Annu. Rev. Astron. Astrophys. 2012. 50: 491-529 Copyright © 2012 by Annual Reviews. All rights reserved

3. HALOS OF OTHER SPIRAL GALAXIES

Halo gas observations for other spiral galaxies provide insight into the uniqueness and location of the Milky Way halo gas and the likely origin of the gas. Extragalactic observations are generally not as deep as Milky Way observations in the case of emission line observations, and limited to a single sightline for absorption line observations. For emission line observations, the break between what is disk gas and what is halo gas depends on the spatial and kinematic model of the galaxy adopted. Because the halo gas usually connects to the disk gas in extragalactic observations, the break is often somewhat arbitrary and varies in the literature. For absorption line observations, we are limited to the projected distance and velocity of the absorber as a means to associate the gas with the halo of a galaxy.

This section provides a brief overview of observations of halo gas in other galaxies with the same sub-divisions as Section 2. The exception in this section is that there is no attempt to discuss disk-halo gas separately, since the physical resolution of extragalactic observations is generally poorer. The spiral galaxies noted in this section are just a few examples, and though they are generally Milky Way-like, they do span a range of total mass and SFR. Differences in intrinsic galaxy properties can lead to different expectations for the amount of halo gas (see Section 4).

3.1. Neutral Hydrogen Halo Gas

HI halo clouds have now been detected beyond the disks of numerous spiral galaxies. The lowest HI mass clouds observed around other spirals is ~ 10⁵ M (Thilker et al. 2004), but smaller clouds are likely to be detected with deeper and higher resolution observations. The detected clouds are generally within ~ 10 kpc of the disk, and deep observations do not show evidence for HI clouds at distances > 50-80 kpc from the disks (Westmeier et al. 2007), Pisano et al. 2007). Sancisi et al. (2008) find that extraplanar features with masses 10⁸ M exist in the halos of ~ 25% of the field galaxies. 10⁸ M is a very large HI feature (on the order of the Magellanic Stream), and if the criteria for a HI feature to be distinct is relaxed to include warps and/or kinematic evidence for deviations from disk rotation, the detected fraction increases to ~ 50% (Sancisi et al. 2008, Haynes et al. 1998). Thus far, the majority of the spiral galaxies observed to deep enough levels show a disk-halo component that lags the rotation of the disk (often referred to as an `HI beard'; (Sancisi et al. 2001) with a typical gradient of magnitude 15-30 km s^-1 kpc^-1 (Fraternali et al. 2002, Oosterloo et al. 2007, Heald et al. 2011). These deep observations also often find very limited amounts of distinct halo gas (Irwin et al. 2009, Zschaechner et al. 2011).

Many of the halo HI features that can clearly be disentangled from the disk of the spiral are potentially linked to satellite accretion or interaction with a companion - e.g., M31 (Westmeier et al. 2005a, Ibata et al. 2007), NGC 891 (Oosterloo et al. 2007, Mouhcine et al. 2010), M33 (Putman et al. 2009c, Grossi et al. 2008), NGC 5055 (Battaglia et al. 2006, Martínez-Delgado et al. 2010), NGC 253 (Boomsma et al. 2005, Beck et al. 1982), NGC 2442 (Ryder et al. 2001); and see Hibbard et al. (2001) and Sancisi et al. (2008) for other examples. As can be seen from the bottom panels in Figure 6, this would also be the case for the Milky Way with only the Magellanic System HVCs being distinct if it were observed externally. Future deep stellar observations may link additional HI halo features to satellite accretion; however, this is not possible when the HI feature overlaps with the disk in projection - e.g., NGC 2403 (Fraternali et al. 2002, Fraternali et al. 2001); NGC 4395 (Heald et al. 2007).

Figure 6. Comparisons of HI halo gas for M31 (left) and NGC 891 (right) with that in the Milky Way if viewed with the same inclination. The exterior MW HI view is reconstructed using the distances to the 13 known HVC complexes (white) and assuming a distance for the MS/LA complex (red) ranging from 50 to 150 kpc using the models of Besla et al. (2010). The MW disk (blue), including the disk column density, flare, and warp, is reconstructed from the work of Levine et al. (2006). Contours are set to match the observations: 0.1, 0.3, 1, 3.2, 10, 32, 100, 316, 1000 × 1019 cm-2 for the Westmeier et al. (2007) M31 data and 0.5, 1, 2, 5, 10, 20, 50, 100, 200, 500 × 1019 cm-2 for the Oosterloo et al. (2007) NGC 891 data, respectively. The large cloud above M31's disk is unlikely to be associated with M31 (Westmeier et al. (2007)), and see Ibata et al. (2007) and Mouhcine et al. (2010) for stellar features around M31 and NGC 891, respectively. Viewed externally, the Magellanic system is clearly the dominant feature in our HI halo and our HI disk is comparable to M31 and NGC 891. We find that, without velocity information, the non-Magellanic HVCs would be hard to distinguish from the flared, warped disk.

The high resolution observations of galaxies designed to detect halo clouds often have a HI column density sensitivity of ~ 10¹⁹ cm^-2. This is only the typical peak column density of a Milky Way HVC and therefore may be missing a significant amount of halo gas. The spiral galaxy that has been mapped to the deepest levels (N_HI < 10¹⁸ cm^-2) at a resolution of a few kpc, and is most similar to the Milky Way, is M31 (Thilker et al. 2004, Westmeier et al. 2005a). NGC 891 has also been mapped to 10¹⁸ cm^-2 levels at this resolution, but this galaxy has a much higher SFR than the Milky Way (Oosterloo et al. 2007). Both of these galaxies are shown in the top panels of Figure 6. The majority of their halo gas is connected to the HI disk (like most other galaxies), and/or could be associated with a satellite interaction. The calculated HI mass in M31's halo is the same as that calculated for the Milky Way without the Magellanic System (3-4 × 10⁷ M; see Section 2.1.1), and presumably these clouds also have a substantial ionized component similar to the Milky Way's HVCs (Fittingoff et al. 2009). NGC 891 shows a larger amount of halo and disk-halo gas, which is consistent with its higher star formation rate (Oosterloo et al. 2007). A diffuse halo medium can be inferred to exist from the head-tail nature of several of the M31 HVCs (Westmeier et al. 2005a). The M31 HVCs also have similar linewidths to Galactic clouds.

3.2. Warm Ionized and Warm-Hot Halo Gas

Similar to the Milky Way, the warm gas in the halos of other spiral galaxies is detected in emission and absorption. The emission line observations primarily detect gas close to the disk of the spiral (within ~ 5 kpc; Hoopes et al. 1999, Collins & Rand 2001), and show that the extent of the extraplanar H emission scales with the galaxy's SFR (e.g., Rand 1996, Rossa & Dettmar 2003). Investigations of the motion of the ionized gas find a similar rotation lag to the HI with increasing z-height (Rand 2000, Heald et al. 2006). Heald et al. (2007) find a typical vertical gradient of magnitude 15-25 km s^-1 per scale height. The gas increases in temperature with z-height (Collins & Rand 2001, Rand et al. 2011), as also found for the Milky Way. See Haffner et al. (2009) for a further review of the warm ionized gas of spiral galaxies detected in emission.

There are abundant absorption line results that are presumed to trace the warm gas in galaxy halos based on their observed kinematic and spatial proximity. Since low column density absorbers (N_HI 10¹⁵ cm^-2) are also detected along large scale cosmic filaments, the association with a galaxy halo is not certain (Bowen et al. 1996, Putman et al. 2006, Penton et al. 2002). Within ~ 300 kpc of a galaxy, Ly absorber results show a covering fraction close to 100% and total gas mass estimates are 10^9-10 M (Bowen et al. 2002, Wakker & Savage 2009, Prochaska et al. 2011, Chen & Prochaska 2000). For the higher column density Ly absorbers (and Mg II absorbers that are generally thought to trace gas with N_HI > 10¹⁶ cm^-2), a relationship to the halos of galaxies is relatively clear as the detections are almost solely in the vicinity of galaxies (Bowen et al. 2002, Penton et al. 2002, Chen et al. 2010a). There are a limited number of these high column density absorbers at z = 0 around spiral galaxies. Studies of Mg II absorbers at z ~ 0.1-1 suggest the covering fraction is > 50% within 100 kpc of the galaxy's center (Chen et al. 2010a, Kacprzak et al. 2008); although the covering fraction is expected to be higher at higher redshifts (Fernández et al. 2012).

Studies of the warm-hot component ( ~ 10^5-6 K) are largely limited to O VI absorbers at z = 0 (Prochaska et al. 2011, Wakker & Savage 2009), but will soon be extended to include C IV (Green et al. 2012). Broad Ly absorbers have also been claimed to be warm-hot gas (Bowen et al. 2002, Savage et al. 2011, Danforth et al. 2010). The low redshift O VI absorbers have significant covering fractions of ~ 70-80% within 200-400 kpc of galaxies (Prochaska et al. 2011, Wakker & Savage 2009). This seems to be consistent with the extensive work at higher redshifts that measure covering fractions close to unity for C IV (z ~ 0.5) and O VI (z > 0.1) within 100 kpc of star forming galaxies (Chen et al. 2001, Tumlinson et al. 2011, Chen & Mulchaey 2009). Star forming galaxies are more likely to have both warm and warm-hot gas in their halos (Tumlinson et al. 2011, Chen et al. 2010b). Tumlinson et al. (2011) estimate a mass of circumgalactic gas of 2 × 10⁹ M(assuming solar metallicity) out to 150 kpc from the abundant O VI detections near galaxies at redshifts z = 0.1-0.36. This is similar to the mass derived from the Ly absorber detections in galaxy halos noted above, but approximately an order of magnitude higher than the mass estimates for ionized gas in the Milky Way halo (Section 2.2). This may be due to several reasons: 1) extragalactic absorption line observations may also capture gas in cosmic filaments near the galaxy, but beyond the virial radius, 2) the assumptions used to calculate total gas mass from extragalactic absorption line results lead to an over-estimate, 3) a range of galaxy types are used for extragalactic measurements, or 4) the Milky Way warm and warm-hot halo gas is less massive or has not yet been fully observed.

3.3. Hot Halo Gas

The x-ray detected hot gas in the halos of Milky Way-like spiral galaxies always extends from the galaxy's disk and is within a 10 kpc radius (Li et al. 2008, Li et al. 2006, Wang et al. 2003, Immler et al. 2003, Otte et al. 2003, Wang et al. 2001, Bregman & Houck 1997). The scale heights are calculated to be 1-2 kpc for many spiral galaxies, and the x-ray features are similar to the extraplanar features detected in H emission (Strickland et al. 2004, Bregman & Houck 1997, Tyler et al. 2004). There are numerous limits on the amount of hot halo gas at large radii from non-detections (e.g. Yao et al. 2010, Anderson & Bregman 2010, Rasmussen et al. 2009). As with the Milky Way, most researchers argue that the hot halo medium cannot account for a large percentage of a spiral galaxy's baryons. If galaxies substantially more massive than the Milky Way are considered, the hot halo gas extends to larger radii (e.g., ~ 50 kpc for NGC 1961 with M_tot > 10¹³ M; Anderson & Bregman 2011). Unlike the Milky Way, there is limited indirect evidence for an extended, diffuse, hot halo medium around other spiral galaxies (although see Section 3.2 for the warm-hot gas). One exception is M31, where the head-tail HI clouds out to ~ 30 kpc from the disk are likely to originate from the movement of cold clouds through a diffuse, hot halo medium (Westmeier et al. 2005a).

3.4. Dust, Metals and Molecules

Direct detections of dust are primarily at the disk-halo interface of spiral galaxies (e.g., Keppel et al. 1991). A survey of nearby edge-on galaxies found dense dust (A_V ~ 1) at least 2 kpc above the plane, and the dust is correlated with the presence of extraplanar diffuse ionized gas (DIG) (Howk & Savage 1999). Ménard et al. (2010) have made a convincing detection of dust in extended galaxy halos using the reddening of SDSS background quasars. They found substantial reddening 20 kpc from the galaxies, with detectable reddening all the way out to 10 Mpc (see also Zaritsky 1994). Under an assumption of SMC-like dust, they claim that L_* galaxies have similar amounts of dust in their halos and disks, consistent with observations of reddening in Mg II absorbers at intermediate redshift (Ménard & Fukugita 2012).

Metals are also present in spiral galaxy halos, as directly evident from their detection with absorption lines (Section 3.2). Though ionization conditions are difficult to assess, Tumlinson et al. (2011) estimate there is ~ 10⁷ M in oxygen within 150 kpc of star forming galaxies at z = 0.1-0.36 using O VI measurements. The abundant detections indicate enriched gas has been fed into halos for some time. Sightlines for absorption line experiments that pass through extragalactic HI halo features are rare, but in the cases where star formation has occurred in the gas, metallicities can be measured with the HII regions. Most of this type of halo gas is related to an interaction and the metallicity of the halo HI features is similar to the outer region of the galaxy (Werk et al. 2011). These interacting systems with halo star formation are also the only ones for which molecular gas has been detected or inferred to exist (e.g. Lee et al. 2002).