and Cosmology

9. The Universe at High Redshift
386
Fig. 9.30. Left: a sketch of the geometry of reionization is
shown: initially, relatively low-mass halos collapse, a first
generation of stars ionizes and heats the gas in these halos.
By heating, the temperature increases so strongly (to about
T ∼ 10 4 K) that gas can escape from the potential wells; these
halos may never again form stars efficiently. Only when more
massive halos have collapsed will continuous star formation
set in. Ionizing photons from this first generation of hot stars
produce HII regions around their halos, which is the onset of
reionization. The regions in which hydrogen is ionized will
grow until they start to overlap; at that time, the flux of ionizing
photons will strongly increase. Right: the average spectrum of
photons at the beginning of the reionization epoch is shown;
here, it has been assumed that the flux from the radiation
source follows a power law (dashed curve). Photons with an
energy higher than that of the Lyα transition are strongly suppressed
because they are efficiently absorbed. The spectrum
near the Lyman limit shows features which are produced by
the combination of breaks corresponding to the various Lyman
lines, and the redshifting of the photons
been detected in the HUDF. The result is that these two
volumes are quite comparable. Hence, it seems that this
galaxy was capable of reionizing “its” volume of the
Universe. Again we should warn that this is a preliminary
conclusion, based on a single object; nevertheless,
it indicates that we might be seeing direct evidence for
early reionization, in accordance with the results from
WMAP.
Prospects for Observing Reionization Directly. We
note that only a small fraction of the baryons needs
to burn in hot stars to ionize all hydrogen, as we can
easily estimate: by fusing four H-nuclei (protons) to
He, an energy of about 7 MeV per nucleon is released.
However, only 13.6 eV per hydrogen atom is required
for ionization.
Furthermore, we point out again that the very dense
Lyα forest seen towards QSOs at high redshift is
no unambiguous sign for approaching the redshift of
reionization, because a very small fraction of neutral
atoms (about 1%) is already sufficient to produce
a large optical depth for Lyα photons. Direct observation
of reionization will probably be quite difficult; an
illustration of this is sketched in Fig. 9.31.
We have confined our discussion to the ionization
of hydrogen, and disregarded helium. The ionization
energy of helium is higher than that of hydrogen, so
that its ionization will be completed later. From the
statistical analysis of the Lyα forest and from the analysis
of helium absorption lines in high-redshift QSOs,
a reionization redshift of z ∼ 3.2 for helium is obtained.
With the upcoming Next Generation Space Telescope
(which has more recently been named the James Webb