Early preheating and galaxy formation

Winds from pregalactic starbursts and 'miniquasars' may pollute the intergalactic medium (IGM) with metals and raise its temperature to a much higher adiabat than expected from photoionization, and so inhibit the formation of early galaxies by increasing the cosmological Jeans mass. We compute the thermal history of the IGM when it experiences a period of rapid, homogeneous 'preheating' at high redshifts, and the impact of such a global feedback mechanism on the IGM ionization state and the subsequent galaxy formation and evolution. Measurements of the temperature of the Lyα forest at redshift z ∼ 3 constrain the redshift and energy of preheating, and rule out models that preheat too late or to too high a temperature, i.e. to TIGM gsim; 106 K at z ≳ 10. The IGM thermal history is used to estimate the effects of preheating on the formation of galaxies at later epochs, allowing us to predict galaxy luminosity functions in preheated universes. The results depend crucially on whether the baryonic smoothing scale in the IGM is computed globally, or in a local, density-dependent fashion (since the IGM temperature can become highly inhomogeneous in the post-preheating epoch). Using a globally averaged smoothing scale, we find that models with excessive preheating produce too few L* and fainter galaxies, and are therefore inconsistent with observational data. More moderate preheating scenarios, with TIGM ≳ 105 K at z ∼ 10, are able to flatten the faint-end slope of the luminosity function, producing excellent agreement with observations, without the need for any local feedback mechanism within galaxies. A density-dependent smoothing scale requires more energetic preheating to achieve the same degree of suppression in the faint-end slope. All models, however, appear unable to explain the sharp cut-off in the luminosity function at bright magnitudes - a problem that is also common to more conventional local feedback prescriptions. Supernova-driven preheating scenarios tend to raise the mean metallicity of the universe well above the minimum levels observed in the Lyα clouds. The high energies associated with preheating cause a sharp drop in the abundance of neutral hydrogen in the IGM and are often sufficient to double ionize helium at high redshift, well before the 'quasar epoch'. We find that ionizing photon escape fractions must be significantly higher than 10 percent in order to explain the low inferred H I fraction at z ≈ 6, particularly when using a globally averaged smoothing scale. While early preheating causes strong suppression of dwarf galaxy formation, we show that it is not able to reproduce the observed abundance of satellite galaxies in the Local Group in detail. The detailed thermal history of the universe during the formative early stages around z = 10-15 remains one of the crucial missing links in galaxy-formation and evolution studies.