The previous sections focus on the accretion of gas onto the Galactic disk. Up until recently the Milky Way was the only galaxy with a known halo cloud population, with the exception of some detections of large halo features similar to the Magellanic Stream (~ 108 M) in nearby galaxies (e.g., Ryder et al. 2001; Hibbard et al. 2001). New receivers and larger allocations of telescope time have allowed deep HI surveys of nearby spiral galaxies to be completed, and these surveys are beginning to indicate cold halo clouds are a common property of spiral galaxies. The Local Group spirals are discussed here, but see also the deep HI observations of other nearby spirals where additional halo clouds are observed (e.g. Fraternali et al. 2001; Oosterloo et al. 2007; Heald et al., this proceedings; Battaglia et al 2006; Boomsma et al. 2005).
M31 is our closest spiral of similar size to the Milky Way, and was found to have numerous halo clouds by Thilker et al. (2004). The population of clouds detected is within 50 kpc of M31 and has a total HI mass of a few × 107 M. Westmeier et al. (2007) later showed there are no additional clouds at larger radii. Several of the M31 halo clouds are in the vicinity of the giant stellar stream (to the south of M31) and NGC 205, while others appear to be completely isolated. The clouds in the vicinity of satellites may be remnants of their gaseous components, but they may also represent clouds formed in instabilities in the hot halo medium created by the passage of the satellite. The isolated M31 halo clouds are most likely condensed clouds, although they could also represent the remnants of clouds stripped from satellites at earlier times. The velocities of the clouds relative to M31 and the known satellites may cast some light on the most likely origin (e.g. see figure 3 of Peek et al. 2008). The total mass in M31's HI halo clouds is within a factor of three of the Galaxy's population if you adopt the distance constraint of < 60 kpc for the Galactic halo clouds and exclude the Magellanic Stream (Putman 2006).
M33 also has clouds in its halo. Figure 5 shows the preliminary GALFA channel maps of M33. Two main halo features can be noted in this figure, though additional clouds at lower contour levels are evident in the data cube. The first feature is most predominant at , d = 1h31m, 31.5° and -240 to -260 km s-1, but continues beyond this velocity range and spatially wraps around to join the northern side of M33's warped disk. The second halo feature is a small cloud with a filament to the main galaxy at , d = 1h32.5m, 29.5° which is most evident at -148 km s-1 (see also Westmeier et al. 2005). The first halo cloud feature potentially has an interesting link to what may be a stellar overdensity in the maps of Ibata et al. (2007); but given the limited coverage of their current survey, it is unclear if this is actually a distinct stellar feature. The potential continuation of this northern HI emission towards M31 (Braun & Thilker 2004) appears to be due to contamination by a cloud that merges with Galactic emission in the GALFA data. The GALFA data extends up to d = -34.5°, and with its sensitivity (3σ ~ 4 × 1018 cm-2 in the 11 km s-1 channels shown in Figure 5) and resolution (3.5' and 0.2 km s-1) reveals the detailed kinematic and spatial structure of numerous low velocity clouds that continue into the Galactic disk. This link to Galactic emission would have been partially obscured in the Braun & Thilker data with the smoothing and exclusion of examining Galactic emission.
Figure 5. Channel maps of the HI in M33 with GALFA data. The velocity is labeled in the upper left. Contours are 0.2, 0.4, 0.8, 1.6, 3.2 K. The lowest velocity channel shows some of the Galactic emission that overlaps with the emission from M33.
The finding of halo clouds around M33 is somewhat surprising given M33 is only ~ 1/10 the mass of the Milky Way (5 × 1010 M; Corbelli 2003). Models do not generally expect a galaxy of this size to have a large amount of mass locked up in its halo at z = 0 (Maller & Bullock 2004). The clouds in the halo of M33 may represent the cold accretion mode that is expected to dominate for galaxies of this mass (Keres et al. 2005), or the structures may represent the gradual destruction of M33 by M31 which is now ~ 200 kpc away. The recent proper motion measurements of M33 by Brunthaler et al. (2005) have been used by several authors to put constraints on the motion of M31 and its impact on M33. Loeb et al. (2005) rule out a large region in parameter space because of the lack of disruption evident in the stellar component of M33. van der Marel & Guhathakurta (2007) use the motion of M33, along with the motions of other Local Group satellites, to calculate the most likely motion of M31. The most likely transverse velocity for M31 has M33 on a tightly bound orbit and signatures of disruption are therefore likely. Muratov and Gnedin at the University of Michigan are completing more detailed orbit calculations and finding similar results. The HI features discussed here are most likely the beginning of this disruption process, and the disruption is currently only evident in the extended gaseous component, similar to the Magellanic Clouds. It is possible M33 will eventually form giant HI features akin to those found in the Magellanic System as it approaches perigalacticon. Both of these systems have a similar total HI mass (1-2 × 109 M) and will provide a substantial amount of fuel to the galaxies.