Early-type stellar systems, such as elliptical galaxies and spiral bulges, are sometimes characterized by a prominent flux shortward of 2000 Å (Fig. 1). The nature of this flux excess (also called UV upturn, UVX phenomenon or UV rising branch) has been a subject of much debate ever since its discovery in the bulge of M31 (Code 1969). Soon after this detection, a variety of hypotheses emerged to explain the FUV rising flux (see the excellent reviews by Burstein et al. 1988, Greggio & Renzini 1990, O'Connell 1999): among them, a non-thermal origin through the galactic nuclear activity, the presence of an unexpected (for longly thought quiescent systems) population of hot young stars (Tinsley 1972), low-mass metal-rich evolved objects (Bressan, Chiosi & Fagotto 1994, Yi et al. 1997) or their metal-poor counterparts (Lee 1994). More recently, it has been analyzed the potential dominant role that subdwarf stars could play in modulating the far-UV properties of the host galaxy, either as members of binary systems (Han, Podsiadlowski & Lunas-Gray 2007, Han, Podsiadlowski & Lunas-Gray 2009) or arising from the evolution of single stellar objects (Napivotski 2008, 2009). A supplemental piece of the puzzle has been provided by the detection of multiple main sequences (MSs) in galactic globular clusters, which reinforces the explanation of UVX phenomenon through presence of helium-rich sub-populations.
Figure 1. Comparison of the mid-UV and far-UV spectral energy distribution of two elliptical galaxies observed with IUE (NGC 1399 and NGC 4649) to illustrate the far-UV excess. The violet and blue curves correspond to the observed spectra, broadened with a Gaussian kernel of 50 Å, and the red flux to a 10 Gyr old theoretical spectra of solar metallicity, calculated with the ingredients described in Chavez et al. (2009).
Over the past four decades evidence has grown in favour of the low-mass star hypotheses. Hills (1971) argued against a non-thermal origin based on the overall shape of the far-UV energy distributions, which more closely resembles that of the Rayleigh-Jeans tail of a thermal source. Additionally, the diffuse distribution of the UV radiation, which can be fitted by a de Vaucouleaurs profile such as the visible light, cannot be explained by the presence of a highly concentrated source, as would be expected if an active nucleus is the origin (Oke, Bertola & Capaccioli 1981, Ohl et al. 1998). In a similar way, the UV imaginery has also worked against the residual star formation hypothesis (which would be implied by the presence of hot MS stars) since, within the detection and resolution limits of several experiments, O and B stars have not yet been detected.
Whilst the currently most accepted picture for the nature of the far-UV excess is that the bulk of UV radiation is dominated by helium burning low mass stars and their progeny, in particular the so-called AGB-manqué, the idea of having on going star formation at very low levels has remained as a still plausible elucidation for the UVX phenomenon (Rich et al. 2005, Rich 2009). This fact might be supported by the prevalence of molecular hydrogen (usually traced by the more accessible signatures of carbon monoxide at millimeter wavelengths) in early-type systems, although in small amounts, indicative of star formation with an efficiency which is in fact similar to that found in spiral aggregates (e.g., Sage, Welch & Young 2007).