Warm molecular Hydrogen at high redshift with the James Webb Space Telescope
P. Guillard, F. Boulanger, M. D. Lehnert, P. N. Appleton, G. Pineau des Forêts
SSF2A 2015
S. Boissier, V. Buat, L. Cambr´esy, F. Martins and P. Petit (eds)
WARM MOLECULAR HYDROGEN AT HIGH REDSHIFT WITH THE JAMES WEBBSPACE TELESCOPE
P. Guillard , F. Boulanger , M. D. Lehnert , P. N. Appleton and G. Pineau des Forˆets , Abstract.
The build-up of galaxies is regulated by a complex interplay between gravitational collapse,galaxy merging and feedback related to AGN and star formation. The energy released by these processeshas to dissipate for gas to cool, condense, and form stars. How gas cools is thus a key to understand galaxyformation.
Spitzer Space Telescope infrared spectroscopy revealed a population of galaxies with weak starformation and unusually powerful H line emission. This is a signature of turbulent dissipation, sustained bylarge-scale mechanical energy injection. The cooling of the multiphase interstellar medium is associated withemission in the H lines. These results have profound consequences on our understanding of regulation ofstar formation, feedback and energetics of galaxy formation in general. The fact that H lines can be stronglyenhanced in high-redshift turbulent galaxies will be of great importance for the James Webb Space Telescope observations which will unveil the role that H plays as a cooling agent in the era of galaxy assembly.Keywords: Galaxies: evolution, interstellar medium, molecular gas, turbulence, accretion, feedback – stars:formation – ISM: turbulence, kinematics, dynamics – Infrared: ISM In the ΛCDM framework, galaxies are assembled from the collapse of gas in virialized dark matter haloes (e.gWhite & Rees 1978; Fall & Efstathiou 1980). The most outstanding question in all contemporary theoreticalstudies of galaxies evolution is what processes regulate the gas content of galaxies, that is, the balance betweenaccretion and mass loss. This balance, and thus the build-up of baryonic mass in galaxies, is regulated by acomplex interplay between gravitational collapse, gas accretion, galaxy merging and feedback related to activegalactic nuclei (AGN) activity and star formation (e.g. Dekel & Birnboim 2006). It is this competition betweenthe rates of inflow, outflow, and star formation that gives the properties and physical characteristics of thegalaxies we observe today. What currently limits our understanding of galaxy formation is how does thegas respond to those feedback mechanisms, which may inject sufficient mechanical energy into the interstellarmedium (ISM) to have a major impact on star formation and galaxy assembly, thus potentially regulating thegrowth of galaxies (Lehnert et al. 2015; Guillard et al. 2015). Those feedback processes will be particularlyimportant during the early phases of galaxy evolution at high redshift, when galaxies were gas-rich and mostof the stars in the universe were formed.The energy injected by feedback processes has to be dissipated for gas to cool and form stars. Observationsof galaxies experiencing strong feedback and turbulence (e.g. galaxy interactions, AGN, cluster cooling flows)show that a significant fraction of this energy cascades down to small scales and is dissipated through lineemission. This turbulent cascade is associated with the formation of multiphase ISM and one of the dominantcooling channel is through H line emission (Guillard et al. 2009, 2012b). H is a natural outcome of gas cooling,and the material from which stars are formed. Being affected by star formation, massive central black holes,and inflows and outflows of gas, H plays an important role in all stages of galaxy formation and evolution Sorbonne Universit´es, UPMC Universit´e Paris 6 et CNRS, UMR 7095, Institut d’Astrophysique de Paris, 98 bis bd Arago,75014 Paris, France. Institut d’Astrophysique Spatiale, UMR 8617, CNRS, Universit´e Paris-Sud, Bat. 121, 91405 Orsay, France NASA Herschel Science Center, Infrared Processing & Analysis Center, California Institute of Technology, Pasadena, CA91125, USA Observatoire de Paris, LERMA, UMR 8112, CNRS, 61 Avenue de l’Observatoire, 75014 Paris, Francec (cid:13)
Soci´et´e Francaise d’Astronomie et d’Astrophysique (SF2A) 2015 a r X i v : . [ a s t r o - ph . GA ] O c t
38 SF2A 2015
Figure 1.
Left:
Spitzer IRS mid-infrared spectrum of the Stephan’s Quintet intra-group medium (Appleton et al. 2006;Cluver et al. 2010).
Right:
Integrated VLT/SINFONI near-infrared spectra of two regions (a compact and extended one)in the overlap area of the Antennae galaxies (Herrera et al. 2011). In the right panel, blue contours mark the aperturefrom which each spectrum was extracted. In both sources, the spectra show prominent H rotational and ro-vibrationallines, respectively. The H emission is originating from the dissipation of turbulent energy driven by large-scale gasdynamics, a galaxy hitting a tidal filament for Stephan’s Quintet and the formation of bound clouds through accretionfor the Antennae overlap region. (Boulanger et al. 2009). In this paper, we stress the unique capability of the James Web Space Telescopeto detect and characterize H line emission at the peak of the star-forming activity of the Universe. Thoseobservations will be key to study the structure and phase distribution of the gas, because they will allow us toestimate the cooling, turbulent dissipation, and dynamical times. line emission as a probe of the energetics of molecular gas Excitation of rotation-vibration levels of H can occur through different mechanisms. Collisional excitationwith atoms and molecules (e.g. Flower 1998; Le Bourlot et al. 2002), absorption of UV photons followed byfluorescence (e.g. Gautier et al. 1976; Black & van Dishoeck 1987), heating by hard X-rays penetrating into themolecular clouds (Maloney et al. 1996; Tine et al. 1997), and cosmic ray heating (Dalgarno et al. 1999), whichhas mainly been discussed for the strong H emission in cooling-flow filaments (Ferland et al. 2008).Rotational and ro-vibrational lines of molecular hydrogen (H ) have become an important diagnostic toolfor shocks in the galactic (e.g. Allen & Burton 1993; Falgarone et al. 2005; Hewitt et al. 2009; Ingalls et al. 2011)and extra-galactic interstellar medium (e.g. Wright et al. 1993; Appleton et al. 2006; Veilleux et al. 2009; Ogleet al. 2010; Beir˜ao et al. 2015). Two examples are given in Figure 1. Pure rotational lines of H (0-0 S(0), 0-0S(1), etc.) are found in the mid-infrared between 3 − µ m and trace warm gas with typical temperatures of afew 100 K to 1000 K. Ro-vibrational lines of H (e.g., 1-0 S(1) at 2.12 µ m) are observed in the near-infrared andtrace hotter gas with temperatures of a few 1000 K. According to shock models, H line ratios can be used toinfer the pre-shock gas characteristics (density, magnetic field) and shock properties (velocity, non-dissociative– C-type – or dissociative – J-type –) (Flower & Pineau des Forˆets 2010; Guillard et al. 2009, 2012b). z ∼ . − . with JWST The
James Webb Space Telescope will be, 15 years after the
Spitzer Space Telescope , the next mission tohave access to rotation-vibration H transitions. As such, it will play a critical role in the context of galaxyevolution since H represents an important, if not dominant, cooling agent in the energetics of galaxy formation.Observations by the InfraRed Spectrograph (IRS) onboard the Spitzer Space Telescope unveiled a significant anddiverse population of low- z objects where the mid-infrared rotational line emission of H is strongly enhanced( L H ∼ − erg s − ), while star formation is suppressed (see Figure 2). This suggest that shocks are theprimary cause of the H emission (Guillard et al. 2009). This sample of H -luminous sources includes galaxiesin several key phases of their evolution, dominated by, for instance, gas accretion onto bright central galaxiesin clusters (Egami et al. 2006), galaxy interactions (Appleton et al. 2006), or galactic winds driven by starformation (e.g. M82 Beir˜ao et al. 2015), and radio-loud AGN (Ogle et al. 2010; Guillard et al. 2012b). In thosesources, the turbulent dissipation time is longer than the dynamical time, the mechanical energy contained inthe molecular phase being dominant over the thermal energy of the gas (e.g. Guillard et al. 2012a).arm molecular Hydrogen at high redshift with JWST 239 M82wind z=2 ULIRGs
PKS1138-26
Figure 2.
Ratio of the mid-IR H line luminosities (summed over S(0) to S(3)) to the PAH 7.7 µ m emission vs. 24 µ mcontinuum luminosity (updated from Guillard et al. 2012b). This ratio indicates the relative contribution of mechanicalheating (shocks) and star-formation (SF) power (UV excitation). The red pentagons are nearby radio galaxies with fast( > i outflows observed with Spitzer IRS (Guillard et al. 2012b). The orange triangles and green ellipsesare samples of radio galaxies (respectively Ogle et al. 2010; Kaneda et al. 2008). These H -luminous galaxies stand outabove SF and AGN galaxies from the SINGS survey (Roussel et al. 2007). The H emission in these sources cannot beaccounted by UV or X-ray photon heating. The blue dashed line shows the upper limit given by the Kaufman et al.(2006) PDR models ( n H = 10 cm − , G UV = 10). For comparison, a few other types of H -luminous galaxies are shown:the Stephan’s Quintet (SQ) and Taffy galaxy collisions (Cluver et al. 2010; Peterson et al. 2012), other Hickson CompactGroups (black squares, Cluver et al. 2013), the ZW 3146 (Egami et al. 2006) and Perseus A (Johnstone et al. 2007)clusters, and the NGC 6240 merger (Armus et al. 2006). The black ellipse shows the Spitzer IRS observations of theM82 wind (Beir˜ao et al. 2015), the black rectangle shows the detection of H in stacked Spitzer spectra of z = 2 ULIRGs(Fiolet et al. 2010), and the purple cross the detection of H in the Spider Web (PKS1138-26) radio galaxy protocluster(Ogle et al. 2012). Constraining the impact of merging and AGN feedback on the formation and evolution of massive galaxiescan only be addressed through direct H line observations at z ∼
2, near the cosmologically most active periodof star formation, galaxy interactions and AGN activity. By analogy to what is observed on local H -luminousobjects, we expect the mid-IR lines to be the dominant cooling lines for warm, 10 − K, gas in the stronglyshocked, highly turbulent, colliding flows in galaxy interactions (e.g. the galaxy-wide shock in Stephan’s Quintet,Guillard et al. 2009), but also, e.g., in AGN-driven outflows. High gas velocity dispersions measured in z ∼ emission is expected to be more frequent and more powerful than at low- z , as suggested byH detections in z ≈ line emission is likely powered by the dissipation of turbulence, which could originate from star formation(supernovae), radiation pressure, or gas accretion.40 SF2A 2015 Figure 3.
Observing H lines at high-redshift with JWST/MIRI. The observed wavelengths of some H lines are shownas a function of the redshift. The colored bars indicate the channels and bands of the Medium Resolution Spectrometer(MRS) of the MIRI instrument. One observation corresponds to four sub-bands, like 1A, 2A, 3A, 4A for instance (seeGuillard, P. 2010, for technical details about the MRS operations). The vertical lines indicate the redshifts of somehigh- z radio-galaxies that might be interesting to look at with MIRI. Covering a wavelength range of 4 . − . µ m, the MIRI Medium Resolution Spectrometer (MRS, Wellset al. 2015) will be the first Integral Field Unit (IFU) instrument to provide the sensitivity and resolvingpower to spatially and spectrally resolve H and forbidden ionized gas lines at rest-frame near-IR and mid-IRwavelengths, out to z = 1 . − . . − µ m in the near-infrared, NIRSPEC (Posselt et al.2004) will allow the detection of ro-vibrational lines at very high sensitivity (0 . × − W m − for 1h) andspectral resolution ( R ≈ >