pith. sign in

arxiv: 2511.15215 · v2 · submitted 2025-11-19 · 🌌 astro-ph.GA

Evolution of action-space coherence in a Milky Way-like simulation

Arunima Arunima , Mark R. Krumholz , Michael J. Ireland , Chuhan Zhang , Sven Buder This is my paper

Pith reviewed 2026-05-17 21:04 UTC · model grok-4.3

classification 🌌 astro-ph.GA
keywords action coherencestellar streamsMilky Way simulationstar cluster progenitorsmoving groupsorbital actionsgalactic disk dynamics
0
0 comments X

The pith

Stars born near each other keep similar orbital actions for up to 0.5 Gyr in a Milky Way-like disk simulation.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper tests whether orbital actions stay conserved enough in a time-varying galactic disk for tracing stars back to dissolved birth clusters. In a high-resolution magnetohydrodynamic simulation, stars formed in close proximity evolve their actions in a correlated manner, preserving similarity over timescales of hundreds of millions of years even as individual actions change substantially. Coherence shows little radial dependence but differs by component, with vertical actions losing correlation beyond a few hundred parsecs while radial and azimuthal actions stay linked across kiloparsec scales. These measurements support a probabilistic framework for estimating the initial sizes of star cluster progenitors from the action distributions of present-day stellar streams, which the authors apply to 438 known moving groups.

Core claim

In the simulation, stars experience significant action evolution over roughly 100 Myr but maintain correlated actions when born nearby, with vertical actions decohering for birth separations greater than a few hundred parsecs and radial and azimuthal actions remaining correlated on kiloparsec scales for up to 0.5 Gyr. These decoherence rates enable a probabilistic model that infers the initial sizes of dissolving clusters from the action spreads observed in stellar streams today.

What carries the argument

The measured rate of action decoherence across different birth separations and components, which is converted into a probabilistic mapping from present-day action distributions back to original cluster sizes.

If this is right

  • Most of the 438 examined moving groups likely originated in compact clusters rather than as resonant or induced structures.
  • The method reduces the need for expensive spectroscopic abundance measurements by first classifying streams according to their likely birth origins.
  • Upcoming higher-precision kinematic data will tighten the constraints on which streams come from clusters versus dynamical processes.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Action coherence measurements could be combined with chemical tagging to cross-validate cluster identifications in the Milky Way.
  • If real galactic perturbations are stronger than in the simulation, coherence times would shorten and more streams might be reclassified as non-cluster structures.
  • The same framework could be run on simulations with varied spiral arm strengths to predict how cluster reconstruction success changes across different galaxy types.

Load-bearing premise

The simulation's time-varying non-axisymmetric features such as spiral arms and giant molecular clouds produce action evolution that matches the real Milky Way disk.

What would settle it

Measuring the actual spread in actions among stars from a known young open cluster and checking whether it matches the spread predicted by the simulation's decoherence rates for that cluster's size and age.

Figures

Figures reproduced from arXiv: 2511.15215 by Arunima Arunima, Chuhan Zhang, Mark R. Krumholz, Michael J. Ireland, Sven Buder.

Figure 1
Figure 1. Figure 1: The distribution of 5th nearest neighbour distances 𝑑5 for coeval stars at different ages (shown in the legend). We show the 30 pc limit that we use to separate cluster stars from dispersed stars as the red dashed vertical line. 2.3 Quantifying action-space coherence The overarching goal of this work is to determine whether stars born together remain clustered in action space even as they drift apart in ph… view at source ↗
Figure 2
Figure 2. Figure 2: Median of absolute change in action difference (row-wise in order: 𝛿absΔ𝐽𝑅, 𝛿absΔ𝐽𝑧 and 𝛿absΔ𝐽𝜙) versus time for coeval pairs of stars binned by stellar separation at birth; birth distance bins are indicated by colour, as shown in the legend. For comparison, the blue dashed line shows the change in the action of single stars relative to their own initial actions as a function of age gap; this result is tak… view at source ↗
Figure 3
Figure 3. Figure 3: Same as [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Median of relative change in action difference (row-wise in order: 𝛿relΔ𝐽𝑅, 𝛿relΔ𝐽𝑧 and 𝛿relΔ𝐽𝜙 in order) for coeval pairs of stars in time with their birth distances in the 0.5 – 2 pc bin. The darker blues represent higher radii. The red curve shows the median of the relative change in action differences for the entire dataset of pairs of coeval stars born within 0.5 – 2 pc of each other. MNRAS 000, 1–14 … view at source ↗
Figure 5
Figure 5. Figure 5: Relative action differences calculated for some of the observed moving groups (markers) placed on the theoretical grid of action-space evo￾lution (coloured lines indicating different birth distance bins as shown in the legend). Rows from top to bottom show 𝛿relΔ𝐽𝑅,𝑧,𝜙, respectively. Each point represents the median of the distribution { 𝛿Δ𝐽obs } and age of a moving group, with error bars showing the 16th a… view at source ↗
Figure 6
Figure 6. Figure 6: Coloured lines show examples of cumulative distributions corre￾sponding to the one-dimensional probability distributions, P1D (𝑥 | 𝑎, 𝜏), of log vertical action difference 𝑥 = log 𝛿relΔ𝐽𝑧 , for different initial separations 𝑑init (as indicated in the legend) and stellar populations of age 𝜏 = 148 Myr. The top and bottom panels show results for the full and truncated distribu￾tions, respectively – see main … view at source ↗
Figure 7
Figure 7. Figure 7: Cumulative distribution corresponding to the marginal posterior distribution, P (𝑎 | {𝑥𝑖 }, 𝜇, 𝜎), for HSC 2282, given its observed action differences {𝑥𝑖 }, age 𝜇 = 148.17 Myr and 𝜎 = 125.64 Myr. This CDF is derived for the non-truncated assumption (see main text). The median logarithmic initial size 𝑎50 = log 𝑑init50 (in units of pc) is shown as the solid green vertical line while the 84th and 97th perce… view at source ↗
Figure 9
Figure 9. Figure 9: Distribution of the 𝐿 1 𝑓 (purple) and 𝐿 1 𝑓 ,trunc (green) values that characterise how well the observed distribution of action differences matches the best-fitting simulation-predicted distribution. The black solid line repre￾sents the 90th percentile of 𝐿 1 𝑓 while the dashed line shows the same for 𝐿 1 𝑓 ,trunc. unfilled histograms show the results if we do not omit these less well-fit streams; the ov… view at source ↗
Figure 8
Figure 8. Figure 8: Distribution of the inferred median logarithmic initial sizes (𝑎50; in units of pc) for all streams in our sample. The shaded histogram highlights the subset of “well-fit” streams, selected by applying a cut on their 𝐿 1 𝑓 values. The top panel shows results derived from the full posterior probability distribution, P (𝑎 | {𝑥𝑖 }, 𝜇, 𝜎), which is the appropriate distribution if the observed samples are compl… view at source ↗
Figure 10
Figure 10. Figure 10: Cumulative distributions of the inferred logarithmic initial sizes (median 𝑎50, 84th percentile 𝑎84 and 97.5th percentile 𝑎97) for the “well-fit” streams in our sample represented by the colours specified in the legend. of ∼ 30 pc. The fact that most inferred birth sizes are tens to hun￾dreds of times smaller suggests that a large fraction of these systems were compact stellar clusters that have since bec… view at source ↗
Figure 12
Figure 12. Figure 12: Distribution of the inferred median logarithmic initial sizes (𝑎50) for all streams in our sample, plotted as a function of their median stellar ages (black points). Blue contours indicate regions of higher density in the (𝑎50, age) plane, highlighting the overall structure of the distribution. find that streams identified as disrupting clusters (those lying to the left of the figure) span the full age ra… view at source ↗
Figure 13
Figure 13. Figure 13: Distribution of the completeness fraction ( 𝑓comp) for all streams in our sample shown in blue. The pink histogram shows the same only for streams whose inferred initial size 𝑑trunc50 < 10 pc so they likely began as compact clusters instead of the probable resonant structures. This separation is useful to explore whether the streams that are more likely to have formed as compact clusters exhibit different… view at source ↗
read the original abstract

Efforts to dynamically trace stars back to the now-dissolved clusters in which they formed rely implicitly on the assumption that stellar orbital actions are conserved. While this holds in a static, axisymmetric potential, it is unknown how strongly the time-varying, non-axisymmetric structure of a real galactic disk drives action drift that inhibits cluster reconstruction. We answer this question using a high-resolution magnetohydrodynamic simulation of a Milky Way-like spiral disc galaxy. We show that, while stars experience significant action evolution over $\lesssim 100$ Myr, they do so in a correlated fashion whereby stars born in close proximity maintain very similar actions for up to 0.5 Gyr. The degree of coherence shows no significant dependence on galactocentric radius, but varies between action components: vertical actions decohere for stars born more than a few hundred parsecs apart (likely due to giant molecular clouds), while radial and azimuthal actions remain correlated on kiloparsec scales (likely influenced by spiral arms). We use our measurements of the rate of action decoherence to develop a probabilistic framework that lets us infer the initial sizes of the star cluster progenitors of present-day stellar streams from their measured action distributions, which we apply to 438 known moving groups. Our results suggest that most of these streams likely originated from compact clusters, but that a significant minority are instead likely to be resonant or dynamically induced structures. This method of classifying streams complements existing methods, optimises the use of expensive spectroscopic abundance measurements, and will be enhanced by the more precise kinematic data that will soon become available from \textit{Gaia} DR4.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper uses a high-resolution MHD simulation of a Milky Way-like disk to track stellar action evolution, showing that stars born nearby maintain correlated actions for up to 0.5 Gyr despite significant drift over ~100 Myr. Vertical actions decohere for birth separations beyond a few hundred parsecs (attributed to GMCs), while radial and azimuthal actions remain coherent on kpc scales (attributed to spiral arms), with no strong galactocentric radius dependence. These empirical decoherence rates are used to construct a probabilistic framework that classifies 438 observed moving groups as likely compact-cluster remnants or resonant/dynamically induced structures.

Significance. If the quantitative scales hold, the work is significant for galactic dynamics: it provides concrete measurements of how non-axisymmetric, time-varying structures erode action conservation, directly addressing a key assumption in dynamical cluster reconstruction. The probabilistic classification offers a practical complement to chemical tagging and abundance studies, optimizing follow-up observations ahead of Gaia DR4. The simulation-based derivation of falsifiable action-spread predictions is a methodological strength.

major comments (2)
  1. [§3] §3 (simulation results): the reported 0.5 Gyr coherence timescale, few-hundred-pc vertical decoherence scale, and kpc-scale radial/azimuthal scales are extracted from particle tracking in a single high-resolution MHD run. No variations across multiple realizations, resolution changes, or controlled sweeps of spiral-arm strength/GMC properties are presented, so the mapping from observed action spreads to progenitor sizes inherits unquantified model dependence that is load-bearing for the classification of the 438 moving groups.
  2. [§5] §5 (probabilistic framework): the framework is built directly from decoherence rates measured in the same simulation used to derive the scales, creating moderate circularity. The manuscript would be strengthened by an explicit discussion of how the inferred cluster sizes would shift under plausible variations in the input rates (e.g., factor-of-two changes in vertical scale) or by calibration against an independent external constraint.
minor comments (1)
  1. [Figure 2] Figure 2 caption: the time intervals and birth-separation bins shown in the panels should be stated explicitly in the caption rather than only in the main text for clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive review and positive assessment of the work's significance. We address each major comment below and have revised the manuscript to incorporate additional discussion of model dependence and a sensitivity analysis for the probabilistic framework.

read point-by-point responses

  1. Referee: [§3] §3 (simulation results): the reported 0.5 Gyr coherence timescale, few-hundred-pc vertical decoherence scale, and kpc-scale radial/azimuthal scales are extracted from particle tracking in a single high-resolution MHD run. No variations across multiple realizations, resolution changes, or controlled sweeps of spiral-arm strength/GMC properties are presented, so the mapping from observed action spreads to progenitor sizes inherits unquantified model dependence that is load-bearing for the classification of the 438 moving groups.

    Authors: We agree that the results derive from a single simulation and that a multi-realization study or controlled parameter sweeps would better quantify uncertainties from specific subgrid choices or initial conditions. Such an exploration is computationally prohibitive at the required resolution. We have added a new paragraph to the revised §3 that discusses the robustness of the key trends: the separation of vertical decoherence (driven by GMCs) from in-plane coherence (driven by spirals) is expected to persist across Milky Way-like models with similar non-axisymmetric structure. We also provide a qualitative estimate of how plausible changes in GMC mass spectrum or spiral strength would affect the reported scales, and we now flag the resulting systematic uncertainty on absolute progenitor sizes in the classification of the 438 groups while arguing that the compact-versus-extended distinction remains informative. revision: partial

  2. Referee: [§5] §5 (probabilistic framework): the framework is built directly from decoherence rates measured in the same simulation used to derive the scales, creating moderate circularity. The manuscript would be strengthened by an explicit discussion of how the inferred cluster sizes would shift under plausible variations in the input rates (e.g., factor-of-two changes in vertical scale) or by calibration against an independent external constraint.

    Authors: We acknowledge the moderate circularity inherent in using the same simulation both to measure decoherence rates and to construct the classification framework. In the revised §5 we have added an explicit sensitivity subsection that recomputes the posterior probabilities for all 438 moving groups after varying the input decoherence scales by factors of two (e.g., vertical scale 200–400 pc, radial/azimuthal scales 1–2 kpc). The analysis shows that the overall conclusion—most groups consistent with compact progenitors, with a significant minority likely resonant or dynamically induced—remains stable, although the precise fractions shift by up to ~15 %. We also outline prospects for future external calibration using chemical-tagging results from upcoming surveys and comparisons with independent simulations. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation uses external simulation benchmark

full rationale

The paper extracts action decoherence rates and coherence scales directly from particle tracking within a single high-resolution MHD simulation, then constructs a probabilistic framework that applies those measured rates to classify external observational data consisting of 438 real moving groups. This constitutes standard model calibration followed by application to independent data rather than any self-definitional loop, fitted prediction on the same dataset, or load-bearing self-citation. The simulation functions as an external dynamical benchmark whose outputs (decoherence timescales and spatial scales) are not redefined or re-derived from the framework itself, and no equations reduce the classification results to the input measurements by algebraic construction.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the simulation's ability to produce realistic non-axisymmetric perturbations and on the assumption that measured decoherence rates can be extrapolated to real Milky Way streams without additional free parameters beyond those fitted in the framework.

free parameters (1)
  • decoherence length scales
    Lengths (few hundred pc for vertical, kpc for radial/azimuthal) are measured from the simulation and used to set the probabilistic model.
axioms (1)
  • domain assumption Orbital actions are approximately conserved except for perturbations from spiral arms and giant molecular clouds in the simulation.
    Invoked in the opening paragraph to motivate the study.

pith-pipeline@v0.9.0 · 5602 in / 1266 out tokens · 28292 ms · 2026-05-17T21:04:12.269557+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    We show that, while stars experience significant action evolution over ≲100 Myr, they do so in a correlated fashion whereby stars born in close proximity maintain very similar actions for up to 0.5 Gyr. ... vertical actions decohere for stars born more than a few hundred parsecs apart ... radial and azimuthal actions remain correlated on kiloparsec scales

  • IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    We use our measurements of the rate of action decoherence to develop a probabilistic framework that lets us infer the initial sizes of the star cluster progenitors

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

1 extracted references · 1 canonical work pages · 1 internal anchor

  1. [1]

    Abdurro’uf et al., 2022, ApJS, 259, 35 Antoja T., et al., 2018, Nature, 561, 360 Antoja T., Ramos P., García-Conde B., Bernet M., Laporte C. F. P., Katz D., 2023, A&A, 673, A115 Armillotta L., Krumholz M. R., Di Teodoro E. M., McClure-Griffiths N. M., 2019, MNRAS, 490, 4401 AroraA.,SandersonR.E.,PanithanpaisalN.,CunninghamE.C.,WetzelA., Garavito-Camargo N...

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1007/978-1-4020-3407-7_5 2022