J/A+A/603/A63 FADO code (Gomes+, 2017)
Fitting Analysis using Differential Evolution Optimization (FADO):
Spectral population synthesis through genetic optimization under
self-consistency boundary conditions.
Gomes J.M., Papaderos P.
<Astron. Astrophys. 603, A63 (2017)>
=2017A&A...603A..63G 2017A&A...603A..63G (SIMBAD/NED BibCode)
ADC_Keywords: Models ; Galaxies, spectra
Keywords: galaxies: evolution - galaxies: star formation - galaxies: starburst -
galaxies: stellar content - galaxies: fundamental parameters -
methods: numerical
Abstract:
The goal of population spectral synthesis (PSS; also referred to as
inverse, semi-empirical evolutionary- or fossil record approach) is to
decipher from the spectrum of a galaxy the mass, age and metallicity
of its constituent stellar populations. This technique, which is the
reverse of but complementary to evolutionary synthesis, has been
established as fundamental tool in extragalactic research. It has been
extensively applied to large spectroscopic data sets, notably the
SDSS, leading to important insights into the galaxy assembly history.
However, despite significant improvements over the past decade, all
current PSS codes suffer from two major deficiencies that inhibit us
from gaining sharp insights into the star-formation history (SFH) of
galaxies and potentially introduce substantial biases in studies of
their physical properties (e.g., stellar mass, mass-weighted stellar
age and specific star formation rate). These are i) the neglect of
nebular emission in spectral fits, consequently, ii) the lack of a
mechanism that ensures consistency between the best-fitting SFH and
the observed nebular emission characteristics of a star-forming (SF)
galaxy (e.g., hydrogen Balmer-line luminosities and equivalent
widths-EWs, shape of the continuum in the region around the Balmer and
Paschen jump). In this article, we present FADO (Fitting Analysis
using Differential evolution Optimization) -- a conceptually novel,
publicly available PSS tool with the distinctive capability of
permitting identification of the SFH that reproduces the observed
nebular characteristics of a SF galaxy. This so-far unique
self-consistency concept allows us to significantly alleviate
degeneracies in current spectral synthesis, thereby opening a new
avenue to the exploration of the assembly history of galaxies. The
innovative character of FADO is further augmented by its mathematical
foundation: FADO is the first PSS code employing genetic differential
evolution optimization. This, in conjunction with various other
currently unique elements in its mathematical concept and numerical
realization (e.g., mid-analysis optimization of the spectral library
using artificial intelligence, test for convergence through a
procedure inspired by Markov chain Monte Carlo techniques,
quasi-parallelization embedded within a modular architecture) results
in key improvements with respect to computational efficiency and
uniqueness of the best-fitting SFHs. Furthermore, FADO incorporates
within a single code the entire chain of pre-processing, modeling,
post-processing, storage and graphical representation of the relevant
output from PSS, including emission-line measurements and estimates of
uncertainties for all primary and secondary products from spectral
synthesis (e.g., mass contributions of individual stellar populations,
mass- and luminosity-weighted stellar ages and metallicities). This
integrated concept greatly simplifies and accelerates a lengthy
sequence of individual time-consuming steps that are generally
involved in PSS modeling, further enhancing the overall efficiency of
the code and inviting to its automated application to large
spectroscopic data sets.
Description:
FADO comes from the Latin word "fatum" that means fate or destiny. It
is also a well known genre of Portuguese music, and by choosing this
acronym for this spectral synthesis tool we would like to pay tribute
to Portugal. The main goal of FADO is to explore the star-formation
and chemical enrichment history (the "Fado") of galaxies based on two
hitherto unique elements in spectral fitting models: a)
self-consistency between the best-fitting star formation history (SFH)
and the nebular characteristics of a galaxy (e.g., hydrogen
Balmer-line luminosities and equivalent widths; shape of the nebular
continuum, including the Balmer and Paschen discontinuity) and b)
genetic optimization and artificial intelligence algorithms.
This document is part of the FADO v.1 distribution package, which
contains two different ascii files, ReadMe and Read_F, and one tarball
archive FADOv1.tar.gz. FADOv1.tar.gz contains the binary (executable)
compiled in both OpenSuSE 13.2 64bit LINUX (FADO) and MAC OS X
(FADO_MACOSX). The former is compatible with most LINUX distributions,
while the latter was only tested for Yosemite 10.10.3. It contains the
configuration files for running FADO: FADO.config and PLOT.config, as
well as the "Simple Stellar Population" (SSP) base library with the
base file list Base.BC03.L, the FADO v.1 short manual Read_F and this
file (in the ReadMe directory) and, for testing purposes, three
characteristic de-redshifted spectra from SDSS-DR7 in ascii format,
corresponding to a star-forming (spec1.txt), composite (spec2.txt) and
LINER (spec3.txt) galaxy. Auxiliary files needed for execution of FADO
(.HIfbound_em.ascii, .HeII_fbound.ascii, .HeIfboundem.ascii,
grfont.dat and grfont.txt) are also included in the tarball. By
decompressing the tarball the following six directories are created:
input, output, plots, ReadMe, SSPs and tables (see below for a brief
explanation).
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
ReadMe 80 . This file
Read_F 512 165 FADO v.1 ReadMe
FADOv1.tar 512 66400 FADO v.1 tarball
--------------------------------------------------------------------------------
Description of file: FADOv1.tar.gz
If you list the files after decompressing you should get:
--------------------------------------------------------------------------------
SIZE (bytes) FILES | Description
--------------------------------------------------------------------------------
7234080 FADO | FADO binary (executable) LINUX
6664 FADO.config | FADO config file for the fitting
7003036 FADO_MACOSX | FADO binary (executable) MAC OS X
66020 grfont.dat | FONTS for PGPLOT
557436 grfont.txt | FONTS for PGPLOT
324993 .HeIfboundem.ascii | Auxiliary file for running FADO
44222 .HeII_fbound.ascii | Auxiliary file for running FADO
21286 .HIfbound_em.ascii | Auxiliary file for running FADO
4096 input | Directory with 3 SDSS-DR7 spectra
4096 output | Empty directory for output
2815 PLOT.config | FADO plotting configuration file
4096 plots | Empty directory for plots
4096 ReadMe | Short manual Read_F and ReadMe
12288 SSPs | SSPs and base file Base.BC03.L
4096 tables | Empty directory for tables
--------------------------------------------------------------------------------
Note on input directory:
Three characteristic SDSS DR7 spectra are two-column ascii format,
corresponding to wavelenghts and fluxes and one ascii table (List)
with the name and luminosity distance in Mpc.
Byte-by-byte Description of file: spectra: spec1.txt spec2.txt spec3.txt
-------------------------------------------------------------------------------
Bytes Format Units Label Explanations
-------------------------------------------------------------------------------
1 - 5 F6.1 0.1nm Wavelength Wavelength
15- 21 F8.2 10-20W/m2 Flux Flux (in 10-17erg/s/cm2)
-------------------------------------------------------------------------------
Byte-by-byte Description of file: List
-------------------------------------------------------------------------------
Bytes Format Units Label Explanations
-------------------------------------------------------------------------------
1 - 8 A9 --- Spectrum Spectrum file name
10- 15 F6.2 Mpc Distance Distance
17- 20 F4.2 0.1nm Resolution FWHM
-------------------------------------------------------------------------------
Acknowledgements:
Jean Michel Gomes, jean (at) astro.up.pt
Polychronis Papaderos, papaderos (at) astro.up.pt
(End) Patricia Vannier [CDS] 30-Mar-2017