Annu. Rev. Astron. Astrophys. 1999. 37: 603-648 Copyright © 1999 by Annual Reviews. All rights reserved

7.4. Cosmic Evolution of the UVX

Since the models predict that the UVX (if dominated by the EHB or PAGB channels) should decrease as the main-sequence turnoff mass increases, there should be strong evolution of the UVX with lookback time (see Section 7.2). If E galaxies are sufficiently homogeneous, there could be a unique lookback beyond which the UVX disappears. Given the uncertainty in the models discussed above, lookback effects should probably be viewed for the present more as a valuable opportunity to refine the models than as a way to age-date the universe.

There are serious technical challenges in making restframe UV measurements at moderate redshifts. The galaxies are faint. At a redshift of 0.5 (a lookback time of 6 Gyr), the distance modulus is 43, implying that the unevolving UVX of a strong upturn source in a typical luminous elliptical would have m(2250 Å) ~ 24.5. Simple detection of far-UV light (e.g. in broad bands) is not sufficient to distinguish a UVX component from the decaying initial burst or late star formation (see Section 5). Multiband photometry or spectroscopy is necessary. Several attempts to observe the UVX at high redshift have been made (e.g. Windhorst et al 1994), but only recently has a detection been claimed in the cluster Abell 370 (z = 0.38) by Brown et al (1998b). Using broad-band filters with HST/FOC, they find four cluster E galaxies to have a range of 1500-V similar to that in local galaxies. If this is UVX light, it implies a high formation redshift (z_F > 4) in the context of most existing models. An absence of UV evolution over the past few Gyr would be inconsistent with some classes of UVX models. It will be especially important to link changes in the UVX of distant galaxies with evolution of the initial burst at optical/IR wavelengths (now detected up to z ~ 0.9, e.g. Stanford et al 1998).