Massive young star clusters blow away their birth clouds in roughly 5 million years, not the 7 to 8 million years that leading theoretical models had predicted. That is the central finding of a study published in Nature Astronomy in June 2026, based on a census of star clusters across four nearby galaxies observed by both the James Webb Space Telescope and the Hubble Space Telescope. The 2- to 3-million-year discrepancy may sound small on cosmic scales, but it reshapes how astrophysicists think about stellar feedback, the survival of planet-forming disks, and the pace at which galaxies recycle their raw materials.
A four-galaxy census with two telescopes
The research team, led by principal investigator Adamo and collaborators within the FEAST (Feedback in Emerging extrAgalactic Star clusTers) program, surveyed star clusters in M51, M83, NGC 628, and NGC 4449, all galaxies close enough for Webb and Hubble to resolve individual clusters. Webb’s infrared cameras can peer through thick dust to spot clusters still wrapped in their natal gas, while Hubble’s optical instruments reveal clusters that have already broken free. By classifying each cluster as embedded, partially emerged, or fully exposed, the researchers built a statistical picture of how long each stage lasts and how that duration changes with cluster mass.
For the heaviest clusters in the sample, packed with dozens of luminous O-type stars, the emergence timescale came in at about 5 million years. “The massive clusters are clearing out their gas envelopes significantly faster than our simulations predicted,” said Adamo, the study’s lead author, in a statement accompanying the Nature Astronomy publication. That means these stellar powerhouses are flooding their surroundings with hard ultraviolet radiation significantly sooner than standard semi-analytic and hydrodynamic feedback models, such as those used in widely cited galaxy-evolution frameworks, had assumed. The finding carries direct consequences: protoplanetary disks forming near such clusters face intense radiation earlier, potentially shortening the window in which they can hold onto the gas needed to build giant planets.
The observations are part of the broader FEAST initiative, cataloged under JWST program 1783. FEAST has already produced peer-reviewed catalogs of emerging clusters across all four galaxies, giving astronomers a consistent framework for comparing cluster ages, masses, and environments. A companion study, available on arXiv, used two-point correlation analysis to map how young clusters trace the fractal architecture of their host galaxies’ interstellar medium, drawing on the same underlying catalogs and reinforcing the age and mass estimates through independent statistical methods. As of June 2026, that companion paper’s peer-review status has not been independently confirmed.
Spectroscopy confirms the gas is being shoved aside
Imaging alone can show where clusters sit relative to their dust cocoons, but it cannot reveal what is physically happening to the gas. A separate FEAST result tackled that gap by pointing Webb’s NIRSpec instrument at emerging clusters in NGC 628 and measuring emission lines sensitive to gas temperature, density, and ionization. The spectra confirmed what the images implied: the regions around young clusters are filled with ionized bubbles, shocked gas, and sharp density contrasts between cleared cavities and the surrounding molecular material.
When infrared imaging, optical imaging, and spectroscopy all converge on the same sequence of embedded-to-exposed clusters in the same galaxies, the inferred timescales become considerably harder to dismiss. Each method probes a different physical property, and the fact that all three tell a consistent story strengthens the case that massive clusters really do shed their envelopes faster than theorists expected.
What the study does not settle
Four galaxies, however nearby and well-studied, are still four galaxies. All sit at broadly similar metallicities and star-formation conditions. Whether the accelerated clearing holds in metal-poor dwarf galaxies or in extreme starburst environments remains untested. Extending the census to those regimes will be critical for determining how universal the 5-million-year benchmark really is.
There is also the question of mechanism. Massive clusters produce radiation pressure, powerful stellar winds, and, within a few million years, the first supernovae. The current study constrains the outcome (how fast the gas disappears) but not the precise recipe of forces responsible. A preprint associated with the Nature Astronomy paper provides extended methods and tables, but competing theoretical groups may interpret the same data differently depending on which feedback channel they emphasize in their simulations.
The implications for planet formation, while physically well-motivated, are also a step beyond what the paper directly measures. Faster cloud dispersal means nearby protoplanetary disks face harsh ultraviolet fields sooner, which could accelerate photoevaporation and cut short giant-planet assembly. But the severity of that effect depends on where individual disks sit relative to the clearing front. A disk shielded behind a dense gas filament or positioned far from the cluster core may experience a very different radiation history from one sitting inside the newly opened cavity. The study does not resolve those spatial relationships at the individual-disk level.
Finally, the clusters driving the headline result are rare, high-mass systems. Most stars in the universe form in smaller, quieter associations with more modest feedback. If those lower-mass environments clear their gas more slowly, then the accelerated emergence seen in FEAST may matter most for the harshest star-forming regions rather than for the typical birthplace of a Sun-like star and its planets.
Why the 2-million-year gap matters beyond star formation
Star formation models do not exist in isolation. They feed into simulations of galaxy evolution, predictions for the cosmic ultraviolet background, and estimates of how quickly heavy elements are mixed back into the interstellar medium. If the most massive clusters punch through their cocoons faster than those models assume, the downstream effects ripple outward: galaxies may regulate their star formation differently, ionizing photons may escape into intergalactic space on shorter timescales, and the chemical enrichment of gas reservoirs may proceed in bursts rather than the smoother ramps some simulations produce.
For planet-formation researchers, the finding sharpens a question that has been simmering for years. Observations from the Atacama Large Millimeter Array have shown that many protoplanetary disks appear to lose their gas within just a few million years. If the radiation environment around massive clusters turns hostile even sooner than previously thought, the overlap between disk lifetimes and intense UV exposure narrows further, tightening the constraints on when and where giant planets can form.
The full imaging and spectroscopic datasets from FEAST are publicly archived at the Mikulski Archive for Space Telescopes, and the preprint lays out every operational definition and statistical technique used to classify clusters and convert those classifications into timescales. That transparency means any research group can reanalyze the data, challenge the assumptions, or extend the methods to new targets. For now, the observational benchmark is set: the most massive star clusters emerge from their birth clouds faster than any widely used model anticipated, and the next generation of feedback simulations will need to explain why.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
