The LIGO–Virgo observatories are three kilometre-scale interferometers that detect tiny spatial displacements (10⁻¹⁸ m) that occur when gravitational waves – ripples in space–time – pass through Earth. The detectors are famous for detecting short, intense pulses of gravitational waves that are emitted in the final moments before pairs of black holes or neutron stars merge. But astrophysicists also want to use these huge facilities to observe much fainter continuous signals from objects such as rotating neutron stars. Meg Millhouse and Lucy Strang at the University of Melbourne and Karl Wette at the Australian National University are members of the Australian Research Council’s Center of Excellence for Gravitational Wave Discovery, or OzGrav, and they talk about their search for continuous gravitational waves.

Who belongs to OzGrav and what are its primary goals?

Karl Wette: OzGrav was funded by the Australian government in 2017 for seven years, with a mission to join with international collaborators in leading the exciting new field of gravitational astronomy, and to inspire the next generation of Australian scientists. It’s made a big difference to the field in Australia. When I did my PhD at Australian National University, well before OzGrav was funded, the number of people working on gravitational waves was a lot smaller, and I was one of the few people working on data analysis. We now have well over 100 students and postdocs working across instrumentation, data analysis, interpretation and astrophysics. We’ve recently submitted a proposal for a new centre to succeed OzGrav, and hopefully that gets funded so that we can continue to grow this new field Down Under.

Why do you expect neutron stars to emit continuous gravitational wave signals?

Meg Millhouse: For a neutron star to emit continuous gravitational waves it must have a time-varying quadrupole moment in its mass. In simple terms the star must be lumpy – not a perfect sphere – and it must be rotating. We believe that neutron stars have rigid crusts and previous research suggests that the crust could be deformed to create “mountains” on the surface. This is actually an overstatement because the deformities are thought to be on the order of millimetres – but that could be enough to create gravitational waves that we can detect. 

We also know that many neutron stars rotate, because we can observe the pulses of radiation that they emit. It is these pulsars that we are studying.

In your latest study, you’ve targeted 15 neutron stars that have recently formed in supernovae. What’s special about these objects?

Lucy Strang: We expect young neutron stars to be lumpier and therefore emit more intense gravitational waves than older stars. Neutron stars are created in supernovae, so we target young supernovae remnants because these are relatively easy to find. 

The LIGO–Virgo detectors are kilometre-scale interferometers. How do you point them at neutron stars? 

LS: We can’t point the LIGO–Virgo detectors at specific parts of the sky – they observe the entire sky. Instead, we have a series of mathematical transformations on LIGO–Virgo data to look for signals that could correspond to the neutron stars we are interested in. So in a sense we are pointing a telescope using mathematics.

Are the signals that you are looking for different from the gravitational-wave signals that have already been seen from merging black holes and neutron stars?

KW: Mergers of black holes and neutron stars are violent events that produce short-lived and relatively intense pulses of gravitational waves. The continuous signals that we expect from slightly deformed and rotating neutron stars are much weaker – more like a very low background hum. While the low intensity makes the signals difficult to find, their continuous nature gives us an advantage. The signals are always humming in the background, so if we keep observing and analysing data over a long period of time, we may be able to extract the weak signals. 

Ideally the search would require vast amounts of computing power – but this just isn’t available to us. So, as well as being important from an astrophysical perspective, our research is also a big data challenge – which is very exciting to work on.

Is LIGO–Virgo very sensitive to the gravitational-wave frequencies that you expect from rotating neutron stars?

KW: Most of the pulsars that we know of spin at about 1 Hz or slower. Unfortunately, LIGO–Virgo is not very sensitive to signals at these low frequencies, where seismic interference from things like human activity are a problem. Instead, we focus on neutron stars spinning at hundreds of hertz, where the detectors are much more sensitive. There are some neutron stars that fit the bill and we hope that we can see continuous waves from them

Have you spotted any continuous wave signals so far?

LS: Sadly, no – but that is not surprising because we know that there is currently a low probability of making a detection. However, we have established an upper limit on the strength of signals from neutron stars and that has allowed us to put constraints on some properties of neutron stars.

What sort of constraints?

KW: One thing we are interested in are r-modes. These are like giant ocean waves on the surface of a rotating neutron star, which could broadcast relatively intense gravitational waves. X-ray observations of the pulsar J0537-6910, for example, provide strong evidence that the neutron star is radiating gravitational waves via r-modes. 

However, we haven’t seen such waves and that allows us to exclude certain models of r-mode emissions. This translates directly into a better understanding of the neutron star equation of state, which relates the radii of neutron stars to their masses. Even though we have not made a detection, we can already say something important about neutron-star physics.

Even though we have not made a detection, we can already say something important about neutron-star physics

LS: Another thing we can constrain is the shape of the neutron star. We talked earlier about how we expect lumpy (or, to put it another way, non-spherical) neutron stars to produce continuous waves. The less spherical and more lumpy the neutron star is, the louder we expect the signal to be. By constraining the size of the signal, we’re also constraining how non-spherical the neutron star must be.

As with the r-modes Karl just discussed, this restriction can be translated into constraints on the neutron star equation of state. The equation of state contains the fundamental physics governing the neutron star and determines its mass, radius and so on. The conditions inside a neutron star are impossible to replicate in a laboratory, so astrophysical observations are our only window into physics in these extreme conditions. With each limit we set, we get a step closer to understanding the underlying physics.

You are targeting neutron stars created in supernovae, but what if you are lucky enough to observe a nearby supernova?

MM: There are several different gravitational-wave signals that could come from a supernova. The explosion itself would create a short burst of gravitational waves, which would be a very exciting thing to detect. The astronomy community could do multimessenger observations of the event, capturing electromagnetic radiation, and possibly neutrinos, along with gravitational waves. This would give us a wealth of information about supernovae.

If there is a neutron star remnant present after the explosion, it should emit continuous gravitational waves. These could be difficult to detect because we expect that the rotational speed would be decreasing very quickly. If we don’t know what frequency we are looking for or how fast the star is spinning down, it can be a tricky observation. 

LIGO–Virgo scans the entire cosmos. Are you also looking for neutron stars that we don’t already know about?

KW: Yes, that’s another strategy that we are pursuing – eyes wide open surveys that scan the entire sky over a range of signal frequencies. Estimates suggest there are approximately one billion neutron stars in the Milky Way, but we only observe a few thousand of them as pulsars. We hope that maybe a few of the billion are radiating gravitational waves that we can detect, and that would provide new insights into neutron stars. 

We’re still looking at some of the data from the latest LIGO–Virgo observations and we are hoping that we will soon have more results to report. Next year will bring further upgrades to the LIGO–Virgo detectors, which will make them more sensitive to continuous waves. So, this is an ongoing story.