Discovery and analysis of three faint dwarf galaxies and a globular cluster in the outer halo of the Andromeda galaxy

We present the discovery of three faint dwarf galaxies and a globular cluster in the halo of the Andromeda galaxy (M31), found in our MegaCam survey that spans the southern quadrant of M31, from a projected distance of similar to 50 to similar to 150 kpc. Though the survey covers 57 deg(2), the four satellites lie within 2 degrees of one another. From the tip of the red giant branch (RGB), we estimate that the globular cluster lies at a distance of 631 58 kpc from the Milky Way and along with a similar to 100 kpc projected distance from M31 we derive a total distance of 175 55 kpc from its host, making it the farthest M31 globular cluster known. It also shows the typical characteristics of a bright globular cluster, with a half-light radius of 2.3 +/- 0.2 pc and an absolute magnitude in the V band of M-V,M-o = - 8.5 +/- 0.3. Isochrone fitting reveals that it is dominated by a very old population with a metallicity of [Fe/H] similar to -1.3. The three dwarf galaxies are revealed as overdensities of stars that are aligned along the RGB tracks in their colour-magnitude diagrams. These satellites are all very faint, with absolute magnitudes in the range -7.3 less than or similar to M-v.0 less than or similar to -6.4, and show strikingly similar characteristics with metallicities of [Fe/H] similar to - 1.4 and half-light radii of similar to 120 +/- 45 pc, making these dwarf galaxies two to three times smaller than the smallest previously known satellites of M31. Given their faintness, their distance is difficult to constrain, but we estimate them to be between 740 and 955 kpc which places them well within the virial radius of the host galaxy. The panoramic view of the MegaCam survey can provide an unbiased view of the satellite distribution of the Andromeda galaxy and, extrapolating from its coverage of the halo, we estimate that up to 45 20 satellites brighter than M-V similar to -6.5 should be orbiting M31. Hence faint dwarf galaxies cannot alone account for the missing satellites that are predicted by A cold dark matter models, unless they reside in dark matter minihaloes that are more massive than the typical masses of 10(7) M-. currently inferred from their central radial velocity dispersion.