🔭 astrophysics

Chemical decoding of kinematic substructures in the Galactic halo

This study utilizes chemical abundances from Gaia and APOGEE data to demonstrate that kinematically defined Galactic halo substructures are significantly contaminated by stars from the Gaia-Sausage-Enceladus merger, the Sagittarius dwarf galaxy, the in situ disc, and $\omega$ Centauri, revealing that none of these structures represent unique stellar populations with a single origin.

Original authors: A. Mori, P. Di Matteo, S. Salvadori, M. Mondelin, S. Khoperskov, M. Haywood, A. Mastrobuono-Battisti

Published 2026-05-21

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: A. Mori, P. Di Matteo, S. Salvadori, M. Mondelin, S. Khoperskov, M. Haywood, A. Mastrobuono-Battisti

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the Milky Way galaxy as a giant, cosmic smoothie. Over billions of years, it has swallowed up smaller "satellite" galaxies, mixing their stars into its own. Astronomers have long tried to taste-test this smoothie to figure out exactly which fruits (galaxies) were added and when.

Traditionally, they tried to sort the stars by how they were moving. If a group of stars was flying in the same direction and speed, they assumed those stars must have come from the same "fruit" (the same original galaxy). They called these groups "substructures," like the Gaia Sausage/Enceladus (GSE), Sequoia, Helmi Stream, and Heracles.

However, this new paper argues that sorting stars just by their movement is like trying to sort a mixed bag of M&Ms only by their shape. Because the galaxy is so chaotic, stars from different origins get jumbled up, and stars from our own galaxy get kicked into weird orbits that look like they came from elsewhere.

Here is how the authors "decoded" the true ingredients of the Milky Way's halo using a new method:

1. The Problem: The "Kinematic" Mix-Up

The authors explain that when a massive galaxy crashes into the Milky Way, it doesn't just leave a neat pile of stars. It scatters them everywhere, like a cannonball hitting a sandcastle. This scattering means that stars from the same original galaxy can end up with very different speeds and directions. Conversely, stars that were born right here in the Milky Way's disk can get "heated up" by these crashes and start moving like halo stars.

So, if you pick a group of stars based only on how they move, you are likely getting a messy cocktail of:

Stars from the original crash (the "accreted" stars).
Stars from our own galaxy that got kicked around (the "heated disk").
Maybe even stars from a different crash that happened to land in the same spot.

2. The Solution: Chemical DNA Testing

To fix this, the authors decided to ignore the stars' movement for a moment and look at their chemical composition instead. Think of chemical abundances (like the amount of Iron, Magnesium, or Carbon in a star) as the star's DNA or fingerprint.

Just as a person's DNA reveals their ancestry regardless of where they are currently living, a star's chemical makeup reveals where it was born. The authors used data from the Gaia satellite (which tracks movement) and APOGEE (which tracks chemistry) to look at 7 different chemical elements for over 160,000 stars.

They used a statistical tool called a Gaussian Mixture Model (GMM). Imagine this as a sophisticated "sorting machine" that doesn't just look at the average color of a pile of marbles, but looks at the entire distribution of colors to see if there are actually two or three different piles mixed together.

3. The Big Discoveries

After running their chemical "DNA tests," the authors found that the neat groups astronomers thought they had identified were actually much more mixed up than previously thought.

Sequoia and GSE are likely the same family: The "Sequoia" group of stars looked chemically almost identical to the massive "GSE" crash. The authors argue that Sequoia isn't a separate galaxy at all, but rather the outer, retrograde (backwards-moving) leftovers of the same massive GSE crash. They are essentially siblings from the same parent galaxy.
Heracles is a mix: The "Heracles" group turned out to be a soup of about 66% GSE stars and 20% stars from our own Milky Way disk that got heated up. It's not a unique foreign galaxy; it's a local mess.
Thamnos has a surprise guest: While Thamnos had some GSE stars, the "cleaned" version (removing the GSE and disk stars) looked suspiciously like ω Centauri, a giant globular cluster. This suggests Thamnos might be the debris of the galaxy that ω Centauri came from.
The Helmi Stream is a Sagittarius impostor: The "Helmi Stream" was thought to be a unique ancient crash. But after removing the GSE and disk stars, the remaining stars looked exactly like the Sagittarius Dwarf Galaxy, which is currently crashing into the Milky Way. The Helmi Stream is likely just a part of the Sagittarius stream that got mixed in.
GSE itself is contaminated: Even the massive GSE group isn't pure. It contains stars from the Sagittarius stream, meaning the Sagittarius galaxy has been contaminating the GSE sample for a long time.

4. The Conclusion: No "Pure" Groups

The main takeaway is that none of these "substructures" are unique, single-origin families.

Instead of finding distinct islands of stars from different galaxies, the authors found that every group they studied is a mixture.

Some are mostly GSE stars with a few disk stars.
Some are mostly disk stars with a few GSE stars.
Some are mixtures of GSE, the Sagittarius stream, and our own disk.

The paper concludes that the Milky Way's history is so complex that you cannot simply draw a line on a map of moving stars and say, "These are all from Galaxy X." The chemical fingerprints show that the galaxy is a true "melting pot," where the debris of massive crashes, the leftovers of smaller crashes, and our own local stars are all thoroughly blended together. To understand the galaxy's history, we have to stop looking at the "shape" of the stars and start reading their "chemical DNA."

Technical Summary: Chemical Decoding of Kinematic Substructures in the Galactic Halo

Problem Statement
In the hierarchical mass assembly framework, the Milky Way's accretion history is reconstructed by identifying stellar substructures in the halo based on kinematic coherence (integrals of motion like energy $E$ and angular momentum $L_z$ ) and chemical homogeneity. However, recent studies indicate that integrals of motion are not strictly conserved during massive mergers due to dynamical friction and time-dependent potentials. Consequently, accreted debris spreads across dynamical spaces, and in situ disc stars can be kinematically heated to halo-like orbits. This leads to significant contamination: kinematically defined samples of minor mergers are likely mixed with debris from massive mergers (specifically Gaia Sausage/Enceladus, or GSE) and heated disc stars. Previous studies often assumed these kinematic samples were pure populations, potentially obscuring distinct chemical signatures and misinterpreting the origin of substructures. The fundamental challenge is determining whether observed substructures represent unique accretion events or are merely mixtures of known components.

Methodology
The authors utilize a cross-matched dataset from Gaia EDR3 (astrometry) and APOGEE DR17 (spectroscopy), comprising over 168,000 Milky Way stars with full 6D phase-space information and high-precision chemical abundances. The study focuses on seven elements: [Fe/H], [Mg/Fe], [Si/Fe], [Ca/Fe], [C/Fe], [Al/Fe], and [Mn/Fe], which probe different nucleosynthetic channels (Type Ia and II supernovae, AGB stars, etc.).

The methodology proceeds as follows:

Sample Selection: Kinematic criteria from the literature (e.g., Horta et al. 2023; Koppelman et al. 2019a,b; Massari et al. 2019) are applied to define samples for GSE, Sequoia, Helmi Stream, Thamnos, Heracles, Sagittarius, and $\omega$ Centauri.
Reference Definition: A "pure" GSE reference sample is constructed by fitting a Gaussian Mixture Model (GMM) to the GSE distribution in a 7-dimensional chemical space. The optimal model (selected via Bayesian Information Criterion) uses four Gaussian components. Stars falling within the bulk of this distribution (excluding the lowest 20th percentile outliers) define the "pure GSE" training set. A similar approach defines a metal-poor disc reference ( $\eta > 0.8$ , [Fe/H] $\le -0.8$ ).
Chemical Compatibility Analysis: The GMM trained on the reference samples is applied to every star in the target substructures. For each star, the probability of belonging to the GSE or disc distribution is calculated.
Quantification: The fraction of chemically compatible stars is computed for each substructure relative to GSE and the metal-poor disc. Uncertainties are derived via bootstrap resampling ( $N=1000$ ).
Cleaning: Stars identified as chemically compatible with GSE or the disc are removed to isolate "cleaned" samples, which are then analyzed for residual chemical patterns.

Key Contributions

Systematic Quantification of Contamination: Unlike previous works that compared median trends (which can mask mixed populations), this study employs a probabilistic GMM approach to quantify the percentage of contamination from GSE and the heated disc in kinematically defined substructures on a star-by-star basis.
Re-evaluation of Substructure Origins: The paper challenges the notion that kinematic substructures are unique populations, demonstrating that many are chemically indistinguishable from a mixture of GSE, the disc, and other known streams.
Discovery of Hidden Associations: By analyzing the "cleaned" samples, the authors identify specific chemical overlaps that were previously obscured, linking Thamnos to $\omega$ Centauri and the Helmi Stream to the Sagittarius dwarf spheroidal galaxy.

Results

GSE Contamination: A significant fraction of stars in various substructures are chemically compatible with GSE.
- Heracles: ~66% compatible with GSE.
- Sequoia: ~60% compatible with GSE; the chemical distributions show a robust overlap, suggesting a shared origin.
- Thamnos: ~50% compatible with GSE.
- Helmi Stream: ~40% compatible with GSE.
- Sagittarius: ~31% compatible with GSE, suggesting GSE samples may be contaminated by Sagittarius debris.
- $\omega$ Centauri: Only ~2% compatible with GSE, contradicting hypotheses that it is the nuclear cluster of GSE.
Disc Contamination:
- Substructures on prograde orbits (Helmi Stream) and low-energy retrograde/radial orbits (Thamnos, Heracles) show significant contamination (15–25%) from the metal-poor disc.
- High-energy retrograde systems like Sequoia show negligible disc contamination (~1%).
Cleaned Substructures:
- Sequoia: After removing GSE and disc stars, a small fraction (~5–10%) remains with lower $\alpha$ -abundances at $[Fe/H] \gtrsim -1$ , potentially representing the chemically less enriched outskirts of the GSE progenitor.
- Thamnos: The cleaned sample (~35% of the original) exhibits chemical patterns (specifically high [Al/Fe]) that strongly overlap with $\omega$ Centauri, suggesting a common progenitor or accretion event.
- Helmi Stream: The cleaned sample shows a striking chemical overlap with the Sagittarius stream, supported by spatial and dynamical proximity, indicating the Helmi Stream is likely a mixture of GSE, disc, and Sagittarius debris.

Significance
The paper concludes that none of the studied kinematic substructures appear to be unique populations with a single origin. Instead, they are mixtures of stars chemically compatible with GSE, the metal-poor disc, Sagittarius, and $\omega$ Centauri. The broad distribution of GSE-compatible stars across the $E-L_z$ space implies that defining independent accretion events based solely on kinematic cuts and chemical averages is insufficient. The study emphasizes the necessity of chemical decoding to disentangle the complex merger history of the Milky Way, revealing that structures like the Helmi Stream and Thamnos may be better understood as overlaps of multiple distinct accretion events rather than singular entities.

1. The Problem: The "Kinematic" Mix-Up

2. The Solution: Chemical DNA Testing

3. The Big Discoveries

4. The Conclusion: No "Pure" Groups

More like this