Gravitational waves probe dark matter

Soumen Roy and collaborators reported on May 12 that they have built a new way to search gravitational-wave data for signs of dark matter around merging black holes. The method, published in *Physical Review Letters*, models how a light scalar field — one candidate form of dark matter — could alter the orbital motion of two black holes and slightly reshape the gravitational-wave signal detected on Earth. The team then applied that model to public data from the LIGO-Virgo-KAGRA collaboration. In 28 of the clearest merger signals from the detectors’ first three observing runs, the researchers found that 27 were consistent with black holes merging in vacuum, while one 2019 event stood out as a possible exception. ### Why would dark matter show up in a black hole merger signal? Dark matter is expected to interact mainly through gravity, which makes it hard to detect directly, and the new work asks whether black holes could concentrate it enough to leave an observable trace. The authors said gravitational interactions near black holes can build up dense scalar configurations, and that those environments can change binary dynamics during inspiral and merger. (journals.aps.org) Josu Aurrekoetxea of MIT said in a university release that black holes can provide a mechanism to enhance dark-matter density, making it possible to search for its effects in the gravitational waves emitted when the objects merge. The paper describes those effects as changes to the waveform shape, which the team modeled and compared against detector data in a Bayesian analysis. ### Which event drew the researchers’ attention? (journals.aps.org) GW190728, a gravitational-wave event detected on July 28, 2019, was the signal the researchers identified as the most interesting case in the catalog they tested. The GW Open Science Center lists the event as GW190728_064510, and MIT’s summary of the work said the signal showed a “preference,” or agreement, with the team’s dark-matter model. (physics.mit.edu) The *Physical Review Letters* abstract said that for GW190728 and GW190814, vacuum lay outside the 95% credible region in the team’s analysis. It added that when the researchers included priors from black-hole superradiance, GW190728 showed tentative evidence for a scalar environment, with a Bayes factor of about ln B ≈ 3.5 and a scalar mass near 10^-12 electron volts. ### What did the researchers actually claim — and what did they not claim? (gwosc.org) The authors did not say they had detected dark matter. MIT’s release and the Phys.org report based on it both said the scientists found possible signs of a dark-matter imprint and described the method as a way to screen gravitational-wave data for candidates that would need further confirmation. The paper itself framed the result as tentative evidence for a scalar environment in one event, not proof that the environment was dark matter. (journals.aps.org) That distinction matters because environmental effects around black holes can be difficult to separate from other astrophysical or modeling uncertainties, and the abstract reports upper limits on scalar densities for most compact binaries rather than a broad positive detection. (physics.mit.edu) ### How does this fit with earlier work on the same idea? A 2024 *Physical Review Letters* paper by Aurrekoetxea, Katy Clough, Jamie Bamber and Pedro Ferreira had already shown in simulations that wave-like dark matter around equal-mass black hole binaries could produce measurable dephasing in the last orbits before merger. That earlier work focused on the theoretical effect and the conditions needed for dark matter near black holes to become dense enough to matter for the waveform. (journals.aps.org) The 2026 paper moves from that theoretical setup to catalog data from LIGO-Virgo-KAGRA. The authors said they developed a semianalytic waveform model, validated it against numerical-relativity simulations, and then used it to test observed events. ### What comes next for this line of research? The LIGO-Virgo-KAGRA network has continued to expand its event catalog, including the GWTC-4.0 release in August 2025 with 128 new significant signal candidates from the O4a observing period. (journals.aps.org) That gives researchers a larger pool of mergers to test with models for scalar environments and other dark-matter scenarios. Physical Review Letters published Roy and colleagues’ paper on May 12, 2026, and the next step is likely to be applying the method to additional events and future observing runs as more public data become available. (journals.aps.org) The GW Open Science Center and the LIGO-Virgo-KAGRA catalog releases are the main sources for those follow-up analyses. (ligo.caltech.edu)