While M31 certainly has both a bulge and a halo, the distinction between these components is somewhat muddled in the literature. Historically, studies of the bulge and halo were driven by the appearance of M31 in wide-field shallow imaging (e.g. Walterbos & Kennicutt 1988), with "bulge&" studies generally focusing on the region within a kpc of the center, and "halo" studies focusing on regions beyond 10 kpc on the minor axis. However, subsequent studies showed that the spheroid looks like a bulge out to ~ 20 kpc in both its surface-brightness profile (Pritchet & van den Bergh 1994) and metallicity distribution (Mould & Kristian 1986, Durrell et al. 1994, Durrell et al. 2001). It was only recently discovered that the surface-brightness profile transitions to a power-law (Guhathakurta et al. 2005, Irwin et al. 2005) and the metallicity drops by ~ 1 dex (Kalirai et al. 2006) beyond 20 kpc, as expected for a halo. For consistency with historical studies, I will refer to all regions beyond 10 kpc on the minor axis as "halo" despite the persistence of some bulge-like properties out to 20 kpc.
4.1. M31 Bulge
Two fields in the M31 bulge were recently imaged in the near-IR at high resolution using adaptive optics on Gemini North (Davidge et al. 2005), resolving the upper 4-5 mag of the RGB and AGB. The resulting H and K CMDs are consistent with a population dominated by stars at an age of ~ 10 Gyr, with a metallicity near solar, although the best-fit models include a minority intermediate-age component that may be spurious (Olsen et al. 2006). As with the center of M33, no observatory in operation or development is capable of obtaining the detailed star formation history of the M31 bulge, because crowding precludes photometry of the low-mass MS stars.
4.2. M31 Halo
Until the past decade, studies of the resolved stellar populations in the M31 halo generally focused on the metallicity distribution (e.g. Mould & Kristian 1986, Durrell et al. 1994, Durrell et al. 2001). Metallicity distributions were usually fit by assuming an old age (> 10 Gyr) and then comparing the stars on the AGB and RGB to isochrones or globular cluster templates, although some studies were deep enough to reach the HB (e.g. Holland et al. 1996). Although the RGB, AGB, and HB distributions are, in principle, sensitive to age in broad age bins (as noted above), several factors prevented these studies from exploring the age distribution, including photometric scatter, insufficient star counts, and contamination from foreground Milky Way dwarfs.
When the Advanced Camera for Surveys (ACS) was installed on HST in 2002, it became possible to obtain photometry of low-mass MS stars in the M31 halo, enabling the exploration of both the metallicity and age distributions. Brown et al. (2003) imaged a field 11 kpc from the nucleus on the southeast minor axis, obtaining photometry of 250,000 stars down to V ~ 30.5 mag in bands similar to V and I (Figure 2). By providing a large number of stars with small photometric errors on the low-mass MS, the catalog was immune to contamination from foreground Milky Way dwarfs. The resulting fit to the CMD found a wide age range in addition to the wide metallicity range that was already known, and speculated that this was due to a significant merger or series of smaller mergers in the galaxy. In the best-fit model (Brown et al. 2006), ~ 40% of the stars are younger than 10 Gyr and more metal-rich than 47 Tuc ([Fe/H] = -0.7), with significant numbers of stars down to ages of 2 Gyr. Besides the field population, there is evidence from integrated colors and spectroscopy that the M31 globular cluster system also extends to intermediate ages (Puzia et al. 2005, Beasley et al. 2005).
Figure 2. Radial velocities (top row; dotted line is M31 systemic velocity), CMDs (middle row; curve shows 47 Tuc ridge line for comparison), and star formation histories (bottom row; area of circles proportional to weight in fit) for four regions in the M31 halo (Brown et al. 2003, Brown et al. 2006, Brown et al. 2007, Brown et al. 2008). The three regions on the minor axis (at 11, 21, and 35 kpc) show a kinematically hot population at the M31 systemic velocity, while the off-axis field shows the kinematically cold stream plunging into M31 toward us from behind the galaxy. Despite their distinct kinematic profiles, the CMDs and associated star formation histories for the 11 kpc field and the stream are very similar, due to the inner halo being polluted by debris from the stream's progenitor (Brown et al. 2006, Fardal et al. 2007, Gilbert et al. 2007). Although the fields further out on the minor axis do not include significant numbers of stars as young and metal-rich as those found in the 11 kpc and stream fields, all of the halo fields exhibit an extended star formation history, consistent with expectations from hierarchical merging.
The recent wide-field imaging surveys of M31 clearly show that it has undergone a violent merger history (Ferguson et al. 2002, Ibata et al. 2007), including a giant stellar stream resulting from the tidal debris of a recent merger event (Ibata et al. 2001). Subsequent studies demonstrated that the inner spheroid of M31 (within ~ 15 kpc) is polluted by material stripped from progenitor satellite of the giant stellar stream. This evidence includes N-body simulations of the satellite disruption that reproduce the morphology of the major substructures in the galaxy (Fardal et al. 2007), kinematical surveys that confirm the motions in the N-body simulations (Gilbert et al. 2007), and followup ACS imaging of the giant stellar stream that shows strong similarities between the star formation histories of the stream and inner spheroid (Figure 2; Brown et al. 2006).
Given the discovery that the M31 halo becomes more like a halo beyond 20 kpc, a subsequent ACS survey explored regions on the minor axis further out, at 21 kpc (in the transition region; Brown et al. 2007) and 35 kpc (where the spheroid clearly exhibits a halo surface brightness profile and metallicity; Brown et al. 2008). Compared to the field at 11 kpc, these fields host far fewer stars at ages younger than 8 Gyr, but the populations clearly do not represent a classical halo formed via monolithic collapse at early times (Figure 2); in the best-fit model, ~ 30% of the stars are younger than 10 Gyr, and only ~ 10% of the stars are ancient ( 12 Gyr) and metal-poor ([Fe/H] -1.5). All regions of the halo explored to date are consistent with a history whereby the galaxy forms over a prolonged period of hierarchical merging.