INTERVIEW: Dr. Xavier Siemens, Pulsar Arrays are Nature’s Gift

TCA: Hi, I’m Sally Lehman and I’m with The Advocate. Today, I’m speaking with Dr. Xavier Siemens of Oregon State University. Dr. Siemens is part of the North American Nanohertz Observatory for Gravitational Waves, also known as NANOGrav – a research group which has a facility at OSU. Thank you for being here today, Dr. Siemens.  

Siemens: It’s my pleasure.   

TCA: Let’s start with a little history. In 1915, Albert Einstein’s general theory of relativity said that what we perceive as the force of gravity arises from the curvature of space and time. And we generally think of curvature as being a physical attribute. How do space and time have curvature? 

Siemens: That’s a really difficult question to start this interview.  

Well, this really arises from the fact that gravity affects all masses in the same way, meaning – and this is Galileo’s experiment of throwing a light object and the heavy object from the Tower of Pisa and them falling at the same rate, basically landing on the ground at the same time. So the fact that that happens, this is called the equivalence principle. The fact that gravity affects all masses in the same way allows you to write the theory of gravity rather than as a force, as a property of space itself. And what general relativity tells us is that matter curves space, and it tells matter how to move and matter, itself, tells space-time how to curve.  

TCA: How can planets and suns change gravity? 

Siemens: Planets essentially deform the space-time around them, and then the Earth modifies space around it and the Moon moves in that space-time around the Earth.  

TCA: And this creates ripples in the fabric of time-space. 

Siemens: It can create ripples. The Moon orbiting around the Earth creates ripples at a period of a month. Which is how long it takes for the Moon to orbit around the Earth. But these are very weak gravitational waves because the Earth and the Moon are not particularly massive objects.  

TCA: Can you explain to us the purpose of NANOGrav? 

Siemens: What NANOGrav was trying to do is to detect very low frequency gravitational waves. These are gravitational waves with frequencies of about a few nanohertz.  

One nanohertz is a wave with a period of 30 years. So a few nanohertz is something like a wave of a period of a few to say 10 years or something like that. So this is a gravitational wave with a very low frequency – does a full cycle up and down in about 10 years.  

So the way we do this is by using these types of stars that are called pulsars. And this is a type of neutron star, which is the remains of a star that has run through the course of evolution. Basically, it’s burned through all of its nuclear fuel. It’s gone supernova. It has collapsed and it has turned into this kind of object called a neutron star.   

Now, neutron stars generally are fast spinning and they have magnetic fields that come out of their poles. Along these magnetic fields, some of them emit beams of radio waves that we can detect. And these magnetic poles are not aligned or not always aligned with the rotational axis of the star.  

So if you imagine a star with magnetic poles going in this direction, spinning about like this, then the radio waves basically sweep out in space, kind of like the light from the beams of a lighthouse. And so if we look out in space with our radio telescopes, we observe them as sources of bursts of radio waves. Whenever the beam of radio waves sweeps past the Earth, we’ll see a little blip – a little burst.  

Now there’s a particular kind of neutron star, which is called a millisecond pulsar, which rotates very, very stably, so much so that we can use them as clocks, meaning they tick very regularly. They emit these bursts of radio waves. We point our radio telescope at them, and we see these bursts of radio waves and they arrive at us in a very stable way.  

So these neutron stars are like hundreds of thousands of light years away from us – they’re kind of spread out throughout the galaxy. Now, imagine a gravitational wave that’s passing through the galaxy. It’s stretching and it’s squeezing the galaxy as it’s doing so, it’s moving these pulsars toward the Earth and away from the Earth.   

So these pulses will be arriving a little earlier when the gravitational wave is squeezing everything together or a little bit later when the gravitational wave is stretching things apart. And so these very stable, rotating neutron stars – these pulsars, millisecond pulsars – can be used as clocks. And so we can measure the times of arrival of these pulses so precisely that when they arrive a little bit earlier or a little bit later, we notice that and we can measure that.  

So we’re effectively building this galactic scale gravitational wave detector. It works on the same physical principles as a LIGO [Laser Interferometer Gravitational-Wave Observatory], for example, which you might have heard of. One of the instruments of LIGO is just up in Hanford, a few hundred miles from here in Washington State. So it’s the same physical principles as LIGO, except that rather than having lasers and mirrors, we have radio telescopes and neutron stars doing our job. 

TCA: So this is why you’re able to create these pulsar timing arrays, is to be able to measure these. 

Siemens: And this is the way the word array comes from the fact that we time many of these pulsars. So it’s not enough to just take one pulsar and measure the fact that the pulsars sometimes arrive a little bit earlier. Sometimes they arrive a little bit later. It’s not enough to have one, because reality is quite complicated.  

What you want to do is measure that same pattern of stretching and squeezing in the correlations among many pulsars. And this is what we call a pulsar timing array.   

Right now, for example, NANOGrav is timing close to 80 pulsars. So it’s kind of like LIGO but with 80 arms instead of two arms. 

TCA: So can you explain what an interferometer is? 

Siemens: So that’s what LIGO is, an interferometer. I worked in LIGO between 2002 and 2017 – so for quite a long time, I was pretty heavily involved with LIGO.  

And so what LIGO does is they have a laser which they shoot into a beam splitter and then the laser light splits. Part of the light goes in one arm, the other part of the light goes down the other arm, and they are at 90 degree angles.  

At the ends, you have mirrors. The light bounces off from the mirrors and then it recombines like a beam splitter. And light is a wave, so the way it recombines has to do with where the crests and the troughs are of this wave. And as the mirrors are moving, that changes where the crests and the troughs are.  

So by monitoring how much light you’re getting at the beam splitter, you can tell what the relative motion is of these mirrors that are at the end by measuring this interference pattern of light.   

So the word interferometer comes from the fact that light is interfering. It’s interfering or the interference pattern is changing because these mirrors are moving due to a gravitational wave or something else.  

TCA: When you talk about crests and troughs, you’re talking about like a sine or cosine wave? 

Siemens: Precisely. In this case, the crests and the troughs are the laser light. But you have two beams, one of them which went out one arm and one which went out the other arm, and now you’re recombining them. They have two light beams, each with its own set of crests and troughs. And depending on where the crest and the trough line up, this is called an interference pattern. And so as the mirrors are moving, that interference pattern is changing. And that’s the physical principle by which LIGO operates. So it’s an interferometer. 

TCA: Is the collision of two black holes a common thing? 

Siemens: Well, I think right now LIGO is up to about one event a week. Well, it depends on what you mean by [common] in a volume of the size of the galaxy. LIGO sees up to half a billion light years away – something like that, or several billion light years away is how far away LIGO can see. [LIGO can see to about 650 billion light years.] So even though these events don’t happen particularly often in any given galaxy, LIGO can see so much volume of the universe that they happen about one a week.  

TCA: You began working with these time space ripples in 2015 in Milwaukee – with NANOGrav, at least.  

Siemens: With NANOGrav, I began in 2008. 

TCA: Ok, how far has science come since then?  

Siemens: We’ve come a really, really long way.  

This is a very exciting area to work in. Just last September, we put out a paper on the archive and it’s been published in The Astrophysical Journal where we found for the first time ever a signal which is common to all of our pulsars. So this is possibly the first hints of a stochastic gravitational background. We don’t know yet because we don’t have the correlations necessary to really prove that.  

The most promising signal that we’re looking for and pulsar timing arrays is this something called a stochastic background. And this signal is the gravitational wave equivalent of the cosmic microwave background. So this is like a hiss or a hum that’s coming from all directions of space. And our case, in the case of pulsar timing arrays, it’s being produced by all the supermassive black holes that are emerging in the universe.   

So there are supermassive black hole binaries across the sky, they’re all merging, all emitting a signal. But the signal is so called an incoherent superposition, meaning you can think of each individual binary is producing a note, a musical note. And then there’s millions of them across the sky, and all the notes are coming to us at once. And so it’s producing a kind of a hum or a hiss.  

That kind of a signal is called a stochastic background. And that’s the most promising signal that we are going to be able to detect using pulsar timing arrays. In each individual pulsar, this signal has some properties, it looks like a form of noise in each individual pulsar, and it’s so-called red noise.  

And what that means is that there’s more power in this signal at low frequencies than at high frequencies – meaning there’s more power or energy in the signal at waves of periods of 10 years. They have more energy or more power in them than waves of periods of one year. Larger wavelengths, lower frequency has more power, than higher frequency, smaller wavelengths like those kinds of processes are called red noise processes. And so we search for these in our data. And we found evidence in this twelve-and-a-half year data set that we submitted back in September for this kind of a noise process in common to all of our pulsars. 

And what common means is that it looks like that process has the same amplitude and spectrum and all of the pulsars. What we were not able to find sufficient evidence of is correlations.  

I mentioned before that what we want to monitor many, many pulsars because we want to make sure that the way in which the pulses arrive at our radio telescopes earlier or later is correlated among pulsars. That’s the real evidence. That’s the kind of smoking gun of gravitational waves is while this pulsars pulses are coming a little bit ahead, this pulsar over here, they should be arriving a little bit later and so on.  

You can do that with all these multiple pairs of pulsars across the sky. And that gives us confidence. Not that we’re just seeing a noise process in our data like there’s some red noise in our pulsars, we don’t know where it’s coming from. This red noise is actually correlated. It’s producing a correlated pattern of advances or delays in the pulses across different pulsars. 

TCA: Are there other types of noise other than red noise?  

Siemens: We have many different kinds of noises that come into our data. Most of these are not correlated and none, as far as we know, are correlated in the specific way that gravitational waves would produce correlations.  

But we have many sources of noise in our data. The so-called white noise is constant across frequencies. There are different kinds of white noise that we put into our models when we search our data, different kinds of red noises that we put into our data. And then there’s this common red noise that we also model in that we found evidence for – but not sufficient evidence for – correlations in that red noise that we are modeling. 

TCA: And during the course of your career, you’ve actually seen the ability to monitor and track that red noise.  

Siemens: Yeah, absolutely. So we were seeing evidence of this noise growing over the years. The first time we unambiguously detected it was in the twelve-and-a-half year data set. And our expectation is that if this common red noise process that we’re seeing is due to gravitational waves, that our next data set, which is the 15 year data set, should show more evidence of correlations. We are working on that analysis right now, and we’ve been working on it for the past year or so.  

TCA: What do you believe will be shown by that analysis? 

Siemens: Well, I can tell you what I’m hoping.  

I’m hoping – I don’t know yet. I mean, I’m hoping that this red nose will turn out to be gravitational waves and we’ll see evidence of these correlations that we’re looking for. But we can’t say at this point until we’ve dotted all our I’s and cross all our T’s. We’re not going to discuss the results, but what I’m hoping is that we will detect gravitational waves  

TCA: And what would that mean to the field of physics if you’re able to prove what you think this will show?  

Siemens: It’s significant in a couple of different ways.   

First of all, gravity is like light in the sense that it has a spectrum. So LIGO observes things, observes gravitational waves, the frequencies of a one hundred hertz or a few hundred hertz, maybe 10 hertz to about a kilohertz or so. So that’s a window onto the gravitational waves spectrum at those frequencies. We’re observing gravitational waves at nanohertz frequencies, that’s 11 orders of magnitude away in frequency. So that’s like the way we do optical or electromagnetic astronomy.  

We have optical telescopes, radio telescopes, X-ray telescopes, gamma ray telescopes. We have all these kinds of telescopes that observe light at these many different light frequencies. But they’re all the same thing – they’re all photons of different light, photons of different frequencies. So we’ll be opening a window like that.   

It’s as though you’d only ever have optical astronomy, and all of a sudden you have radio astronomy and you can do astronomy that way. And what that does is open a window onto a whole new set of sources.   

When you look at very different frequencies, you’re seeing very different physical objects that are producing these gravitational waves. So LIGO, for example, can measure gravitational waves from black holes of a few to a few tons of solar masses. Our most promising candidates are supermassive black hole binaries. Those are supermassive black holes that are billions of solar masses, and they live at the centers of galaxies and they merge with galaxies themselves. By measuring the form of the signal that we receive, we are able to characterize the sources that produce them.  

In our case, we have supermassive black holes that we can attribute a stochastic background to. So, what can we learn? Well, the way we think this stochastic background is produced is by all the supermassive black holes that are merging in the universe – supermassive black hole binaries that are merging.  

And so the way we think galaxy evolution proceeds is by mergers. So early on in the universe small galaxies that merged to form slightly larger ones, these larger ones merge again and so on. In this hierarchical process, that results in the galaxies that we see around us today – large structured objects.   

Each time one of these mergers happens, the supermassive black holes at the centers will meet and coalesce. And so by measuring the entire stochastic background, the position of all of these, we can talk about the systems that led to that. So we can solve for the merger rate of galaxies, for example. How often do galaxies merge? How big are the black holes that live inside galaxies? That’s still an open question that we want to answer. What are the environments that are around supermassive black holes at the cores of galaxies?  

So all the physics that comes into figuring out what the signal looks like when we measure it, all these astrophysical predictions, we can turn those around once we make a measurement and we can say, ‘oh, well, we’ve measured this.’ So that means that this is the physics that’s consistent with the signal that we’ve already measured. So this is called solving the inverse problem. 

TCA: How close is the closest combined galaxy to our galaxy? 

Siemens: You mean the closest merging galaxy?  

TCA: Yeah, merging galaxy. 

Siemens: Well, we’re merging with Andromeda right now, so that’s pretty close.  

A few years ago, there was this article in the news about [how] Andromeda was going to collide with the Milky Way, because we’re actually on a collision course of one of these mergers that happens. So we’re on a collision course with Andromeda. That’s not going to happen for a very long time. We don’t need to worry about that. The nearest merging black hole, binary black hole, is something like a few hundred million light years away.  

TCA: What led you to study this?  

Siemens: Well, I was originally a cosmology kind of theorist, and I got a job – my first postdoc job – in 2002 working for LIGO. I had done my thesis on gravitational waves from cosmic strings – my thesis was on that. And so I knew a little bit about gravitational waves at that point.  

I got hired to work in LIGO and I worked on that starting in 2002 until 2017 or so. And then I learned about pulsar timing arrays in [2008 or 2009]. And the reason I thought this field was so appealing is that in LIGO – well, LIGO was an amazing experiment where you build all the mirrors and lasers and tunnels and the vacuum tubes where you put the lasers and so on. But it’s all made by people. The entire instrument is made by people.  

The reason I found pulsar timing so interesting and so beautiful is that a very important part of the apparatus is nature’s gift. Like we just happen to have these pulsars out there that are very stable rotators and they’re like clocks. And all we have to do to build a gravitational wave detector is to build a radio telescope and point at these objects. I just found that concept really, really interesting and really just very beautiful  

TCA: An elegant solution to a lot of things.  

Siemens: And it’s the only way we can measure these very low frequency gravitational waves. So that’s kind of amazing. Amazing coincidence, beautiful concept,  

TCA: So I guess the most important question right now is what is the secret to overseeing a collection of high level scientists? 

Siemens: I mean, it depends on the group of scientists that you’re talking about. NANOGrav is honestly a really great place to work. That’s another reason I like working in NANOGrav, is it’s a relatively small collaboration. There are maybe 200 people or so – that may not sound small to some, but compared to other collaborations, it is a smaller collaboration.   

We have maybe 100 undergraduates who work on Pulsar. They operate radio telescopes and they look for us, they search for more pulsars. We have 100 other people which are graduate students, postdocs, and faculty people like myself. And they’re just a group of very, very smart and very kind people that are a pleasure to work with.   

Honestly, I think if I were to say what the secret is to working – and that’s something that’s possibly not scalable to larger numbers of people, we generally operate by consensus. So we communicate very frequently. We have several meetings a week when there are decisions to be made. We all sit around and talk about it and we have a discussion. And, you know, it’s not like we don’t have disagreements, of course, but we settle the disagreements through conversations and we eventually arrive at a consensus and we move forward. And this model has worked very well for us.   

TCA: Your enthusiasm for all of this is infectious. And I’ve really had an interesting time looking into it and reading a little about it. So thank you for your time and for explaining so much of it to me. I really appreciate it.  

Siemens: Thank you for your interest.  

By Sally K Lehman