A new classification of the Milky Way’s globular clusters

Globular clusters: galactic fossils with an erased memory

Globular clusters are among the oldest inhabitants of our Galaxy. These dense spheres, made up of hundreds of thousands of stars bound together by gravitational forces, are veritable cosmic fossils. Studying them enables us to trace the history of the Milky Way and understand how it has formed and enriched over time.

Some of these clusters were born in situ, directly in our young Galaxy around 12 billion years ago (spectral shift of z = 3), while others are ex situ, formed in satellite galaxies before being captured by the Milky Way around 10 billion years ago (z = 2) during accretions and mergers (cf. Figure 1).

Distinguishing between these two families today (z = 0) is not straightforward. Traditionally, astronomers have used the current orbital properties of clusters, such as their speed and position, to try to trace their origin. But these methods have their limits: over billions of years, the Milky Way’s turbulent past has blurred the traces of these objects. Its mass has increased, its gravitational field has evolved, and ex situ clusters have gradually lost the memory of their initial trajectory, making them difficult to identify.

An unprecedented cosmological approach

In order to better trace the history of globular clusters, a European team led by the LIRA, in collaboration with researchers from the Institute for Astrophysics Potsdam (Germany) and the Lund Observatory (Sweden), has developed a novel method: simulating the evolution of galaxies similar to our own in order to study the history of their cluster populations over several billion years.

To do this, the researchers simulated nearly 18,000 globular clusters in around 200 galaxies, from the early Universe (z = 3) to the present day. Their model takes into account the main physical processes influencing the evolution of these systems, such as galaxy mergers, gravitational friction and the loss of cluster mass. By reproducing the dominant effects of galactic evolution on such a vast sample, the team obtains a statistically robust view of the formation and evolution of globular clusters, rather like unrolling a film retracing the history of our Galaxy.

By combining the results of the simulations with real data from Gaia, which provides the positions and velocities of 161 clusters in our Galaxy, the team has revealed an almost equal proportion between in situ and ex situ clusters in the Milky Way - a result that contrasts sharply with previous estimates, highlighting the role of galactic accretions and mergers in the construction of our Galaxy (see Figure 2). The study also confirms the ex situ origin of certain emblematic systems, such as Omega Centauri or the clusters associated with the Sagittarius dwarf galaxy, demonstrating the remarkable resistance of these clusters to tidal forces. However, the researchers point out that the distinction between in situ and ex situ clusters is blurrier than previously thought. While in situ clusters are mainly observed in the central region of the galaxy, ex situ clusters occupy all of galactic space, creating a mixing zone where the two populations overlap. As a result, the origin of some clusters remains ambiguous, and could be clarified by future spectroscopic measurements.

JWST and Euclid: a new era in the study of globular clusters

Apart from the revolution brought about by Gaia, which has provided information of unprecedented precision on globular clusters, this data concerns just one example: our own.
It is therefore essential to extend our observations to clusters located in other galaxies in order to enrich the available statistics (see Figure 3). This approach is crucial for testing the models of formation and evolution of globular clusters, but also for refining our cosmological understanding, by validating or discarding certain alternative dark matter scenarios - clusters are in fact excellent tracers of dark matter [1].

The Euclid mission will paint an unprecedented portrait of globular cluster populations by detecting potentially half a million extragalactic clusters up to nearly 100 Mpc. Unlike the Hubble Space Telescope, which has a small field of view, Euclid’s wide survey will cover vast areas of the sky in a single image, making it possible to characterise clusters around galaxies of all types, in high or low density environments, and even to locate clusters that are very far from their host galaxy, well beyond our current capabilities.

At the same time, the James Webb Space Telescope (JWST) will be observing forming clusters at very high redshift, with exceptional sensitivity and spatial resolution. Combined with the magnifying effect of natural gravitational lenses, JWST is already making it possible to distinguish structures of a few tens of parsecs at z > 4, and even to identify individual stellar clusters - potential future globular clusters. These observations pave the way for the direct study of the formation of globular clusters in the young Universe, while reinforcing and consolidating our models of cluster formation.

Thanks to JWST, Gaia and Euclid will together be able to offer a coherent vision, from clusters forming in the early Universe to mature populations in the local Universe.

Contact: Pierre Boldrini

Article: Boldrini, P., Di Matteo, P., Laporte, C. et al. A new classification of ex situ and in situ Galactic globular clusters based on a method trained on Milky Way analogues in the TNG50 cosmological simulations, A&A (2025).