Some Ancient Starbursts Spin Like Disks, and Galaxy Models Can’t Explain It Yet

Across a 15-arcsecond-wide patch of sky, at a redshift of 2.41, five galaxies are turning gas into stars at a combined rate of roughly 7,400 Suns per year. That makes HATLAS J0849 one of the most highly star-forming protocluster cores detected so far. The question is not whether these galaxies are starbursts. It is whether they are wrecking themselves to do it.

A crowded core of early galaxy growth

The system is HATLAS J084933.4+021443, or HATLAS J0849. It was first spotted as an unusually bright infrared source, then found to be a dense grouping of dusty, gas-rich galaxies at a redshift of 2.41. That redshift marks a time billions of years ago, when galaxy clusters were still assembling. Astronomers call such a dense region an overdensity or protocluster core.

HATLAS J0849 contains five known starburst galaxies: W, T, C, M and E. Their star formation rates range from about 600 to 3,500 solar masses per year. W and T are hyper-luminous infrared galaxies; C, M and E are ultra-luminous. W also hosts an active supermassive black hole. These are the kind of extreme objects proposed to grow into the most massive elliptical galaxies seen in groups and clusters today.

But how they grow matters. If mergers trigger the starbursts, the cold gas should move chaotically. If smooth inflows feed the galaxies, the gas should settle into rotating discs. The new study set out to tell those histories apart.

Reading cold gas like a traffic map

The team used ALMA, the Atacama Large Millimeter/submillimeter Array, to observe the five galaxies at 0.15-arcsecond resolution, fine enough to map internal motion.

ALMA measured the CO(4-3) and [C I] 1-0 lines, which trace cold molecular gas and neutral carbon, plus dust-continuum emission. By measuring how the line emission shifts across each galaxy, the researchers built velocity maps. A rotating disc leaves a smooth gradient and an S-shaped position-velocity diagram (a plot of gas speed versus position). A merger can leave knots, twists and mismatches between the visible gas and dust morphology and the gas motion.

W spins; T looks like a merger

The clearest result is HyLIRG-W. It is a rotating disc, but an unusual one: lopsided, with one dominant spiral arm. Its outer rotation velocity is about 520 kilometers per second, placing it among the fastest rotating cold-gas discs measured at such distances. W is strongly rotation-dominated, with a rotation-to-dispersion ratio near 10—that is, ordered rotation is about ten times faster than the gas's random motions.

W also shows non-circular radial motions. The team finds these are consistent with a bar, an elongated central structure that can funnel gas inward. The evidence is kinematic, not a direct image of a stellar bar, but it fits the pattern. The study finds no sign of a large-scale molecular outflow from W, despite its extreme star formation.

ULIRG-C is the second rotating disc. It is smaller and less rotationally supported, with a lower limit on its rotation-to-dispersion ratio of about 2.2; the authors note it may yet prove rotation-dominated. It too shows features consistent with bar-driven streaming.

HyLIRG-T is the surprise. Earlier, lower-resolution studies had classified it as a rotating disc. The new data show a velocity field whose major axis is nearly 85 degrees away from the galaxy's visible gas and dust axis. It also splits into two clumps with distinct kinematics. The authors consider a polar ring, an outflow, and a late-stage merger. They find the merger scenario most plausible: a smaller galaxy is likely falling in from the northeast.

Lensing cannot explain the mismatch. The foreground lenticular galaxy seen in Hubble images is too weak to twist T's velocity field. Upcoming JWST observations should test the merger interpretation directly.

The remaining galaxies are messier. ULIRG-E is most consistent with a merger, while ULIRG-M could be a merger, a disturbed disc, or a system where only the central, dispersion-dominated gas is visible. In total: two rotating discs, two likely mergers, and one uncertain case among five.

Too many orderly discs for models

The five galaxies are not alone. The authors combined their results with other resolved cold-gas studies of massive starbursts in distant overdensities. They find that at least 67 percent are rotating discs, and at least 42 percent are rotation-dominated discs. The 42 percent is likely a lower limit. That fraction is hard for many galaxy-evolution models, which tend to make extreme starbursts through mergers or violent disc instabilities that should scramble ordered rotation.

The study also finds a striking environmental pattern. Beyond redshift 1, disc galaxies with rotation velocities above 400 kilometers per second appear almost exclusively in overdensities. That supports a picture in which dense environments assemble mass earlier and faster than the field.

Turbulence stays modest, and the progenitors line up

If starbursts are fed by violent inflows, their cold gas should be more turbulent than in ordinary star-forming discs. It is not. W and C fall on the same relationship between velocity dispersion and star-formation rate as lower-rate field galaxies. Simple models in which stellar feedback drives turbulence match the data. Models that add turbulence from radial gas transport overshoot.

The authors suggest an explanation: gas may stream inward along ordered structures, such as bars, without converting much of its kinetic energy into random motion. The starburst burns hot, but the disc-wide turbulence stays surprisingly modest.

Finally, the team compared the distant discs with local early-type galaxies that still have cold gas discs. When they include the distant galaxies' gas mass along with their stars, the rotation velocities and total baryonic (star plus gas) masses line up. These young starbursts already sit in halos comparable to those of the most massive ellipticals at the centers of nearby groups and clusters.

That does not prove a one-to-one ancestry. But it strengthens the case that extreme starbursts in dense, early environments are the progenitors of today's giant ellipticals.

What could still change the picture

This is a preprint submitted for peer review, so the conclusions are not yet final. The sample is small: five galaxies here, plus a heterogeneous literature sample. Observations of cold gas also favor the brightest, gas-richest systems, which may not represent every starburst.

Some classifications remain open. The merger interpretation for T needs JWST confirmation. M is uncertain. The bar evidence is kinematic and consistent with bars but not conclusive. The 42 percent figure is likely a lower limit, and the authors note that simulations still need to reproduce the high fraction of orderly discs.

Still, the result changes the question. In a protocluster core where galaxies are forming stars at furious rates, order and chaos coexist. Some galaxies spin; others collide. The future of the most massive ellipticals may depend on that mix.