AN OPTICAL STUDY OF THE FAINT END OF THE STELLAR LUMINOSITY FUNCTION

We implement a new method by which to study the faint end of the field star luminosity function. The method relies on deep, multicolor photometry of fields projected against highly obscured, nearby molecular clouds. The clouds act as nearly opaque screens and delimit a well-defined survey volume which is in principle free of the problem of distinguishing nearby, intrinsically faint dwarf stars from more distant red giants. This study is based upon deep photographic and CCD photometry at optical (V, R, I) bandpasses toward the most highly obscured portions of the Taurus and Ophiuchus molecular clouds. The total volume delimited by the clouds is approximately 200 pc3. Within this region our survey is complete for all stars brighter than M(V) = 16-17 mag; at R and I, the survey is complete down to the lowest mass stars capable of sustaining core hydrogen burning, Color-color criteria are used to distinguish between background, highly reddened stars and stars located in front of the clouds, and the method of photometric parallax is used to deduce the absolute magnitudes and spectral types of those stars found to lie in front of the clouds. To help further resolve foreground/background ambiguities, we have made proper motion measurements of the candidate foreground stars using the photographic and CCD data, and the Palomar Sky Survey. We estimate the faint end of the field star luminosity function for the composite Taurus and Ophiuchus foreground sample and find that it resembles the local luminosity function down to M(V) approximately 16. At still fainter magnitudes we find more stars than do photometric parallax studies of the polar regions. This difference widens dramatically if even the simplest correction for incompleteness is applied to our data. We therefore tentatively conclude that the luminosity function rises beyond M(V) approximately 16; even if we discard our attempts to correct for incompleteness in the faintest magnitude bins, the luminosity function at least remains flat for the lowest mass stars. Our provisional finding that the luminosity function rises again beyond its well-known peak at M(V) approximately 12-13, implies that the IMF probably rises beyond the turnover point associated with this peak. Even if our most conservative estimate for the faint end of the luminosity function is used-in which no corrections are made for incompleteness-the IMF must at least remain flat down to the edge of the hydrogen-burning main sequence.