A specific search for galaxies large in Hi has been executed by Broeils and van Woerden (1994) and Broeils and Rhee (1997), using short observations with the WSRT. Broeils and Rhee reported an interesting correlation between the Hi mass and the Hi size, the latter defined as an isophotal radius at the level of 1 M⊙ pc−2. More recently, a number of surveys have been executed on various telescopes, typically with sample sizes of order 10 − 60 galaxies. Wang et al. (2016) collected Hi sizes as defined by Broeils and Rhee (1997) for 437 galaxies — spread over 14 projects — although not for every galaxy an Hi size was determined, and there might be a slight overlap in the sense that several galaxies are in more than one sample. They found again a very tight relationship between the Hi mass and the Hi size. It is instructive to examine for each survey they considered the distribution of the ratio of Hi to optical size, which is shown in Fig. 17. The survey done by Broeils and Rhee (1997) shows a peak at RHI / Ropt ∼ 1.5, and they found relatively few galaxies with extended Hi disks. The statistics based on the data by Wang et al. (2016) show that the proportion of extended Hi disks is not too different for spirals compared to irregulars (even though statistically there are more irregulars with large RHI / Ropt), and the fraction of very extended disks with an Hi size larger than, e.g., three times the optical size is about 10%.
 |
Figure 17. Statistics of the ratio of Hi diameter, evaluated at ΣHI = 1 M⊙ pc−2, to the optical diameter, following Wang et al. (2016), except for the Bluedisk sample, for which Wang et al. (2013) data are used. The samples are: BR-1997 − Broeils and Rhee 1997; WHISP(S) − Swaters et al. 2002; LVHIS − Koribalski 2008; Westmeier et al. 2011, 2013; Bluedisk − Wang et al. 2013; Diskmass − Martinsson et al. 2016; THINGS − Walter et al. 2008; Ursa Major − Verheijen and Sancisi 2001; VIVA − Chung et al. 2009; WHISP(ET) − Noordermeer et al. 2005; Atlas3D − Serra et al. 2012, 2014; LITTLE THINGS − Hunter et al. 2012; K09 − Kovač et al. 2009; L14 − Lelli et al. 2014; FIGGS − Begum et al. 2008; PVB − Ponomareva et al. 2016. Note that several large galaxies are missing from the LVHIS, THINGS, LITTLE THINGS, FIGGS and VIVA samples, while for the WHISP samples a maximum diameter of 400′′ is imposed, on account of missing flux when comparing the interferometric data with single-dish observations or too large in extent compared to the primary beam. For the SPARC sample (Lelli et al. 2016). I determined the 3.6 µm radius from the tables associated with their publication, by interpolating those radial luminosity profiles which reached a depth of 1.0 L⊙pc−2, which could be done for 84 of the 174 galaxies in that sample. The bottom row refers to the total number of spirals (left) and irregulars (numerical Hubble type ≥ 8.5, right) in the samples studied by Wang et al. (2016). |
Even though Broeils and Rhee (1997) found a very good relation between the Hi mass and the size of the Hi disk, indicating a roughly constant mean Hi surface density, there is still no clear indication about when we can expect an Hi disk which is much larger than the optical disk. This is due to the fact that the amount of Hi in the extended Hi disk is usually only a modest fraction of the total Hi mass, so that differences are drowned in the “bigger things are bigger in many quantities” effect.
Not all surveys peak at the same Hi-to-optical size ratio. This is to be expected for the Virgo cluster survey, VIVA (Chung et al. 2009), since in the centre of that cluster the Hi gas in galaxy outskirts is subject to ram-pressure stripping against the hot intergalactic medium (IGM) probed by the X-ray emission. Indeed, several galaxies have been reported undergoing stripping, such as NGC 4522 (Kenney et al. 2004), NGC 4388, where a long Hi tail has been observed (Oosterloo and van Gorkom 2005), NGC 4254 and its Hi tail, including an Hi dwarf almost devoid of stars (Minchin et al. 2007), and NGC 4569, where Boselli et al. (2016) report very extended Hα emission.
Studies have shown that Hi sizes of galaxies in a group environment could also be affected by the presence of the IGM. Obviously this is the case for compact groups (Verdes-Montenegro et al. 2001), where an evolutionary scenario is suggested depending on the state of the merging.
For loose groups, the effects are rather more subtle. In a recent extensive study, Wolfinger et al. (2016) analysed data for the Ursa Major region, and constructed a complete sample of 1209 galaxies limited in systemic velocity (300 ≤ VLG ≤ 3000 km/s), absolute magnitude (Mr ≤ −15.3) and stellar mass (M* ≥ 108 M⊙). They identified six groups with more than three members, centred around NGC's 4449, 4258, 4278, 4026, 3938 and 5033, while 74% of the galaxies in their sample reside outside those groups. The high resolution interferometer data on the “Ursa Major cluster” (Verheijen and Sancisi 2001) and the CVn region (Kovač et al. 2009) are part of this sample, but for other parts of the region single-dish data from Jodrell Bank and Arecibo were used. There are Hi-deficient galaxies associated with the densest parts of some of the groups, but not all, as, e.g., NGC 4449 (shown in Fig. 15), which does not seem Hi-deficient. There are also galaxies with excess Hi mass, usually in regions of low galactic density.
For the Sculptor group, Westmeier et al. (2011) find that the deep Hi image of NGC 300 is asymmetric in the very outer parts, and attribute this to a possible interaction with the IGM in that group. The process of the formation of warps due to the interaction with the IGM has been explored further by Haan and Braun (2014). Deep Hi imaging data on M83 with the KAT-7 telescope (1092016Heald et al. 2016) shows sharp edges in the outer Hi there as well, which can be either due to a ram pressure effect with respect to the surrounding IGM, or photoionisation, as discussed below.
The Void Galaxy Survey (Kreckel et al. 2011, 2012) investigated the properties of Hi disks in galaxies in voids, and found that on average the statistics of the size ratio RHI / Ropt are roughly similar to those observed by Swaters et al. (2002) for the late-type WHISP sample. The galaxy with the largest ratio, VGS_12, has an Hi disk in the polar direction of a small S0-like galaxy (cf. Stanonik et al. 2009). No star formation is associated with this Hi gas. However, apart from this, there is nothing outstanding about the properties of void galaxies with respect to the field galaxies studied in most other surveys.
The outermost Hi is prone to be ionized by the ultraviolet background radiation, as already suggested by Sunyaev (1969). An early attempt to get limits on this has been done for NGC 3198 by Maloney (1993), using an unpublished deep VLA Hi image. The realisation that there ought to be detectable ionized gas emission around Hi envelopes, which could be used to probe the dark matter at larger radii, has led to some further studies of this gas using Fabry-Pérot Hα observations, but these are technically very challenging (cf. Bland-Hawthorn et al. 1997).
Hlavacek-Larrondo et al. (2011a, b) report deep Hα observations of three Sculptor group galaxies with the wide-field Fabry-Pérot system on the 36 cm Marseille Telescope in La Silla. For NGC 247, the Hα and Hi data extend out to ∼13.5 arcmin, barely beyond the Holmberg radius of 12.2 arcmin. For NGC 300, the field of view was limited, and the region of the outer, warped Hi disk not imaged in Hα. For NGC 253, the Hα disk goes out to similar radii (11.5 arcmin) as the Hi disk seen with the VLA, but faint [Nii] emission has been detected out to 19.0 arcmin at the southwest part of the galaxy. The kinematic data seem to indicate a declining rotation curve, but the galaxy is heavily perturbed there as seen on recent deep optical images, and recent KAT-7 Hi data show the presence of extraplanar cold gas (Lucero et al. 2015). Earlier Hα observations of the Sculptor group galaxy NGC 7793 with the same instrument (Dicaire et al. 2008) led to the conclusion that the ionizing sources of the Hα disk of this galaxy are likely to be internal, rather than the UV background.
For more on the interface between the cold Hi gas and the hotter circumgalactic medium, see the review by Chen (this volume).