MAGPHYS: a publicly available tool to interpret observed galaxy SEDs

Transcription

1 The Spectral Energy Distribution of Galaxies Proceeings IAU Symposium No. 284, 2011 R.J. Tuffs & C.C. Popescu, es. c International Astronomical Union 2012 oi: /s MAGPHYS: a publicly available tool to interpret observe galaxy SEDs Elisabete a Cunha 1,Stéphane Charlot 2, Loretta Dunne 3, Dan Smith 4, an Kate Rowlans 3 1 Max Planck Institute for Astronomy, Königstuhl 17, Heielberg, Germany 2 UPMC Univ. Paris 6/CNRS, UMR 7095, Institut Astrophysique e Paris, France 3 School of Physics & Astronomy, Nottingham University, University Park Campus, Nottingham NG7 2RD, UK 4 Centre for Astrophysics, Science & Technology Research Institute, University of Hertforshire, Hatfiel, Herts, AL10 9AB, UK cunha@mpia.e Abstract. We present a simple, physically-motivate moel to interpret consistently the emission from galaxies at ultraviolet, optical an infrare wavelengths. We combine this moel with a Bayesian metho to obtain robust statistical constraints on key parameters escribing the stellar content, star formation activity an ust content of galaxies. Our moel is now publicly available via a user-frienly coe package, MAGPHYS at Wepresentan application of this moel to interpret a sample of 1400 local (z <0.5) galaxies from the H- ATLAS survey. We fin that, for these galaxies, the iffuse interstellar meium, powere mainly by stars oler than 10 Myr, accounts for about half the total infrare luminosity. We iscuss the implications of this result to the use of star formation rate inicators base on total infrare luminosity. Keywors. ust, extinction galaxies: ISM galaxies: stellar content galaxies: statistics. 1. A simple moel to interpret galaxy SEDs Multi-wavelength surveys of large samples of galaxies both in the local an highreshift Universe have become wiely available in recent years. To unerstan these observations in the framework of galaxy formation an evolution, we must be able to extract key physical parameters from their observe spectral energy istributions (SEDs). In a Cunha, Charlot & Elbaz (2008), we have evelope a simple moel to interpret consistently the ultraviolet, optical an infrare emission from galaxies in terms of their star formation histories an ust content Description of the moel We compute the spectral evolution of stellar populations in galaxies using the state-ofthe-art population synthesis moel of Bruzual &Charlot (2003). This moel is base on the propertythat stellar populations with any star formation history can be expane in a series of instantaneous bursts, simple stellar populations (SSPs). The spectral energy istribution of a galaxy is then compute by aing the iniviual spectra of all SSPs weighte by the star formation rate over time since the galaxy was forme. The spectra of galaxies also contain valuable information about the interaction of starlight with interstellar gas an ust, an the physical properties of the interstellar meium (ISM), such as ust content. Following Charlot & Fall (2000), we escribe the 292

2 MAGPHYS: a tool to interpret observe galaxy SEDs 293 Birth clous (young stars) iffuse interstellar meium (oler stars) Figure 1. Schematic view of the two-component ISM by Charlot & Fall (2000) (top-right): stars are born in ense molecular clous birth clous an later (after 10 7 yr) migrate to the ambient iffuse ISM. The left-bottom plot shows an example total ust emission SED (soli line) in the infrare range, constructe using the emission components escribe in a Cunha, Charlot & Elbaz (2008). The contributions by the birth clous an the ambient ISM are plotte with short-ashe an long-ashe lines, respectively. ISMof galaxies inourmoel usingtwo main components: the ambient (iffuse) interstellarmeium an the star-forming regions (birth clous). Starsare bornin ense molecular clous which issipate typically on a time-scale of 10 7 years. As a result, the non-ionizing continuum emission from young OB stars an line emission from their surrouning Hii regions may be absorbe by ust in these birth clous an then in the ambient ISM, while the light emitte by stars oler than 10 7 yr propagates only through the iffuse ISM. This simple moel successfully accounts for the ifferent attenuation of line an continuum emission in star-forming galaxies. We use this prescription to compute the total energy absorbe by ust in the birth clous an in the ambient ISM. We efine the total ustluminosity re-raiate byustin the birthclousanin the ambient ISM as L BC an L ISM, respectively. The total luminosity emitte by ust in the galaxy is L tot = L BC + L ISM. We istribute L BC to 1000 µm using four main ust components: (i) the emission from polycyclic aromatic hyrocarbons (PAHs); (ii) the mi-ir continuum from hot ust; (iii) the emission from warm ust (30 60 K) in thermal equilibrium; (iv) the emission from col ust (15 25 K) an L ISM in wavelength over the range from 3 in thermal equilibrium. In stellar birth clous, the relative contributions to L BC by PAHs, the hot mi-infrare continuum an warm ust are kept as ajustable parameters. These clous are assume not to contain any col ust. In the ambient ISM, the contribution to L ISM by col ust is kept as an ajustable parameter. The relative proportions of the other 3 components are fixe to the values reproucing the mi-infrare cirrus emission of the Milky Way. We fin that this minimum number of components is require to

3 294 E. a Cunha et al. Figure 2. Summary of the Bayesian approach we use to compare observe galaxy SEDs with our moel an erive statistical constraints on the physical parameters of observe galaxies (see a Cunha et al for etails). account for the infrareseds of galaxies in a wie range of star formation histories (see a Cunha et al for etails). An example moel SED is shownin in Fig Statistical constraints of physical parameters Ourmoel is optimize to erive statistical constraints of funamental parameters relate to star formation activity an ust content (e.g. star formation rate, stellar mass, ust attenuation, ust temperatures) of large samples of galaxies using a wie range of multiwavelength observations (e.g. a Cunha et al. 2008, 2010). We use a Bayesian approach, summarize in Fig. 2, to interpret the SEDs all the way from the ultraviolet/optical to the far-infrare. A similar approach has been previously use mostly to interpret optical galaxy spectra from the ultraviolet to the near-infrare, i.e. not incluing the ust emission. This approach allows us to unerstan in etail our parameter constraints by ientifying egeneracies an exploring what observations are require to constrain each parameter. 2. The MAGPHYS package The a Cunha, Charlot & Elbaz (2008) moel is publicly available as the MAG- PHYS package at MAGPHYS - Multi-wavelength Analysis of Galaxy Physical Properties - is a self-containe, user-frienlymoel package to interpret observe spectral energy istributions of galaxies in terms of galaxy-wie physical parameters pertainingto the stars anthe interstellar meium. The analysis of the spectral energy istribution of an observe galaxy with MAGPHYS is one in two steps: (i) The assembly ofa comprehensive libraryofmoel SEDsat the same reshiftanin the same photometric bans as the observe galaxy, for wie ranges of plausible physical parameters pertaining to the stars an ISM. (ii) The buil-up of the marginalise likelihoo istribution of each physical parameter of the observe galaxy, through the comparison of the observe SED with all the moels in the library (Fig. 2).

4 MAGPHYS: a tool to interpret observe galaxy SEDs 295 Figure 3. Star formation rate, SFR, plotte against total ust luminosity, L tot, for 1404 galaxies from the H-ATLAS survey (a Cunha et al., in prep.). Each galaxy is colour-coe (on a grey-scale) accoring to the fraction of total infrare luminosity contribute by the iffuse ISM, f µ. The otte line shows, for reference, the conversion between total IR luminosity an SFR from Kennicutt (1998). 3. Dust heating by stars oler than 10 Myr in H-ATLAS galaxies The H-ATLAS survey is etecting thousans ofgalaxies in the Herschel/SPIREbans over a large area of the sky. Thanks to the overlap with other multi-wavelength surveys, complete UV-to-IR SEDs are available for large samples of galaxies (Eales et al. 2010). We have use MAGPHYS to extract statistical constraints on star formation rates (SFRs), stellar masses, ustluminosities anustmasses of a sample of 250-µm etecte galaxies at z<0.5 from their observe SEDs (Smith et al., submitte). We are investigating the use of the IR luminosity as a SFR inicator in these galaxies. The total IR is often use as a SFR tracer but observational evience shows that ustin galaxies is not exclusively heate by newly-forme stars (e.g. Beno et al. 2010). We fin that the SFR erive from fits to the total SED with our moel, which inclues heating by stars oler than 10 7 yr in the iffuse ISM, is not exactly trace by the ust luminosity (as, for example, when using the Kennicutt 1998 IR to SFR conversion; see Fig. 3). A significant fraction (typically 50%) of L tot in H-ATLAS galaxies comes from the iffuse ISM, mainly powere by stars oler than 10 Myr. This implies that using the total IR as a SFR tracer may lea to overestimating the SFR unless the contribution by the iffuse ISM to the total IR is properly taken into account (see also e.g. Kennicutt 2009). Acknowlegements We thank the H-ATLAS team for proviing the photometry for the work escribe in Section 3; the H-ATLAS website is References Beno G. J. et al. 2010, A&A, 518, L65 Bruzual G. & Charlot S. 2003, MNRAS, 344, 1000 Charlot S. & Fall S. M. 2000, ApJ, 539, 718 a Cunha E., Charlot S., & Elbaz D. 2008, MNRAS, 388, 1595 a Cunha E., Eminian C., Charlot S., & Blaizot J. 2010, MNRAS, 403, 1894 Eales S. et al. 2010, PASP, 122, 499 Kennicutt R. C., Jr. 1998, ARA&A, 36, 189 Kennicutt R. C., Jr. et al. 2009, ApJ, 703, 1672

5 296 E. a Cunha et al. Discussion Maen: Coul you escribe further the parameters of the ust components in your moel? For example for the moifie black boies you use for the FIR-submm, are the beta parameter an temperature fixe? Do you solve for PAHs, or o you use fixe templates from moels/obervations? Da Cunha: The emission by ust in thermal equilibrium in our moel is escribe by a set of moifie black boies. In the stellar birth clous, we aopt an emissivity inex of β =1.5 to escribe the emission by warm ust; the temperature of this ust can vary between 30 an 60 K. Col ust in the iffuse ISM is escribe by β =2 an the equilibrium temperature can vary between 15 an 25 K. The emissivity inex is fixe for simplicity, to keep the number of ajustable parameters of the moel minimal. The PAH emission is escribe by an empirical template for simplicity: we use the mi-ir spectrum of the M17 PDR. The contribution of PAHs to the total IR of the iffuse ISM is fixe accoring to the observe SED of the Milky Way cirrus emission. In the birth clous the contribution of PAHs to the total IR is allowe to vary (more etails are given in a Cunha, Charlot & Elbaz 2008). Beno: Diyou test your moel fittingcoe on moel spectra without some SEData points (similar to the tests performe with observational ata)? Da Cunha: Yes, we i such tests in a Cunha, Charlot & Elbaz (2008), to test how well we can recover known moel parameters from a set of photometric observations. We have not one this in the context of higher reshift galaxies (z 2) an two-imensional pfs as I showe in this talk, but this is efinitely a test we plan to o in the near future. Schaerer: I have a comment regaring the contribution from ol stars to the IR emission. One has to be careful: I suppose that by ol you mean younger than 10 Myr? Remember that stars oler than this still contribute to the UV. That may explain the very large contribution of ol stars you fin. Da Cunha: Inee that is a goo point. By ol stars I mean stars oler than 10 Myr, which have migrate from the stellar birthclous into the iffuse ISM component. These stars still raiate a large amount of non-ionizing UV raiation which heats the ust. But even stars with little UV emission (typically oler than 100 Myr) still contribute to the ustheating in a significant way.buti agree that if we consiere a large time-scale, the fraction contribute to the IR luminosity might be smaller. The main point of the plot I showe was to emphasize that a non-negligible fraction of the total IR is not powere by recent star formation, an this shoul be taken into account when eriving SFR from the observe IR luminosity.