Observational Signatures of the First Galaxies
aa r X i v : . [ a s t r o - ph . C O ] M a r Observational Signatures of the First Galaxies
Jarrett L. Johnson
Max-Planck-Institut für extraterrestrische Physik, Giessenbachstraße, 85748 Garching, GermanyTheoretical Modeling of Cosmic Structures Group
Abstract.
Detection of the radiation emitted from some of the earliest galaxies will be madepossible in the next decade, with the launch of the James Webb Space Telescope (JWST). Asignificant fraction of these galaxies may host Population (Pop) III star clusters. The detection ofthe recombination radiation emitted by such clusters would provide an important new constrainton the initial mass function (IMF) of primordial stars. Here I review the expected recombinationline signature of Pop III stars, and present the results of cosmological radiation hydrodynamicssimulations of the initial stages of Pop III starbursts in a first galaxy at z ∼
12, from which the time-dependent luminosities and equivalent widths of IMF-sensitive recombination lines are calculated.While it may be unfeasible to detect the emission from Pop III star clusters in the first galaxies at z > 10, even with next generation telescopes, Pop III star clusters which form at lower redshifts (i.e.at z < 6) may be detectable in deep surveys by the JWST. Keywords: cosmology, galaxy formation, H II regions, halos, high-redshift, intergalactic medium PACS: INTRODUCTION
In this contribution, I will address three key questions pertaining to the observationalsignatures of the first galaxies, and in particular to the prospects for the detection andidentification of Pop III stars. These questions are the following: • How can Population (Pop) III stars be identified observationally and their initialmass function (IMF) constrained? • How long did Pop III star formation continue after the epoch of the first stars? • Will observational facilities in the coming years, and in particular the
James WebbSpace Telescope (JWST), be able to detect and identify Pop III stellar populations?The next three sections are devoted to addressing these three questions in the order givenabove; in the final section I will briefly summarize the main conclusions.
THE SPECTRAL SIGNATURE OF POP III STAR FORMATION
It is well-known that Pop III stars are likely to be considerably hotter than present-daystars, for a given stellar mass (e.g. [3, 6, 9, 10]). The high surface temperatures of Pop IIIstars lead to enhanced emission of ionizing radiation, in particular photons with energiesabove 54.4 eV which can ionize He II . This, in turn, implies that the photoionized regionssurrounding Pop III stars should shine brightly in He II recombination lines, principallyin He II l IGURE 1.
The escape fraction of H I -ionizing photons ( left panel ) and that of He II -ionizing photons( right panel ), from a Pop III star cluster formed in a first galaxy at z ∼
12. Each line corresponds to adifferent choice of IMF and total number of stars, as labeled. There is a tight anticorrelation between theescape fraction of H I -ionizing photons and the emission in hydrogen recombination lines (e.g. Ly a andH a ; see Figs. 2 and 3). For most cases, however, the negligible escape fraction of He II -ionizing photonsleads to a tight correlation between the luminosity emitted in the He II l geared toward detecting this recombination line as an indicator of Pop III star formation[19, 21] (see also [4, 7, 11, 23, 27, 30]), although to date no definitive detections on PopIII star formation have been reported. Nonetheless, the He II l ⊙ or 100 M ⊙ ). However, constrainingthe IMF using the ratio of the observed fluxes in He II l a or Ly a posessome challenges, owing to the evolution of the flux emitted in these lines. While theevolution of the massive stars in a cluster will alter its spectral characteristics [25, 26],the photoheating of the gas surrounding the cluster will also lead to similar evolution[17]. In particular, as the gas surrounding the cluster is photoheated it expands, therebyallowing for the escape of ionizing photons into the intergalactic medium (IGM); asmore ionizing photons escape, fewer are available to ionize the dense gas from whichrecombination lines are emitted, and the luminosity in those lines correspondingly drops.The escape fraction of ionizing photons from Pop III stellar clusters forming in a firstgalaxy at z ∼
12, as calculated from the simulations presented in [17], are shown inFigure 1. As this Figure shows, the escape fractions of H I - and He II -ionizing photonscan differ greatly, leading to evolution of the ratio of the luminosities emitted in H I andHe II recombination lines, thus complicating the use of such ratios as indicators of theIMF. For two simulations presented in [17], the evolution of the flux visible in H a isshown in Figure 2; as the escape fraction of H I -ionizing photons generally increases IGURE 2.
The flux in H a per square arcsecond, emitted from Pop III stellar clusters at z ∼
12, asobserved on the sky at z = 0, assuming a spectroscopic resolution of R = 1000. Shown here are the twomost massive of the four simulated stellar clusters presented in [17], one containing 25 M ⊙ stars ( bottompanels ), the other containing 100 M ⊙ stars ( top panels ). From left to right, the clusters are shown at 10 yr, 1 Myr, and 3 Myr after formation. The highest total fluxes occur at the earliest times, before the H II region has broken out of the galaxy; hence, the youngest stellar clusters are the most readily observed. with time, the total flux is highest at early times, making the youngest clusters the mosteasily observed.Figure 3 shows the equivalent width (EW) of three prominent recombination lines foreach of the four simulations presented in [17]. While the ratio of the fluxes in He II andH I recombination lines can be a problematic indicator of the stellar IMF, this Figureshows that the EW of He II l POP III STAR CLUSTERS FORMED AFTER REIONIZATION
It is likely that even surveys to be carried out by the JWST will not be deep enough todetect the first stars or galaxies at z ≥
10 [1, 13, 17, 24], although Pop III supernovae (e.g.[14, 28, 31, 33]) and perhaps stars powered by dark matter (DM) annihilation [12, 34]may still be detected. This provides motivation to consider whether Pop III star clustersmay form also at lower redshifts, where IMF-sensitive recombination lines may be morereadily detected.In regions of the universe that undergo reionization at sufficiently early times (i.e. z ∼ z ≤
6. Figure 4 showsthe predicted abundance of such clusters, for various assumptions on the reionizationhistory, the minimum mass of halos which may host star formation after reionization,and the speed of external metal enrichment by neighboring galaxies [16]. As this Figureshows, although likely to be very rare, such Pop III clusters may be abundant enough to
IGURE 3.
The rest frame EWs of Ly a , H a , and He II l dotted lines show the observed EWs of galaxies from two differentsurveys carried out at z = 4.5 and z ≥ II l ⊙ starsis always higher than that for 25 M ⊙ stars, regardless of the total stellar mass in the clusters; hence, weconclude that the EW of this line is a robust indicator of a very top-heavy IMF (see [17]). be detected in the Deep-Wide Survey (DWS) to be carried out by the JWST [13, 32]. PROSPECTS FOR DETECTION IN JWST DEEP SURVEYS
If the number density of Pop III stellar clusters at low redshift (e.g. z ≤
6) is indeed highenough for some of them to lie within the area to be surveyed by the JWST, the questionremains whether these clusters would be bright enough to be detected. Figure 5 showsthe monochromatic Ly a flux predicted for such clusters, for different assumptions onthe IMF, the minimum halo mass for star formation, and the star formation efficiency[16]. As the Figure shows, for a very top-heavy IMF and/or a high star formation effi-ciency, the JWST NIRCam may detect these clusters in the planned DWS. Furthermore,spectroscopic follow-up with NIRSpec may detect the He II l l z ≤
6, and furthermore perhaps only in halos within typical distances of ∼ z ∼
20 [16].
IGURE 4.
The number density n III of Pop III star clusters formed in reionized regions of the universe,taking into account enrichment of the IGM by galactic winds, and normalized to a cluster lifetime t cluster = 2 Myr. The solid lines correspond to a minimum halo circular velocity for star formation of 20 km s − ,while the dashed lines correspond to a minimum of 30 km s − . For each series of lines, the top ( black )line is a model which neglects external metal enrichment, while the colored lines correspond to differentmetal-enriched wind velocities, as labeled. The number density at which one cluster per unit redshift isexpected to be within the planned JWST Deep-Wide Survey area is shown by the dotted line (see [16]). SUMMARY
In closing, I would like to highlight the following conclusions corresponding to the threekey questions addressed in the work presented here: • The Pop III IMF can be constrained with detection of helium recombination emis-sion (particularly the He II l • In rare regions which are reionized at early times, Pop III star formation may extendwell beyond the epoch of the first stars and galaxies. • If this is so, then planned JWST surveys may detect Pop III stellar clusters atredshifts z < 6 and allow for constraints to be placed on the IMF. ACKNOWLEDGMENTS
I would like to thank the conference organizers, Dan Whalen, Naoki Yoshida, and VolkerBromm, for hosting a most enjoyable and productive event, as well as for allowing meto present this work. I am also grateful for support from the Theoretical Modeling ofCosmic Structures (TMoX) Group at MPE.
REFERENCES
1. Barkana R., Loeb A. 2000, ApJ, 531, 6132. Barton E. J., Davé R., Smith J.-D. T., Papovich C., Hernquist L., Springel V. 2004, ApJ, 604, L1
IGURE 5.