5. GAS: TRANSPARENCY OF THE LYMAN CONTINUUM

Star-forming galaxies are the dominant contributor to the non-ionizing UV radiation field in the universe. Are they a significant component of the ionizing background as well? Simple arguments might suggest otherwise. An H I column density of ~ 1 × 10¹⁸ cm^-2 is sufficient to absorb essentially all the ionizing radiation. Since the measured extinctions imply column densities that are three or four orders of magnitude higher than this, it might appear that essentially no ionizing radiation can escape. However, the porosity of the ISM seen in the non-ionizing continuum ( > 912 Å) could very well extend below the Lyman edge and may dominate the shape of the emergent spectrum. The situation is sufficiently complex that the only way to determine the escape fraction f_esc of the ionizing radiation is via a direct measurement.

Attempts to measure f_esc fall into two categories: observations of local galaxies with a far-UV detector, or measurements using galaxies at cosmological redshift, which are accessible from the ground with 8-m class telescopes. Either technique has its advantages and disadvantages. The "local" approach faces the obvious challenge of extreme UV observations, whereas the "distant" measurement is affected by the radiative transfer in the intergalactic medium. In addition, a somewhat less direct method is to determine the Lyman continuum opacity from a comparison of the H and UV luminosity functions in the local universe. Table 1 gives a summary of recent results. Except for the last entry in this table, all quoted studies find more or less stringent upper limits on f_esc both in the low- and high-redshift universe. The ISM in the observed galaxies is highly opaque, and very little stellar ionizing radiation leaks out.

Table 1. Determinations of the Lyman escape fraction in galaxies.

Reference	Instrument	Objects	fesc
Leitherer et al. (1995)	HUT	4 galaxies; z 0	< 3%
Hurwitz et al. (1997)	HUT	4 galaxies; z 0	< 19%
Deharveng et al. (1997)	H / UV	local luminosity function	< 1%
Deharveng et al. (2001)	FUSE	Mrk 54 z 0	< 5%
Giallongo et al. (2002)	FORS2	z = 2.96, 3.32	< 16%
Fernández-Soto et al. (2003)	WFPC2	HDF; 1.9 < z < 3.5	< 4%
Malkan et al. (2003)	STIS	1.1 < z < 1.4	< 1%
Steidel et al. (2001)	Keck/LRIS	29 galaxies; z 3.4	~ 100%

Steidel et al. (2001) detected significant Lyman-continuum flux in the composite spectrum of 29 Lyman-break galaxies with redshifts z = 3.40 ± 0.09. If the inferred escaping Lyman-continuum radiation is typical of galaxies at z 3, then these galaxies produce about 5 times more H-ionizing photons per unit comoving volume than quasars at this redshift, with the obvious cosmological implications. Haehnelt et al. (2001) fitted the composite spectrum by a standard stellar population and no intrinsic H I opacity. Therefore, f_esc must be close to 100% for the observed sample. Confirmation or rejection of this striking result will be a major objective of observational cosmology.