The reionization of the intergalactic medium and its observational consequences

We have calculated the thermal and ionization evolution of a uniform density intergalactic medium (IGM) composed of H and He in a postrecombination Friedmann universe as sources inject ionizing radiation and energy into it. We have solved nonequilibrium rate equations for ionization and recombination, together with the equations of energy conservation, including the effects of cosmological expansion, radiative and Compton cooling, and the diffuse flux emitted by the gas, and radiative transfer, in this coarse-grained-average description of the IGM. For the radiative transfer, we also include the mean effect of gas clumps (the quasar absorption-line clouds [QALCs]) embedded in the smoothly distributed ambient gas. We focus on the presumed transition within the IGM from cold neutral gas to a highly ionized state. We have considered the effect of a metagalactic ionizing radiation background, such as would be contributed by quasars and primeval galaxies, as well as the possibility that hydrodynamical processes deposit thermal energy in the IGM. We describe our numerical method and apply this method here for the purpose of elucidating the minimum requirements for reionizing the IGM by z approximate to 5 to the extent required by the hydrogen Gunn-Peterson (GP) constraint, by the energy release associated with cosmological structure formation. These minimum requirements are a significant constraint on theories of the origin of structure in the universe. For a photoionized universe, we determine the minimum required ionizing photon emissivity of the universe at z > 4. We find that the required emissivity exceeds that due to the observed quasar population unless the IGM density Omega(IGM)h(2) < 4-6 X 10(-3). We consider the consequences for primordial galaxy luminosity density and metal production if early-type stars are the photoionization source instead. Such reionization by stars implies that a substantial metallicity per galactic baryon was generated at high redshift (i.e., z > 3). We also consider scenarios in which the IGM is primarily collisionally ionized as a result of heating to temperatures T> 10(5) K by bulk hydrodynamical heating processes. In this case, the GP constraint may be satisfied with no violation of current COBE Limits on the Compton y-parameter. If the source of this heating is the supernova explosions of early-type stars, then there is a substantial metallicity production at high redshift implied that is similar to that of the photoionized case. If such bulk heating occurs in conjunction with photoionization, however, then the average ionizing photon emissivity from sources at high z required to photoionize the universe enough to satisfy the GP constraint is substantially reduced. Our calculations yield a set of observational diagnostics of the nature and source of the ionization of the IGM and of quasar absorption-line clouds. This includes predictions of the H, He I, and He II Gunn-Peterson Ly alpha absorption troughs from the smoothly distributed IGM, dependent on the nature of the ionizing source and the physical state of the Ly alpha cloud absorbers. We compare our results with the latest observational constraints.