7.3. Effects of the IGM: Hints of the Gunn-Peterson Effect?

The ability of astronomers to calculate reliable photometric redshifts at z > 3 is largely due to the neutral hydrogen absorption in the intergalactic medium (IGM). Absorption below the redshifted Lyman limit, Ly, and Ly strongly modulates the observed optical spectra of the objects we are interested in. At sufficiently high redshifts, the continuum depression blueward of Ly (the Ly forest) dominates over those decrements associated with higher members of the Lyman series. Oke & Koryanski (1982) define the D_A parameter to measure the strength of the Ly decrement

(3)

(see Fig. 12). The "predicted" denominator in equation (3) is the extrapolated continuum observed at longer wavelengths carried to the 1050-1170 region. The question is: at what redshift does the Ly forest become an impenetrable "jungle"?

Figure 12. Values of the continuum depression blueward of Ly (DA; Oke & Koryanski 1982) plotted as a function of redshift from several observational samples (Dey et al. 1998; Franx et al. 1997; Henry et al. 1994; Hu et al. 1999; Kennefick et al. 1995; Schneider, Schmidt, & Gunn 1989; Schneider, Schmidt, & Gunn 1991a, 1991b; Smith et al. 1994; Songaila et al. 1999; Spinrad et al. 1998; Weymann et al. 1998; Zuo & Lu 1993). The Madau (1995) model (solid line) and several parameterizations from Zhang et al. (1997) (dotted lines) are also plotted (see text). The measurement by Zuo & Lu (1993) may be systematically offset from the other measurements. At z 5 are we perhaps seeing the hints of hydrogen absorption in excess of what is expected from the Ly forest alone? Could this be the first indications of the long-anticipated Gunn-Peterson effect?

This is an intriguing question, as we are now observing galaxies at early enough cosmic epoch that we may perhaps encounter the edge of the "Dark Era" - prior to the complete reionization of the IGM when the universe is still optically thick below the rest wavelength of Ly. Theoretical estimates suggest that the dark era ends between z = 5 and z = 20 (Shapiro 1995; Rees 1995; Miralda-Escude & Rees 1997; Madau et al. 1998). Observationally, we would expect a (Gunn-Peterson) trough of complete absorption for some interval shortward of Ly, corresponding to the epoch when the universe was completely opaque to Ly photons.

How can we measure the Gunn-Peterson effect? For objects of I ~ 26, the obvious problem is one of signal-to-noise ratio. Combining the results of our spectrophotometry with deep HST images, we measure rather large values of D_A for HDF 3-951.0 (z = 5.34; D_A = 0.88; Dey et al. 1998) and HDF 4-473.0 (z = 5.60; D_A = 0.91-0.96; Weymann et al. 1998). Our concern about the break amplitude arises from its strength: the Madau (1995) theoretical estimate of the contribution of the Ly forest to D_A (solid line in Fig. 12) is only 0.79 at z = 5.34 and 0.83 at z = 5.60. This extrapolation assumes a distribution of high and low optical depth foreground Ly clouds causing Lyman series absorption in the spectrum of a distant quasar or galaxy. The dotted lines in Figure 12 represent different models of D_A from Zhang et al. (1997). Using the data of Steidel & Sargent (1987) and Zuo & Lu (1993), they estimate the mean intergalactic Ly opacity as a function of redshift using two parameterizations of : = A(1 + z)^3.46 and = Ae^(1+z). The scatter around the theoretical curves is substantial, even at lower redshifts, so the high values of D_A at z > 5 may simply reflect the usual scatter observed in that parameter. However, the possibility to directly measure the epoch of reionization is not to be overlooked. The best chance for this will be with the discovery of a bright(er) object at z > 5, either in the form of a quasar, or a magnified galaxy behind a rich galaxy cluster.