The smallest galaxies orbiting the Milky Way are barely visible. Some host only a few thousand stars; their total stellar mass is about a millionth of our Galaxy’s. Yet a new theoretical study argues that these inconspicuous companions, known as ultra-faint dwarf galaxies, are not just minor relics but precision instruments for probing the universe at an age it is otherwise impossible to access directly.

The result, published in Monthly Notices of the Royal Astronomical Society by Brown et al. as part of the LYRA collaboration, is based on a new suite of cosmological zoom-in simulations carried out on the COSMA 8 supercomputer at Durham University’s Institute for Computational Cosmology. The work was led by Dr Azadeh Fattahi at the Oskar Klein Centre in Stockholm, with first author Shaun Brown, and brings together teams in Sweden, the United Kingdom, and the United States.

Why ultra-faint dwarfs matter

Within the standard cosmological model (ΛCDM), dark matter is the gravitational scaffolding upon which galaxies form. Massive galaxies such as the Milky Way live inside large dark matter halos; ultra-faint dwarfs sit inside the smallest halos that can hold on to baryons long enough to ignite stars at all. They therefore mark a physical threshold — the boundary between a dark matter clump that becomes a galaxy and one that remains forever dark.

That threshold is governed by a delicate competition. Gravity tries to pull baryons into the halo’s potential well; cosmic radiation tries to heat that gas and prevent it from cooling, fragmenting, and forming stars. The lower the halo mass, the easier it is for radiation to win.

The Lyman-Werner background, explained

The radiation field that the LYRA team isolates is the Lyman-Werner background, photons in the 11.2–13.6 eV range that are energetic enough to dissociate molecular hydrogen (H₂) but not energetic enough to ionise atomic hydrogen. In the early universe, before the first generations of stars produced significant ionising radiation, H₂ was the principal coolant capable of bringing primordial gas down to the temperatures needed for star formation in mini-halos. Destroying H₂, even partially, raises the minimum halo mass that can host a first generation of stars, with cascading consequences for the population of galaxies that emerge later.

The team ran their simulations under two different assumptions about how strong this Lyman-Werner background was during the first ~500 million years after the Big Bang. The contrast is stark: the smallest dwarf galaxies are extraordinarily sensitive to the assumption, whereas Milky Way–scale systems are essentially indifferent. In some configurations, the same dark matter halo produces a faint but observable galaxy; in others, the gas never cools enough to make stars, and the halo remains entirely dark.

A meteorology of the cosmos

Brown describes the situation through a useful agricultural metaphor: just as the yield of a summer crop encodes information about the spring weather that preceded it, the present-day stellar properties of an ultra-faint dwarf encode information about the radiation field that was bathing its proto-halo more than 13 billion years ago. The galaxies we observe in the local volume today are, in effect, a fossilised barometer of cosmic dawn.

This is a remarkable claim because it inverts the usual direction of cosmological inference. Traditionally, the early universe is studied by looking outward and backward in time — the cosmic microwave background, high-redshift galaxies, the 21-cm signal. The LYRA work suggests that the same epoch can be reconstructed by looking inward, at the faintest fossils orbiting our own Galaxy.

Connection to recent JWST surprises

The timing of the result is not incidental. The James Webb Space Telescope has, over the past three years, uncovered galaxies in the early universe that appear unexpectedly massive and luminous for their epoch. Whatever physics is responsible for those high-redshift surprises — whether modifications to the initial mass function, more efficient star formation, or non-standard variations in the radiation environment — should also leave traces in local ultra-faint dwarfs. Two complementary windows on the same physics, one at the cosmological frontier, the other in the Galactic backyard.

The Vera C. Rubin window

Why now? Because the observational pipeline is about to widen dramatically. The Vera C. Rubin Observatory, on Cerro Pachón in Chile, will conduct the Legacy Survey of Space and Time (LSST), and is expected to deliver something close to a complete census of Milky Way satellite galaxies down to extremely low surface brightness. The LYRA simulations indicate that this census will not just refine the count of small galaxies; it will function as a constraint on cosmic dawn physics, an inferential bridge between the local universe and the first 500 million years of cosmic history.

Computational scale

The technical accomplishment is also worth noting. The full simulation suite required more than six months of computation on COSMA 8, hosted by Durham’s Institute for Computational Cosmology on behalf of the UK’s DiRAC High Performance Computing Facility, and produced roughly 300 terabytes of data. Several of the analysis pipelines had to be refactored from older tools that were not designed to handle dataset volumes of this scale, an increasingly common challenge as galaxy formation simulations approach the resolution required to model the smallest dwarfs.

Outlook

Looking forward, the LYRA team plans to use the same simulation suite to address two open questions in galaxy and structure formation: where the first generation of stars (Population III) might be hiding today, and what the detailed properties of ultra-faint dwarfs imply about the particle nature of dark matter. Both lines of inquiry treat the local universe as a laboratory for high-redshift physics — a perspective that is increasingly central to modern near-field cosmology.

The paper, «LYRA ultra-faints: The emergence of faint dwarf galaxies in the presence of an early Lyman-Werner background,» is available in Monthly Notices of the Royal Astronomical Society under DOI 10.1093/mnras/stag439.

© 2026 SKYCR.ORG | Homer Dávila Gutiérrez, FRAS. All rights reserved. Reproduction in whole or in part is prohibited without express authorization. Original source: Royal Astronomical Society