[Paper Review] Galactic Chemical Evolution Models Favor an Extended Type Ia Supernova Delay-Time Distribution

Liam O. Dubay, Jennifer A. Johnson+1 more|arXiv (Cornell University)|Apr 11, 2024

Gamma-ray bursts and supernovae4 citations

TL;DR

This study uses multi-zone galactic chemical evolution models with radial stellar migration to test delay-time distributions (DTDs) for Type Ia supernovae (SNe Ia). By comparing predictions to APOGEE data, it finds that an extended DTD—producing fewer prompt SNe Ia—better fits the observed [α/Fe] distribution in the Milky Way, favoring a DTD flatter than the fiducial t⁻¹ power law.

ABSTRACT

Type Ia supernovae (SNe Ia) produce most of the Fe-peak elements in the Universe and therefore are a crucial ingredient in galactic chemical evolution models. SNe Ia do not explode immediately after star formation, and the delay-time distribution (DTD) has not been definitively determined by supernova surveys or theoretical models. Because the DTD also affects the relationship among age, [Fe/H], and [$α$/Fe] in chemical evolution models, comparison with observations of stars in the Milky Way is an important consistency check for any proposed DTD. We implement several popular forms of the DTD in combination with multiple star formation histories for the Milky Way in multi-zone chemical evolution models which include radial stellar migration. We compare our predicted interstellar medium abundance tracks, stellar abundance distributions, and stellar age distributions to the final data release of the Apache Point Observatory Galactic Evolution Experiment (APOGEE). We find that the DTD has the largest effect on the [$α$/Fe] distribution: a DTD with more prompt SNe Ia produces a stellar abundance distribution that is skewed toward a lower [$α$/Fe] ratio. While the DTD alone cannot explain the observed bimodality in the [$α$/Fe] distribution, in combination with an appropriate star formation history it affects the goodness of fit between the predicted and observed high-$α$ sequence. Our model results favor an extended DTD with fewer prompt SNe Ia than the fiducial $t^{-1}$ power law.

Motivation & Objective

• To determine which delay-time distribution (DTD) for Type Ia supernovae best reproduces observed stellar abundance patterns in the Milky Way.
• To assess the impact of radial stellar migration on chemical evolution model predictions using hydrodynamical simulation analogues.
• To evaluate whether the DTD, combined with realistic star formation histories, can explain the bimodal [α/Fe] distribution observed in APOGEE data.
• To test multiple DTD forms against observed interstellar medium abundance tracks and stellar age distributions.

Proposed method

• Implemented multiple DTD forms (including t⁻¹, exponential, and extended power-law) in VICE-based multi-zone chemical evolution models.
• Incorporated radial stellar migration using a Gaussian sampling scheme derived from the h277 hydrodynamical simulation, with σ_R_M scaling as τ^0.33 × R_form^0.61.
• Used APOGEE DR17 data as observational constraints for [α/Fe], [Fe/H], and stellar age distributions.
• Fitted a sech² function to vertical phase-space distributions from h277 to model scale height evolution with age and radius.
• Mapped h277 star particles to VICE stellar populations using formation radius, age, and migration parameters to preserve realistic abundance trends.
• Compared predicted abundance tracks and distributions to observed APOGEE data using statistical goodness-of-fit metrics.

Experimental results

Research questions

• RQ1 Does the delay-time distribution (DTD) of Type Ia supernovae significantly affect the predicted [α/Fe] distribution in the Milky Way?
• RQ2 Can an extended DTD—producing fewer prompt SNe Ia—better reproduce the observed bimodal [α/Fe] distribution in APOGEE data?
• RQ3 How does radial stellar migration, as modeled from hydrodynamical simulations, influence the predicted abundance distributions in chemical evolution models?
• RQ4 To what extent can the DTD alone, or in combination with star formation history, explain the observed [α/Fe] trends in the Milky Way's disk?

Key findings

• The DTD has the largest impact on the [α/Fe] distribution, with more prompt SNe Ia leading to a lower median [α/Fe] ratio in the stellar population.
• An extended DTD with fewer prompt SNe Ia provides a significantly better fit to the high-α sequence in the observed [α/Fe] distribution compared to the fiducial t⁻¹ DTD.
• The model results favor a DTD that is flatter than the t⁻¹ power law, indicating a longer delay time for a larger fraction of SNe Ia.
• Radial migration, modeled via a time- and radius-dependent Gaussian dispersion (σ_R_M ∝ τ^0.33 × R_form^0.61), reduces artificial clumping in abundance distributions and improves model realism.
• The inclusion of realistic migration and star formation histories is essential for accurately reproducing the observed [α/Fe] distribution, as DTD alone cannot explain the bimodality.
• The scale height of stellar populations (h_z) evolves with age and radius, following h_z ∝ exp(τ/τ_s + R_final/R_s), which improves vertical abundance distribution modeling.