Annu. Rev. Astron. Astrophys. 2000. 38: 667-715 Copyright © 2000 by Annual Reviews. All rights reserved

3. STARS

The nature of the faint stellar component of the Galaxy is critical to determining the stellar luminosity function and the composition of halo dark matter. In spite of the small angular sizes of the HDFs, their depth enables detection of low-luminosity objects to large distances. If the halo dark-matter is mostly composed of low-mass stars, there should be several in the HDF [Kawaler 1996, Kawaler 1998]. Although HST resolution permits star-galaxy separation down to much fainter levels than ground-based observations, many distant galaxies also appear nearly point-like. Color criteria are effective at identifying likely red subdwarfs, but for I₈₁₄ > 25 and V₆₀₆ - I₈₁₄ < 1 the possibility of galaxy contamination becomes significant. The counts of red main-sequence stars have been compared with galactic models by [Flynn et al. 1996], [Elson et al. 1996] and [Mendez et al. 1996]. More detailed results have subsequently emerged from studies that have combined the HDF-N with other HST images to increase the sky coverage [Gould et al. 1996, Gould et al. 1997, Reid et al. 1996, Mendez & Guzman 1997, Kerins 1997, Chabrier & Mera 1997]. A general conclusion is that hydrogen-burning stars with masses less than 0.3 M account for less than 1% of the total mass of the Galactic halo. The overall HST database indicates that the Galactic disk luminosity function experiences a decided downturn for magnitudes fainter than M_V = 12 (M 0.2 M). For the halo, the constraints on the luminosity function are not as good because of the limited sky coverage. However, the luminosity function clearly cannot turn up by the amount it would need to for main-sequence stars to be an important constituent of halo dark matter.

In addition to the nine or so candidate red dwarfs in the HDF-N, to a limiting magnitude of I = 28 the field has about 50 unresolved objects with relatively blue colors. Although in principle these could be young hot white dwarfs (WDs), this possibility has appeared unlikely, as the lack of brighter point sources with similar colors would require all to be at very large distances, greater than 10 kpc away. However, recent work by [Hansen 1998, Hansen 1999] has shown that at low metallicity, molecular hydrogen opacity causes the oldest, lowest luminosity WDs to become blue as they cool. [Harris et al. 1999] have recently discovered such an object in the Luyten proper motion survey. This result changes the way in which colors should be used to discriminate faint point-like sources in the HDFs, and may lead to the reclassification of some faint galaxies as WDs.

Proper motions are an unambiguous way to distinguish between Galactic stars and galaxies. The HDFs serve as excellent first-epoch data for detection of changes in position or brightness for any objects. A second-epoch set of images of the HDF-N were obtained 2 years after the initial HDF-N campaign [Gilliland et al. 1999], and were analyzed by [Ibata et al. 1999] to search for object motion in the images. Of 40 identified point sources with I > 28, five were found to have proper motions that were > 3 above the measurement uncertainty (~ 10 mas year^-1), with two of the objects having proper motions exceeding 25 mas/yr and the remaining three near the detection limit. The five objects are all faint (I ~ 28) and of neutral color (V - I < 0.9), and realistic velocities require that they have distances d < 2 kpc. Although only a very small fraction of the total sources in the HDF-N, these objects represent a large increase over the number of stars expected from standard models of the Galaxy, which predict less than one star in the range 27 < V < 29 with V - I < 1.0.

Separate evidence may be emerging that is consistent with the identification of faint blue stars in the HDFs. The HDF-S (l = 328, b = - 49) points closer to the Galactic center than HDF-N (l = 126, b = 55), and hence samples a larger path length through the Galactic halo. There should be more stars in HDF-S than in HDF-N. [Mendez & Minniti 1999] found roughly double the number of blue point-like sources in the HDF-S than its northern counterpart (Fig. 1). This supports the hypothesis that a significant fraction of these sources are stars. A number of these stars are too faint and too blue to be on the main sequence, and could represent the old, low luminosity, blueish halo WDs proposed by [Hansen 1998].

Figure 1. Color-magnitude diagrams for point sources in the Hubble deep fields (HDF). (Dash-dot line) Locus of disk M-dwarfs at a distance of 8 kpc; (solid star symbols) main-sequence disk and halo stars. (Horizontal line) The (15) detection limit of I = 27 magnitude; (filled circles) galactic stars and unresolved distant galaxies. (Open star symbols) Point sources which are too faint to be on the main sequence at any reasonable distance, and which may be white dwarfs (WD). (Dashed line) The 0.6 M WD cooling track from Hansen (1999) for a distance of 2 kpc. (Adapted from Mendez & Minniti 2000)

Although these developments are exciting, the evidence is far from compelling. The sources identified by [Mendez & Minniti 1999] are brighter than those found to have proper motions by [Ibata et al. 1999]. The two studies thus appear to be inconsistent, in that the closer WDs ought to have larger proper motions. Also, the enhancement in faint blue point-like objects in HDF-S relative to HDF-N is sensitive to the magnitude limit chosen. If [Mendez & Minniti 1999] had included objects down to I = 29 in their sample, they would have found more objects in HDF-N than in HDF-S (Fig. 1).

Although still very tentative, the identification of WDs in the HDFs is potentially extremely important. The results of the MACHO project [Alcock et al. 1997] suggest that a substantial fraction of the halo mass is due to objects with WD masses (although this is not a unique interpretation of the microlensing statistics) ([Sahu 1994]). If WDs do contribute significantly to the halo mass, then the early stellar population of the halo must have formed with a peculiar initial mass function (IMF) [Reid et al. 1996], deficient in both high-mass stars [to avoid over-enrichment of the halo by metals from supernovae (SN) and planetary nebulae] and low mass stars, (because stars with M < 0.8 M would still be on the main sequence today). The presence or absence of WDs in deep fields will become more definite within the next few years via forthcoming third-epoch proper motion measurements.