For years, astrophysicists have argued over a deceptively simple question: how do pairs of black holes find each other, spiral inward, and collide? A new analysis of more than 150 binary black hole mergers detected by the LIGO-Virgo-KAGRA (LVK) gravitational-wave network now argues that the answer is not one story but three. Published as a preprint in May 2026, the study applies a parametrized mixture model to the latest gravitational-wave catalog and concludes that three distinct astrophysical formation channels are needed to explain the full spread of masses and spins in the data.
The finding challenges the hope that a single dominant pathway could account for most observed mergers. Instead, it points to a universe where black hole binaries are assembled through at least three different cosmic processes, each leaving its own fingerprint on the signals that reach Earth.
The catalog behind the claim
The foundation for this work is GWTC-4.0, the Gravitational-Wave Transient Catalog compiled by the LVK Collaboration. It incorporates validated detections from the first through fourth observing runs (O1 through the first half of O4, known as O4a), making it the largest standardized dataset of confirmed gravitational-wave events to date. A detailed catalog description from the collaboration explains how each candidate event was identified, vetted for instrumental artifacts, and assigned measured parameters such as component masses, spins, and distance.
With that catalog in hand, a team of researchers posted a study titled “On the Astrophysical Origin of Binary Black Hole Subpopulations: A Tale of Three Channels?” Their mixture-model analysis assigns each detected merger a probability of belonging to one of three formation channels, based on its mass, spin, and redshift properties. The three channels are:
- Isolated binary evolution: Two massive stars born together in a galactic field evolve, collapse into black holes, and eventually spiral together. These systems tend to produce lower-mass black holes with spins roughly aligned to their shared orbital plane.
- Dynamical formation in dense star clusters: Black holes that form independently get gravitationally shuffled together inside globular clusters or nuclear star clusters. Because they were not born as a pair, their spins can point in nearly any direction relative to the orbit.
- Hierarchical mergers: A black hole remnant left over from a previous merger pairs up and merges again, roughly doubling the mass scale with each generation. This process can build black holes heavier than any single star could produce on its own.
The study argues that collapsing all observed events into a single-origin model obscures real structure in the data, structure that becomes visible only when the population is allowed to split into subgroups.
Independent confirmation from a second framework
A separate research effort using a mixture-model framework called Vamana reached compatible conclusions. Analyzing the same catalog of more than 150 binary black hole signals, the Vamana team reported multiple distinct peaks in the mass distribution, including peaks separated by roughly a factor of two. That spacing matters because it is exactly what physicists predict when hierarchical mergers are at work: two black holes of similar mass collide, the remnant is about twice as heavy, and if that remnant finds another partner and merges again, the next remnant is heavier still. In a large population, this process stamps a ladder of preferred masses into the distribution rather than a single smooth hump.
A third, peer-reviewed study covering the earlier O1, O2, and O3 observing runs provides additional support. That work quantified branching fractions for different origin channels and identified likely environments for hierarchical mergers, including nuclear star clusters and the gas-rich disks surrounding supermassive black holes in active galactic nuclei (AGN). Crucially, it showed that hierarchical mergers were already motivated by pre-O4 data, not introduced after the fact to patch a gap in the newer catalog.
Taken together, the three analyses converge on a consistent picture. The lightest black hole binaries, with modest masses and aligned spins, fit the profile of isolated binary evolution. Heavier, spin-misaligned systems point toward dynamical assembly in dense stellar environments. And the heaviest outliers, especially those sitting at mass scales that single stars cannot reach, are best explained as products of successive mergers.
Where the picture gets blurry
The three-channel model is well-motivated, but it is not yet settled science. The primary study proposing three subpopulations remains a preprint as of May 2026 and has not completed formal peer review. In astrophysics, preprints circulated on arXiv carry real weight because the community routinely shares results before journal publication. Still, the absence of a finalized referee report means the statistical methodology, the choice of priors, and the model-comparison criteria have not been independently vetted through that formal process. Revisions could shift the inferred importance of each channel.
A deeper challenge is overlap. Isolated binary evolution and dynamical formation can both produce black holes in similar mass ranges. When that happens, the mixture model must lean on subtler clues, such as spin alignment and mass ratios, to sort events into channels. The peer-reviewed hierarchical-merger study flags this difficulty explicitly, noting that branching fractions can shift depending on assumptions about the environments where mergers occur. Nuclear star clusters and AGN disks, for instance, predict different spin distributions for hierarchical remnants, and current gravitational-wave data cannot yet cleanly distinguish between them. The result is that any individual event might be classified as “field” or “dynamical” with non-trivial probability, depending on the model’s assumptions.
Selection effects add another layer of uncertainty. Gravitational-wave detectors are more sensitive to heavier, closer binaries, so the observed catalog is not a simple mirror of the true cosmic population. All three analyses attempt to correct for this bias, but small differences in how each team models the detection threshold, noise background, or redshift evolution can ripple through to different inferred channel fractions. The tails of the distribution, where the rarest and most massive events live, are especially vulnerable to these systematics.
There is also the question of metallicity. Lower-metallicity stars shed less mass to stellar winds during their lifetimes and can therefore produce heavier black holes at death. If future gravitational-wave detections come paired with electromagnetic observations that reveal host-galaxy properties, astronomers could test whether the three channels map onto distinct metallicity regimes. That test, however, depends on detector upgrades and multi-messenger follow-up capabilities that are still under development.
What sharpens next
The strongest takeaway from this body of work is not the specific number “three” but the growing evidence that the mass distribution of merging black holes contains real substructure, structure that a single formation story cannot reproduce. Multiple peaks, factor-of-two mass spacings, and a spread of spin orientations all point toward a universe that builds black hole binaries through more than one assembly line.
The most decisive tests will come from larger samples. As the LVK network completes the full O4 observing run and moves toward planned sensitivity upgrades, the number of detected mergers is expected to climb into the hundreds and eventually thousands. With that volume, subtle features in the mass and spin distributions should sharpen, making it easier to separate overlapping channels and to track how their relative contributions change across cosmic time. Future observatories, including the planned space-based detector LISA and proposed ground-based instruments like the Einstein Telescope and Cosmic Explorer, would extend sensitivity to lower masses and higher redshifts, filling in parts of the population that current detectors cannot reach.
For now, the three-channel framework represents the most detailed attempt yet to decompose the gravitational-wave population into its astrophysical ingredients. It is a framework still being stress-tested, not a final verdict. But the days of assuming a single origin story for merging black holes appear to be over.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
