THE THREE-mm ULTIMATE MOPRA MILKY WAY SURVEY. II. CLOUD AND STAR FORMATION NEAR THE FILAMENTARY MINISTARBURST RCW 106

The ThrUMMS survey (Barnes et al. 2015) covers the Galactic plane between $300^\circ \leqslant l\leqslant 360^\circ$ and $-1^\circ \leqslant b\leqslant 1^\circ$ in the ¹²CO (1–0), ¹³CO (1–0), C¹⁸O (1–0), and CN spectral lines at frequencies of 115.27, 110.20, 109.78, and 113.60 GHz, respectively. The survey was conducted using the new Very Fast Mapping (VFM) technique developed for use with the 22 m Mopra telescope.¹⁴ Half-Nyquist sampling rates were obtained along the scanning direction and between scan rows, yielding a 72'' resolution. Compared to the traditional on-the-fly mapping technique, the VFM technique is capable of observing a larger area in the same amount of time if Nyquist-sampled maps are not needed. As well as being more efficient, it provides a nearly uniform sensitivity per unit area. The data were converted from antenna temperature ${T}_{\mathrm{ant}}$ to main-beam temperature, ${T}_{\mathrm{mb}}$, by dividing ${T}_{\mathrm{ant}}$ by the main-beam efficiencies ${\eta }_{\mathrm{mb}}$ of 0.42 for ¹²CO (1–0), and of 0.43 for both ¹³CO (1–0) and C¹⁸O (1–0).

From the Southern Galactic Plane Survey (SGPS; McClure-Griffiths et al. 2005), we extracted the H i atomic line and 21 cm continuum data over the range of $330^\circ \leqslant l\leqslant 335^\circ$ and b = ±1°. The survey was conducted using the Australia Telescope Compact Array interferometer and was complemented with the Parkes 64 m telescope (FWHM = 15') for short spacings. The observations were performed simultaneously in a spectral line mode with 1024 channels across a 4 MHz bandwidth centered at 1420 MHz and in a continuum mode with 32 channels across a 128 MHz bandwidth centered at 1384 MHz (McClure-Griffiths et al. 2001, 2005). For our study, we used data which have a FWHM 22, line rms 1.6 K km s⁻¹, and continuum rms $\lt 1$ mJy beam⁻¹ (see Table 1).

We used 160 and 250 μm images from the Herschel Space Observatory to examine more deeply the properties of the cold dust near RCW 106. The fields were observed as part of the Herschel Infrared Galactic Plane Survey (Hi-GAL; Molinari et al. 2010) at 70/160 μm with the Photodetector Array Camera and Spectrometer (PACS; Poglitsch et al. 2010) and at 250/350/500 μm with the Spectral and Photometric Imaging REceiver (SPIRE; Griffin et al. 2010). The Hi-GAL data were taken in parallel mode with a fast scanning speed of 60'' s⁻¹. The raw (level-0) data of each individual scan from both PACS and SPIRE were calibrated and deglitched using HIPE¹⁵ version 11.0. The SPIRE and PACS level-1 data were then fed to version 18 of the Scanamorphos software package¹⁶ (Roussel 2013) to regrid and create the final maps. For our region, three Hi-Gal fields of $2\buildrel{\circ}\over{.} 5\times 2\buildrel{\circ}\over{.} 5$ were combined. The resultant 160 and 500 μm images have angular resolutions of 12'' and 37'', and 1σ rms of 0.08 and 1.0 MJy sr⁻¹.

To trace the warm dust emission associated with high-mass star formation, we used the 24 μm image from the Multi-band Imaging Photometer for Spitzer Galactic Plane Survey (MIPSGAL; Carey et al. 2009).

In the 21 cm continuum image (Figure 1) the giant H ii region RCW 106 (Rodgers et al. 1960) is the largest ($1\buildrel{\circ}\over{.} 2\times 1^\circ$) and brightest (∼287 Jy) area. RCW 106 is an elongated structure in $332\buildrel{\circ}\over{.} 6\lt l\lt 333\buildrel{\circ}\over{.} 8$ and $-0\buildrel{\circ}\over{.} 8\lt b\lt 0\buildrel{\circ}\over{.} 1$, and angled roughly 60° with respect to Galactic east within the $l,b$ plane (hereafter referred to as the "eastern" part). In the region $330\buildrel{\circ}\over{.} 5\lt l\lt 331\buildrel{\circ}\over{.} 7$ and $-0\buildrel{\circ}\over{.} 2\lt b\lt 0\buildrel{\circ}\over{.} 3$ there is another similarly angled loose structure containing bright 21 cm continuum sources (hereafter the "western" part). Between these two complexes of thermal sources there is a gap around l ∼ 332°. Projected in this gap is an OB association R103B ($332\buildrel{\circ}\over{.} 08$, $-0\buildrel{\circ}\over{.} 08$) (Mel'Nik & Efremov 1995) and also another further to the south, R103A ($331\buildrel{\circ}\over{.} 89$, $-0\buildrel{\circ}\over{.} 93$). These are at spectroscopic distances of 3.22 and 3.00 kpc, respectively, and so in the foreground. The eastern part is located at a distance of ∼3.6 kpc and the western part is located at a distance of ∼5 kpc.

Zoom In Zoom Out Reset image size

We have further divided the map in Figure 1 into 31 subregions based on a contour corresponding to the 3σ noise level above the local background. Most of the sources are H ii regions whose properties are classified according to their morphologies in Appendix. Additionally, there are four confirmed and three candidate supernova remnants (SNRs; Green 2014). For full details of each subregion, refer to the Appendix and the summary provided in Table 3.

In Figure 1 we also mark the positions of OH masers and H ii regions found in large scale surveys (Caswell et al. 1980; Jones & Dickey 2012). The positions of masers and H ii regions detected from radio recombination lines coincide with the bright radio continuum sources and help to distinguish substructures and to quantify the cloud properties (see Section 4).

To investigate the relative contributions of different clouds superimposed along the line of sight, we first plot the average ¹²CO spectrum of the entire region along with the distinctive spectra of the two main structures, eastern and western, defined in Section 3.1 (Figure 2). In the average spectrum there are three distinct peaks within the velocity range −140–10 km s⁻¹. The spectra for both main structures exhibit significant peaks at roughly the same velocities, in the ranges [−112, −80], [−72, −60], and [−60, −35] km s⁻¹ centered on V_LSR = −91, −63, and −48 km s⁻¹, respectively. These ranges are highlighted as green, blue, and red sectors in Figure 2. For the western region, the three spectral maxima are roughly the same, on average ∼7 K. For the eastern region, there are two peaks in the lowest velocity range, the brighter one at V_LSR ∼ −55 km s⁻¹ and the other at ∼−45 km s⁻¹ (see Figure 2). The peak at ∼−55 km s⁻¹ coincides with the ¹³CO peak that Bains et al. (2006) attributed to the cloud surrounding the RCW 106 cluster.

Zoom In Zoom Out Reset image size

In Figures 3(a)–(c) and (d)–(f) we display maps of the ¹²CO and H i emission integrated over three velocity ranges: [−112, −80], [−80, −60], and [−60, −40] km s⁻¹. Additionally in Figure 3, the positions of OH masers and H ii regions mostly coincide both spatially and spectrally with the molecular gas peaks.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figure 3(a) (corresponding to the green velocity range in Figure 2) shows extended CO emission spanning the entire longitude range. Its main peak is between 331° and 332° in longitude and extends just $0\buildrel{\circ}\over{.} 2$ below the mid-plane; it covers the western region and might extend to $l=333\buildrel{\circ}\over{.} 4$.

The CO map integrated from −80 to −60 km s⁻¹ (the middle white and blue velocity ranges in Figure 2), shown in Figure 3(b), is dominated by the strongest emission from the western region, but is otherwise relatively free from emission from the central plane (i.e., b ∼ 0°). On the other hand, the map shown in Figure 3(c) integrated from −60 to −40 km s⁻¹ (the middle white and red velocity ranges in Figure 2) is much brighter near the RCW 106 H ii region and has a diffuse filamentary component in the western region.

The H i emission in Figures 3(d)–(f) does not appear to correlate very well with either ¹²CO or 21 cm continuum emission, indicating that star formation activity does not correlate with this extended phase of the gas. However, in Figures 3(d)–(e), H i seems to form in the outer envelope of the MCCs.

We have integrated the ¹²CO, ¹³CO, and H i data cubes in Galactic latitude (from $-0\buildrel{\circ}\over{.} 5$ to $0\buildrel{\circ}\over{.} 5$) to produce position-velocity (${\ell }\;-\;V$) diagrams (Figure 4). The main emission in the ${\ell }\;-\;V$ maps is from −72 to −35 km s⁻¹ as in the averaged spectra (Figure 2). Another feature of strong though less prominent emission is from −112 to −80 km s⁻¹ at low ℓ (western end). This is detached from the lower velocity structure supporting our division of the emission into two different complexes: in the East is the RCW 106 MCC spanning the velocity range from −72 to −40 km s⁻¹, the West is the complex MCC 331.0 + 0.0 (V_LSR = −90 km s⁻¹) (hereafter referred to with the informative name MCC331–90 ) spanning the velocity range from −112 to −80 km s⁻¹.

Zoom In Zoom Out Reset image size

Figure 5 shows images of the thermal dust emission obtained in the mid-IR by Spitzer and in the far-IR and submillimeter by Herschel (Section 2.3).¹⁷ These provide insight into the dust content, cloud morphology, and radiation field. In particular, data at 24 μm (Figure 5(a)) allow us to trace local dust heating from the UV photons from young OB stars. Far-IR and submillimeter data probe cooler dust as well, both in the cloud and outside, and away from the influence of OB stars (Figures 5(b)–(c)).

Zoom In Zoom Out Reset image size

The morphology of far-IR emission is generally similar to the CO emission (Figure 3) suggesting that the gas and dust coexist within the MCCs. We also detect infrared dark clouds (IRDCs), seen as absorption in the mid-IR data but as emission in the far-IR data. These IRDCs are potential sites of massive star formation (Hennebelle et al. 2001; Rathborne et al. 2006; Simon et al. 2006; Peretto et al. 2010; Nguyen Luong et al. 2011a). In the dust emission at all wavelengths we see two main concentrations of dust at 331° and 333° whose detailed morphology matches well with that in the angled structures seen in the radio continuum and in ¹²CO. This correlation indicates that both the 21 cm continuum and the dust heating are generated by the massive stars in the RCW 106 and MCC331–90 MCCs. The Herschel far-IR data also reveal finger-like protrusions at the bottom of the structure linked to MCC331–90. These features are plausibly formed by radiation pressure from numerous OB stars in the local interstellar medium (e.g., Krumholz & Matzner 2009; Gritschneder et al. 2010; Tremblin et al. 2013).

In summary, the dust emission data from 8–500 μm confirm that this region has two distinct MCCs that are both extremely active in forming massive stars, including hosting a young stellar cluster.

Because an O7V star with a mass of 25 ${M}_{\odot }$ has a lifetime of $\sim 2\times {10}^{5}$ yr starting from its first appearance on the main-sequence diagram, we can calculate the star formation efficiency ${\epsilon }_{\mathrm{cloud}}^{\mathrm{past}}$, the star formation rate ${\mathrm{SFR}}_{\mathrm{cloud}}^{\mathrm{past}}$, and the SFR density ${{\rm{\Sigma }}}_{\mathrm{SFR},\mathrm{cloud}}^{\mathrm{past}}$ of the entire RCW 106 cloud complex for this "past" time interval as

$$\begin{eqnarray}&&{\epsilon }_{\mathrm{cloud}}^{\mathrm{past}}=\frac{{M}_{*}}{{M}_{\mathrm{cloud}}+{M}_{*}}={0.008}_{-0.003}^{+0.004},\end{eqnarray} \tag{ 6 }$$

$$\begin{eqnarray}&&{\mathrm{SFR}}_{\mathrm{cloud}}^{\mathrm{past}}=\frac{{M}_{*}}{2\times \ {10}^{5}\;\mathrm{yr}}={0.25}_{-0.023}^{+0.09}\;{M}_{\odot }\;{\mathrm{yr}}^{-1},\end{eqnarray} \tag{ 7 }$$

$$\begin{eqnarray}&&{{\rm{\Sigma }}}_{\mathrm{SFR},\mathrm{cloud}}^{\mathrm{past}}=\frac{\mathrm{SFR}}{{A}_{\mathrm{cloud}}}={9.5}_{-0.9}^{+3.4}\;{M}_{\odot }\;{\mathrm{yr}}^{-1}\;{\mathrm{kpc}}^{-2}.\end{eqnarray} \tag{ 8 }$$

with the total stellar mass ${M}_{*}=4.8\times {10}^{4}\;{M}_{\odot }$, the total gas mass ${M}_{\mathrm{cloud}}=5.9\times {10}^{6}\;{M}_{\odot }$ calculated over an area of ${A}_{\mathrm{cloud}}=2.6\times {10}^{4}\;{\mathrm{pc}}^{2}$.

The estimated ${\epsilon }_{\mathrm{cloud}}^{\mathrm{past}}$ is in the middle of the range for giant molecular clouds (GMCs), 0.002–0.2, as derived from the ionizing flux of young stars (Murray 2011). However, the SFR density of RCW 106 is high and for its surface density of 220 ${M}_{\odot }$ pc⁻² (Section 5) it is well above the trend in the Schmidt–Kennicutt relation (see Figure 7). This shows that RCW 106 has been very active in forming massive stars during the last 2 × 10⁵ yr.

Zoom In Zoom Out Reset image size

The large amount of gas remaining in this complex suggests that the burst of star formation activity in the RCW 106 molecular cloud complex might not yet be finished. We investigated this by estimating the future star-formation activity using the dense clump and core populations.

First, we make an assumption that the dense clump mass of $2.2-2.7\times {10}^{5}\;{M}_{\odot }$ derived from ¹³CO line emission (Bains et al. 2006) or 1.2 mm continuum emission (Wong et al. 2008) represents the clump mass of the entire RCW 106 complex. This assumption is justified since most of the dense gas concentrates in the region considered by Bains et al. (2006) and Wong et al. (2008; see for example Figure 5). We estimate the clump formation efficiency, ${\epsilon }_{\mathrm{cloud}\to \mathrm{clump}}$,²⁰ as the ratio of the total clump mass to the total cloud mass (Eden et al. 2012, 2013; Louvet et al. 2014):

$$\begin{eqnarray}&&{\epsilon }_{\mathrm{cloud}\to \mathrm{clump}}=\frac{{M}_{\mathrm{clump}}}{{M}_{\mathrm{cloud}}}\;\sim \;0.042\pm \;0.013.\end{eqnarray} \tag{ 9 }$$

The estimated ${\epsilon }_{\mathrm{cloud}\to \mathrm{clump}}$ of the RCW 106 complex is about 10 times higher than that of the famous massive star-forming region Cygnus X and almost equal to that of W43 (Nguyen Luong et al. 2011b). Thus the RCW 106 complex is forming massive clumps almost as efficiently as the W43 star-forming region despite the mass surface density of RCW 106 being four times lower.

Second, a cloud's efficiency at converting mass into dense clumps subsequently affects the star formation efficiency (Eden et al. 2012; Louvet et al. 2014). If the dense clumps (r ∼ 1 pc) in the RCW 106 complex form dense cores (r ∼ 0.1 pc) with a mass transfer efficiency ${\epsilon }_{\mathrm{clump}\to \mathrm{core}}$, and dense cores will form protostars with a mass transfer efficiency ${\epsilon }_{\mathrm{core}\to *}$, the future star formation efficiency of the entire cloud will be

$$\begin{eqnarray}&&{\epsilon }_{\mathrm{cloud}\to *}={\epsilon }_{\mathrm{cloud}\to \mathrm{clump}}\times {\epsilon }_{\mathrm{clump}\to \mathrm{core}}\times {\epsilon }_{\mathrm{core}\to \ *}.\end{eqnarray} \tag{ 10 }$$

If we adopt typical values ${\epsilon }_{\mathrm{clump}\to \mathrm{core}}$ ∼ 0.1–0.3 (Parmentier & Pfalzner 2013; Louvet et al. 2014) and ${\epsilon }_{\mathrm{core}\to *}\;\sim$0.3–0.8 (Alves et al. 2007; Parmentier & Pfalzner 2013; Myers 2014), we will have a future ${\epsilon }_{\mathrm{cloud}\to *}$ ∼0.004 ± 0.004.

Because the massive protostellar phase has a timescale of $\sim 2\times {10}^{5}$ yr (Motte et al. 2007; Russeil et al. 2012), similar or OB stars, we calculate the "future $2\times {10}^{5}$ yr" ("future") SFR and SFR density as

$$\begin{eqnarray}&&{\mathrm{SFR}}_{\mathrm{cloud}}^{\mathrm{future}}=\frac{{\epsilon }_{\mathrm{cloud}\to *}}{2\times {10}^{5}\;\mathrm{yr}}{M}_{\mathrm{cloud}}=0.12\pm 0.1\;{M}_{\odot }\;{\mathrm{yr}}^{-1}\end{eqnarray} \tag{ 11 }$$

and

$$\begin{eqnarray}&&{{\rm{\Sigma }}}_{\mathrm{SFR},\mathrm{cloud}}^{\mathrm{future}}=\frac{{\mathrm{SFR}}_{\mathrm{cloud}}^{\mathrm{future}}}{{A}_{\mathrm{cloud}}}=4.8\pm 3.8\;{M}_{\odot }\;{\mathrm{yr}}^{-1}\;{\mathrm{kpc}}^{-2}.\end{eqnarray} \tag{ 12 }$$

This SFR density and the surface density also combine to place the RCW 106 complex in the starburst regime of the Schmidt–Kennicutt relation (Kennicutt 1998; see Figure 7). It should be possible to refine and confirm the future SFR density of the RCW 106 complex by counting the massive dense cores identified using the Herschel data. This will be the topic of a subsequent paper.

These SFR densities are similar to the future SFR density of W43 and higher than that of Cygnus X, reinforcing the conclusion that RCW 106 is undergoing a ministarburst at present. The similarity between past and future SFR densities implies that this ministarburst event is not yet finished but is in the declining phase. It is also much higher than the SFR density of CMZ, the region of the galactic plane within a few degrees of the Galactic centre, well-known for its inefficiency in star formation despite being 100 times more massive than the RCW 106 complex (Immer et al. 2012; Kruijssen et al. 2013).

While this is consistent with the large mass reservoir still available in the RCW 106 complex, there are still outstanding questions, including what the final conversion efficiency will be and its relationship to the virial parameter (Section 6).

The calculations in Sections 7.1 and 7.2 should involve the general uncertainties from the assumptions of the IMF, the timescales, and the mass transfer efficiencies. Several authors suggest that the IMF of the extreme star formation environment is top-heavy (Stolte et al. 2002, 2005; Chabrier et al. 2014). However, the validity of this statement is still largely debated (Kim et al. 2006; Brandner et al. 2008). If we use an IMF with a steeper slope, the SFR likely increases by about 20%. We assume a burst period timescale of 2 × 10⁵ yr, which might have variations within a factor of two; but this is the most appropriate assumption since this is the measured timescale of an OB star (Martins et al. 2005) and an MDC (Motte et al. 2007). The third uncertainty is the mass transfer efficiencies from clouds to clumps, to cores and to stars. Our assumptions of these quantities are the average results of the current state-of-the-art theoretical and observational studies, therefore the true values might vary at most by 10%–30%. We already take these three uncertainties into account in our calculations.

The uncertainties from Lyα photons and the clumps are low since we focus only on the brightest regions in the radio continuum map (in the case of estimating the Lyα photons) and in millimeter continuum or ${}^{13}$CO maps (in case of estimating the clump mass). For the later case, the clump mass is verified with an independent measurement from NH₃ emission (Lowe et al. 2014).

The H ii region complex RCW 106 spans approximately $1\buildrel{\circ}\over{.} 5$ in the range $332^\circ \leqslant l\leqslant 333\buildrel{\circ}\over{.} 5$ and so at a distance of 3.6 kpc (Lockman 1979) is of "giant" extent, 94 pc. The structure lies predominately below the Galactic mid-plane. We separated the structure into six components which are large themselves, as illustrated in Figure 8 and described in the brief notes below.

• (RCW 106a). Large in size at nearly 33 pc. It has many known H ii regions, star clusters, and SNR G333.6–00.2 which is located at the brightest peak.
• (RCW 106b). An extended structure of ∼17 pc in length. It has an irregular shape with prominent protrusions. The main intensity peak is crescent shaped.
• (RCW 106c). An irregular H ii region with a fairly large main peak and size of almost 9 pc.
• (RCW 106d). An irregular H ii region with a fairly large main peak and a tail.
• (RCW 106e). An irregular structure with three main peaks loosely connected and a size of about 17.5 pc.
• (RCW 106f). Diffuse H ii region with four dominant peaks.

A number of the smaller H ii regions in Figure 1 appear to be fairly "circular" in structure. These are illustrated in Figure 9 and described below.

• The compact H ii region contains the single IRAS source 16226–4900 at its center.
• An ultracompact H ii region coincides with an IR star cluster [DBS 2003] 170.²¹ As discussed by Dutra et al. (2003), the composition is a bit ambiguous but there are a few bright stars. The H ii region GAL 334.71–00.67 found in this box is 17.4 kpc away as determined by its positive velocity (Russeil et al. 2005).
• The compact H ii region that we detect is roughly 5.8 pc in size and probably coincides with the H ii region GAL 334.68–00.11. It has a nearby cluster of IR sources (Robitaille et al. 2008).
• This compact H ii region is roughly 1.6 pc in size, which almost makes it an ultracompact candidate.
• There is a YSO on the border of the compact H ii region. The star cluster [MCM 2005b] 79 (Mercer et al. 2005) is also nearby. This source has a faint protrusion in the lowest contour, not so significant that we describe it as irregular.
• This compact H ii region has a few YSOs nearby.
• This compact H ii region contains the IR star cluster [DBS 2003] 102. It also coincides with the known H ii regions GAL 332.98+00.79 and [WHR97] 16112–4943. The latter is classified as UCH ii (Walsh et al. 1997).
• This compact grouping encompasses the known H ii regions [KC 97c] G332.5–00.1, GRS 332.54–00.11, and GAL 332.54–00.11 (Kuchar & Clark 1997) and the star clusters [DBS 2003] 160 and 161. This clump is a potential OB cluster as seen from the numerous OB stars within the region. There are also some YSOs and outflow regions within the clump.
• This is the compact H ii region IRAS 16119-5048 driving an outflow.
• An H ii region hosting a star cluster [MCM 2005b] 74. It has a cometary shape and is significantly brighter with an off-center peak in its circular structure. It is located near the supernova remnant, SNR G332.4-0.4.
• Most likely the compact H ii region GAL 331.36+00.51, the structure is elliptical with a slight protrusion along the longitudinal axis and is 8.6 pc in size.
• A possible compact H ii region. It is ∼5 pc in size with the known H ii regions [KC 97c] G331.3–00.2, GAL 331.28–00.19, and GAL 331.26–00.19 within its peak.
• A compact H ii region, 2.9 pc in size and associated with the IR cluster [DBS 2003] 153. It also hosts the know H ii region IRAS 16037–5223 and GAL 330.31–00.39.
• A compact H ii region probably powered by GAL 330.04–00.05.

The remaining H ii regions as seen in Figure 10 are "irregular" in outline as compared those in Figure 9. Descriptions of these follow.

• An extended H ii region of ∼10 pc in size. It has numerous IR star clusters and known H ii regions. It displays a comet-like structure with its brightest and largest feature at $331\buildrel{\circ}\over{.} 33$. There are two smaller peaks trailing off from the tail. They are diffuse and seemingly attached to the main structure.
• A compact H ii region with the IR star cluster, [DBS 2003] 158. It is elongated and has double peaks. The main peak is off-center in one of the protrusions where the known H ii regions are located. It is also near the supernova candidate MSC331.8+0.0 whose structure may be related.
• The structure is elongated with three main compact peaks that may all be connected. Together they form an elongated structure that is ∼11 pc in length.
• Both GAL 331.03–00.15 and PMN J1610–5150 coincides with this H ii region. There are numerous OB associations spread around the main intensity peak.
• This has been subdivided into three regions 28A, 28B, and 28C from largest to smallest with sizes ∼12, 4, and 3 pc, respectively. The H ii region 28A hosts numerous known H ii regions and the IR star cluster [DBS 2003] 155 while 28B appears to be a compact H ii region with IR star cluster [DBS 2003] 154. 28C is as yet undefined. Individually, 28A has a comet-like shape with two tails while 28B and 28C are circular in nature.

There are four confirmed and two candidate SNRs (Figure 11). However, we do not see 6332.4+0.6 (Stupar et al. 2008). These are seen in projection and are not necessarily associated with the two main star-forming complexes.

• The SNR candidate G333.9+00.0 seems very faint with many separated clumps.
• The SNR G332.4+0.1 has three main peaks that may be part of one shell judging from the contour lines. From OH absorption it is thought to be in the range of 7–11 kpc (Caswell & Haynes 1975).
• The SNR G332.4–0.4 seems to have a circular shell. A higher intensity portion situated near the bottom is the RCW 103 star cluster. It has a kinematic distance of 3.1 kpc from H i absorption lines (Green 2009).
• The shell of SNR G332.0+0.2 is a crescent and is thought to be at at least 7 kpc (Caswell & Haynes 1975).
• The SNR candidate MSC 331.8+0.0 shows a high intensity portion with scattered filaments.
• The SNR G330.2+01.0 has a minimum distance of 4 kpc (McClure-Griffiths et al. 2001).