what are the chances that this was just a noise event and not a real signal.

And then, once we had all of that, even though we continued to take data, we basically froze our analysis at that point and said, that's the block of data we're going to publish this event on. We wanted to get it out to the world as quickly as possible.

Yes, but you were very careful with the analysis—one didn’t get the impression that this was rushed out.

Whitcomb: Yes, I know. We continued to take data until nearly the end of October and chose to analyze that, and we got this out as quickly as we could in as responsible a way as possible. We felt like this first detection claim was going to get extra scrutiny, and we wanted to be sure that we had asked all of the questions of ourselves that we thought the outside world might be asking of us. So that's why it took so long. It also happens that we ran across the holiday period, and it’s really hard to get people to concentrate and come to committee meetings over the holidays. But, anyway, we got it out as quickly as we could.

“What's happening now is, we've stopped taking data and we've started to make a next set of tweaks to the instruments.”

We continued to run and take data until about the middle of January…. What's happening now is, we've stopped taking data and we've started to make a next set of tweaks to the instruments. So they'll be turning up the laser power a little bit, going into the interferometer, adjusting some of the control loops, trying to find some lower noise configurations for some of those things. And we will come back online later in this year.

And what sort of sensitivity are you targeting?

Whitcomb: Our goal is a sensitivity that's a little bit over one and one-half  times better than the current sensitivity—that’s in terms of the amplitude of the wave that we see, not the power in the wave. So when we improve things by a factor of two, we can see two times farther, because the amplitude of the wave goes down as 1/r, not 1/r².

So when we come back on line later this year, we'll get an improvement of one-and-one-half times or maybe even as much as two times improvement in the sensitivity, and that means we see a volume as much as 2³ larger. So those sensitivity improvements are really important to us, because that's what the rate [of observations] is tied to—the volume that we see.

So if we come back on line—okay; I'll be a little bit optimistic—if we come back on line with a factor-of-two better sensitivity in our next observing run, we would see events at eight times the rate that we would see them in this first observing run. So let's say that [the event in September] was the only event we saw in run 1, in run then we should see something like eight events. So I think we expect to start to see meaningful numbers of events even as soon as our next run.

Doing gravitational-wave astronomy

And what happens when those events start streaming in? What will this new world of “gravitational-wave astronomy” actually look like?

Whitcomb: Let me give you three examples of things that we will be able to do. The first is kind of a “physics-y” thing. The object that we observed in September was a binary black hole system—two black holes orbiting each other; they're moving closer together, spiraling to each other until they touch, merge to form a single black hole object, ring down and go quiescent.

And the waveform that you saw carries information about all the phases of that. During the in-spiral phase, the portion at the beginning when they're still separate objects in-spiraling, that carries information about the individual masses of the two objects, it carries information about their spins, and so on. The final stages when it's ringing down, that carries information about the mass of the final object and its spin, its angular momentum. And so we can actually do significant tests of general relativity by comparing different parts of that waveform with the predictions from numerical relativity.

So not just testing the broad prediction of general relativity that gravitational waves exist, but actually digging down into the equations on a more granular level.

Whitcomb: Exactly. We did something where we determined the masses from the early phase of the in-spiral and compared it with the mass determined from the ring-down portion at the end—to show that, in fact, those are consistent with the predictions of general relativity. Because this is our first event we've only done some of those tests, and while none of them show any deviation from general relativity, they are not super-high-precision tests yet. Once we get more events, and once we get better signal-to-noise ratios, we'll actually be able to do those tests in a much more stringent way.

A second thing is there's another type of source that we expect to see, and it's called a binary neutron star. Instead of having two black holes that spiral together, merge and form a single black hole, you could have the same system with two neutron stars; they spiral together the same way. When they touch and merge, they will typically form a black hole at the end, but there will be an accretion disk around it, and those kinds of systems are the systems that are believed to be responsible for these things called short gamma-ray bursts—short GRBs.

“[B]y doing comparisons combining the gravitational wave information with the electromagnetic information, we can learn a little bit more about the short GRBs.”

If we can measure binary neutron star in-spirals and mergers, and alert our astronomer friends to quickly look and see what they see, we'll be able to start to do comparisons with the electromagnetic spectrum. So the GRBs produce these jets of gamma-rays, but also then a longer-period afterglow of X-rays and hard UV. And by doing comparisons combining the gravitational wave information with the electromagnetic information, we can learn a little bit more about the short GRBs.

So how do you make those comparisons? How do you combine the gravitational-wave information with the information from more conventional astronomy?

Well, for example, the gravitational wave information gives us—very accurately, in fact—the orientation of that binary system to us. Are we seeing it face-on; are we seeing it at an inclination angle, and so on. That's important, because the jet of gamma-rays is believed to be beamed. And so we would be able to actually tell, by seeing these [collision] events, if we see a bunch of them at 5-degree orientation where we see the gamma rays, and then we detect others that they're inclined at 30-degrees where we don't see the gamma rays, then we know that the beaming angle is someplace between 5- and 30-degrees. So we get to really combine our observations with astronomy observations there.

The final thing that I'll mention is something that is still going to take a little bit of luck, probably, and that's a core-collapse supernova. The predictions are that when Advanced LIGO is fully operational at full sensitivity, we may have a distance range for core-collapse supernova of, I think, only about 1 or 2 megaparsecs, so it's got to be a fairly close by our galaxy. But if we can get a core-collapse supernova that is close enough, the information that we’ll see is actually of the collapse of the core—it’s the bulk motions of that core collapse … the mass motions that are going on there.

New observatories in the mix

In looking for these events, and constraining them, presumably it will help that we have a number of other gravitational-wave observatories coming on the scene—there’s the Advanced VIRGO project in Italy, which as I understand it is coming on line this year; the project in Japan, KAGRA; and, I believe, others as well. What does that larger number of observatories do for the effort?

Whitcomb: First of all, let me share with you a piece of good news in case you hadn't heard it. We got word just yesterday [17 February] that another gravitational wave project, called LIGO-India has been approved by the Indian government. So in addition to Advanced VIRGO and KAGRA, we will soon—in something like 8 years—have another LIGO detector, in India.

Why is that good news for us? The answer really is that the gravitational waves that we're looking at have wavelengths that are typically of order 3,000 km. So our ability to locate them on the sky is basically determined by diffraction—just λ/d. So basically the farther apart our detectors are, the better we are at locating sources on the sky. So that was one of the factors in the two Advanced LIGO detectors; we wanted them as far apart as they could be inside the U.S. But now having the Advanced VIRGO detector in Italy gives us another very long baseline; LIGO-India will give us a very long baseline, and KAGRA as well.

“[U]nlike the electromagnetic spectrum, the gravitational wave sources are expected to be highly polarized; polarization is really important to us.”

There's yet another factor that I think maybe some people don't appreciate as much. These things really are quadrupole antennas, so they are sensitive to one polarization of gravitational waves. There's two polarizations of gravitational waves, and for every direction on the sky, for any one of these L-shaped detectors there's one linear polarization that we're completely insensitive to. And unlike the electromagnetic spectrum, the gravitational wave sources are expected to be highly polarized; polarization is really important to us.

So being able to squeeze out the polarization from the observation is hugely important for getting a complete understanding of the sources. And so what do we need for that? Well, we need these L-shaped detectors with many different orientations. If they're flat on the ground, you can do a couple of independent orientations, but it's really hard to get one where you have got one arm sticking straight up or straight down for four kilometers!

So the only way to get some of the orientations is to go a quarter of the way around the world, and on a detector that's built parallel to the Earth's surface you naturally get those other orientations. So having enough detectors that you scatter them around the world so you get decent polarization coverage is an important part of it.

A long-term commitment

This project has really been an amazing, decades-long sustained commitment—both for the scientists and for the public, in terms of the funding commitment. Any parting thoughts to share about that—what it’s been like to stay the course with this project over decades, and the importance of that kind of long-term commitment?

Whitcomb: Well, the National Science Foundation has done a phenomenal job of supporting this. We are funded through what's called a cooperative agreement, and it really is cooperative; they're working to help us. They ask hard questions, but we appreciate that they're doing it for the right reasons—not to trip us up, but to be sure that we've thought of all of the things that we should be thinking of. I think it's a tribute to the NSF staff.

For those of us who have been working on the project, I have to say that it's hard to maintain long-term enthusiasm, to be running a marathon like this with only the final goal as your motivation. It's really important that we're able to take whatever small pleasure we can at various times from the intermediate things that we've accomplished, and that we actually have a very good team of people to work with. We all enjoy working together, and that's really important for sustaining that long-term commitment.