🔭 astrophysics

Impact of selection criteria on the structural parameters of the Galactic thin and thick discs

This study demonstrates that while chemical and age-based selection methods provide the cleanest separation of the Milky Way's thin and thick discs, all tested classification techniques consistently reveal a flaring thin disc with a scale length between 2.3 and 3.0 kpc and a non-flaring thick disc with a scale length of approximately 2.0 kpc.

Original authors: Simon Alinder, Thomas Bensby, Paul McMillan

Published 2026-06-15

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Simon Alinder, Thomas Bensby, Paul McMillan

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, spinning cosmic pizza. For decades, astronomers have agreed that this pizza has two main layers of toppings: a thin, crispy crust (the thin disc) and a thicker, fluffier layer underneath (the thick disc).

But here's the problem: these two layers aren't perfectly separated like distinct slices of cheese and pepperoni. They are more like a marbled cake where the chocolate and vanilla swirl into each other. Because they mix, it's very hard to tell which star belongs to which layer.

This paper is essentially a "taste test" to see how different ways of sorting the stars change our picture of the galaxy. The authors tried five different "sorting recipes" to separate the stars into the thin and thick discs and then measured the shape of the galaxy based on those sorted groups.

The Five Sorting Recipes

The researchers used data from a massive telescope survey called APOGEE (which looks at the chemical makeup of stars) and age estimates from a computer program called astroNN. They tried five different methods to sort the stars:

The Chemical Recipe (Magnesium & Aluminum): They looked at the ratio of magnesium to aluminum in the stars. Think of this like checking the specific spice blend in a soup. Some stars have a "thick disc spice mix," while others have a "thin disc spice mix." This method was very good at telling the difference between the galaxy's own stars and stars that were stolen from other galaxies (accreted stars).
The Chemical Recipe (Alpha & Iron): Similar to the first, but they compared "alpha" elements (like magnesium) to iron. This is like checking if the soup is salty or spicy. It's a classic way to separate the layers but sometimes gets confused by the "stolen" stars.
The Movement Recipe (Kinematics): Instead of chemistry, they looked at how fast and in what direction the stars were moving. Thick disc stars are like rowdy teenagers running around the kitchen (high speed, erratic paths), while thin disc stars are like calm adults walking in a straight line. However, because the "rowdy" and "calm" groups often run in the same direction, this method got a lot of the groups mixed up.
The Orbit Recipe (Dynamics): This is a more advanced version of the movement recipe. It calculates the stars' orbital paths (like calculating the exact trajectory of a thrown ball). While this sounds precise, the paper found that because the populations are so mixed, this method also struggled to draw a clean line between the two discs.
The Age Recipe: They simply asked, "How old is this star?" The idea is that the thick disc is the "old generation" and the thin disc is the "young generation." However, the paper found that the "young" thin disc and the "old" thick disc have ages that overlap significantly, making it hard to draw a hard line in the sand.

What They Found

After sorting the stars with these five different methods, they measured the shape of the galaxy. Here is what they discovered:

The "Flaring" Effect: No matter which sorting method they used, they found that the thin disc gets "flared" as you move away from the center of the galaxy. Imagine the pizza crust getting thicker and puffier as you move from the center to the edge. The thick disc, however, stays roughly the same thickness everywhere (about 1,000 light-years thick).
The Size Difference: Every single method agreed on one thing: the thin disc stretches out much farther from the center of the galaxy than the thick disc does. If the thick disc is a small, dense island, the thin disc is a vast, sprawling archipelago.
The "Messy Middle": The chemical methods (looking at what the stars are made of) gave the cleanest separation, like sorting M&Ms by color. The movement and age methods were messier, like trying to sort M&Ms by how fast they roll or how long they've been in the bag. Because the populations are so well-mixed, these methods often accidentally put stars from the thick disc into the thin disc group and vice versa.

The Big Takeaway

The main conclusion is that how you choose to sort the stars changes the map you draw.

If you use a chemical recipe, you get a very clear picture of two distinct layers, but you need expensive telescope data to do it. If you use a movement or age recipe, you can sort more stars easily, but your map will be fuzzier and might show a slightly different shape for the galaxy.

The authors warn that we can't just pick one method and say, "This is the true shape of the Milky Way." The "thick disc" and "thin disc" aren't perfectly separate physical boxes; they are overlapping populations. Depending on which "lens" (chemical, movement, or age) you look through, the galaxy looks slightly different.

In short: The Milky Way is a complex, mixed-up family. Depending on whether you sort your family members by their last names, their height, or their age, you might end up with slightly different family trees, even though they are all the same people.

Technical Summary: Impact of Selection Criteria on the Structural Parameters of the Galactic Thin and Thick Discs

Problem Statement
The dual nature of the Milky Way disc, comprising distinct thin and thick components, is a cornerstone of Galactic structure studies. However, these components exhibit significant overlap in their chemical, kinematic, and spatial distributions, particularly at higher metallicities. While various classification methods exist—ranging from geometric and kinematic to chemical and chronological—there is no consensus on whether these different definitions yield equivalent stellar populations. Consequently, the choice of selection criteria may significantly bias the inferred structural parameters (such as scale heights and scale lengths) of the Galactic disc. This paper investigates how different classification methodologies influence the determination of these structural properties.

Methodology
The authors utilized red giant stars from the APOGEE DR17 survey, supplemented with stellar age estimates from the astroNN catalogue. The sample was restricted to Galactocentric radii ( $R_{gal}$ ) between 4 kpc and 14 kpc to avoid the non-axisymmetric Galactic bar and regions with low star counts. A selection function was applied to ensure the sample was representative of the true population.

Five distinct selection methods were applied to categorize stars into thin disc, thick disc, and halo components:

Chemical ([Mg/Mn]-[Al/Fe]): Separates populations based on the distribution in the magnesium-manganese versus aluminum-iron plane, distinguishing accreted stars from in-situ disc stars.
Chemical ([Mg/Fe]-[Fe/H]): Uses a cut in the alpha-element-to-iron versus metallicity plane to separate high- $\alpha$ (thick) and low- $\alpha$ (thin) populations.
Dynamical ( $J_\phi$ - $J_Z$ ): Classifies stars based on orbital actions (azimuthal and vertical) within a model of the Galactic gravitational potential.
Chronological (Age): Separates populations based on estimated stellar ages, using an 8 Gyr threshold to distinguish the older thick disc from the younger thin disc.
Kinematic: A likelihood-based method using 3D stellar velocities to assign membership probabilities, restricted to the solar neighborhood ( $7 < R_{gal} < 9$ kpc).

For each method, the authors derived relative density profiles as functions of height above the Galactic plane ( $|Z|$ ) and Galactocentric radius. These profiles were fitted to a simple two-exponential disc model (plus a constant halo term) using a multinomial likelihood approach and Markov chain Monte Carlo (MCMC) sampling. Scale heights ( $H$ ) and scale lengths ( $R$ ) were derived for each component.

Key Results

Separation Purity: Methods based on chemical abundances or age data produced the cleanest separations between disc components. In contrast, kinematic and dynamical methods suffered from higher contamination due to the difficulty of separating well-mixed populations in velocity and action space.
Scale Heights:
- The thin disc scale height exhibits clear flaring, increasing with Galactocentric radius from $\sim$ 100–400 pc in the inner disc to $\sim$ 1–2 kpc in the outer disc.
- The thick disc scale height remains approximately constant at $\sim$ 1 kpc across most radii for all methods, showing no significant flaring.
- The age-based method yielded a slightly different trend for the thin disc scale height, remaining flatter ( $\sim$ 400–500 pc) before rising in the outermost bins.
Scale Lengths: All methods consistently found the thin disc to have a longer scale length than the thick disc.
- Assuming a thick disc scale length of 2.0 kpc, the thin disc scale length was found to range between 2.3 and 3.0 kpc, depending on the selection method.
- The chemical selection methods yielded the largest difference between the scale lengths of the two discs.
Disc Fractions: The fraction of thick disc stars in the Galactic plane decreases exponentially with radius, dropping from 4–14% in the inner disc to less than 2% at $R_{gal} \ge 12$ kpc.

Significance and Conclusions
The paper concludes that the method chosen to distinguish between Galactic components significantly impacts the resulting view of the Galaxy's structure.

Trade-offs: Abundance-based approaches offer high purity and clear separation but require spectroscopic data, which is not as widely available as photometric or astrometric data. Conversely, kinematic and dynamical methods can be applied to larger samples but suffer from less distinct separations due to the mixing of populations.
Comparison to External Galaxies: Unlike studies of external galaxies where thick discs often appear to have longer scale lengths than thin discs (based on geometric luminosity profiles), this study finds the opposite for the Milky Way when using chemically or kinematically defined components.
Methodological Impact: The authors emphasize that inconsistent classifications across methods raise questions about the interpretation of disc structure. They suggest that while no single method provides a perfect separation due to the well-mixed nature of the stellar populations, combining multiple approaches (e.g., chemical and dynamical) may offer a superior selection strategy. The study highlights the necessity of acknowledging methodological biases when deriving structural parameters of the Milky Way.

The Five Sorting Recipes

What They Found

The Big Takeaway

More like this