BEAGLE (for BayEsian Analysis of GaLaxy sEds, pre-print available here) is a new-generation tool to model and interpret galaxy spectral energy distributions (SEDs) developed by Jacopo Chevallard (ESA) and Stephane Charlot (IAP). This tool incorporates the consistent modelling of stellar radiation and its transfer through the interstellar and intergalactic media, allowing one to build mock galaxy catalogues as well as to interpret, in a Bayesian framework, any combination of photometric and spectroscopic galaxy observations. The current version of the tool includes, among other specificities: the emission from stars and photo-ionized $\textnormal{H}\,\textsc{ii}$ regions; different parametrizations of the star formation and chemical enrichment histories of galaxies, including predictions from sophisticated galaxy formation models; the effect of changes in the α-element to iron abundance ratio on galaxy spectral properties; and the accounting, within a Bayesian framework, of different instrumental effects.

BEAGLE builds upon the modelling efforts deployed within the ERC-funded NEOGAL (NEw frOntiers in Galaxy spectrAl modeLing) project at the IAP. The BEAGLE tool is currently under active development (with the recent strong implication of Emma Curtis-Lake from the NEOGAL team), and in the (near) future the code will be released as an open-source project.

What's specific about BEAGLE?

Several codes for the modelling and interpretation of galaxy SEDs exist in the literature (see www.sedfitting.org for a comprehensive list), but none of them includes all the characteristics that a generic framework for galaxy SEDs modelling should possess. In particular, many existing codes rely on simple $\chi^2$ minimisation techniques, which do not allow the computation of realistic uncertainties in the derived physical parameters.  In addition, standard SED interpretation tools are usually tied to specific choices of built-in physical ingredients (e.g., stellar evolution and dust attenuation prescriptions), which cannot easily be changed, nor tested, nor extended to include new physical ingredients (e.g., emission from $\textnormal{H}\,\textsc{ii}$ regions and from an AGN, neutral ISM absorption). An additional specificity of most current tools focused on the interpretation of photometric (rather than spectroscopic) galaxy SEDs is that these tend to be optimised for either redshift estimation  or the determination of galaxy physical parameters, but not both simultaneously.

The limitations of existing SED fitting codes mentioned above are the main motivation for our development of the BEAGLE tool. This tool incorporates the most recent  prescriptions for stellar and nebular emission (and the dependence of these components on chemical composition), attenuation by dust, IGM absorption, etc., in a physically consistent and highly flexible way: the modular design of the BEAGLE tool allows any of these prescriptions to be easily replaced by an alternative one. It also enables the straightforward implementation of additional physical ingredients (e.g. dust and AGN emission, neutral ISM absorption). The Bayesian approach adopted in the tool allows the user to rigorously quantify the uncertainties and degeneracies affecting model parameters. Finally, the use of MCMC techniques and a compiled language (Fortran 2003) makes the exploration of complex, multi-dimensional parameter spaces much more efficient, and much less memory-demanding, than in conventional, grid-based approaches.

Subject