[Paper Review] VOSA: Virtual Observatory SED Analyzer. An application to the Collinder 69 open cluster

A. Bayo, C. Rodrigo+5 more|ArXiv.org|Aug 2, 2008

Stellar, planetary, and galactic studiesPhysics and Astronomy34 references233 citations

TL;DR

This paper introduces VOSA, a Virtual Observatory-compliant tool for automating spectral energy distribution (SED) fitting to derive physical parameters of stellar objects. Applied to 167 candidate members of the Collinder 69 open cluster, VOSA enabled high-throughput analysis of photometric data against theoretical models, yielding an upper-age limit of 12.3–16 Myr, consistent with prior studies, and demonstrating the scalability of VO tools for large-scale stellar population analysis.

ABSTRACT

The physical properties of almost any kind of astronomical object can be derived by fitting synthetic spectra or photometry extracted from theoretical models to observational data. We want to develop an automatic procedure to perform this kind of fittings to a relatively large sample of members of a stellar association and apply this methodology to the case of Collinder 69. We combine the multiwavelength data of our sources and follow a work-flow to derive the physical parameters of the sources. The key step of the work-flow is performed by a new VO-tool, VOSA. All the steps in this process are done in a VO environment. We present this new tool, and provide physical parameters such as T$_{ m eff}$, gravity, luminosity, etc. for $\sim$170 candidate members to Collinder 69, and an upper-limit for the age of this stellar association. This kind of studies of star forming regions, clusters, etc. produces a huge amount of data, very tedious to analyse using the traditional methodology. Thus, they are excellent examples where to apply the VO capabilities.

Motivation & Objective

• To develop an automated, VO-compliant workflow for deriving physical parameters of stellar cluster members from multiwavelength photometry.
• To address the computational burden of fitting large datasets of observed SEDs against extensive grids of theoretical models.
• To apply this methodology to the Collinder 69 open cluster to estimate age, mass, effective temperature, gravity, and luminosity for candidate members.
• To validate the method by comparing results with existing classifications and age estimates from the literature.
• To demonstrate the scalability and efficiency of Virtual Observatory tools in handling complex, data-intensive astrophysical analyses.

Proposed method

• The workflow integrates multiwavelength photometric data from various surveys into a unified SED for each source.
• VOSA computes synthetic photometry by interpolating between theoretical model grids (DUSTY, NextGen, Kurucz) using VO-compliant services.
• A χ² minimization technique is applied to compare observed SEDs with synthetic SEDs across multiple photometric bands.
• Effective temperature, gravity, and bolometric luminosity are derived by minimizing the χ² statistic between observed and synthetic SEDs.
• Isochrones and evolutionary tracks from Baraffe et al. (1998) and others are interpolated to estimate age and mass, using the derived T_eff and log g.
• The process is fully automated and executed within the Virtual Observatory framework, enabling seamless data and model access.

Experimental results

Research questions

• RQ1 What is the most accurate and efficient method for deriving physical parameters (T_eff, log g, L_bol, mass, age) for a large sample of stellar cluster members using photometric data?
• RQ2 How can the Virtual Observatory infrastructure be leveraged to automate and scale the fitting of observed SEDs to theoretical model grids?
• RQ3 What age constraints can be placed on the Collinder 69 open cluster using SED fitting and isochrone comparison across a representative sample of its members?
• RQ4 How do the results from automated SED fitting compare with previous classifications and age estimates for Collinder 69?
• RQ5 To what extent does the inclusion of disk-affected sources (e.g., with infrared excess) bias age estimates in cluster population studies?

Key findings

• VOSA successfully derived physical parameters for 167 candidate members of the Collinder 69 open cluster, including T_eff, gravity, bolometric luminosity, and inferred mass and age.
• The method independently confirmed the non-member status of 16 sources previously classified as such by Barrado y Navascués et al. (2007b), and identified one additional possible non-member (LOri162).
• For the three sources best fit by DUSTY models, an upper-limit age of 5 Myr was derived, consistent with the presence of circumstellar disks.
• Among the 117 sources best fit by NextGen or Kurucz models, 90% were younger than 16 Myr, with the 3rd quartile at 12.3 Myr, yielding an upper-age limit of 12.3–16 Myr for the cluster.
• The entire workflow, including SED construction, synthetic photometry, χ² fitting, and isochrone interpolation, was completed in approximately 20 minutes, demonstrating high computational efficiency.
• The results are consistent with previous literature estimates, validating the robustness of the automated VO-based approach.