H I Emission and Absorption in the Southern Galactic Plane
Survey¹
N. M. McClure-Griffiths ,
John M. Dickey ,
B. M. Gaensler ,
A. J. Green ,
R. F. Haynes ,
M. H. Wieringa
, PASA, 18 (1), in press.

Next Section: H I Self-Absorption (HISA)
Title/Abstract Page: H I Emission and Absorption
Previous Section: Introduction
Contents Page: Volume 18, Number 1

H I Shells

We here present a small gallery of H I shells detected in the Test Region and the Parkes data. These shells represent a selection of the broad range of structures displayed by shells. They range in diameter from about 30 pc to 600 pc. H I shells provide an interesting environment in which to study the ISM because they are among the largest structures visible in the Galaxy. In addition, they are among the few discrete objects which are apparently deterministic in nature. They result from relatively well understood phenomena such as a supernova explosion for the smallest shells, or the combined effects of stellar winds and many supernovae for the largest shells. As the fossils of extremely energetic events they allow us to study how energy is injected into the ISM, and subsequently how that energy affects the ISM as it undergoes the transition from deterministic to turbulent in nature.

Local Shell

One of the smallest shells visible in the Test Region is extremely local. Figure 1 shows a greyscale of a large angular diameter shell centered at about

$l=330\mbox{$.\!\!^\circ$}5$ ,

$b=2\mbox{$.\!\!^\circ$}2$ ,

$v_{\sc{lsr}} \approx 2.1$ km ${\rm s}^{-1}$ . Because of its local velocity, distance estimates are extremely uncertain. However, we used a standard rotation curve for the Galaxy (Fich, Blitz & Stark 1989), to estimate the kinematic distance to this shell to be in the range 350 pc to 1 kpc. At these distances the diameter is between 15 and 50 pc. There is also tentative evidence for slow expansion on the order of 4 km ${\rm s}^{-1}$ , though no front or back caps are detected. It should be noted, though, that random H I cloud motions are on the order of 6 km ${\rm s}^{-1}$ , making it difficult to place much confidence in the expansion velocity estimate. The small size of this shell suggests that it may be an old supernova remnant (SNR) which no longer radiates in the continuum. It is interesting to note, however, that the shell is extremely circular, far more so than a typical SNR.

**Figure 1:** Grey-scale image of an H I channel map at v=2.12 km ${\rm s}^{-1}$ showing an apparent H I shell approximately
$2\mbox{$.\!\!^\circ$}5$ in diameter in the local ISM at

$l=330\mbox{$.\!\!^\circ$}5$ ,

$b=2\mbox{$.\!\!^\circ$}2$ . The grey-scale is logarithmic to emphasize the shell walls. The brightness temperature scale in Kelvins is displayed in the wedge on the right.
$\begin{figure} \begin{center} \psfig{file=local.ps,width=7cm,angle=-90} \end{center} \end{figure}$

Terminal Velocity Shell

There is another small shell at the terminal velocity of v=-110 km ${\rm s}^{-1}$ . This shell, as shown in Figure 2, is centered on

$l=329\mbox{$.\!\!^\circ$}9$ ,

$b=0\mbox{$.\!\!^\circ$}4$ and is at a kinematic distance of $\sim 7.4$ kpc. This distance translates into a physical diameter of $\sim 51$ pc. The shell stands out as the only feature in the H I channel maps near the terminal velocity that persists over many velocity channels. It is detectable from beyond the terminal velocity at v=-120 km ${\rm s}^{-1}$ , where there is very little gas, to v=-105 km ${\rm s}^{-1}$ , where it appears as a void surrounded by a great deal of H I emission. The structure remains changes very little over the entire velocity range. There are no obvious front or back caps and there is no evidence of continued expansion. With no obvious expansion it is difficult to hypothesize about the origins of the shell. Its size indicates that it, too may be an old supernova remnant, but its curious position at the terminal velocity suggests that it could be gas displaced from circular rotation by any number of possible influences.

**Figure 2:** Grey-scale image of an H I channel map at v=-108 km ${\rm s}^{-1}$ showing an apparent H I shell in the ISM. A small shell, of diameter
$\sim 0\mbox{$.\!\!^\circ$}4$ is located at

$l=329\mbox{$.\!\!^\circ$}9$ ,

$b=0\mbox{$.\!\!^\circ$}4$ . This shell is identified by the bright ring of emission surrounding an H I void. The grey-scale is linear as shown in the wedge on the right.
$\begin{figure} \begin{center} \psfig{file=term.ps,width=7cm,angle=-90} \end{center} \end{figure}$

RCW 94 Shell

There is an H I shell surrounding the H II region RCW 94 at

$l=326\mbox{$.\!\!^\circ$}3$ ,

$b=0\mbox{$.\!\!^\circ$}4$ . The shell is itself surrounded by a ring of H I depletion. Figure 3 shows the H I channel map at v=-38 km ${\rm s}^{-1}$ with continuum contours overlaid. The H I absorption distance for this H II region is $\sim 3.0$ kpc, which agrees well with a distance of $\sim 3.1$ kpc determined by Caswell & Haynes (1987) using recombination line velocities. The distance implies a shell diameter of $\sim 25$ pc. There is evidence for a small expansion velocity of $\sim 5$ km ${\rm s}^{-1}$ . We speculate that this shell was formed in a molecular cloud, where the molecular hydrogen was dissociated by the H II region, RCW 94. The depletion ring exterior to the shell is most likely an effect of the high molecular gas densities.

**Figure 3:** Grey-scale plot of an H I channel map at v=-38 km ${\rm s}^{-1}$ with 21-cm continuum contours overlaid. The grey scale is linear and runs from 0 to 80 K, while the contours are at 10 K intervals from 20 K to 400 K. The small shell, centered at
$l=326\mbox{$.\!\!^\circ$}3$ ,

$b=0\mbox{$.\!\!^\circ$}4$ is identified by the ring of emission around the H II region. The grey-scale is linear and is shown in the wedge on the right.
$\begin{figure} \begin{center} \psfig{file=rcw94.ps,height=7cm,angle=-90} \end{center} \end{figure}$

Supershells

Another class of H I shells are so-called ``supershells'', with diameters on the order of hundreds of parsecs. Two large supershells have been discovered in the Parkes data (McClure-Griffiths et al. 2000b). These shells, GSH 277+0+36 and GSH 280+0+59, have diameters of 620 pc and 430 pc, respectively, in addition to showing evidence of break-outs above and below the Galactic plane (see Figures 4 and 5). These shells appear to lie on the far edge of the Carina arm in the interarm region. The energy required to form a shell of this size is extremely large, on the order of

10⁵²-10⁵³ ergs. At that level of energy input, it is clear that supershells are one of the driving sources of energy in the ISM and have a significant impact on the structure of the Galaxy.

Figure 4: Grey-scale image showing three orthogonal slices through the H I cube containing the supershell GSH 277+0+36. The greyscale is linear and runs from 0 K (white) to 35 K (black). The largest panel is a channel map at v=+36 km ${\rm s}^{-1}$ . The panel to the right is the b-v slice at the position marked by the cross-hair in the l-b slice. Similarly, the bottom figure is the l-v slice at the same position. Several chimney-like extensions to high latitudes are visible in both the channel map and b-v slice (McClure-Griffiths et al. 2000b).

$\begin{figure} \psfig{file=shell1all.ps,height=7cm,angle=-90} \end{figure}$