Because EHB stars are the source of the UV upturn, the UV upturn is expected to fade dramatically with increasing redshift as one looks to younger elliptical galaxies (Greggio & Renzini, 1990; Tantalo et al., 1996). We have been undertaking a series of observations with HST to map the evolution of the UV upturn as a function of redshift, by observing galaxy clusters at 0.3 < z < 0.6 (figure 4). Although early measurements were consistent with a relatively flat evolution out to z 0.4 and a rapid fading at higher redshifts (Brown et al., 1998a; Brown et al., 2000a), recent observations (Brown et al., 2003) at z = 0.33 show a UV upturn as weak as that at z = 0.55. Because the earliest observations, at z = 0.375, were the only ones that did not use a solar-blind camera, they might be systematically in error; setting aside the z = 0.375 measurements, the remaining observations show a UV upturn that is weaker than that in the present epoch, but a relatively flat evolution with increasing lookback time. Taking the models at face value implies large variations in the formation epoch of giant elliptical galaxies in clusters, which is implausible (Brown et al., 2003). Although the onset of the UV upturn occurs at ~ 6 Gyr in these models, the formation of EHB stars is tied to a wide range of poorly constrained parameters (mass loss, metallicity, binary fraction, etc.), with the elliptical galaxies presenting the aggregate behavior at any one epoch. The "floor" in the UV emission seen at increasing redshift might indicate a wide dispersion in the parameters that govern EHB formation, or it might be that another source of UV emission is becoming dominant at increasing redshift as the EHB stars disappear (e.g., residual star formation). With our current understanding of EHB formation, the UV upturn remains a poorly calibrated age indicator. However, if galaxy ages are determined by independent methods, these surveys of the UV upturn could instead be used to constrain theories of EHB formation.
Figure 4. The evolution of the UV upturn as a function of redshift (triangles), with the expected evolution of this emission in giant elliptical galaxies (Tantalo et al., 1996), assuming three different formation redshifts (zf; labeled).