Astron. Astrophys. 327, 890-900 (1997)

4. Summary and conclusions

We have shown that the IGM in the line of sight of HE 2347-4342 is "patchy" in the sense that there are "voids" and "troughs". In the voids, the He II opacity can be understood as Ly forest line opacity with either = 100 and and no diffuse medium or Ly forest line opacity with = 45 and plus a diffuse medium with = 0.3. The He II opacity seen in the voids is consistent with what has been found in HS 1700+6416 (Davidsen et al. 1996) or in the Q 0302-003 void by Hogan et al. (1997). In the "troughs" we need in addition to the He II Ly forest opacity a continuous opacity of 4.8 ( is a strict lower limit). A natural explanation for this dichotomy is the assumption of delayed He II ionization, as predicted by e.g. Madau & Meiksin (1994), where in the phase of reionization of the universe there are still not yet ionized He II regions between the expanding He III regions. In that case h would be necessary. The absorption troughs have the shape and widths as predicted by Meiksin & Madau (1993).

The patchy He II opacity appears to be the "missing link" between the at in HS 1700+6416 and the high opacity in Q 0302-003 at : The "voids" in HE 2347-4342 have optical depths comparable to the mean opacity observed in HS 1700+6416 and represent the reionized IGM, while the "troughs" represent the IGM with delayed He II reionization which may also be the case for Q 0302-003 and PKS 1935-692 at . It is important to repeat the He II opacity measurement in case of Q 0302-003 with STIS because the background subtraction procedure applied to the GHRS data appears doubtful at such low flux levels.

Our He II observations are in accordance with Songaila and Cowie's (1996) finding from the observed C IV /Si IV ratio as a function of redshift that around there appears to be a transition from a soft ionizing UV background () to relatively hard ionizing radiation (). If confirmed by further observations, the delayed He II ionization gives important constraints on the evolution of the UV background for .

The same ionization model of Madau & Meiksin (1994) which predicts incomplete He II ionization for predicts incomplete H I ionization for . After our observation of delayed He II ionization there appears to be a realistic chance to observe directly the phase of reionization of the universe if QSOs exist since the latter might show H I blacked-out troughs similar to what we have seen in He II. One has to keep in mind, however, that HE 2347-4342 is just one line of sight, and that before generalizations are possible, more objects have to be probed.

The absence of a measurable proximity effect in HE 2347-4342 might be explained by shielding of  Å radiation by the clouds responsible for the strong associated system seen toward this quasar.

Observations of the He II troughs in HE 2347-4342 with STIS in low resolution can be expected to yield an improved lower limit to the diffuse density in the yet unionized medium, and HE 2347-4342 is even bright enough in the 1160 Å void for FUSE to resolve directly the He II 304 Å forest.

© European Southern Observatory (ESO) 1997

Online publication: April 6, 1998
helpdesk.link@springer.de