Annu. Rev. Astron. Astrophys. 1997. 35: 389-443
Copyright © 1997 by . All rights reserved

Next Contents Previous

4.2. The Role of Low Surface Brightness Galaxies

Although intimately connected with uncertainties in the local LF, the presence (or otherwise) of an abundant population of low surface brightness galaxies (LSBGs) is best considered as a separate issue affecting the interpretation of faint data. (McGaugh 1994, 1996) has shown how the presence of systems with central surface brightnesses fainter than µB = 23 arcsec-2 covering a wide range of luminosities would seriously affect determinations of the LF particularly if the local data were plagued by isophotal losses, as discussed in Section 4.1. McGaugh (1994), Phillipps & Driver (1995), Ferguson & McGaugh (1995), Babul & Ferguson (1996) have explored this uncertainty and proposed the existence of an abundant population of local LSBGs that could be faded remants of blue star-forming systems identified fainter than B > 22. The idea stems from Babul & Rees's (1992) suggestion that the excess seen in the counts might arise from a separate population of dwarfs whose initial star-formation era is delayed until the UV ionizing background drops below a critical value. Rapid fading would produce a large present-day abundance of red LSBGs. The role of LSBGs in these suggestions is thus twofold; first, as an additional uncertainty in the local LF and second, as remnants of the faint blue galaxies.

Quantifying the contribution of LSBGs to the local LF will remain controversial until a suitable catalog exists for which rigorous selection criteria have been applied and redshifts determined. Most of the available field data is angular-diameter limited from photographic plates (Impey et al 1988, 1996, Schombert et al1992, Sprayberry et al 1996) and, although illustrating the range of scale lengths and surface brightnesses possible, cannot easily be converted into volume-limited data. McGaugh (1996) convincingly argues that the presence of only a small number of LSBG examples implies a significant correction must be made to the faint end slope, although several assumptions are made in calculating the survey volume as a function of limiting surface brightness.

Although CCDs have been used for LSBG searches in clusters, Dalcanton et al (1996) recently analyzed 17.4 deg2 of deep transit scan CCD data and identified seven LSBGs with µv > 23 arcsec-2. Of importance, spectroscopic data for this sample have provided an estimate of the volume density of LSBGs of known physical size. The large mean distance for these LSBGs indicates a LSBG volume density, though still uncertain, that is comparable to that of normal galaxies. As expected, the contribution to the overall luminosity density is very small. Dalcanton et al (1997) argued, as did McGaugh (1996), that part of the normalization change in the local LF (discussed in Section 4.1) may arise from the selective loss of these systems in the bright photographic data.

Although LSBGs undoubtedly exist, perhaps quite abundantly and with a range of properties, it seems unlikely that they dominate the luminosity density or that the bulk of them represent the faded remants of a faint blue population. The fading required to push a typical faint blue galaxy below typical local detection thresholds would have to be 2.5-4 mag, depending on the true faint end slope of the local LF (Phillipps & Driver 1995). To be effective, given the narrow time interval involved, the end of the star-forming era would also have to be very abrupt. Moreover, most well-studied LSBGs are gas-rich and blue (de Blok et al 1996) and quite unlike the postulated faded remnants. Neither is there an obvious correlation between LSBG central surface brightness and color (McGaugh & Bothun 1994). An abundant population of faded remants would be detectable as a significant Euclidean upturn in the number of red and infrared sources at faint limits, which has not yet been seen (Babul & Ferguson 1996).

Next Contents Previous