A substantial fraction of all ionizing photons originate from radiative recombinations. However, in radiative transfer calculations this recombination radiation is often assumed to be absorbed 'on-the-spot' because for most methods the computational cost associated with the inclusion of gas elements as sources is prohibitive. We present a new, CPU and memory efficient implementation for the transport of ionizing recombination radiation in the TRAPHIC radiative transfer scheme. TRAPHIC solves the radiative transfer equation by tracing photon packets at the speed of light and in a photon-conserving manner in spatially adaptive smoothed particle hydrodynamics simulations. Our new implementation uses existing features of the TRAPHIC scheme to add recombination radiation at no additional cost in the limit in which the fraction of the simulation box filled with radiation approaches 1. We test the implementation by simulating an HII region in photoionization equilibrium and comparing to reference solutions presented in the literature, finding excellent agreement. We apply our implementation to discuss the evolution of the HII region to equilibrium. We show that the widely used case A and B approximations yield accurate ionization profiles only near the source and near the ionization front, respectively. We also discuss the impact of recombination radiation on the geometry of shadows behind optically thick absorbers. We demonstrate that the shadow region may be completely ionized by the diffuse recombination radiation field and discuss the important role of heating by recombination radiation in the shadow region.

It is often assumed that local sources of ionizing radiation have little impact on the distribution of HI in the post-reionization Universe. While this is a good assumption for the IGM, analytic arguments suggest that local sources may typically be more important than the background radiation for high column density absorbers (N_HI > 10^17 /cm^2). We post-process cosmological simulations with accurate radiation transport to investigate the impact of local stellar sources on the HI distribution. We demonstrate that the limited numerical resolution and the simplified treatment of the ISM that are typical of current cosmological simulations provide significant challenges, but that many of the problems can be overcome by taking two steps. First, using ISM particles rather than stellar particles as sources results in a much better sampling of the source distribution. Second, by rescaling the source luminosities so that the amount of radiation escaping into the IGM agrees with that required to produce the observed background, many of the results become insensitive to errors in the predicted fraction of the radiation that escapes the immediate vicinity of the sources. By adopting this strategy and by varying the assumptions about the structure of the unresolved ISM, we conclude that we can robustly estimate the effect of local sources for column densities N_HI << 10^21 /cm^2. However, neither the escape fraction of ionizing radiation nor the effect of local sources on the abundance of N_HI >~ 10^21 systems can be predicted with confidence. We find that local stellar radiation is unimportant for N_HI << 10^17, but that it can affect Lyman Limit and DLA systems. For 10^18 < N_HI < 10^21 the impact of local sources increases with redshift. At z = 5 the abundance of absorbers with N_HI >> 10^17 is substantially reduced, but at z = 0 the reduction only becomes significant for N_HI >~ 10^21 /cm^2.

We use a set of cosmological simulations combined with radiative transfer calculations to investigate the distribution of neutral hydrogen in the post-reionization Universe. We assess the contributions from the metagalactic ionizing background, collisional ionization and diffuse recombination radiation to the total ionization rate at redshifts z=0-5. We find that the densities above which hydrogen self-shielding becomes important are consistent with analytic calculations and previous work. However, because of diffuse recombination radiation, whose intensity peaks at the same density, the transition between highly ionized and self-shielded regions is smoother than what is usually assumed. We provide fitting functions to the simulated photoionization rate as a function of density and show that post-processing simulations with the fitted rates yields results that are in excellent agreement with the original radiative transfer calculations. The predicted neutral hydrogen column density distributions agree very well with the observations. In particular, the simulations reproduce the remarkable lack of evolution in the column density distribution of Lyman limit and weak damped Ly\alpha\ systems below z = 3. The evolution of the low column density end is affected by the increasing importance of collisional ionization with decreasing redshift. On the other hand, the simulations predict the abundance of strong damped Ly\alpha\ systems to broadly track the cosmic star formation rate density.

An ionization front expanding into a neutral medium can be slowed-down significantly by recombinations. In cosmological numerical simulations the recombination rate is often computed using a 'clumping factor', that takes into account that not all scales in the simulated density field are resolved. Here we demonstrate that using a single value of the clumping factor significantly overestimates the recombination rate, and how a local estimate of the clumping factor is both easy to compute, and gives significantly better numerical convergence. We argue that this lower value of the recombination rate is more relevant during the reionization process and hence that the importance of recombinations during reionization has been overestimated.

The Durham GALFORM semi-analytical galaxy formation model has been shown to reproduce the observed rest-frame 1500\AA\ luminosity function of galaxies well over the whole redshift range z=5-10. We show that in this model, this galaxy population also emits enough ionizing photons to reionize the Universe by redshift z=10, assuming a modest escape fraction of 20 per cent. The bulk of the ionizing photons is produced in faint galaxies during starbursts triggered by galaxy mergers. The bursts introduce a dispersion up to ~ 5 dex in galaxy ionizing luminosity at a given halo mass. Almost 90 per cent of the ionizing photons emitted at z=10 are from galaxies below the current observational detection limit at that redshift. Photo-ionization suppression of star formation in these galaxies is unlikely to affect this conclusion significantly, because the gas that fuels the starbursts has already cooled out of their host halos. The galaxies that dominate the ionizing emissivity at z=10 are faint, with M_{1500, AB} ~ -16, have low star formation rates, \dot{M_{*}} ~ 0.06 h^{-1} M_sun yr^{-1}, and reside in halos of mass M ~ 10^9 h^{-1} M_sun.