GASTRO Library Chemodynamical Simulations

• GASTRO Library is a suite of controlled galaxy simulations that model chemodynamical signatures and merger impacts in Milky Way analogs.
• It systematically varies supernova feedback, satellite properties, and orbital parameters to quantify effects on stellar halo and disk evolution.
• The framework demonstrates how merger kinematics, feedback processes, and clumpy star formation drive chemical bimodality and stellar migration in galaxies.

The GASTRO Library is a suite of controlled, idealized smoothed particle hydrodynamics (SPH) plus N-body galaxy simulations developed to paper the chemodynamical signatures of merger events in Milky Way analogs, specifically focusing on the Gaia-Sausage/Enceladus (GSE) accretion and its consequences for the Milky Way's stellar halo and disk. The framework delivers a systematic approach for isolating internal, external, and secular processes in galaxy evolution by combining variable feedback prescriptions, meticulously constructed initial conditions, and diagnostic techniques based on chemistry and stellar dynamics. The GASTRO Library is divided into two principal releases: GASTRO I, which targets GSE-like halo debris, and GASTRO II, which expands to disk chemodynamics, chemical bimodality, and the interplay between clumpy star formation and mergers.

1. Design Goals and Framework

The GASTRO Library suite's central objectives are: (i) to map how variations in satellite structure, orbital parameters, and sub-grid feedback physics propagate into observable chemodynamical features in the Galaxy, and (ii) to discern which halo or disk substructures can originate from a single massive progenitor as opposed to requiring multiple distinct accretion events (Amarante et al., 2022, Amarante et al., 1 Dec 2025). The public GASTRO configurations systematically vary:

• Supernova feedback efficiency: Implemented as a fraction (ε_fb = 0.2 or 0.8) of supernova energy coupled to the ISM, regulating the satellite’s internal structure, star formation history, and metal distribution.
• Satellite models: Three major types (d1, d2, d3) each with unique mass, scale lengths, gas temperature, and particle numbers, assembled using GalactICS methodology.
• Host parameters: Milky Way analogs with NFW (Navarro–Frenk–White) halos (M_200 = 10^12 M_⊙), coronal gas components, and realistic spin parameters.
• Merger orbits: Spanning a range of initial radii (150–200 kpc), inclinations (15°), and circularities (η = 0.3 to 0.5, positive for prograde and negative for retrograde).
• Disk and clumpy phases: GASTRO II incorporates both smooth SFH and “clumpy” disk growth triggered by low SN feedback and Toomre instability.

The experimental library is constructed using the GASOLINE code suite, which couples SPH hydrodynamics, cooling (primordial and metal-line), star formation, and a “blastwave” feedback algorithm (Amarante et al., 1 Dec 2025, Amarante et al., 2022).

2. Numerical Implementation and Chemodynamical Modeling

All GASTRO simulations adopt the following constellation of numerical schemes:

• Hydrodynamics: Density calculated via kernel sums, ρ_i = ∑_j m_j W(|r_i−r_j|, h_i), using a cubic spline kernel and adaptive smoothing lengths.
• Gas and star formation physics: A pressure floor (p_floor = 3Gε^2ρ^2) enforces minimum support. Star formation activates for cold (T < 15000 K) and dense (n > 1 cm⁻³) gas exhibiting convergent flows, following either the Stinson et al. or Schmidt–Kennicutt-like laws: $\dot{\rho}_* = c_*\rho_{\rm gas}/t_{\rm ff}$, with c_* = 0.05 and $t_{\rm ff} = \sqrt{3\pi/(32G\rho_{\rm gas})}$.
• Feedback and metal diffusion: Each SN injects $E_{\rm SN}=10^{51}$ erg with ε_fb controlling the fraction coupled as thermal energy; "blastwave" formalism disables radiative cooling after blasts to prevent numerical over-cooling. Turbulent metal diffusion follows Shen et al. (2010).
• Chemical evolution: Full tracking of H, He, O (α-element), and Fe (iron-peak), with source yields from SN II, SN Ia, and AGB stars. Mass-fraction evolution is solved by

$$\frac{dX_k}{dt} = -X_k\dot{m_*} + y_k\dot{m_*} + \sum_{\text{sources}}\mathcal{R}_k(t).$$

Abundance ratios are defined as $$[\alpha/\mathrm{Fe}] = \log_{10}[(\alpha/\mathrm{Fe})/(\alpha/\mathrm{Fe})_\odot]$$ and $$[\mathrm{Fe}/\mathrm{H}] = \log_{10}[(\mathrm{Fe}/\mathrm{H})/(\mathrm{Fe}/\mathrm{H})_\odot]$$.

• Stellar actions and energies: Diagnostic calculations include energy ($E = \frac{1}{2} v^2 + \Phi(r)$ where Φ is the total gravitational potential) and orbital actions (radial $J_R$, azimuthal $J_\phi = L_z$, vertical $J_z$) using the Stäckel fudge (Vasiliev 2019).

3. Simulation Campaigns: GASTRO I and II

GASTRO I concentrates on the chemodynamical footprints in the stellar halo resulting from a single GSE-like accretion event (Amarante et al., 2022). The suite explores:

• How different SN feedback efficiencies alter the satellite’s post-merger spatial distribution, star formation, and chemical gradients (Table 1 in (Amarante et al., 2022)).
• The dynamical and chemical profiles of very retrograde high energy (VRHE) stars, which are the most metal-poor and may be mistaken for distinct merger events (e.g., Sequoia/Arjuna/I’itoi).
• The “Heracles issue,” where the most bound debris is the most metal-rich in GASTRO, contrary to the observed Milky Way’s innermost metal-poor, α-rich stars, suggesting a second massive accretion or an early in-situ component is required.

GASTRO II extends the focus to disk chemodynamics, targeting the origin of the observed [α/Fe] bimodality in the Milky Way disk (Amarante et al., 1 Dec 2025). Methodological innovations include:

• Implementation of both “clumpy” and “non-clumpy” disk assembly modes, set by varying SN feedback from 20% to 80% efficiency.
• Direct comparison of the effects of GSE-like mergers on prograde and retrograde orbits. Retrograde encounters induce strong SFR suppression and open a gap in [α/Fe], while prograde passage leaves the SFR relatively unaffected and fails to generate chemical bimodality.
• Detailed decomposition of stellar populations by formation radius ($R_{\rm form}$) and orbital properties to distinguish the contributions of inner and outer disk stars to chemical structure.

4. Principal Findings and Chemodynamical Diagnostics

The GASTRO Library identifies several key phenomena:

• Supernova feedback modulates satellite survival and chemodynamical signature: Low-feedback dwarfs undergo strong central concentration, develop steep metallicity gradients, and ultimately deposit most of their stars onto radial, low-angular-momentum orbits, whereas high-feedback dwarfs remain diffuse and more rotationally supported.
• Origin of VRHE stars: VRHE stars in all models are primarily drawn from the satellite’s metal-poor outskirts. Their distinct chemo-orbital profile mimics multiple progenitors but is shown to arise from a single satellite.
• Chemical bimodality in disks: Both an internally driven ("clumpy phase") SFR drop and a retrograde GSE-like merger can produce a bimodal [α/Fe] distribution. Clumpy disks generate short-lived (≈0.1–0.3 Gyr) massive clumps ($\sim10^8 M_\odot$), drive an intense starburst, and—following quenching—allow for a low-α sequence to emerge.
• Inner vs. Outer Disk Migration: Chemical bimodality originates in the outer disk (R_form > 4 kpc), with radial migration of inner disk stars amplifying but not generating the split—contradicting purely analytic radial-mixing models.
• Old α-poor disk stars: Only the clumpy-mode runs yield a significant population (10–20%) of old (>11 Gyr), low-α, near-circular-orbit disk stars, as recently observed.

The table below summarizes several key simulation configurations and their outcomes:

GASTRO Run	Feedback ε_fb	Key Phenomenon	Phenomenological Outcome
GASTRO I FB20_d2	0.2	Low-feedback dwarf	Steep Fe/H gradient, VRHE stars, GSE debris
GASTRO I FB80_d1	0.8	High-feedback dwarf	Diffuse remnant, broader Lz, fewer VRHE
GASTRO II: Clumpy	0.2	Intense clumpy SF phase	SFR drop, [α/Fe] bimodality, old α-poor disk
GASTRO II: Merger (retro)	0.8	Retrograde GSE merger	SFR suppression, [α/Fe] bimodality
GASTRO II: Merger (prograde)	0.8	Prograde GSE merger	No SFR quench, unimodal [α/Fe]

5. Interpretation and Implications for Galactic Archaeology

These findings provide a powerful interpretive framework for Galactic structure observations:

• Accreted substructures that appear chemically or dynamically distinct (e.g., Sequoia, Arjuna, I’itoi) may be unified as VRHE debris from a single GSE-like accretion (Amarante et al., 2022).
• The “Heracles” discrepancy—real Milky Way’s low-E, metal-poor innermost stars—signals the necessity for an additional early massive merger or a predominantly in-situ origin.
• The origin of disk [α/Fe] bimodality can result from either early clumpy disk evolution or retrograde mergers, quantitatively matching features observed in the Milky Way (Amarante et al., 1 Dec 2025).
• The formation and migration history of inner/outer disk stars can now be dissected via simulated action distributions and metallicity measurements, enabling critical discrimination between radial-mixing, in-situ, and accreted scenarios.

A plausible implication is the critical role of feedback strength and merger kinematics in reconstructing the Milky Way's early assembly history.

6. Broader Significance and Future Directions

The GASTRO Library offers a controlled laboratory for comparative paper of internal, external, and secular mechanisms in galaxy formation. Its ability to connect observed chemodynamical features with specific physical drivers—feedback-driven clumpiness, orbital geometry of mergers, star formation thresholds—allows for robust confrontation with survey data.

The sequence established by GASTRO I and II provides a roadmap for:

• Quantitative spectral and kinematic tagging of halo substructure origins.
• Establishing the physical conditions under which chemical bimodality in disks appears.
• Predicting the orbital and evolutionary characteristics of clump-driven stellar populations.
• Integrating hydrodynamical simulation outputs with analytic models and cosmological runs for a multiscale approach.

A plausible implication is that future extensions could incorporate variations in cosmological accretion histories, gas fraction, environmental parameters, and AGN feedback to further generalize GASTRO's findings to broader disk galaxy populations.

• GASTRO library I: the simulated chemodynamical properties of several GSE-like stellar halos (Amarante et al., 2022)
• GASTRO library II: Exploring Chemical Bimodalities in Disk Galaxies with GSE-like Mergers and Massive Star-forming Clumps (Amarante et al., 1 Dec 2025)