Hubble dates the Milky Way's oldest known merger to 11.8 billion years ago

The Milky Way absorbed a dwarf galaxy roughly 11.8 billion years ago, when the universe was less than 15 percent of its current age, according to a study published in Nature Astronomy. Drawn from data collected by the NASA and ESA Hubble Space Telescope, the finding marks the earliest known merger in the galaxy's history and pushes that record back by 1.8 billion years.

Our home is the Milky Way galaxy, but we do not know how our house was built. In this paper we discover where the first significant batch of bricks came from: a dwarf galaxy that we call LKH.

The researchers named the absorbed galaxy Low-energy-Kraken-Heracles, or LKH, a nod to three earlier papers that had championed the idea of a merger early in the galaxy's life, including the Kraken proposal of Kruijssen and colleagues and the Heracles proposal of Horta and colleagues. They estimate the dwarf carried stars totaling roughly 500 million times the mass of the Sun, a stellar mass close to that of Gaia-Sausage-Enceladus and a significant fraction of the Milky Way's mass at the time. Most of that material, the team found, settled in the galaxy's inner regions rather than scattering into the outer halo.

What Hubble and Gaia measured

The evidence came not from the dwarf galaxy itself, which dissolved into the Milky Way long ago, but from globular clusters, roughly spherical collections of tens of thousands to a few million stars that survive for billions of years. Astronomers combined Hubble with ESA's Gaia telescope to gauge the age, chemical composition and orbital path of 39 clusters spread across the galaxy's innermost 20,000 light-years, a region packed with the oldest material the Milky Way holds.

Sorting the clusters by those properties, the team found three distinct populations. One traces the proto-Milky Way, the stars born natively in the galaxy. A second matches the Gaia-Sausage-Enceladus merger, already well documented. A third sat in between, older than the Gaia-Sausage-Enceladus clusters but younger than the native ones, regardless of metal content, and that middle group pointed to a separate, earlier merger no one had confirmed.

Hubble is observing globular clusters that have never been studied before, and this will help us characterize the merger events that are far back in time in the Milky Way galaxy's history.

"The key finding of our research is the unambiguous discovery of the first merger event experienced by our galaxy in its infancy," Massari said. Both observations and simulations had suggested that a large merger preceded the two already on the record, but the specifics were heavily debated. The paper describes the third cluster sequence as the trace of at least one such merger.

A galaxy built from other galaxies

The Milky Way today is a spiral home to hundreds of billions of stars, among them the Sun, but it did not begin at that scale. It grew in two ways: by forming fresh stars out of collapsing gas clouds, and by pulling in stars, gas and dark matter from smaller galaxies caught in its gravity. Each capture added mass and left marks on the structure that survives in the disk and halo today.

Growth by accretion was already established for the galaxy's more recent past. What the LKH finding changes is the timeline. The Milky Way was drawing in neighboring galaxies within its first two billion years, far earlier than the roughly 10-billion-year horizon scientists had been able to confirm before.

Every single one of these mergers had an influence on the events that have led to the formation of our own solar system and, ultimately, of Earth.

Reconstructing that sequence, Massari said, feeds a larger question about human origins. Some earlier work had held that the Milky Way's first chapters were written by stars born only within it, a purely internal affair. The new data contradict that reading, showing that stars formed in other galaxies must also be counted when astronomers tally where the material of the early galaxy came from.

Chiara Zerbinati, a co-author, said the measurement rested on the quality of the imaging. "Thanks to the high resolution and depth of Hubble imaging, we could measure the age and the metal content of these clusters with unprecedented precision," she said, adding that the Hubble figures were coupled with measurements from Gaia.

Three mergers, three eras

The LKH event joins two other confirmed major mergers in the Milky Way's record, each falling in a different era of cosmic time. Set side by side, they sketch a galaxy assembled in stages rather than born whole, with long stretches of quieter growth in between the large collisions.

Gaia-Sausage-Enceladus, absorbed about 1.8 billion years after LKH, is credited with greatly reshaping the disk of stars that gives the galaxy its present form. The Sagittarius dwarf galaxy, the most recent of the three, has been folding into the Milky Way for roughly the last six billion years and continues to merge with it even now, its stars still being drawn inward.

Smaller captures happened between these landmarks, the researchers note, and the full census is nowhere near complete. LKH stands out only because it is the earliest event that can now be pinned to a specific set of clusters and a specific moment in the galaxy's youth, giving it a fixed place on the timeline.

Most of LKH's debris settled within the inner 6,000 parsecs of the galaxy, roughly the same crowded central region the surveyed clusters occupy.

How the clusters were read

Globular clusters work as fossils because their stars formed together, at the same time and out of the same gas, and so carry a shared chemical fingerprint of their birthplace. A cluster that formed inside a separate galaxy keeps that signature even after its host has been torn apart and swallowed. Age supplies a second clock, and orbital motion a third, letting astronomers read a cluster's history long after the galaxy that made it is gone.

By plotting all three properties against one another, the team could separate clusters that predate a known merger from those that arrived with it. The LKH clusters proved consistently older than the Gaia-Sausage-Enceladus group yet younger than the native ones, regardless of how many heavy elements they contained. That gap in age could only be explained by a distinct, earlier event.

The earliest chapter of the galaxy

An artist's rendering released with the study shows LKH colliding with a young, still-forming Milky Way about 12 billion years ago, its stars streaming into a galaxy that had not yet grown its familiar spiral arms. The image, meant to illustrate rather than photograph the event, was produced by NASA, ESA and Joseph Olmsted of the Space Telescope Science Institute in Baltimore.

For the astronomers, the collision reads less as a catastrophe than as a foundation. The stars and gas that LKH delivered became part of the raw material from which the galaxy, its later generations of stars, and eventually the solar system took shape over the billions of years that followed.

What the team plans next

The group intends to keep surveying globular clusters, including many that Hubble has never examined closely, to build a fuller map of every major merger in the galaxy's past. The aim is a complete accounting of which stars are homegrown and which arrived as the spoils of an ancient collision, and roughly when each population joined.

The study drew on Hubble, in orbit for more than 35 years, alongside Gaia's survey of some two billion stars. Hubble is a NASA and ESA collaboration managed from Goddard Space Flight Center in Maryland, with science operations at the Space Telescope Science Institute in Baltimore.

Massari led the study. Its co-authors include Chiara Zerbinati and Cristiano Fanelli, working across institutions in Italy, Spain, the Netherlands and the United States. It appeared in Nature Astronomy on August 17, 2026, under the digital identifier 10.1038/s41550-026-02931-5.

The finding lands amid a broader effort to trace the Milky Way's ancient mergers one debris field at a time. In March 2026, a team led by Federico Sestito of the University of Hertfordshire reported evidence of a dwarf galaxy nicknamed Loki, after the Norse trickster god, whose metal-poor stars turned up unexpectedly within the galactic disk rather than in the outer halo where such remnants usually settle. That work paired Gaia astrometry with spectroscopy from the Canada-France-Hawaii Telescope.

Why the timing matters

Placing a major merger just two billion years after the Big Bang fits the established view that the largest galaxies grow piece by piece, through repeated accretion of smaller systems rather than in a single formative burst. For the Milky Way, it means the process was underway almost from the start, and that some of its oldest material came from a galaxy that no longer exists.