COSMIC DAWN SIMULATIONS A&G • June 2015 • Vol. 56 • www.astrongeo.com 3.31 T he frst billion years of cosmic evolu- tion remains the least well under- stood cosmic epoch. During this key period the the very frst stars and galaxies formed, and the cosmic web of structures we observe today frst started taking shape. Better knowledge of these epochs is there- fore of key importance for understand- ing galaxy formation and evolution. The ionizing ultraviolet radiation from these frst galaxies initiated the process of cosmic reionization, which eventually ionized and heated the entire universe to about ten thousand degrees Kelvin, in which ionized state it still remains. This process has had profound effects on the formation of cosmic structures and has left a lasting impression on them. This reionization process is inherently multiscale. Under the cold dark matter paradigm, nonlinear structures form from initially small-amplitude density fuctuations, by a continuous hierarchical sequence of mergers and infall, with the smallest galaxies forming frst and merg- ing to yield ever larger ones, eventually forming the cosmic web of structures we see today. It is thus generally believed that the reionization process is driven primar- ily by stellar radiation from low-mass galaxies, which cluster on large scales and collectively create very large ionized patches whose eventual overlap completes the process. The star forma- tion inside such galaxies is strongly affected by complex radiative and hydrodynamic feedback effects, including ionizing and non-ionizing UV radiation, shock waves, gas cooling and heating, stellar winds and enrichment by heavy elements. Under- standing the nature of the frst galaxies and how they affect the progress, proper- ties and duration of the cosmic reioniza- tion requires detailed modelling of these complex, nonlinear interactions. These processes are not fully understood even in the present-day universe, with its wealth of observational constraints, and in the early epochs they are essentially still a mystery. It is therefore not yet possible to model the epoch of reionization from frst principles. Observational data The main obstacle to further progress is the scarcity of observational data. Cur- rently this mostly probes the tail-end of reionization: both Lyman-α source surveys and intergalactic medium absorption lines probe low neutral fractions, as a result of the high optical depth to resonant scatter- ing of such radiation by even small amounts of neutral hydrogen. Other data such as the cosmic microwave background optical depth and polarization, kinetic Sunyaev–Zel’dovich effect and near- infrared background are integral measures of the reionization history (described by Wilkins and Stanway on pages 3.21–3.24). The detection of the redshifted 21 cm signal produced by the spin-fip hyperfne split- ting of the ground energy state of hydrogen promises to provide full 3D tomographic observations of the intergalactic medium throughout, and possibly even before, reionization (as described by Jonathan Pritchard on pages 3.25–3.30). This shortage of observational Simulating the cosmic dawn “The main obstacle to further progress is the scarcity of observational data” 1 The formation of early gas dynamical structures (at z = 18) using the code RAMSES (Teyssier 2002), which uses an adaptive mesh refinement technique. (Left): Slice of the gas density in the full, 23 comoving Mpc simulation volume. The highly resolved region in the centre has up to 23 levels of refinement, for an effective resolution of 2.7 proper pc. (Right): A zoomed region of the simulation, centred on a small halo of the kind in which the first stars in the universe formed. Ilian T Iliev, David Sullivan and Keri L Dixon outline two different approaches to simulating reionization and explain how new data will improve modelling of the early universe. Downloaded from https://academic.oup.com/astrogeo/article-abstract/56/3/3.31/230105 by guest on 26 May 2020