A Dwarf Galaxy’s Second Act: Tracing Multiple Crossings of Gaia–Enceladus Through the Milky Way

One of the most important discoveries from the Gaia mission is that a large fraction of the Milky Way’s stellar halo did not form in place, but instead came from ancient merger events. Among these, the Gaia–Enceladus/Sausage (GES) merger stands out as one of the most massive and transformative. In this paper, Berni et al. investigate whether GES was accreted in a single event or through multiple passages across the Milky Way disc. By combining stellar motions (“kinematics”) with detailed chemical information, the authors test the idea that GES contains two distinct stellar populations that were stripped from the progenitor galaxy at different times.

Background: Chemical Tagging and the Gaia–Enceladus System

The paper begins by setting the broader context of Galactic archaeology: stars preserve the chemical composition of the gas clouds in which they formed, even after they are dynamically mixed throughout the Galaxy. This makes chemical abundances a powerful tool for “chemical tagging,” where stars with similar chemical patterns are linked to a common origin. Previous studies suggested that GES may itself be chemically and dynamically complex, possibly reflecting multiple crossings of the Milky Way. In particular, earlier work proposed that stars from the outer regions of the GES progenitor were stripped first, followed later by stars from its more tightly bound inner regions.

Data and Methods: Selecting and Classifying GES Stars

To explore this idea, the authors construct their dataset using stars from the APOGEE spectroscopic survey. They apply strict quality cuts and focus on metal-poor stars ([M/H] < −1) to minimize contamination from the Milky Way disc. A machine-learning algorithm called CREEK, developed in earlier work by Berni et al., is then used to identify stars belonging to GES based on both their orbital properties and their chemical abundances. This approach reveals two clear substructures within GES: a low-energy population (Pop 1), which is more tightly bound to the Galaxy, and a high-energy population (Pop 2), which is more loosely bound and extends to higher orbital energies.

Chemical Differences Between the Two Populations

The chemical characterization of these two populations shows subtle but important differences. The authors examine elements such as magnesium, silicon, oxygen, aluminum, and calcium, so-called α-elements that are mainly produced by massive stars and released by core-collapse supernovae. While the overall abundance trends overlap, Pop 1 consistently shows slightly higher [X/Fe] ratios and broader distributions than Pop 2 at the same metallicity. To demonstrate that these differences are real and not due to random scatter, the authors apply several non-parametric statistical tests, all of which confirm that the two populations are chemically distinct for most α-elements.

Interpreting the Results With Chemical Evolution Models

To understand what might have caused these differences, the authors turn to galactic chemical evolution models. These models describe how gas accretion, star formation, stellar nucleosynthesis, and gas outflows shape the chemical history of a galaxy. By varying key parameters such as the gas accretion timescale and star formation efficiency, the authors test whether a single formation scenario can reproduce both populations. They find that Pop 1 is best matched by a model with a very short gas accretion timescale, while Pop 2 requires a much longer one, even though both populations formed with similarly low star formation efficiencies.

Implications: An Inside-Out Galaxy and Multiple Disc Crossings

This result strongly supports an “inside-out” formation scenario for the GES progenitor galaxy. In this picture, the inner regions of the dwarf galaxy formed stars rapidly and became more chemically enriched, while the outer regions evolved more slowly. During the first passage through the Milky Way disc, stars from the outer, less enriched regions (Pop 2) were stripped away. A later passage then removed stars from the inner, more enriched regions (Pop 1). Importantly, this conclusion emerges naturally from the models without assuming a metallicity gradient in advance.

Conclusions: Reconstructing a Complex Accretion History

In summary, Berni et al. provide compelling chemical and dynamical evidence that Gaia–Enceladus was accreted through multiple disc crossings rather than a single event. The presence of two distinct stellar populations, combined with successful chemical evolution modeling, paints a picture of a dwarf galaxy with an internal structure that survived long enough to be stripped in stages. This work reinforces the idea that major merger events leave behind layered chemical fingerprints, offering a detailed glimpse into the Milky Way’s violent early history and the complex lives of its building blocks.

Source: Berni