A Keck/DEIMOS spectroscopic survey of the faint M 31 satellites And XV and And XVI

We present the results of a spectroscopic survey of the recently discovered M31 satellites And XV and And XVI, lying at projected distances from the centre of M31 of 93 and 130 kpc, respectively. These satellites lie to the south of M31, in regions of the stellar halo which wide-field imaging has revealed as relative voids (compared to the ~degree-scale coherent stream-like structures). Using the Deep Imaging Multi-Object Spectrograph mounted on the Keck II telescope, we have defined probable members of these satellites, for which we derive radial velocities as precise as ~6 km s^(−1) down to i ~ 21.5. While the distance to And XVI remains the same as previously reported (525 ± 50 kpc), we have demonstrated that the brightest three stars previously used to define the tip of the red giant branch in And XV are in fact Galactic, and And XV is actually likely to be much more distant at 770 ± 70 kpc (compared to the previous 630 kpc), increasing the luminosity from M_V ≈ −9.4 to −9.8. The And XV velocity dispersion is resolved with v_r = −339^(+7)_(−6) km s^(−1) and σ_v = 11^(+7)_(−5) km s^(−1). The And XVI dispersion is not quite resolved at 1σ with v_r = −385^(+5)_(−6) km s^(−1) and σ = 0^(+10)_(−indef) km s^(−1) . Using the photometry of the confirmed member stars, we find metallicities of And XV (median [Fe/H] = −1.58 , interquartile range ±0.08) and And XVI (median [Fe/H] = −2.23, interquartile range ±0.12). Stacking the spectra of the member stars, we find spectroscopic [Fe/H] = −1.8 (−2.1) for And XV (And XVI), with an uncertainty of ~0.2 dex in both cases. Our measurements of And XV reasonably resolve its mass (~10^8 M_⊙) and suggest a polar orbit, while the velocity of And XVI suggests it is approaching the M31 escape velocity given its large M31 centric distance.