The last five years have seen the revival of an old idea with modern technology: determining distances from the ``graininess'' of a galaxy image. The basic idea is simple. Galaxies are made up of stars. The discrete origin of galaxian luminosity is detectable in the pixel-to-pixel intensity fluctuations of the galaxy image. Such fluctuations derive from Poisson statistics of two sorts: (1) photon number fluctuations (N / N), and (2) star number fluctuations (N / N)s. The first is distance-independent, but (N / N)s decreases with distance, as the solid angle subtended by a pixel encompasses more and more individual stars. Consequently, the pixel-to-pixel intensity fluctuations in a nearby galaxy are greater than in a more distant galaxy. If this effect can be calibrated, it can be used as a distance indicator.
Though originally proposed by Baum (1955), it was not until the late 1980s that this idea has been put into practice, made possible by the advent of CCD detectors and telescopes with improved seeing. Tonry and coworkers (Tonry & Schneider 1988; Tonry et al. 1989, 1990; Tonry & Schecter 1990; Tonry 1991; Tonry et al. 1997) have pioneered this technique, which has come to be known as the Surface Brightness Fluctuation (SBF) method. The method can, in principle, be applied to any type of galaxy. In practice, late-type ( Sb) galaxies have too many sources of fluctuations over and above Poisson statistics, such as spiral structure and dust lanes, to apply the method to them. The method is thus preferentially applied to ellipticals and the bulges of early-type spirals.
The distance at which SBF may be applied goes inversely with the seeing. It is possible to measure distances out to ~ 4000 km s-1 with a 2.4-meter telescope, ~ 2-hour exposures, and ~ 0.5 arcsecond seeing. This is an effective limit for current ground-based observations. As half-arcsecond seeing is infrequently achieved at even the best sites (such Mauna Kea and Las Campanas), 3000 km s-1 is a practical limit for complete SBF surveys. In principle, the HST is capable of yielding SBF distances for objects as distant as 10,000 km s-1. However, the required exposure times are such that few galaxies at such distances are likely to be observed for this purpose.
Tonry, Dressler, and coworkers have been conducting an SBF survey of ~ 400 early-type galaxies within ~ 3000 km s-1 over the last six years (Dressler 1994; Tonry et al. 1997). As of this writing, the survey is nearly complete. The data suggest that median SBF distance errors are ~ 8% within this distance range; the most well-observed objects have distance errors of ~ 5%. Such accuracy is considerably better than most of the secondary distance indicators discussed here, with the possible exception of Type Ia supernovae (Section 6).