[Paper Review] CoRoT/ESTA-TASK 1 and TASK 3 comparison of the internal structure and seismic properties of representative stellar models: Comparisons between the ASTEC, CESAM, CLES, GARSTEC and STAROX codes

Y. Lebreton, J. Montalbán+3 more|ArXiv.org|Jan 7, 2008

Stellar, planetary, and galactic studies39 references27 citations

TL;DR

This paper compares stellar evolution models from five codes—ASTEC, CESAM, CLES, GARSTEC, and STAROX—across CoRoT/ESTA TASK 1 and TASK 3, focusing on internal structure and seismic properties. Despite overall good agreement, key discrepancies arise from differences in equation of state, opacity treatments, and convective boundary modeling, with frequency differences reaching up to 1 μHz in evolved models, highlighting critical sensitivities for asteroseismic interpretation.

ABSTRACT

We compare stellar models produced by different stellar evolution codes for the CoRoT/ESTA project, comparing their global quantities, their physical structure, and their oscillation properties. We discuss the differences between models and identify the underlying reasons for these differences. The stellar models are representative of potential CoRoT targets. Overall we find very good agreement between the five different codes, but with some significant deviations. We find noticeable discrepancies (though still at the per cent level) that result from the handling of the equation of state, of the opacities and of the convective boundaries. The results of our work will be helpful in interpreting future asteroseismology results from CoRoT.

Motivation & Objective

• Assess the consistency of stellar evolution codes in predicting internal structure and oscillation frequencies for CoRoT mission targets.
• Identify numerical and physical sources of discrepancies among models with identical input physics and global parameters.
• Improve confidence in asteroseismic interpretations by quantifying code-to-code differences in seismic and structural properties.
• Highlight numerical and physical simplifications that affect model precision, especially in regions of strong chemical gradients and convection.
• Support the development of high-precision models needed for future space missions like Kepler and CoRoT, where frequency accuracy reaches 10−7 Hz.

Proposed method

• Selected representative stellar models from five independent stellar evolution codes (ASTEC, CESAM, CLES, GARSTEC, STAROX) with nearly identical global parameters (age, luminosity, radius).
• Compared physical structure, chemical composition profiles, and oscillation frequencies (p and g modes, ℓ = 0–3) across models.
• Computed relative differences in key quantities (density, pressure, sound speed, chemical abundances) and tracked convective boundary evolution.
• Analyzed the impact of microscopic diffusion on surface helium and metal depletion, particularly in main-sequence and evolved models.
• Examined frequency separation ratios and large frequency separations to isolate near-surface versus deep-structure effects.
• Used a common set of physical constants, element mixtures, and input physics to isolate code-specific numerical and algorithmic differences.

Experimental results

Research questions

• RQ1 How do stellar evolution codes differ in predicting the internal structure of models with identical global parameters and input physics?
• RQ2 What are the dominant sources of discrepancy in oscillation frequencies across codes, and how do they relate to physical or numerical choices?
• RQ3 How do differences in convective boundary treatment and chemical gradient formation affect the accuracy of seismic modeling?
• RQ4 To what extent do variations in equation of state and opacity tables influence model consistency?
• RQ5 How do the inclusion of microscopic diffusion and semiconvection affect the reliability of surface and core structure predictions?

Key findings

• Differences in the equation of state and opacity treatments lead to relative discrepancies of up to 1% in key physical quantities like density and pressure.
• Convective boundary treatment and numerical handling of the µ-gradient result in up to 30% differences in convective core mass, especially in pre-main sequence and low-mass main-sequence models.
• Surface helium abundance differences reach 15–30% in 1.3 M⊙ models on the main sequence due to divergent diffusion formalisms.
• Oscillation frequency differences reach 0.1–0.2 μHz for ℓ=0 modes in models with diffusion, and up to 1 μHz in evolved models (TAMS/SGB), primarily due to structure differences near the core and in the second helium ionization zone.
• Frequency separation ratios show excellent agreement across codes, indicating that observed frequency differences are dominated by near-surface effects rather than deep-structure variations.
• Massive models (3–5 M⊙) show the smallest frequency differences (<0.2 μHz), suggesting higher robustness in high-mass evolution, while mixed modes in evolved models are sensitive to µ-gradient features and boundary location.