[Paper Review] Stellar populations of galaxies in the ALHAMBRA survey up to $z \sim 1$. I. MUFFIT: A Multi-Filter Fitting code for stellar population diagnostics

L. A. Díaz-García, A. J. Cenarro+30 more|arXiv (Cornell University)|May 28, 2015

Astronomy and Astrophysical Research2 references24 citations

TL;DR

This paper introduces MUFFIT, a multi-filter fitting code designed for stellar population analysis in galaxies using photometric data from the ALHAMBRA survey up to redshift z ~ 1. It accurately derives age, metallicity, and stellar mass by fitting observed photometry to synthetic spectral energy distributions, demonstrating strong agreement with spectroscopic measurements and enabling robust, high-statistics studies of galaxy evolution across cosmic time.

ABSTRACT

We present MUFFIT, a new generic code optimized to retrieve the main stellar population parameters of galaxies in photometric multi-filter surveys, and we check its reliability and feasibility with real galaxy data from the ALHAMBRA survey. Making use of an error-weighted $χ^2$-test, we compare the multi-filter fluxes of galaxies with the synthetic photometry of mixtures of two single stellar populations at different redshifts and extinctions, to provide through a Monte Carlo method the most likely range of stellar population parameters (mainly ages and metallicities), extinctions, redshifts, and stellar masses. To improve the diagnostic reliability, MUFFIT identifies and removes from the analysis those bands that are significantly affected by emission lines. We highlight that the retrieved age-metallicity locus for a sample of $z \le 0.22$ early-type galaxies in ALHAMBRA at different stellar mass bins are in very good agreement with the ones from SDSS spectroscopic diagnostics. Moreover, a one-to-one comparison between the redshifts, ages, metallicities, and stellar masses derived spectroscopically for SDSS and by MUFFIT for ALHAMBRA reveals good qualitative agreements in all the parameters. In addition, and using as input the results from photometric-redshift codes, MUFFIT improves the photometric-redshift accuracy by $\sim 10$-$20\%$, and it also detects nebular emissions in galaxies, providing physical information about their strengths. Our results show the potential of multi-filter galaxy data to conduct reliable stellar population studies with the appropiate analysis techniques, as MUFFIT.

Motivation & Objective

• To develop a reliable, optimized photometric fitting code tailored for multi-filter surveys like ALHAMBRA to study stellar populations in galaxies at z ~ 1.
• To address the challenge of deriving accurate stellar population parameters (age, metallicity, mass) from photometric data alone, especially in the absence of spectroscopic coverage for large samples.
• To validate the performance of the code by comparing its photometric results with spectroscopic measurements from SDSS, ensuring consistency and reliability.
• To explore the implications of the derived stellar population parameters for galaxy formation and evolution, particularly in the context of the downsizing scenario.
• To enable large-scale, statistically robust studies of galaxy stellar content using photometric data, leveraging the high number statistics of modern multi-filter surveys.

Proposed method

• MUFFIT performs full-spectral energy distribution (SED) fitting by comparing observed ALHAMBRA photometry to a grid of synthetic SEDs generated from stellar population synthesis models.
• The code uses a multi-filter approach with 20 contiguous photometric bands covering the 350–900 nm range, providing high spectral resolution for accurate stellar population diagnostics.
• It employs a χ² minimization technique to find the best-fit combination of stellar population models, including varying metallicities, star formation histories, and dust extinction.
• The method accounts for dust attenuation using a Calzetti extinction law and includes a flexible treatment of star formation histories through linear combinations of simple stellar populations (SSPs).
• The code is optimized for computational efficiency and robustness, enabling application to thousands of galaxies with consistent and reliable results.
• Validation is performed by cross-comparing MUFFIT results with spectroscopic measurements from the SDSS MPA/JHU catalog, using a common sample of galaxies with overlapping apertures and redshifts.

Experimental results

Research questions

• RQ1 How accurately can MUFFIT recover stellar population parameters (age, metallicity, mass) from photometric data in galaxies at z ~ 1?
• RQ2 To what extent do the photometric stellar population results from MUFFIT agree with spectroscopic measurements from SDSS for the same galaxies?
• RQ3 What are the systematic differences between photometric and spectroscopic diagnostics, and what physical or observational factors (e.g., aperture size, metallicity gradients) drive them?
• RQ4 How do the derived stellar population properties of galaxies vary with mass, and what do they imply for the downsizing scenario of galaxy formation?
• RQ5 Can MUFFIT reliably enable large-scale, high-statistics studies of galaxy stellar content in upcoming multi-filter surveys like J-PLUS and J-PAS?

Key findings

• MUFFIT achieves a typical RMS scatter of σ_z^SDSS ~ 0.008 in redshift, σ_Age^SDSS ~ 1.6 Gyr in age, σ_[Fe/H]^SDSS ~ 0.2 dex in metallicity, and σ_M⋆^SDSS ~ 0.19 dex in stellar mass when compared to SDSS spectroscopic measurements.
• A systematic offset of ~2 Gyr is observed between MUFFIT photometric ages and SDSS spectroscopic ages, primarily due to differences in modeling: MUFFIT uses SSP mixtures while SDSS employs τ-models.
• Galaxies with apertures ≤4 arcsec in ALHAMBRA show better agreement with SDSS metallicities (Δ[Fe/H] < 0.05 dex) than those with larger apertures (Δ[Fe/H] < 0.15 dex), indicating that aperture mismatch and metallicity gradients contribute to the observed discrepancies.
• Stellar masses derived by MUFFIT show a minor shift of ~0.18 dex relative to SDSS, which is consistent with the observed age offset and can be explained by the age-metallicity degeneracy.
• More massive early-type galaxies (≥10^11 M☉) formed earlier (ΔAge ~ 3 Gyr) and have higher metallicities (Δ[Fe/H] ~ 0.3 dex) than lower-mass counterparts, supporting the downsizing scenario of galaxy formation.
• The strong agreement between MUFFIT and spectroscopic results, despite photometric-only inputs, confirms that MUFFIT is a reliable tool for large-scale stellar population studies in ALHAMBRA-like surveys and future surveys such as J-PLUS and J-PAS.