Modeling the Ages and Chemical Abundances of Elliptical Galaxies

arXiv:2407.07971 · astro-ph.GA · Submitted 2024-07-10 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: We're looking at the paper: "Modeling the Ages and Chemical Abundances of Elliptical Galaxies".

Jocelyn: Here's the gist of the paper. Spectroscopic studies of elliptical galaxies show that their stellar population ages, mean metallicity, and α-enhancement traced by

Mg/Fe: all increase with galaxy stellar mass or velocity dispersion Modeling the Ages and Chemical Abundances of Elliptical Galaxies.

Vera: Modeling the Core Relationships The authors use one-zone galactic chemical evolution (GCE) models with a flexible star formation history (SFH) to model the age,

Mg/H: , and

Mg/Fe: inferred from simple stellar population (SSP) fits to observed ellipticals at z ∼ 0 and z ∼ 0.

Jocelyn: They show that an SSP fit to the spectrum computed from a full GCE model gives ages and abundances close to the light-weighted, logarithmically averaged values of the composite stellar population, ⟨age⟩, ⟨

Mg/H: ⟩, and ⟨

Mg/Fe: ⟩ The predicted ⟨

Mg/H: ⟩-⟨age⟩ and ⟨

Mg/Fe: ⟩-⟨age⟩ relations are surprisingly insensitive to SFH parameters, but the star formation efficiency and outflow efficiency affect the early and late evolution of ⟨

Mg/H: ⟩, respectively.

Vera: Input Yields and Model Assumptions The authors adopt values for supernova Mg and Fe yields motivated by Milky Way stellar populations They use core collapse supernova (CCSN) yields calibrated to the plateau

Mg/Fe: cc ≈ 0.45 observed in many Milky Way studies, finding that predicted ⟨

Mg/Fe: ⟩ ratios of ellipticals are underpredicted by 0.05-0.1 dex, which they suggest increasing the core collapse yield ratio to

Mg/Fe: cc = 0.55 might improve.

Paper summary: Vera: The main point is that they use these one-zone galactic chemical evolution models with a flexible star formation history to model the age,

Mg/H: , and

Mg/Fe: inferred from simple stellar population fits to observed ellipticals at redshift zero and redshift zero point seven.

Jocelyn: They show that when you fit a spectrum from a full GCE model, you get ages and abundances that are close to the light-weighted, logarithmically averaged values of the composite stellar population, which they call the mean age, age, and

Mg/H: and

Mg/Fe: .

Subrahmanyan: They also found that the predicted relations between

Mg/H: and age, and between

Mg/Fe: and age, are surprisingly insensitive to the star formation history parameters.

Vera: But they do find that the star formation efficiency and outflow efficiency affect the early and late evolution of

Mg/H: , which is interesting because it points to how different processes shape these populations.

Jocelyn: Their input yields come from supernova Mg and Fe yields motivated by Milky Way stellar populations, and they use core collapse supernova yields calibrated to the plateau

Mg/Fe: cc about zero point four five observed in many Milky Way studies.

Subrahmanyan: They found that these inputs lead to predicted

Mg/Fe: ratios of ellipticals being underpredicted by zero point zero five to zero point one dex, and they suggest increasing the core collapse yield ratio to

Mg/Fe: cc = zero point five five might improve this match <ref:2407.07971#pg3>.

Conclusion: Vera: So, what does this paper ultimately tell us about these elliptical galaxies? It suggests that massive quiescent galaxy populations exhibit well-defined trends of stellar age,

Mg/H: , and

Mg/Fe: with galaxy velocity dispersion or stellar mass.

Jocelyn: The components they look at are the star formation history, the star formation efficiency timescale tau*, the outflow efficiency eta, and the IMF-averaged Mg and Fe yields from core collapse supernovae, as well as supernova Ia rates.

Subrahmanyan: They set up a normalization using Rodriguez et al. two thousand twenty-three's determination of the mean observed Fe yield from core collapse supernovae, choosing

Mg/Fe: cc based on the plateau seen in metal-poor Milky Way stars, and choosing y IaFe/y ccFe so that GCE models with a continuous, slowly declining star formation history evolve to

Mg/Fe: about zero at late times <ref:2407.07971#pg1>.

Vera: The authors conclude that the tension between their models and the data is mild enough that it might be explained by systematic uncertainties in the SSP spectral fits, for example, from imperfect stellar libraries.

Jocelyn: They also suggest a promising direction where using broad-band SED measurements in addition to the continuum-normalized absorption spectra modeled by alfac might allow them to constrain their star formation history parameters with the broad-band SED and let absorption spectra better constrain the yields and GCE parameters.

Nicole Marcelina Gountanis, David H. Weinberg, Aliza G. Beverage, Nathan R. Sandford, Charlie Conroy, Mariska Kriek

Department of Astronomy and Center for Cosmology and AstroParticle Physics, The Ohio State University · Astronomy Department, University of California, Berkeley · Department of Astronomy and Astrophysics, University of Toronto · Center for Astrophysics—Harvard & Smithsonian · Leiden Observatory Leiden University

astro-ph.GA

Submitted: 2024-07-10

Updated: 2026-10-27

Journal ref: The Astrophysical Journal, Volume 985, Issue 2, id.184, 25 pp. June 2025

DOI: 10.3847/1538-4357/adc4e6

Code: https://github.com/nmgountanis/fanCE

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 85/100

The gist: Spectroscopic studies of elliptical galaxies show that their stellar population ages, mean metallicity…

Key concepts

One-zone galactic chemical evolution (GCE) models
These are mathematical frameworks used to track how elements like Magnesium and Iron are created and dispersed within a single galaxy over time. The model takes into account the rate at which stars form, die, and enrich the interstellar medium with these elements.
Star Formation History (SFH)
This describes *when* star formation occurred in the galaxy, modeled here as a 'rise-fall' pattern. It includes parameters defining when star formation started, how quickly it increased initially, how long it remained high, and when it stopped entirely.
Mg/Fe Ratio
This is a measure of the relative abundance of Magnesium compared to Iron in a galaxy's stars. The ratio is sensitive to the type of supernova that enriched the gas; higher ratios are often linked to early, rapid star formation events, while lower ratios suggest later enrichment.
Star Formation Efficiency (SFE)
This parameter defines how efficiently the available gas is converted into stars over time. It relates the rate of star formation to the amount of gas in the galaxy, helping model how quickly a galaxy builds its stellar population and chemical composition.

Terminology

Summary

Spectroscopic studies of elliptical galaxies show that their stellar population ages, mean metallicity, and α-enhancement traced by [Mg/Fe] all increase with galaxy stellar mass or velocity dispersion Modeling the Ages and Chemical Abundances of Elliptical Galaxies. p. 0

Where the paper says it

Box 9513, 2300 RA Leiden, The Netherlands ABSTRACT Spectroscopic studies of elliptical galaxies show that their stellar population ages, mean metallicity, and α-enhancement traced by [Mg/Fe] all increase with galaxy stellar mass or velocity dispersion. We use one-zone galactic chemical evolution (GCE) models with a flexible star formation history (SFH) to model the age, [Mg/H], and [Mg/Fe] inferred from simple stellar population (SSP) fits to observed ellipticals at z ∼ 0 and z ∼ 0.

Modeling the Core Relationships

The authors use one-zone galactic chemical evolution (GCE) models with a flexible star formation history (SFH) to model the age, [Mg/H], and [Mg/Fe] inferred from simple stellar population (SSP) fits to observed ellipticals at z ∼ 0 and z ∼ 0. They show that an SSP fit to the spectrum computed from a full GCE model gives ages and abundances close to the light-weighted, logarithmically averaged values of the composite stellar population, ⟨age⟩, ⟨[Mg/H]⟩, and ⟨[Mg/Fe]⟩ The predicted ⟨[Mg/H]⟩-⟨age⟩ and ⟨[Mg/Fe]⟩-⟨age⟩ relations are surprisingly insensitive to SFH parameters, but the star formation efficiency and outflow efficiency affect the early and late evolution of ⟨[Mg/H]⟩, respectively. p. 0

Where the paper says it

We show that an SSP fit to the spectrum computed from a full GCE model gives ages and abundances close to the light-weighted, logarithmically averaged values of the composite stellar population, ⟨age⟩, ⟨[Mg/H]⟩, and ⟨[Mg/Fe]⟩. With supernova Mg and Fe yields fixed to values motivated by Milky Way stellar populations, we find that predicted ⟨[Mg/H]⟩-⟨age⟩ and ⟨[Mg/Fe]⟩-⟨age⟩ relations are surprisingly insensitive to SFH parameters: older galaxies have higher ⟨[Mg/Fe]⟩, but the detailed form of the SFH has limited impact.

Input Yields and Model Assumptions

The authors adopt values for supernova Mg and Fe yields motivated by Milky Way stellar populations They use core collapse supernova (CCSN) yields calibrated to the plateau [Mg/Fe]cc ≈ 0.45 observed in many Milky Way studies, finding that predicted ⟨[Mg/Fe]⟩ ratios of ellipticals are underpredicted by 0. p. 0

Where the paper says it

With core collapse supernova yields calibrated to the plateau [Mg/Fe]cc ≈ 0. 45 observed in many Milky Way studies, our models underpredict the observed ⟨[Mg/Fe]⟩ ratios of ellipticals by 0.

05-0.1 dex, which they suggest increasing the core collapse yield ratio to [Mg/Fe]cc = 0. p. 0
Where the paper says it

45 observed in many Milky Way studies, our models underpredict the observed ⟨[Mg/Fe]⟩ ratios of ellipticals by 0. Increasing the core collapse yield ratio to [Mg/Fe]cc = 0.

55 might improve. They rely principally on analytic GCE models (Weinberg et al. p. 1
Where the paper says it

If not, what modifications might be required to the stellar yields or other model assumptions, and what do they imply about the distinctive features of elliptical galaxies? We rely principally on analytic GCE models (Weinberg et al.

2017) because they are sufficient for their purposes here. p. 1
Where the paper says it

We rely principally on analytic GCE models (Weinberg et al. 2017, hereafter WAF) because they are sufficient for our purposes here.

Star Formation History and Efficiency

The SFH is parameterized by a rise-fall star formation history, SFH ≡ M˙⋆(t) ∝ (1 − e −t/τ1) · e −t/τ2 The star formation efficiency, SFE ≡ M˙∗/Mgas ≡ τ−1⋆, defines another important timescale. They consider four distinct timescales in the star formation history: the time at which star formation begins, the duration of an initial linear rise, the e-folding timescale of a subsequent decline, and the time at which star formation ceases entirely. p. 1

Where the paper says it

While the high metallicities and α-enhancements of ellipticals are usually attributed to “rapid star formation,” this phrase is imprecise. In our models, there are four distinct timescales in the star formation history: the time at which star formation begins, the duration of an initial linear rise, the e-folding timescale of a subsequent decline, and the time at which star formation ceases entirely.

Impact of Model Parameters on Observables

The analysis investigates how model parameters affect the observables ⟨[Mg/H]⟩, ⟨[Mg/Fe]⟩, and ⟨age⟩ at different epochs. They find that the predicted ⟨[Mg/H]⟩-⟨age⟩ tracks are insensitive to the individual values of τ1, τ2, and tcut The high SFE needed to get early enrichment in ⟨[Mg/H]⟩ tends to drive down ⟨[Mg/Fe]⟩, making it difficult to explain the high observed values of both quantities in the oldest, high-σ ellipticals. p. 0

Where the paper says it

With supernova Mg and Fe yields fixed to values motivated by Milky Way stellar populations, we find that predicted ⟨[Mg/H]⟩-⟨age⟩ and ⟨[Mg/Fe]⟩-⟨age⟩ relations are surprisingly insensitive to SFH parameters: older galaxies have higher ⟨[Mg/Fe]⟩, but the detailed form of the SFH has limited impact. The star formation efficiency and outflow efficiency affect the early and late evolution of ⟨[Mg/H]⟩, respectively; explaining observed trends requires higher star formation efficiency and lower outflows in more massive galaxies.

Future Directions

One promising direction is to use broad-band SED measurements (photometry or spectrophotometry) in addition to the continuum-normalized absorption spectra modeled by alf This might allow them to constrain their SFH parameters with the broad-band SED and allow absorption spectra to better constrain the yields and GCE parameters. Another direction is to extend this work to model other element abundances whose abundance can be estimated by full spectral fitting. p. 3

Where the paper says it

To alter the abundances of the 19 individual elements in these models, alf uses metallicity- and age-dependent synthetic response functions to determine the fractional change in spectra due to the variation of each element. alf models the continuum-normalized spectrum, so it does not incorporate constraints from the overall shape of the spectral energy distribution (SED).

Conclusions

Massive quiescent galaxy populations exhibit welldefined trends of stellar age, [Mg/H], and [Mg/Fe] with galaxy velocity dispersion or stellar mass The components of their GCE models are the star formation history (SFH), the star formation efficiency timescale τ∗ = Mgas/M˙⋆, the outflow efficiency η = M˙out/M˙⋆, and the IMF-averaged Mg and Fe yields from CCSN and SNIa. Following WGJT24, they normalize yccFe using the Rodr´ıguez et al. p. 5

Where the paper says it

Theoretically predicted yields have uncertainties associated with supernova models, black hole formation, and the stellar IMF. We instead base our choice of yields on the empirical analysis of WGJT24, who anchor the overall scale of yields to the Rodr´ıguez

(2023) determination of the mean observed Fe yield of CCSN, choose yccMg/yccFe based on the plateau of [Mg/Fe] seen in metal-poor Milky Way stars, and choose yIaFe/yccFe so that GCE models with a continuous, slowly declining SFH evolve to [Mg/Fe] ≈ 0 at late times. p. 6
Where the paper says it

45 as our reference value based on the abundance plateau observed for low metallicity Milky Way disk and halo stars. We also consider a higher CCSN yield ratio, corresponding to [Mg/Fe]cc = 0.

The authors conclude that the tension between models and the data is mild enough that it might be explained by systematic uncertainties in the SSP spectral fits, e. p. 3
Where the paper says it

With the flexible GCE models used in this paper, it would be possible to compute composite spectra for each model and compare them to data directly in spectral space. However, in this paper we adopt the simpler procedure of computing the light-weighted average quantities from models and comparing them to the empirical values from SSP fits.

g., from imperfect stellar libraries. p. 3
Where the paper says it

The SSP model ingredients include metallicitydependent MIST isochrones (Choi et al. 2016) and the MILES and Extended-IRTF empirical spectral libraries (S´anchez-Bl´azquez et al.

p. 3
Where the paper says it

The SSP model ingredients include metallicitydependent MIST isochrones (Choi et al. 2016) and the MILES and Extended-IRTF empirical spectral libraries (S´anchez-Bl´azquez et al.

Appendix Details

Appendix A gives instructions for accessing fanCE Appendix B gives details of their treatment of terminating bursts of star formation Appendix C tests the relation between SSP fits and the light-weighted average quantities of composite stellar populations.

References

The references include Abazajian et al. 2009, Andrews et al. p. 0

Where the paper says it

2005; Schiavon 2007; Graves et al. 2009; Smith et al.

p. 0
Where the paper says it

2005; Schiavon 2007; Graves et al. 2009; Smith et al.

2017, Beverage et al. p. 3
Where the paper says it

2017; Conroy et al. Therefore in this analysis, we adopt the updated SDSS measurements from Beverage et al.

p. 3
Where the paper says it

2017; Conroy et al. Therefore in this analysis, we adopt the updated SDSS measurements from Beverage et al.

2023, and others. p. 1
Where the paper says it

2017, hereafter WAF) because they are sufficient for our purposes here. We provide a Python implementation of these analytic models, fanCE1 , so that others can easily adapt them to interpret other galaxies or stellar populations or explore alternative assumptions (see Appendix A).

p. 1
Where the paper says it

2017, hereafter WAF) because they are sufficient for our purposes here. We provide a Python implementation of these analytic models, fanCE1 , so that others can easily adapt them to interpret other galaxies or stellar populations or explore alternative assumptions (see Appendix A).

The paper also cites works by Conroy et al. p. 3
Where the paper says it

2010; Conroy et al. With the flexible GCE models used in this paper, it would be possible to compute composite spectra for each model and compare them to data directly in spectral space.

p. 3
Where the paper says it

2010; Conroy et al. With the flexible GCE models used in this paper, it would be possible to compute composite spectra for each model and compare them to data directly in spectral space.

, van Dokkum & Conroy, and Weinberg et al. p. 2
Where the paper says it

More details and comparisons between methods can be found in Section 3 of Beverage et al. Stellar Population Fitting The stellar population fitting is accomplished using the full-spectrum absorption line fitting code, alf (Absorption-Line Fitter) (Conroy & van Dokkum 2012a;

p. 2
Where the paper says it

More details and comparisons between methods can be found in Section 3 of Beverage et al. Stellar Population Fitting The stellar population fitting is accomplished using the full-spectrum absorption line fitting code, alf (Absorption-Line Fitter) (Conroy & van Dokkum 2012a;

Data and Mock Tests

The study uses observational data from the SDSS (at z ≈ 0) and LEGA-C (at z ≈ 0.7) to compare their predictions with model results. p. 2

Where the paper says it

Section 3 describes our GCE models, with particular attention to our flexible SFH parameterization, our choice of empirically motivated Mg and Fe yields, and the impact of model parameters on the evolution of the mean age, [Mg/H], and [Mg/Fe] of stellar populations. In Section 4 we compare our model predictions to the observed trends.

They perform two-population tests in Appendix C, showing that an SSP fit recovers the lightweighted, logarithmically averaged age, [Mg/H], and [Mg/Fe] of a composite stellar population. p. 0
Where the paper says it

We use one-zone galactic chemical evolution (GCE) models with a flexible star formation history (SFH) to model the age, [Mg/H], and [Mg/Fe] inferred from simple stellar population (SSP) fits to observed ellipticals at z ∼ 0 and z ∼ 0. We show that an SSP fit to the spectrum computed from a full GCE model gives ages and abundances close to the light-weighted, logarithmically averaged values of the composite stellar population, ⟨age⟩, ⟨[Mg/H]⟩, and ⟨[Mg/Fe]⟩.

Final Findings on Discrepancy

The authors find that their fiducial models with [Mg/Fe]cc = 0.45 underpredict the observed ⟨[Mg/Fe]⟩ values by ∼ 0. p. 0

Where the paper says it

With core collapse supernova yields calibrated to the plateau [Mg/Fe]cc ≈ 0. 45 observed in many Milky Way studies, our models underpredict the observed ⟨[Mg/Fe]⟩ ratios of ellipticals by 0.

08 dex at z = 0 and ∼ 0. p. 4
Where the paper says it

SSP value, e. 1 dex for populations that differ by 0.

05 dex at z = 0. p. 4
Where the paper says it

SSP value, e. 1 dex for populations that differ by 0.

7. Agreement with the data can be improved by adopting a higher [Mg/Fe]cc = 0. p. 0
Where the paper says it

Increasing the core collapse yield ratio to [Mg/Fe]cc = 0. 55 improves the agreement, though the models still lie below the data.

55, or by considering a terminating burst of star formation. p. 0
Where the paper says it

55 improves the agreement, though the models still lie below the data. We discuss potential resolutions of this discrepancy, including the possibility that many ellipticals terminate their star formation with a self-enriching, terminating burst that reduces the light-weighted age and boosts ⟨[Mg/Fe]⟩.

Summary of Key Results

The results suggest that adding a final burst of star formation to the end of a galaxy’s smooth SFH can bridge the gap between predicted and observed ⟨[Mg/Fe]⟩ ratios, bringing models with [Mg/Fe]cc = 0.45 within the range of the data. p. 5

Where the paper says it

Our model SFHs have four adjustable timescale parameters — tstart, τ1, τ2, and tcut — and tcut is the one with the largest impact on the predicted mean stellar age and abundances. The form of Equation 2 allows a much wider range of behavior than shown in Figure 3, including a linearly rising SFH, a pure exponential, and a rise followed by a constant SFR plateau.

The effects of the burst on ⟨age⟩ and on ⟨[Mg/Fe]⟩ are about equally important. p. 8
Where the paper says it

In addition, our burst model adds a terminating burst of star formation to a smooth SFH at t = tcut and contains parameters that describe gas accretion FD and the fraction of gas consumed in the burst fg. curve to later t or younger ⟨age⟩ by tstart, despite the effects of light-weighting.

Figure Interpretations

Figure 12 demonstrates the difficulty of reproducing observed abundance trends in ellipticals with smooth star formation histories and yields that are empirically motivated by abundance trends in the Milky Way. The blue points in Figure 12c show the ⟨[Mg/H]⟩ and ⟨age⟩ of the observed LEGA-C galaxy bins at z = 0. p. 3

Where the paper says it

Figure 1 shows the observed spectrum and best-fit model spectrum for one velocity dispersion stack from the SDSS (top) and an individual LEGA-C galaxy (bottom). We also show the spectrum for a model in which we increase the total metallicity, [Z/H] by 3σ while keeping other parameters fixed.

7, with velocity dispersion increasing from the light blue to the dark blue points. p. 7
Where the paper says it

As a result, the stellar ⟨age⟩ at the time of observation is younger for galaxies with a slowly declining SFH versus a quickly declining SFH, as shown by the difference between the blue and orange dashed curves in Figure 4b. Light-weighting, represented by the solid curves, gives more weight to younger stars, lowering the ⟨age⟩ for both models by a couple of Gyr at late times, with deviations from the mass-weighted ⟨age⟩ rapidly increasing at times close to tobs.

The intersection of the horizontal lines and model curves indicate the value of tcut required to obtain the observed ⟨age⟩. p. 9
Where the paper says it

to the SFH, except that the exponential SFH, which has its maximum SFR at t = tstart, achieves higher ⟨[Mg/H]⟩ at the oldest ages. For ⟨[Mg/Fe]⟩ the model curves are more separated as a function of ⟨age⟩ than they were as a function of tcut because both the τ1 = 0 and τ2 = 4 Gyr models have older ages at a given time (Figure 6b).

Final Comparison

The models show that for [Mg/Fe]cc = 0.55, they come close to the observed ⟨[Mg/H]⟩-⟨age⟩ and ⟨[Mg/Fe]⟩-⟨age〉 trends at z = 0. p. 0

Where the paper says it

We use one-zone galactic chemical evolution (GCE) models with a flexible star formation history (SFH) to model the age, [Mg/H], and [Mg/Fe] inferred from simple stellar population (SSP) fits to observed ellipticals at z ∼ 0 and z ∼ 0. We show that an SSP fit to the spectrum computed from a full GCE model gives ages and abundances close to the light-weighted, logarithmically averaged values of the composite stellar population, ⟨age⟩, ⟨[Mg/H]⟩, and ⟨[Mg/Fe]⟩.

7. The discrepancy with the data is mild enough that it might be explained by systematic uncertainties in the SSP spectral fits, e. p. 3
Where the paper says it

With the flexible GCE models used in this paper, it would be possible to compute composite spectra for each model and compare them to data directly in spectral space. However, in this paper we adopt the simpler procedure of computing the light-weighted average quantities from models and comparing them to the empirical values from SSP fits.

g., from imperfect stellar libraries.

Additional Notes

The paper also discusses the importance of sharpening empirical constraints on IMF-averaged yields, in the Milky Way and in other local galaxies. The authors note that elliptical galaxies differ from disk galaxies in their star formation histories and potentially in their outflows and other galactic scale phenomena. p. 1

Where the paper says it

The population-averaged Mg and Fe yields of CCSN and the population-averaged Fe yield and delay time distribution (DTD) of SNIa are crucial inputs to our models. We adopt values that are empirically motivated by observations of Milky Way stellar abundances and cosmic SNIa rates, implicitly assuming that the stellar scale astrophysics that determines these quantities is similar in elliptical galaxies despite their distinctive star formation histories.

Acknowledgments

They thank Adam Wheeler, Todd Thompson, Liam Dubay, and Jennifer Johnson for helpful discussions. This work is supported by NSF grant AST-2307621. p. 6

Where the paper says it

We take η = 0. 3 as a reference value, but in line with previous work, we will assume that η decreases with increasing M⋆ to reproduce the observed mass-metallicity trend.

p. 6
Where the paper says it

We take η = 0. 3 as a reference value, but in line with previous work, we will assume that η decreases with increasing M⋆ to reproduce the observed mass-metallicity trend.

NMG acknowledges the hospitality of the Kavli Institute for Theoretical Physics (KITP), supported by NSF PHY-2309135, during the completion of this work. This work is dedicated to the loving memory of Marcelina Kroenke

Appendix A Details

Appendix A gives instructions for accessing fanCE Appendix B gives details of their treatment of terminating bursts of star formation Appendix C tests the relation between SSP fits and the light-weighted average quantities of composite stellar populations.

References List

The references list includes Abazajian et al. 2009, Andrews et al. 2017, Beverage et al. 2023, and others. The paper also cites works by Conroy et al., van Dokkum & Conroy, and Weinberg et al.

They thank Adam Wheeler, Todd Thompson, Liam Dubay, and Jennifer Johnson for helpful discussions. This work is supported by NSF grant AST-2307621. NMG acknowledges the hospitality of the Kavli Institute for Theoretical Physics (KITP), supported by NSF PHY-2309135, during the completion of this work This work is dedicated to the loving memory of Marcelina Kroenke. p. 6

Where the paper says it

We take η = 0. 3 as a reference value, but in line with previous work, we will assume that η decreases with increasing M⋆ to reproduce the observed mass-metallicity trend.

They thank Adam Wheeler, Todd Thompson, Liam Dubay, and Jennifer Johnson for helpful discussions. This work is supported by NSF grant AST-2307621 NMG acknowledges the hospitality of the Kavli Institute for Theoretical Physics (KITP), supported by NSF PHY-2309135, during the completion of this work This work is dedicated to the loving memory of Marcelina Kroenke. p. 6

Where the paper says it

We take η = 0. 3 as a reference value, but in line with previous work, we will assume that η decreases with increasing M⋆ to reproduce the observed mass-metallicity trend.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved model parameter sensitivity analysis: The AI system can perform targeted sensitivity tests on GCE model parameters (like SFE timescale tau star and outflow efficiency eta) to determine their specific impact on predicted observables like [Mg/H] and [Mg/Fe], moving beyond the general conclusions of Section 5.1.

  2. Enhanced comparison between SSP fits and light-weighted averages: The system can automatically generate mock Composite Stellar Population (CSP) spectra using continuous GCE models and then fit these mock spectra with SSP models to quantify the residual inaccuracies in predicting light-weighted quantities, as detailed in Section 4.2 and Figure 13.

  3. Automated exploration of terminating burst effects: The AI can systematically test the impact of terminating bursts by varying dilution factors (FD) and gas fractions (fg) within a leaky box chemical evolution framework to predict the resulting changes in final light-weighted [Mg/Fe] ratios, as shown in Figure 15 and 17.

  4. Integrated yield uncertainty assessment: The system can evaluate how variations in IMF-averaged yields, specifically changing the high-mass slope from -2.3 to -2.0, affect the predicted [Mg/Fe]cc ratio across different regimes of galactic mass and metallicity, as discussed in Section 5.1 regarding IMF differences.

  5. Constraint generation via SED-spectrum combination: The AI can be developed to integrate broad-band Spectral Energy Distribution (SED) measurements with continuum-normalized absorption spectra to constrain star formation histories, allowing for the probing of quenching physics beyond effective population age defined by SSP fits, as suggested in Section 5.2.

Abstract

Spectroscopic studies of elliptical galaxies show that their stellar population ages, mean metallicity, and α-enhancement traced by [Mg/Fe] all increase with galaxy stellar mass or velocity dispersion. We use one-zone galactic chemical evolution (GCE) models with a flexible star formation history (SFH) to model the age, [Mg/H], and [Mg/Fe] inferred from simple stellar population (SSP) fits to observed ellipticals at z about 0 and z about 0.7. We show that an SSP fit to the spectrum computed from a full GCE model gives ages and abundances close to the light-weighted, logarithmically averaged values of the composite stellar population, <age>, <[Mg/H]>, and <[Mg/Fe]>. With supernova Mg and Fe yields fixed to values motivated by Milky Way stellar populations, we find that predicted <[Mg/H]>-<age> and <[Mg/Fe]>-<age> relations are surprisingly insensitive to SFH parameters: older galaxies have higher <[Mg/Fe]>, but the detailed form of the SFH has limited impact. The star formation efficiency and outflow efficiency affect the early and late evolution of <[Mg/H]>, respectively; explaining observed trends requires higher star formation efficiency and lower outflows in more massive galaxies. With core collapse supernova yields calibrated to the plateau [Mg/Fe] cc about0.45 observed in many Milky Way studies, our models underpredict the observed <[Mg/Fe]> ratios of ellipticals by 0.05-0.1 dex. Increasing the core collapse yield ratio to [Mg/Fe] cc = 0.55 improves the agreement, though the models still lie below the data. We discuss potential resolutions of this discrepancy, including the possibility that many ellipticals terminate their star formation with a self-enriching, terminating burst that reduces the light-weighted age and boosts <[Mg/Fe]>.

Sources

Related papers