The large-scale observational signatures of low-mass galaxies during reionization

• Published in: Monthly notices of the Royal Astronomical Society 2016-03, Vol.456 (3), p.3011-3029
• Main Authors: Dixon, Keri L., Iliev, Ilian T., Mellema, Garrelt, Ahn, Kyungjin, Shapiro, Paul R.
• Format: Article
• Language: English
• Subjects: Astronomy; Computer simulation; cosmology: theory; Feedback; Galaxies; galaxies: formation; intergalactic medium; Ionization; Ions; large-scale structure of Universe; Normal distribution; Radiative transfer; radio lines: galaxies; Signatures; Simulation; Stars & galaxies; Statistics; Symbols
• ISSN: 0035-8711; 1365-2966; 1365-2966
• DOI: 10.1093/mnras/stv2887

Observations of the epoch of reionization give us clues about the nature and evolution of the sources of ionizing photons, or early stars and galaxies. We present a new suite of structure formation and radiative transfer (RT) simulations from the PRACE4LOFAR project designed to investigate whether the mechanism of radiative feedback, or the suppression of star formation in ionized regions from UV radiation, can be inferred from these observations. Our source halo mass extends down to 108 M⊙, with sources in the mass range 108–109 M⊙ expected to be particularly susceptible to feedback from ionizing radiation, and we vary the aggressiveness and nature of this suppression. Not only do we have four distinct source models, we also include two box sizes (67 and 349 Mpc), each with two grid resolutions. This suite of simulations allows us to investigate the robustness of our results. All of our simulations are broadly consistent with the observed electron-scattering optical depth of the cosmic microwave background and the neutral fraction and photoionization rate of hydrogen at z ∼ 6. In particular, we investigate the redshifted 21-cm emission in anticipation of upcoming radio interferometer observations. We find that the overall shape of the 21-cm signal and various statistics are robust to the exact nature of source suppression, the box size, and the resolution. There are some promising model discriminators in the non-Gaussianity and small-scale power spectrum of the 21-cm signal.