NEUTRAL HYDROGEN AT HIGH REDSHIFTS AS A PROBE OF STRUCTURE FORMATION .1. POST-COBE ANALYSIS OF CDM AND HDM MODELS

The structures that form in the Universe at redshifts z less than or s imilar to 10 can be detected and studied using the redshifted 21-cm li ne emission from neutral hydrogen. We compute the expected comoving nu mber density, N, of protocondensates that will emit a flux higher than S, at various redshifts, in the CDM and HDM models. The models are no rmalized using COBE results. Our results are compared with the present and expected future sensitivities of various telescopes for the detec tion of protocondensates. In the CDM models the predicted maximum flux es at a redshift z congruent-to 3.3 are about (1.5-3) mJy and N congru ent-to (10(-8)-10(-7)) Mpc-3. These protocondensates cannot be detecte d with present sensitivities, but will become detectable in the near f uture with improved sensitivities. At lower redshifts, the detectabili ty of these structures critically depends on their neutral hydrogen co ntent. In the redshift range around z congruent-to 5, individual proto condensates will not be detectable. The excess variance due to fluctua tions with small density contrasts will, however, be detectable with s omewhat large (say, about 60-h) integration time. At still higher reds hifts, it will be virtually impossible to see any signal, even with su ch a large integration time. Biased CDM models predict larger fluxes, but somewhat lower abundances. Finally, the HDM models - when normaliz ed using COBE results - do not lead to a detectable number of sources ('pancakes') at redshifts z greater than or similar to 2.