The Local Group of galaxies is dominated by two massive spirals, M31 and the Milky Way. The Local Group (LG) is our immediate cosmic neighborhood, a sparse galaxy group with a zero-velocity radius of 1 Mpc (Karachentsev et al. 2002a). Two dominant galaxies also characterize a number of other nearby galaxy groups, and zero-velocity radii of 1 to 1.2 Mpc are typical for these poor groups. Thirty-eight galaxies are currently known to be located within the LG's radius, but the galaxy census of the LG is still incomplete. While all of its luminous members have been detected, new faint member candidates with very low surface brightnesses are still being discovered (e.g., Zucker et al. 2004a). In spite of its small size and its small number of galaxies the LG plays a prominent role in astrophysical research. For reviews of the LG, see Grebel (1997, 1999, 2000, 2001), Mateo (1998), and van den Bergh (1999, 2000). In the present contribution, I will mainly concentrate on the dwarf galaxies of the LG.
The proximity of the galaxies of the LG permits us to study them at the highest possible resolution, enabling us to resolve these galaxies into individual stars down to very faint magnitudes. For example, with the Hubble Space Telescope, point sources down to an apparent magnitude of 30 have been detected in M31 (Brown et al. 2003), corresponding to the magnitudes of stars well below the Population II main-sequence turn-off at that distance. To date, the LG is the only location where we can observe the lowest stellar masses, measure the oldest stellar ages, determine metallicities and element abundance ratios of individual stars of high ages, and measure detailed stellar kinematics. In other words, the LG is unique in affording the highest level of detail and accuracy for stellar population studies. Much of this research has only become possible in recent years thanks to the superior angular resolution and sensitivity of the Hubble Space Telescope (HST), and thanks to the advent of powerful ground-based optical/near infrared 8 to 10m class telescopes, which provide sufficient sensitivity for medium and high-resolution stellar spectroscopy of individual stars in nearby galaxies. It is now becoming possible to uncover the evolutionary histories of these galaxies in unprecedented detail.
Furthermore, the LG contains a variety of different galaxy types. While massive early-type galaxies are absent, the LG still provides us with galaxies covering five orders of magnitude in galaxy masses, a variety of different morphological types, very different star formation histories, and a range of different dominant ages and metallicities. Within the small volume of the LG, galaxies occur in environments ranging from immediate proximity to massive galaxies to fairly isolated locations without nearby neighbors. Hence we can expect to eventually obtain first-hand measurements of the impact of environment on galaxy evolution, at least on small scales.
All in all, by studying the stellar populations of galaxies in the LG and by combining them with detailed knowledge of the other components of these galaxies such as gas, dust, and non-luminous matter we can learn about galaxy properties at the highest possible level of detail. The LG permits us to study stellar evolution at a range of different metallicities and ages not afforded in that combination in the Milky Way; e.g., the evolution of young, massive stars at low metallicities. Moreover, the LG permits us to carry out tests of cosmological galaxy evolution theories, for instance with respect to time scales and number and magnitude of accretion events. Hence here we can conduct "near-field cosmology" (see Freeman & Bland-Hawthorn 2002). The LG is particularly well-suited to learn more about the formation and evolution of disk galaxies, whose theoretical foundations are still poorly understood. Understanding all of these processes in an environment where we can learn about them in detail is the precondition for understanding distant, unresolved galaxies.
1.1. The galaxy content of the Local Group
We currently know of 38 galaxies within the zero-velocity surface of the Local Group. Whether the outermost galaxies are indeed gravitationally bound to the LG remains unclear. Indeed, reliable orbits are not yet known for any of the LG galaxies, and proper motion measurements exist only for the closest Milky Way companions. The uncertainties of these measurements are still large, but this situation should change once data from the planned astrometric satellite missions - Gaia (an ESA mission that will scan repeatedly the entire sky) and the Space Interferometry Mission (SIM, a NASA mission that will perform pointed observations with even higher accuracy) - become available. These missions will probably not start before 2011.
As mentioned before, the LG galaxy census is still incomplete. Apart from missing orbital information, galaxies of very faint surface brightness are still being detected as sky surveys become more sensitive or as data mining techniques are being improved. During the past six years, this led to the discovery of one fairly isolated new dwarf spheroidal (dSph) galaxy in the Local Group (Cetus; Whiting, Irwin, & Hau 1999), and to the discovery of four new dSph companions of M31 (Armandroff, Davies, & Jacoby 1998; Armandroff, Jacoby, & Davies 1999; Karachentsev & Karachentseva 1999, Grebel & Guhathakurta 1999, Zucker et al. 2004a, Harbeck et al. 2005). The conclusion of data mining existing photographic all-sky surveys (Karachentsev et al. 2000; Whiting, Hau, & Irwin 2002) makes the existence of additional faint, Draco- or Ursa Minor-like dSphs seem unlikely unless they are hidden behind regions of high Galactic extinction, particularly in the zone of avoidance. On the other hand, the discovery of And IX (Zucker et al. 2004a) with an unusually low surface brightness of only µV,0 ~ 26.8 mag arcsec-2 indicates that additional very sparse, very faint dwarfs may yet have to be discovered. At present, though, there is little reason to believe that there may still be hundreds of yet to be detected faint objects present in the LG.
Very faint, sparse, low-surface-brightness objects like And IX, which appear to be dominated by ancient, metal-poor stars (Harbeck et al. 2005) are of immense interest from the point of view of galaxy evolution. Do they show signatures of cosmic re-ionization squelching? Do they contain the purest, most metal-poor, most ancient populations known? Or do they show evidence for population gradients in their old population as were found in a number of other ancient dSphs (Harbeck et al. 2001)? Do these objects show large metallicity spreads as measured in other seemingly purely old dSphs (Shetrone, Côte, & Sargent 2001), which in turn indicate extended episodes of star formation (Ikuta & Arimoto 2002)? How did these objects survive? Are they entirely dominated by dark matter? Are they regular dSph galaxies that simply are fainter and sparser than the previously known members of this galaxy class? Yet from a cosmological point of view, objects of this type may appear to be of little interest since they may not significantly contribute toward solving the "missing satellite problem" (Klypin et al. 1999; Moore et al. 1999) unless detected in vast numbers.
Not only low surface brightness, scarcity of stars, high foreground extinction, and unknown orbits for distant LG member candidates contribute to an incomplete LG galaxy census, but also accretion events. We still know little about the actual number of accretion events, about the times when these events occurred, and about the nature of the accreted galaxies (see also the Section 1.4 and the contribution of Majewski in these proceedings). The number of 38 probable LG members comprises all galaxies listed in Grebel, Gallagher, & Harbeck (2003) including the Sagittarius dwarf galaxy, which is currently being accreted by the Milky Way. In addition, it includes the giant stream around M31 (Ibata et al. 2001), which is very likely the remnant of a disrupted galaxy, and the newly discovered M31 satellite And IX (Zucker et al. 2004a). Not included are other possible streams and accretion remnants such as the Monoceros feature (Newberg et al. 2002) and Canis Major (Martin et al. 2004a) since their nature as dwarf galaxy remnants remains disputed (see Section 1.4). Depending on the nature of these and other stellar overdensities and kinematically identified features in the Milky Way and in M31, the LG census may need to be increased by at least up to six or more galaxies if one wishes to account for still measurable accretion events. If these features are confirmed as minor mergers, then it becomes a matter of definition whether their progenitors should be included in the current LG member census.
1.2. Galaxy types in the Local Group
Approximately 90% of the LG's luminosity is contributed by its three spiral galaxies, but the vast majority of the LG galaxy population are satellite and dwarf galaxies. The boundaries between larger galaxies and dwarf galaxies are poorly defined. We adopt here the usual convention of calling every galaxy with an absolute magnitude of MV > -18 mag a dwarf galaxy. This simple magnitude criterion then leaves us with 34 dwarf galaxies (though admittedly the absolute magnitude of the progenitor of M31's "great stream" is yet unknown, and that of the Sagittarius dSph is poorly constrained). Apart from the three spirals, only the Large Magellanic Cloud is more luminous than the above luminosity threshold. Yet M32 is often not considered a dwarf galaxy despite its lower luminosity since it exhibits similar structural properties as giant ellipticals.
Most of the LG galaxies belong to one of four basic classes: Spirals, dwarf irregular galaxies, dwarf elliptical galaxies, and dSphs. In the following paragraphs the main characteristics of the dwarf galaxy classes are summarized following Grebel (2001):
Dwarf irregular galaxies (dIrrs) are gas-rich galaxies with an irregular optical appearance usually dominated by scattered H II regions. They typically have V-band surface brightnesses of µV 23 mag arcsec-2, H I masses of MHI 109 M, and total masses of Mtot 1010 M. The stellar populations of dIrrs range from ancient stars with ages > 10 Gyr to ongoing star formation. Star formation appears to have proceeded largely continuously, although amplitude variations in the intensity or rate of star formation may reach a factor of three. The H I distribution is usually clumpy and more extended than even the oldest stellar populations. When considering only the distribution of older stellar populations (i.e., of red giants and red clump stars, which trace populations older than ~ 1 Gyr), irregulars and dIrrs exhibit a highly regular distribution shaped like an (elongated) disk (LMC; van der Marel 2001) or like a spheroid (example: Small Magellanic Cloud (SMC); Zaritsky et al. 2000). In low-mass dIrrs gas and stars may exhibit distinct spatial distributions and different kinematic properties. Metallicities tend to increase with decreasing age in the more massive dIrrs, indicative of enrichment and of an age-metallicity relation (see also Fig. 3 in Grebel 2004). For low-mass dIrrs detailed measurements are still lacking. Solid body rotation is common among the more massive dIrrs, whereas low-mass dIrrs seem to be dominated by random motions without evidence of rotation. DIrrs are found in galaxy clusters, groups and in the field.
Dwarf elliptical galaxies (dEs) are spherical or elliptical in appearance, tend to be found near massive galaxies (in the Local Group they are all companions of M31), usually have little or no detectable gas, and tend not to be rotationally supported. Note that examples of rotating dEs beyond the Local Group have been found (e.g., Pedraz et al. 2002). DEs are compact galaxies with high central stellar densities. They are typically fainter than MV = -17 mag, have µV 21 mag arcsec-2, MHI 108 M, and Mtot 109 M. DEs may contain conspicuous nuclei (nucleated dEs or dE,N) that may contribute up to 20% of the galaxy's light. The fraction of dE,N is higher among the more luminous dEs. An example for a non-nucleated dE in the Local Group is NGC 185, whereas NGC 205 is a dE,N. Sérsic's (1968) generalization of a de Vaucouleurs r1/4 law and exponential profiles describe the surface density profiles of nucleated and non-nucleated dEs and dSphs best (Jerjen et al. 2000). DEs typically contain old and intermediate-age populations (i.e., populations older than 10 Gyr and populations in an age range of ~ 2 to 10 Gyr), but the fractions of these populations vary, and even present-day star formation may be observed.
Dwarf spheroidal galaxies (dSphs) are diffuse, gas-deficient, low-surface-brightness dwarfs with very little central concentration. They are not always distinguished from dEs in the literature. DSphs are characterized by MV -14 mag, µV 22 mag arcsec-2, MHI 105 M, and Mtot ~ 107 M. They include the optically faintest galaxies known. Their stellar populations tend to be either almost purely old or a mix of old and intermediate-age populations. The luminosity functions (and by inference the mass functions) of dSphs have been found to be "normal" and in excellent agreement with those of Galactic globular clusters (Grillmair et al. 1998; Wyse et al. 2002). DSphs are usually found in close proximity of massive galaxies (counterexamples in the LG: Cetus, Tucana) and are generally not supported by rotation. Since most published measurements concentrated on the central regions of dSphs, the possibility of slowly rotating dSphs cannot be excluded. The central velocity dispersions of dSphs indicate the presence of a significant dark component when virial equilibrium is assumed. However, not all dSphs are in virial equilibrium (e.g., Ursa Minor shows indications of being tidally distorted by the Milky Way; for instance, Palma et al. 2003). Interestingly, the radial velocity dispersion profiles of dSphs show a marked drop at large radii (Wilkinson et al. 2004). Its interpretation and the apparent existence of a kinematically cold stellar population at the outermost radii is not yet understood. The metallicity-luminosity relations of dSphs and dIrrs show the usual trend of increasing metallicity with increasing galaxy luminosity, but the relations are offset from each other: DSphs have higher mean stellar metallicities at a given optical luminosity (Grebel et al. 2003 and Section 3.1).
1.3. Morphological segregation
As already indicated, the distribution of the different types of galaxies within the LG is not random, but is determined by the location of the two most massive spiral galaxies in the LG, M31 and the Milky Way. A sketch of the three-dimensional distribution of galaxies within the LG is shown in Grebel (1999; in Fig. 3). In Fig. 1 the distribution of gas-poor, low-mass dwarfs and of gas-rich, higher-mass dwarfs is shown. The gas-poor dwarfs - primarily dSphs - are strongly clustered around the closest massive spiral, whereas the gas-rich dwarfs (dIrrs) exhibit less of a tendency toward close concentration around massive galaxies and are more widely distributed. The differing distribution of different morphological galaxy types is also known as morphological segregation and is observed in nearby galaxy groups and galaxy clusters as well.
Figure 1. Morphological segregation in the number distribution N of different types of galaxies in the Local Group (solid histograms) and in the M81 and Centaurus A groups (hashed histograms) as a function of distance D to the closest massive primary (updated version of Fig. 1 in Grebel 2004).
Similarly, the H I mass of dwarf galaxies tends to increase with increasing distance to a massive primary (Grebel et al. 2003; their Fig. 3). These trends indicate that environment may have a significant impact on the evolution of low-mass galaxies. Indeed, it is tempting to speculate that these trends are the result of morphological transformations due to the influence of massive galaxies, e.g., via tidal or ram pressure stripping (see Mayer et al. 2001 for simulations). One vital ingredient to verify this hypothesis is once again information about the orbits of the companion galaxies.
1.4. Direct evidence for harassment and accretion
If accretion is the primary mechanism for the growth and evolution of massive galaxies as predicted by cosmological models, then we should be able to find evidence for these processes in our immediate neighborhood. (1) The study of the structural properties of nearby galaxies can reveal whether external tidal forces are distorting them. (2) The detection of extratidal stars and streams around and within massive galaxies is evidence for ongoing harassment and accretion events. (3) The stellar content, population properties, and chemistry of nearby galaxies allow us to constrain to what extent these kinds of objects could have contributed as building blocks to more massive galaxies.
The clearest evidence for ongoing accretion are the extended tidal stream of the Sagittarius dSph galaxy (Ibata, Gilmore, & Irwin 1994), which has now been traced around the entire Milky Way using M giants identified in the Two Micron All-Sky Survey (2MASS) (Majewski et al. 2003), and the giant stream of metal-rich giants in the halo of M31 (Ibata et al. 2001; Ferguson et al. 2002). Additional stellar overdensities have been detected in the Milky Way using various photometric data sets including 2MASS and the Sloan Digital Sky Survey (SDSS): The Monoceros feature (Newberg et al. 2002; Yanny et al. 2003), which may be a tidal tail connected with the Canis Major overdensity (Martin et al. 2004a). The interpretation of Canis Major is disputed; suggestions include that it is part of the Galactic warp or flare (Momany et al. 2004) or indeed the center of another possibly disrupted dSph within the Milky Way (Martin et al. 2004a, 2004b). Additional Galactic stellar overdensities have been identified (for instance, Triangulum-Andromeda; Rocha-Pinto et al. 2004, which may be part of the tidal tail of a more distant disrupted dwarf). Other suggestions of evidence of dwarf galaxy accretion are based on the identification of moving groups and radial velocity surveys (e.g., Gilmore, Wyse, & Norris 2002). However, not only disrupted dwarf galaxies, but even globular clusters in advanced stages of accretion may produce extended stellar tidal tails (Odenkirchen et al. 2001; 2003). These tidal features are valuable also as tracers of the Galactic potential. The most luminous and most massive Galactic globular cluster, Centauri, contains a range of different ages and a large metallicity spread. A popular explanation for its unusual properties is that Cen may be the core of an accreted dwarf galaxy (see van Leeuwen, Hughes, & Piotto 2002 for details). - Possible additional tidal features in and near M31 have been reported by Morrison et al. (2003) and Zucker et al. (2004b).
The Magellanic Clouds and the Milky Way are interacting with each other as evidenced by, e.g., the gaseous Magellanic Stream and the Magellanic Bridge (e.g., Brüns et al. 2005), although the interpretation of these features as being primarily due to tidal (e.g., Putman et al. 1998) or ram pressure effects (Mastropietro et al. 2005) remains controversial. The twisted isophotes of the M31 companions M32 and NGC 205 may be caused by tidal interaction with M31 (Choi, Guhathakurta, & Johnston 2002). The nearby Galactic dSph satellite Ursa Minor shows a distorted, S-shaped surface density profile possibly caused by tidal interaction with the Milky Way (Palma et al. 2003). On the other hand, the drop in the velocity dispersion profiles of Draco and Ursa Minor at large radii (Wilkinson et al. 2004) and the lack of a large depth extent of Draco (Klessen, Grebel, & Harbeck 2003) would seem to argue against ongoing tidal disruption.