A New Search Strategy for Gravitational Waves from Eccentric Neutron-Star Binaries

A rotating neutron star could emit a very faint, persistent gravitational-wave signal if its mass distribution is slightly uneven. Finding such a signal could offer clues about the star’s interior. But if the star orbits a companion, its motion also shifts the signal’s frequency over time. That shift is particularly tricky to model when the orbit is highly eccentric—more oval than circular.

The study by M. Ball and colleagues asks whether gravitational-wave data can be searched efficiently for signals from neutron stars in wide, eccentric binaries identified as possible targets by the Gaia space telescope. Gaia observations help constrain where these systems are in the sky and how their orbits work. The neutron stars’ rotation rates, however, are not known, so the search must look across a range of possible signal frequencies.

The researchers focus on 21 candidate systems identified in earlier work combining Gaia observations with follow-up measurements. Their orbits have periods of roughly 189 to 1,046 days and eccentricities from 0.12 to 0.80. These values put them outside the ranges handled well by many earlier searches, which were designed for circular or nearly circular orbits.

Following a signal through an eccentric orbit

The team developed a way to space out the trial signal models—often called templates—so that they cover uncertainties in the measured orbital parameters. Their “frequency metric” estimates how much a small change in an orbital parameter shifts the predicted signal frequency. The method uses Gaia’s measurements to narrow the range of orbital possibilities, while still accounting for their uncertainties.

There is a computational hurdle: calculating the frequency shift caused by an eccentric orbit requires solving an equation repeatedly. The authors built on the FastTracks search framework and reduced the number of calculations by evaluating the orbital correction at less frequent time intervals, then interpolating between them. They also used a hybrid template strategy: a grid for the signal frequency and its rate of change, with orbital parameters sampled according to Gaia-informed distributions.

To test the approach, they added artificial signals, or injections, to both simulated Gaussian noise and real LIGO data from the second part of the fourth observing run (O4b). The analysis used data from the Hanford and Livingston detectors. It tested two of the 21 candidates, selected because one required comparatively few templates and the other represented a more computationally demanding case. The injections covered two narrow frequency ranges near the low and high ends of the study’s 50–200 hertz search band.

The authors report that the injected signals could be recovered using their procedure. Tests in real detector data gave results consistent with tests in simulated noise, supporting the claim that the method is workable under realistic data conditions. This is a demonstration of search feasibility, not a detection of gravitational waves from the Gaia candidates. The paper does not report that any candidate’s rotation or gravitational-wave emission has been measured.

A promising method, with open questions

The study is an early feasibility assessment, and its results have important boundaries. The injection campaign concentrates on two candidates, not all 21, and on a limited set of injected frequencies and signal properties. Its conclusions therefore support the method’s potential; they do not establish how sensitive a completed search would be for every candidate or every possible signal.

The authors also note that their frequency metric treats parameter spacings separately and does not account fully for correlations among orbital parameters. Those correlations could affect how templates should be placed. In addition, the paper’s computational estimates and sensitivity depend on choices about the search, including its frequency range and candidate-selection threshold. A future search would need to apply follow-up procedures and distinguish genuine signals from instrumental features that can resemble them.

The work lays groundwork for combining Gaia’s measurements with continuous-gravitational-wave searches. A real search could test these systems using detector data, but its outcome is not known from this feasibility study. The paper is a preprint, so its findings should be read as research in progress rather than as a settled result; the source version may not have undergone peer review.