2.1. Resolved stellar populations
Even in an era of Extremely Large Telescopes (ELTs), we will only be able to resolve individual stars and stellar populations out to Virgo (Local Volume), but this may be enough to ultimately unravel the processes involved in galaxy formation, particularly when supported by observations of the high-redshift universe. Some of the most revealing insights have come from the Local Group with its two dominant rajahs and numerous courtiers. The Galaxy and M31 have comparable mass, and show evidence of substructure in most stellar components (e.g. Juric et al 2005). Both have extended disks and stellar haloes, although the former has a small bulge, while M31 has a prominent bulge.
Some of the most remarkable observations have come from the Hubble Space Telescope (HST) which targetted the spheroid, outer disk and the major tidal stream in M31 (Brown et al 2006). All fields of view were found to have an extended star formation history, with most stars being in the age range of 4 to 10 Gyr. The inner spheroid is intermediate age (6−9 Gyr) and metal rich and slightly older than the Stream, and may well comprise mostly stream material. This is in stark contrast to the mostly old Galactic spheroid. The outer disk in M31 is younger (4−8 Gyr) but still older than the thin disk population in the Solar neighbourhood.
Resolved stellar population studies are now achieving much fainter effective surface brightness levels than are possible with observations of diffuse stellar light. The disks in M31 and M33 have been detected at µV = 30 mag arcsec−2 and are seen to extend to more then 10 scale lengths. The metal-poor stellar halo in M31 is now thought to extend to at least 160 kpc (Guhathakurta et al 2005), i.e. a halo subtending 20∘ across the sky! The halo of M33 has now been detected and found to be comparable in metallicity to M31 and the Galaxy (McConnachie et al 2006), which is interesting in light of the fact that it has no detectable bulge. At the present time, stellar haloes have not been detected in either of the Magellanic Clouds (e.g. Gallart et al 2004).
2.2. Thick disks & outer disks
It is well established that the thick disk is 10−12 Gyr old and therefore provides us with an ancient snapshot of what took place in the early universe (q.v. Reddy et al 2006). A recent development is that thick disks may be relatively common in disk galaxies (Yoachim & Dalcanton 2005).
The character of the Galactic thick disk is now known to be distinct from the thin disk in essentially all measurable parameters. The dynamically hot population is characteristically older and more metal poor than the thin disk, with an unexpectedly strong rotational lag. There may be evidence that the thick disk has a longer scale length (Robin et al 1996) and a distinct abundance gradient (Brewer & Carney 2006).
Beyond the Local Group, the closest galactic disk is NGC 300 at a distance of 1.95 ± 0.05 Mpc. This galaxy is a virtual twin of M33 but showing a different behaviour in the outer parts. The red stellar disk is exponential over 10 scale lengths with no evidence for a truncation of any kind (Bland-Hawthorn et al 2005). In contrast, M33 shows dramatic truncation close to the Holmberg radius with no evidence of tidal streams (see below), quite unlike its dominant neighbour.
Pohlen & Trujillo (2006) have obtained SDSS photometry on 90 galaxies to find that 60% truncate in the outer parts, 10% remain exponential to the limits of the data, and 30% appear to show flattening in the outermost parts. While these data do not reach the same effective surface brightness as the resolved stellar surveys, it is clear that outer disks are telling us something important about the build-up of disk material over billions of years. The picture is not as tidy as we used to imagine (e.g. Fall & Efstathiou 1980).
2.3. Abundance patterns and gradients
Stellar abundances are discussed by S. Feltzing; the discussion here is limited to a few key points. The Hipparcos survey has allowed a very clean kinematic separation of the thick disk from the thin disk (Bensby et al 2005; Reddy et al 2006). This has revealed a very clear distinction between the chemistry of the thick and thin disk, particularly in the α elements.
By now, we are used to the idea of nebular abundance gradients in spiral galaxies determined from HII regions. In contrast, stellar populations can be aged and these appear to show an overall abundance gradient that is flattening with time. For example, open clusters appear to show that the Galactic abundance gradient was -0.1 dex/kpc 8 Gyr ago, softening to -0.04 dex/kpc at the present day (Daflon & Cunha 2004; Salaris et al 2004). This may be consistent with new evidence of young metal-poor cepheid variables in the outer disk (12−17 kpc) with enhanced [α/Fe] and [Eu/Fe] (Yong et al 2006), suggestive of recent accretion. An interesting new development is spatial abundance maps for a single population (e.g. Luck et al 2006).
Elemental abundances have long been argued as a key constraint on the accretion history of satellites onto the Galaxy (Unavane, Wyse & Gilmore 1995). A new development is the concept of chemically tagging stars to a parent population from the element abundance patterns (Freeman & Bland-Hawthorn 2002; Bland-Hawthorn & Freeman 2004). The basis for tagging is that stellar clusters are highly uniform in certain chemical elements (e.g. De Silva et al 2006). has now been demonstrated for the moving group HR 1614 originally identified by Eggen. What is remarkable here is that HR 1614 covers most of the sky which has led others to question the integrity of Eggen's group. In an era of Gaia, chemical tagging will be enormously powerful in that it will provide an independent confirmation of the integrity of a stellar group, or allow us to disentangle cells in phase space.
Simulations show that dwarf galaxies spiral into larger ones, where they are torn apart to produce the star streams observed in the big galaxies. But the patterns of heavy elements from UVES observations at the VLT (DART: Tolstoy et al 2003) indicate that no major component of the Galaxy could have been assembled largely by accretion of dwarfs of the kind observed today. M31 and the Galaxy could have formed by merging of dwarfs in the early universe; the curious thing is that the dwarfs that were left behind to be observed as dwarfs today have to be substantially different (Robertson et al 2005).
This topic is discussed by A. Helmi so we limit our discussion here. One of the major developments since the mid 1990s is the discovery of the infalling Sgr dwarf (Ibata et al 1994) from radial velocity observations at the AAT, using stellar identifications towards the Galactic bulge identified in UK Schmidt plates. This was the first detection of a disrupting galaxy within the orbit radius of the Magellanic Stream which led to a resurgence of interest in Galactic studies. Since then, other streams have been identified by the SDSS and 2MASS surveys although these may be associated with either the Sgr stream (Newberg et al 2003) or the outer disk (Ibata et al 2003).
We have become used to hearing about substructure out of the plane, but the Hipparcos survey reveals kinematic substructure even within the thin disk (Dehnen 1998; Fux 1997). The ages of individual clumps are quite distinct (Famaey et al 2005) such that they are likely to be dynamical in origin rather than due to patchy star formation.
The overlap of the Hipparcos survey with the Geneva-Copenhagen survey allowed Navarro et al (2004) to identify a possible moving group associated with Arcturus. They suggest an infalling group ∼ 8 Gyr ago. This shows the potential for astrometric surveys used in combination with wide-field kinematic surveys.
Interestingly, the SDSS has identified complex substructure in the thick disk (Juric et al 2005). Tomographic slices through the galaxy were obtained using photometric distances. Gilmore et al (2002) observed 2000 F/G stars with the 2dF at the AAT that were chosen to extend to high Galactic latitude. They found that the rotation of the thick disk is half the expected value of 180 km/s. A possible interpretation is that the thick disk arose from a satellite merger 10−12 Gyr ago.
Looking further afield, wide-field observations from the INT reveal complex system of streams in the outer disk of M31 (Ibata et al 2005), something that is not seen in M33 (Ferguson et al 2006). That these are discrete subcomponents has now been confirmed kinematically using LRIS on the Keck telescope (Chapman et al 2006).