Lick Indices and Spectral Energy Distribution Analysis based on an M31 Star Cluster Sample: Comparisons of Methods and Models

TL;DR

This study compares different spectral fitting methods and models to derive physical parameters of M31 globular clusters, finding combined Lick indices and SED fitting yields more reliable results with smaller uncertainties.

Contribution

It systematically evaluates multiple models and fitting techniques on M31 GCs, demonstrating the advantages of combined spectral and photometric analysis.

Findings

Combined Lick indices and SED fitting improves parameter accuracy.

Different models yield varying results, highlighting model dependence.

Spectral energy distribution fitting reduces uncertainties.

Abstract

Application of fitting techniques to obtain physical parameters---such as ages, metallicities, and $\alpha$-element to iron ratios---of stellar populations is an important approach to understand the nature of both galaxies and globular clusters (GCs). In fact, fitting methods based on different underlying models may yield different results, and with varying precision. In this paper, we have selected 22 confirmed M31 GCs for which we do not have access to previously known spectroscopic metallicities. Most are located at approximately one degree (in projection) from the galactic center. We performed spectroscopic observations with the 6.5 m MMT telescope, equipped with its Red Channel Spectrograph. Lick/IDS absorption-line indices, radial velocities, ages, and metallicities were derived based on the $\rm EZ\_Ages$ stellar population parameter calculator. We also applied full spectral fitting with the ULySS code to constrain the parameters of our sample star clusters. In addition, we performed $\chi^2_{\rm min}$ fitting of the clusters' Lick/IDS indices with different models, including the Bruzual & Charlot models (adopting Chabrier or Salpeter stellar initial mass functions and 1994 or 2000 Padova stellar evolutionary tracks), the GALEV, and the Thomas et al. models. For comparison, we collected their $UVBRIJK$ photometry from the Revised Bologna Catalogue (v.5) to obtain and fit the GCs' spectral energy distributions (SEDs). Finally, we performed fits using a combination of Lick/IDS indices and SEDs. The latter results are more reliable and the associated error bars become significantly smaller than those resulting from either our Lick/IDS indices-only or our SED-only fits.