Gravitational-wave astronomy has moved fast enough that "just another catalog update" now means hundreds of new black-hole collisions to sift through. But a paper posted to arXiv on July 24, 2026, argues that buried in that flood of data is something more structured than researchers expected: not one population of merging black holes, but four.
The paper, "Revealing Four Subpopulations of Binary Black-Hole Mergers with the Fifth Gravitational-Wave Transient Catalog," comes from Nir Guttman, Paul D. Lasky and Eric Thrane. It draws on GWTC-5, the fifth Gravitational-Wave Transient Catalog compiled by the LIGO-Virgo-KAGRA (LVK) collaboration, and asks a fairly blunt statistical question: does a single, smooth distribution of black-hole masses and spins actually describe what LIGO has detected, or is the data secretly a mixture of separate groups?
The answer, according to the authors, is a mixture — and a four-part one at that.
What the Four Groups Look Like
The dominant group is also the least exotic. It accounts for roughly 70 percent of all detected mergers and is centered on black holes of around 10 solar masses. The authors tie this component to the most mundane origin story available: massive stars that collapse directly into black holes without the fireworks of a visible supernova, sometimes called "failed supernovae." If that interpretation holds, most of what LIGO hears is the quiet aftermath of dying stars that never got their explosive send-off.
The second group is defined less by mass than by mismatch — an unequal-mass branch, where the two black holes in a merging pair are noticeably different sizes rather than close twins.
The third component sits in sharp contrast to the second: a nearly-equal-mass branch, with both black holes weighing in around 30 to 35 solar masses. This is close to the mass range of some of the first black-hole mergers LIGO ever detected, back when the instrument was proving gravitational-wave astronomy could work at all.
The fourth group is the rarest and, for astrophysicists, the most tantalizing: a hierarchical-merger component. These systems show broad, scattered mass ratios paired with unusually large spins — a signature long predicted for black holes that are not first-generation stellar remnants at all, but the product of an earlier black-hole merger that then went on to merge again. In other words: black holes with a merger already in their family history, spinning fast because their progenitors approached each other still tumbling with angular momentum from their own violent past.
Statistically, the authors report that this four-component model is preferred over standard single- or simpler multi-population models by a natural-log Bayes factor of 19.2 — a large enough margin, in Bayesian model-comparison terms, to count as decisive rather than suggestive.
Why It Matters
Every black-hole merger LIGO detects carries a physical record of how the two objects got there — their masses, their spins, and the imprint of whatever violent history shaped them. As Universe Today's Carolyn Collins Petersen has laid out in coverage of the broader trend, gravitational-wave signatures effectively let astronomers read the biography of colliding black holes off the wave pattern itself.
That matters because black holes don't leave light behind. Unlike the neutron-star mergers that produced visible kilonovae, most black-hole collisions are dark from birth to death — gravitational waves are often the only messenger. Sorting the detected population into distinct subgroups, rather than treating it as one statistical blob, is a step toward reconstructing formation channels that no telescope can see directly: which black holes came from lone dying stars, which came from binary star systems that stayed paired all the way to collision, and which came from a black hole that already had a merger under its belt before merging again.
The hierarchical-merger component is the piece with the most riding on it. Confirming that some LIGO detections are second-generation black holes — built in dense stellar environments like globular clusters or galactic nuclei where repeat mergers are more plausible — would be direct evidence for a specific formation environment, not just a formation mechanism. It's the kind of result that connects gravitational-wave data to decades of theoretical work on dynamical black-hole assembly in dense clusters.
None of this happens without the raw material. LIGO Laboratory's own reporting on GWTC-5.0 confirms the catalog added 161 new events recorded between April 2024 and January 2025, pushing the cumulative gravitational-wave detection count to 390. Caltech's summary of the catalog release also cites evidence for second-generation black holes and improved sky localization — both of which feed directly into the kind of subpopulation modeling Guttman, Lasky and Thrane just published.
What Comes Next
A log Bayes factor of 19.2 is a strong statistical statement, but it's a statement about population-level structure in the current dataset, not a confirmed astrophysical mechanism for each group. Follow-up work will need to stress-test whether the four-component split holds up as GWTC-6 and beyond add more detections, and whether the hierarchical-merger component's spin signature survives scrutiny as the sample grows. For now, though, the black-hole population LIGO has been quietly cataloging looks less like a single family and more like four.
