The G305 star forming complex: Wide-Area molecular mapping of NH3 and H2O masers
L.Hindson, M.A. Thompson, J.S. Urquhart, J.S. Clark, B. Davies
aa r X i v : . [ a s t r o - ph . GA ] J un Mon. Not. R. Astron. Soc. , 1–14 (2010) Printed 23 October 2018 (MN L A TEX style file v2.2)
The G305 star forming complex: Wide-Area molecular mapping ofNH and H O masers
L.Hindson , ⋆ , M.A. Thompson , J.S. Urquhart , J.S. Clark , B. Davies , Centre for Astrophysics Research, Science and Technology Research Institute, University of Hertfordshire, College Lane, Hatfield, AL10 9AB, UK Australia Telescope National Facility, CSIRO Astronomy and Space Science, PO Box 76, Epping, NSW, 1710, Australia Department of Physics and Astronomy, The Open University, Walton Hall, Milton Keynes MK7 6AA Center for Imaging Science, Rochester Institute of Technology, 54 Lomb Memorial Drive, Rochester, NY 14623, USA School of Physics & Astronomy, University of Leeds, Woodhouse Lane, Leeds LS2 9JTAccepted 07 / / / / ABSTRACT
We present wide area radio (12 mm) Mopra Telescope observations of the complex and richmassive star forming region G305. Our goals are to determine the reservoir for star formationwithin G305 using NH to trace the dense molecular content, and thus, the gas availableto form stars; estimate physical parameters of detected NH clumps (temperature, columndensity, mass etc); locate current areas of active star formation via the presence of H O andmethanol masers and the distribution of YSOs and ultra compact HII regions associated withthis region. This paper details the NH (J,K), (1,1), (2,2) and (3,3) inversion transition and 22GHz H O maser observations. We observed a ∼ . ◦ × ◦ region with ∼ ′ angular resolutionand a sensitivity of ∼
60 mK per 0.4 km s − channel. We identify 15 NH (1,1), 12 NH (2,2) and 6 NH (3,3) clumps surrounding the central HII region. The sizes of the clumpsvary between < . > M ⊙ and find the total molecular mass of the complex to be ∼ × M ⊙ . We note the positions of 56 star formation tracers, and discover a high degree ofcorrelation with detected NH clumps. We have detected 16 H O masers, find they correlatewith the detected ammonia clumps and in general are found closer to the NH clump coresthan star formation tracers of later evolutionary stages. Key words:
Ammonia – Molecular clouds – Massive Star Formation – Maser
High mass stars ( > ⊙ ) have a significant e ff ect on their sur-rounding volume, injecting a large amount of energy into the natalcloud. This feedback energy takes the form of high velocity winds,outflows, expanding HII regions, high energy radiation and even-tually supernovae. These processes have a profound e ff ect, poten-tially triggering new star forming events (Elmegreen & Lada 1977;Habing & Israel 1979; Elmegreen 2002) or alternatively disruptingfuture formation by dispersing the natal cloud (Elmegreen & Lada1977).The G305 star forming region lies in the Scrutum Crux armwithin the galactic plane at l = ◦ , b = ◦ . At a distance of ∼ >
31 deeply ⋆ E-mail: [email protected] embedded 07V stars (Clark & Porter 2004). G305 provides an ex-cellent laboratory in which to study massive star formation as it isone of the closest sites of massive star formation in the galaxy andcontains multiple sites and di ff erent epochs of star formation overits large scale, allowing the investigation of evolution and environ-mental e ff ects such as triggering.In Fig. 1, we present a three colour composite image of G305,constructed using images obtained from the Galactic Legacy In-frared Mid-Plane Survey Extraordinaire (GLIMPSE) (Benjaminet al. 2003; Churchwell et al. 2009). The 8 µ m filter is of partic-ular interest as it is dominated by polycyclic aromatic hydrocarbon(PAH) features, these are excited in the surface layers of molecularclouds that are exposed to a significant amount of UV radiation andtherefore an excellent tracer of photo-dominated regions (PDRs)(Leger & Puget 1984). For this reason PDRs are a good signpostfor molecular regions that are undergoing a strong interaction witha nearby HII region (Urquhart et al. 2003). In this image we showthe positions of Danks 1 & 2 and WR 48a which are believed tobe the driving force behind the central di ff use HII region (Clark &Porter 2004).Here we report on our first observations of NH (1,1), (2,2), c (cid:13) L.Hindson
Figure 1.
Wide-field 3 colour image ( 8 µ m (red), 4.5 µ m (green) and 3.6 µ m (blue)) GLIMPSE image of the G305 complex, with the positions of Danks 1 &2 over-plotted with boxes. Our scan area is shown by light grey boxes. (3,3) and the 22 GHz H O maser line, using the Mopra radio tele-scope. Previous imaging of G305 by the H O Southern GalacticPlane Survey (HOPS) (Walsh et al. 2008) has revealed the presenceof NH hotspots in G305, but lacked the depth to trace less massiveclumps and moderate column densities. Using NH which has ane ff ective critical density of 10 − cm − (Ho & Townes 1983),we are able to trace the dense gas structure of the region. Surveysof such dense regions are critical to relating the properties of themolecular gas to the star-forming properties of the cloud. We in-tend to make use of these low resolution NH maps as a pathfinderfor future high resolution interferometric observations. We also re-port the positions of 22 GHz H O maser emission, which have beenshown to be good tracers of the early stages of low and high massstar formation (Furuya 2003). In high mass YSOs, H O masershave been found to be associated with hot molecular cores, molec-ular outflows and jets. H O maser emission is believed to highlightan evolutionary phase before the onset of a UC HII region aroundan embedded star (Furuya 2003). In low mass YSOs, H O masersare found to be associated with predominantly class 0 YSOs and inexcited shocks of protostellar jets in the vicinity of the star. Thus,they are an excellent signpost of the early stages of star formation.In Section 2 we present details of the observations and the datareduction processes, in Section 3 we present our observational re- sults and derive basic parameters, Section 4 presents our discussionand in Section 5 we present a summary of our results.
Observations were made using the Australia Telescope NationalFacility (ATNF) telescope Mopra. This is a 22m antenna located26km outside the town of Coonabarrabran in New South Wales atan elevation of 866 meters above sea level and at a latitude of 31degrees south.The telescope is equipped with a 12 mm receiver with a fre-quency range of 16 to 27.5 GHz. The UNSW Mopra spectrometer(MOPS) is made up of four 2.2 GHz bands which overlap slightlyto provide 8 GHz continuous bandwidth. The Mopra spectrometerhas two possible modes, narrow and broadband modes. The narrowband or ”zoom” mode of MOPS was used and allowed us to ob-serve 16 spectra simultaneously, with a bandwidth of 137.5 MHzover 4096 channels in each zoom window. This gives a channelspacing of 34 kHz, corresponding to a velocity resolution of ∼ . − at 24 GHz. We set our zoom windows to match those usedin the HOPS survey, see Table 1 in Walsh et al. (2008) for detailsof targeted lines. Our line detections are summarised in Table 1. Inthis paper we focus upon the analysis of the NH and H O maser c (cid:13) , 1–14 he G305 star forming complex: Wide-Area molecular mapping of NH and H O masers Table 1.
Lines detected, central frequency and the size of the Mopra beamat the given frequency.Line Freq HPBW(GHz) ( ′′ )H O (6-5) 22.235 144NH (1,1) 23.694 136NH (2,2) 23.722 136NH (3,3) 23.870 136HC N (3,2) 27.294 120H69 α OH (3 - 3 ) 24.928 131CH OH (4 - 4 ) 24.933 131CH OH (2 - 2 ) 24.934 131CH OH (6 - 6 ) 24.959 131CH OH (7 - 7 ) 25.018 131 detections. We postpone the analysis of the remaining detections(H69 α , CH OH and HC N) to future publications combining thisdata with forthcoming observations of the cm-wave radio contin-uum (Hindson et al 2010, in prep) and multi-line studies of CH OHand HC N.In order to cover the 1.5 × × ′ × ′ maps with a 3 ′ overlap between adjacent maps (see Fig.1). All maps were obtained using the fly mapping mode, in whichthe telescope scans along the sky taking spectra at regular intervalsalong lines of constant latitude and longitude. The maps were sam-pled with 51 ′′ spacing to give Nyquist sampling at the highest fre-quency of 27.4 GHz and better than Nyquist at lower frequencies.The fastest scan rate of 2s per point was used, an o ff source posi-tion was observed periodically to remove sky emission. Combinedwith the limited scan speed and large map size a single map took ∼ ∼
100 K,with a variation in T sys of no more than 20% during the course ofany particular map. During times of dense cloud cover and rain,observations were halted as these conditions caused system tem-peratures of well over 120 K.The data were reduced using the packages LIVEDATA andGRIDZILLA, both are AIPS ++ packages written by Mark Calabr-reta for the Parkes radio telescope and adapted for Mopra. LIVE-DATA performs a bandpass calibration for each row using the o ff -source data followed by fitting user specified polynomial to thespectral baseline. GRIDZILLA calculates the pixel value from thespectral values and weights using the weighted mean estimation.We use a Gaussian smoothing kernel of 2 ′ with a cut o ff radius of0 . ′ . This produces a map with minimal smoothing giving a pixelsize of 0 . ′ × . ′ and a sensitivity of ∼
60 mK per 0.4 km s − channel, approximately twice as deep as the ongoing HOPS survey(Walsh et al. 2008). We make use of a recent study into the e ffi -ciency of Mopra (Urquhart et al. 2010) to convert the NH antennatemperatures into main beam temperatures using the e ffi ciency at23 GHz of 0.64. http: // / computing / software / livedata.html In this section, we describe the distribution of NH clumps as wellas the locations of detected H O masers in the G305 complex. Wedefine individual NH clumps using the clump finding algorithmF ellwalker and present their source averaged spectra. We derivethe basic physical properties of the detected clumps from their NH emission. O masers
In Fig. 2, we present peak temperature contour plots of NH (1,1)and (2,2) emission, the noise in the map is ∼ .
06 K. We showthe location of H O maser emission detected towards G305 (bluecrosses). The ammonia contours are overlaid onto a grayscaleGLIMPSE 5.4 µ m image, which also shows PAH emission high-lighting the PDR. We use the 5.4 µ m image because it is not asbright as the 8 µ m band and so avoids the problem of saturationcommonly encountered with the 8 µ m band. We detect a numberof NH clumps in four distinct regions around the central HIIregion. Using WR 48a as a reference, we define the four regionsof emission as (Galactic) north east (NE), north west (NW), southeast (SE) and west (W). Within these regions, we distinguish 15NH (1,1), 12 NH (2,2) clumps and 6 NH (3,3) clumps. Themajority of clumps are located around the periphery of the centralcavity, with the higher transitions located mainly towards the PDRboundary. We detect 16 H O masers distributed throughout theregion, all but two are found to be within the bounds of NH emission. Of the 14 masers associated with NH emission, elevenare found close to the clump cores, no NH core is associated withmore than one maser detection. Weak NH clumps (excludingclump 11) in the western region appear to be devoid of H Omaser emission. Detected clumps have been defined and numberedaccording to the output generated by the clump finding algorithmF ellwalker (see section 3.1.1) and H O masers are numberedaccording to Table 5.
NW Region:
Within this region we detect three NH (1,1) clumpsand one NH (2,2) and (3,3) clump centred within clump 4. Wedetect three H O masers in this region within the boundaries ofclump 4 and 7, these are close to but o ff set from the peak emissionof the ammonia clumps. Ammonia clumps are connected by asurrounding region of low emission in the (1,1) transition andclump 4 appears to be elongated in the direction away from thecentral exciting stars. Unlike the other clumps in the complex,the emission peak in clump 4 is coincident with strong 5.4 µ memission towards the centre, indicating that the direction of thepowerful (possibly ionising) radiation is impacting from ourperspective, face on with the cloud, this area of strong PAH emis-sion is also coincident with a HII (see Clark & Porter (2004) Fig. 2). NE Region:
This region contains two NH clumps positionedbehind one another, with clump 1 positioned further to the north,both clumps exhibit NH (1,1), (2,2) and (3,3) emission. We detecttwo H O masers in this region both are located near the cores ofthe detected clumps but again they are slightly o ff set from theNH peak. Clump 1 is host to the strongest peak and averagemain beam temperatures in the complex in both NH (1,1), (2,2)and (3,3) emission lines (1.21, 0.66 and 0.42 K respectively)and it is the largest single clump detected (10.1 pc). The clumpsare bounded by a region of low emission and are separated bya region of lower density. The two clumps are o ff set from what c (cid:13) , 1–14 L.Hindson
Figure 2.
Contour map of the peak temperature NH (1,1) and (2,2) emission towards G305 in black and red contours respectively. Emission is over-plottedonto a GLIMPSE 5.4 µ m (greyscale) image, detected H O masers are shown by blue crosses and numbered according to Table 5. Contours begin at 0.15 Kand increment by 0.1 K, the noise in the map is 0.06 K. The clump numbers are shown for reference within circles and clump areas are shown as light greyoutlines. Following our nomenclature we separate the complex into four regions of emission with grey boxes. appears to be a bubble of intense PAH emission to the west that isalso coincident with a HII region (see Clark & Porter (2004) Fig. 2).
SE Region: NH (1,1) emission is observed in an unbroken stripparallel to the PDR and central cavity boundary, we observe threeseparate clumps within this strip, all of which are coincident withNH (1,1) and (2,2) emission at the clump centres with NH (3,3)detected within clump 2 and 8. We detect four water masers allwithin the bounds of the NH (1,1) emission, three of these masersare found in close proximity to the detected ammonia clump peaks.Clump 2 appears to be extended in the direction away from thecentral clusters. We also note two strong PAH features on theperiphery of clump 2 and in the region separating clump 8 and 6 onthe inward side of the ammonia emission with the region betweenclump 8 and 6 also being the location of a HII region. W Region:
The region to the west of WR 48a shows a di ff erentmorphology to previous regions. Here we find 7 isolated clumpsof weak emission extending away from the centre of G305 asopposed to large clumps. The strongest NH (1,1) source and theonly source of visible NH (2,2) and (3,3) emission are locatedsome distance from the boundary PDR in clump 5. Six H O maserswere detected, five of which are in close proximity to NH clumps.We note that the PDR in this region appears to be dispersed tothe north and south leaving only a small area of PAH emissioncoincident with clump 14. F ellwalker Clumpfind
We have chosen to use the STARLINK tool CUPID to automat-ically detect clumps in our 2D map, in particular the F ellwalker algorithm developed by David Berry (Berry et al. 2007). The F ell - walker algorithm works by finding the paths of steepest gradientfrom each pixel in the image. Starting with the first pixel in the im-age each of the surrounding pixels are inspected to locate the pixelwith the highest ascending gradient, this process continues until apeak is located (i.e. a pixel surrounded by flat or descending gra-dients). The pixels along this path are assigned an arbitrary integerto represent their connection along a path. All pixels in the im-age are inspected in a similar process and the image is segmentedinto clumps by grouping together all paths that lead to the samepeak value. The pixels belonging to paths that lead to a single peakare then defined as belonging to that particular clump. For a fullerdescription of this process see Berry et al. (2007). The F ellwalker algorithm is best used on images that have had the background sub-tracted, this has been performed using the background tracing andsubtraction algorithm Findback in the CUPID package to subtractthe uniform background of 0.06 K from the peak temperature, mo-ment integrated image. The parameters of the detected NH (1,1),(2,2) and (3,3) clumps can be found in Tables 2, 3 and 4. F ell - waker is not a widely used clump finding algorithm and so we car-ried out a number of checks using better known algorithms such as http: // starlink.jach.hawaii.edu / starlink / CUPIDc (cid:13) , 1–14 he G305 star forming complex: Wide-Area molecular mapping of NH and H O masers GAUSSCLUMPS and CLUMPFIND to check for consistency, aswell as performing manual aperture photometry checks. The F ell - walker algorithm gives very similar results to GAUSSCLUMPS,CLUMPFIND and manual aperture photometry with a significantlyhigher robustness to varying input parameters (see e.g. Pineda etal. 2009, for a discussion of the e ff ect of varying the step sizein CLUMPFIND). Due to the simplicity and robustness of F ell - walker we chose to use this algorithm to segment our images intoclumps and extract the photometry for each clump. clump spectra In Fig. 3, we present source averaged NH (1,1) and (2,2) spec-tra for each clump region defined by F ellwalker (shown as a greyoutline surrounding the contours in Fig.2). In Fig. 4, we presentthe NH (3,3) spectra averaged over the (3,3) emission region. Allspectra have been Hanning smoothed to improve the noise to ∼ .
01K per ∼ . − channel. We have applied the same NH (1,1)clump area results to the NH (2,2) emission, therefore the (1,1) and(2,2) clump spectra are averaged over the same area, this allows usto make comparisons between the transitions. We should only ex-pect to see seven strong detections in the NH (2,2) and even fewerin the (3,3) transition that are coincident with NH (1,1) emission,as seen in the peak temperature contour map Fig. 2. This is notthe case, Fig. 3 clearly shows faint NH (2,2) emission in twelveclumps and we see 6 NH (3,3) detections in Fig. 4, this is due tothe improved signal to noise from applying Hanning smoothing.We note the velocity of the NH (2,2) and (3,3) emission appearsat approximately the same velocity as the NH (1,1) emission inall cases, for the particularly weak NH (2,2) and (3,3) emissionthis provide supporting evidence that it is indeed real. The NH (3,3) emission area is much smaller in geometric size than the (1,1)emission, therefore averaging the (3,3) spectra over the (1,1) areasignificantly lowers the temperature of the (3,3) emission due tobeam dilution. We make use of NH (3,3) emission to highlight ar-eas of higher temperature and for this reason we have created NH (3,3) source averaged spectra shown in Fig. 4 in order to betterhighlight the warmer clumps.The hyperfine nature of the NH (1,1) inversion transition pro-vides a good indicator of optical depth through the ratio of inten-sity between the main and satellite lines. For this reason we haveapplied hyperfine fitting to all the NH (1,1) spectra. Due to insuf-ficient signal to noise, we are unable to resolve hyperfine structurein the NH (2,2) emission and so we have applied Gaussian fitting.We only fit a Gaussian to the NH (2,2) and (3,3) data if we can beconfident that it is a true detection and the peak emission is > σ ,thus only clumps 10, 11 and 15 are left with no (2,2).Profiles were fitted using the IDL functions CURVEFIT forNH (1,1) and GAUSSFIT for the NH (2,2) and (3,3) lines. Themodel line profiles are over plotted in red in Fig. 3 and 4. TheGAUSSFIT function computes a non-linear least-squares fit to thedata. The CURVEFIT function uses a gradient-expansion algorithmto compute a non-linear least squares fit to a user-supplied functionwith an arbitrary number of parameters. Iterations are performeduntil the chi square changes by a specified amount, or until a max-imum number of iterations have been performed. To fit the ammo-nia profiles we assume all the components have equal excitationtemperatures and that the ammonia line widths and separations areidentical to laboratory values.We find line widths range from 1.7 km s − to 10.0 km s − . Thethermal part of the line width accounts for only 0.28 km s − (forT kin =
30 K) we therefore assume the line widths are largely dom- inated by non-thermal turbulent motion. It can be clearly seen thatseveral spectra exhibit blended main line and first satellite lines,this may be caused by emission from several unresolved clumpswith di ff ering velocities. The NH (3,3) line widths are found to bewider than the lower transition counterparts for all clumps with theexception of clump 5. O masers
Sixteen H O masers were detected towards G305, their positions,velocity (LSR), velocity range and peak flux can be found in Table5. It is important to note that due to the large size of the Mopra beam( ∼ ′ ), we are currently unable to determine the location of theemission to better than a few 10s of arcseconds. However follow upobservations currently being obtained by the HOPS survey (Walshet al. 2008) will accurately determine the locations to within a fewarc-seconds. We make use of a recent study into the e ffi ciency ofMopra by Urquhart et al (submitted 2010), specifically Table 3, toconvert H O maser antenna temperatures to Jy using the conversionfactor of 12.4. We find 6 new H O masers in the region, Walsh et al.(2008) show a detection of two H O maser at l = . , b = − . l = . , b = + .
07 which we did not detect. The onlyapparent explanation for this is H O maser variability, since ourobservational setup mimicked the HOPS survey and we obtaineddeeper, less noisy maps we should expect to detect all of the H Omasers reported in Walsh et al. (2008) and more. Future HOPS ob-servations should clarify this ambiguity. H O maser emission wasdetected across a broad velocity range between -120.8 and -10.4km s − with a peak intensity varying between 2.4 and 1087.8 Jy,these values are characteristic of H O maser emission (Breen et al.2007). The peak V
LSR and velocity range match well with the ve-locity of the NH emission, it is clear that all detected H O masersare associated with G305 with the exception of maser 12, whichhas a peak velocity of -97.6 km s − and a velocity range that doesnot coincide with the G305 complex. In the following section, we present a standard analysis of the am-monia clumps using the spectra shown in Fig. 3.
Using the hyperfine structure, we extract the optical depth from theratio of satellite to main line intensity, using Eq. (1) from Ho &Townes (1983): △ T ∗ α (J , K , m) △ T ∗ α (J , K , s) = − e − τ (J , K , m) − e − ατ (J , K , m) (1)where T ∗ α (J , K , m) is the observed antenna brightness tempera-ture, (m,s) refer to the main and satellite hyperfine components, τ (J , K , m) is the optical depth of the main component and α is theratio of intensity for the satellite compared to the main component[a = ff erent hyperfine components. This is a reasonableassumption because of the very close energy separations and thesmall probability of special excitation mechanisms that di ff erenti-ate between the hyperfine components (Ho & Townes 1983). c (cid:13) , 1–14 L.Hindson
Figure 3.
Clump averaged spectra of the 15 NH clumps towards G305. Clumps have been integrated spatially over the clump area defined by F ellwalker and Hanning smoothed to provide a sensitivity of ∼ .
01 K per ∼ . − channel. Hyperfine fitting is applied to the (1,1) emission and Gaussian fitting tothe (2,2), shown as red overplots. c (cid:13) , 1–14 he G305 star forming complex: Wide-Area molecular mapping of NH and H O masers Table 2. NH (1,1) clump properties, showing central position, velocity, diameter (geometric average), peak main beam temperature and clump averaged mainbeam temperature. We can only resolve clump diameters of > . (1,1) clump region defined by F ellwalker Region Clump No Peak V
LSR
FWHM ∆ V Diameter Peak T MB (1,1) Avg T MB (1,1)Galactic ( l ) Galactic ( b ) (km s − ) (km s − ) (pc) (K) (K)NE 1 305.23 0.29 -40.4 3.8 10.1 1.21 0.333 305.19 0.21 -42.1 4.6 5.2 0.56 0.22NW 4 305.36 0.20 -38.8 5.6 7.6 0.55 0.227 305.42 0.25 -39.8 3.1 7.4 0.46 0.199 305.54 0.34 -36.4 4.8 3.4 0.32 0.17SE 2 305.26 -0.02 -32.3 4.9 7.3 0.74 0.288 305.19 0.01 -34.0 7.1 5.2 0.43 0.156 305.14 0.061 -37.4 3.4 7.0 0.45 0.21W 5 305.82 -0.11 -42.7 3.0 6.0 0.60 0.2210 305.83 -0.19 -34.7 3.5 3.2 0.28 0.1211 305.89 0.02 -35.1 3.2 3.9 0.27 0.1612 305.69 0.04 -37.4 1.7 2.9 0.34 0.1813 305.77 -0.25 -32.3 4.1 3.9 0.34 0.1414 305.56 0.02 -40.4 3.9 3.8 0.34 0.1615 305.48 -0.10 -39.3 2.6 2.9 0.31 0.19 Table 3. NH (2,2) clump properties, showing central position, velocity, diameter (geometric average), peak main beam temperature and clump averaged mainbeam temperature, these values are taken from the spectra averaged over the NH (1,1) clump region defined by F ellwalker .Region Clump No Peak V LSR
FWHM ∆ V Diameter Peak T MB (2,2) Avg T MB (2,2)Galactic ( l ) Galactic ( b ) (km s − ) (km s − ) (pc) (K) (K)NE 1 305.23 0.28 -39.9 3.8 6.3 0.66 0.173 305.20 0.21 -41.6 6.2 4.3 0.32 0.14NW 4 305.36 0.19 -38.6 6.2 4.5 0.34 0.147 305.42 0.25 -39.2 2.9 < . < . < . < . < . < . Once the optical depth of the NH line is known the excitation tem-perature T ex of the NH (1,1) inversion can be derived using Eq. 2in Ho & Townes (1983):T B ( ν ) = T T / T ex − − T / . − ! (1 − e ( − τν ) ) (2)where T = h ν/ k , T MB is the main beam telescope brightness tem-perature and T ex is the source excitation temperature. We assumea background temperature of 2.7 K. This approach yields very lowexcitation temperatures T ex ∼ ex introduces significant error dueto being so close to the background temperature and leads to lowerlimits for the mass of the detected clumps. We therefore assume amore representative excitation temperature of T ex =
10 K for therest of our analysis (Harju et al. 1993). We can make an estimateof the excitation by assuming a beam filling factor, it is likely thatthe beam filling factor is < ∼ . ex =
10 K.
The column density, N u , of the molecules in the upper state of thetransition u → l can be expressed as a function of the integratedoptical depth of the line using Eq. (3) from Harju et al. (1993):N u = ǫ π | µ ul | F(T ex ) Z τ ∆ V (3)where | µ ul | is the transition dipole moment (in Cm) and the functionF T ex is defined by: F(T ex ) = T / T ex − T = h ν / k , the integration of τ is with respect to velocity.The application of the formula above depends on whether we havebeen able to determine τ , e.g., from a hyperfine fit as in the case ofthe (1,1) transition. Z τ dv = √ π √ ln ∆ V τ (5)where ∆ V is the FWHM of the line profile. For the inversion transi-tions of NH (J , K) the squares of the transition dipole moments canbe written as: c (cid:13) , 1–14 L.Hindson
Table 4. NH (3,3) clump properties, showing central position, velocity, diameter (geometric average), peak main beam temperature and clump averaged mainbeam temperature. Clump values are derived from the NH (3,3) source averaged spectra Gaussian fitRegion Clump No Peak V LSR
FWHM ∆ V Diameter Peak T MB (3,3) Avg T MB (3,3)Galactic ( l ) Galactic ( b ) (km s − ) (km s − ) (pc) (K) (K)NE 1 305.22 0.28 -40.2 4.8 5.7 0.42 0.203 305.20 0.21 -42.0 7.6 3.8 0.36 0.20NW 4 305.35 0.20 -38.6 7.1 4.9 0.36 0.20SE 2 305.25 -0.02 -32.2 5.9 4.4 0.42 0.158 305.19 0.00 -33.6 10.0 3.4 0.23 0.08W 5 305.81 -0.1 -41.9 1.4 < . Figure 4.
Source averaged spectra of the 6 NH (3,3) clumps towards G305. Clumps have been integrated spatially over the NH (3,3) emission area, spectrahave been Hanning smoothed to provide a sensitivity of ∼ .
01 K per ∼ . − channel. Gaussian fitting is applied, shown as red overplots. Table 5. H O maser positions, peak velocity, velocity range and peak flux. Asterisks above the maser number show new detections over those identified byWalsh et al (2008). Previously detected H O masers are detailed in Table 2 in Walsh et al (2008)Region Maser Peak V
LSR
Velocity PeakNumber Galactic Galactic Peak Range Flux( l ) ( b ) (km s − ) (km s − ) (Jy)NE 1 ∗ ∗ ∗ ∗ ∗ ∗ (cid:13) , 1–14 he G305 star forming complex: Wide-Area molecular mapping of NH and H O masers | µ ul | = K J(J + µ (6)The column density N u refers to the upper transition level, we makeuse of the Boltzmann equation with the assumed T ex to estimate thetotal column density N(1 ,
1) assuming that both levels are evenlypopulated: N(1 , = N u + N l = N u (1 + e hv / kT ex ) (7)the metastable inversion transitions, i.e., J = K gives the most reli-able estimate of the rotational temperature T . Using the (1,1) and(2,2) excitation levels, we make use of the rotational temperatureequation from Wilson et al. (1993):T = . .
57 (T / T )] (8)where T and T is the main beam brightness temperatures of theNH (1,1) and (2,2) transition. We do not detect NH (2,2) emissionin clumps 10, 11 and 15, and therefore we are unable to calculatethe rotational temperature for these clumps. However, we are ableto calculate a range of values for each parameter of interest such asH column and number densities and clump masses, using a rangeof rotational temperatures typical of similar clumps (10 - 40 K; e.g.,Wu et al. 2006, Tieftrunk et al. 1998). This range of temperaturescorresponds to at most a factor of two di ff erence in column den-sity and so will not a ff ect our results considerably. In order to ob-tain H column densities we have assumed an ammonia fractionalabundance of 3 × − (Wu et al. 2006; Harju et al. 1993). The H densities and clump masses have been calculated using the clumparea defined by the F ellwalker algorithm. We have made use ofthe metastable transitions and this may not resemble a true Boltz-mann distribution, nevertheless, T is a good indicator of the gastemperature T kin for temperatures less than 20 K, although T un-derestimates T kin for temperature over 20 K (see e.g. Walmsley &Ungerechts 1983; Danby et al. 1988). We are able to calculate thekinetic temperature using:T kin = T × + T kin . ! × ln (cid:16) + . × e( − . / T kin ) (cid:17) (9)We estimate the total column density of NH within the clumpsusing the following equation assuming that only metastable levelsare populated:N(NH ) = N(1 ,
13 e . / T + +
53 e − . / T +
143 e − . / T ! (10)We can test the stability of the cores against collapse by calculatingthe virial mass, which we derive using the standard equation (e.g.,Evans 1999): M vir ≃ core h ∆ V i (11)where R core is the ammonia clump radius (geometric average di-mensions given in Table 2) in parsecs (assuming distance of 4 kpc)and ∆ V is the FWHM line width. The resulting parameters for eachclump are presented in Table 6. The spatial distribution and morphology of the detected ammoniaclumps are given in section 3.1, here we describe the basic param-eters. We find NH column densities on the order of 10 cm − thehighest values are found closest to the PDR boundary. Unfortu-nately we are unable to derive the gas density directly from ourmeasurements, as the derivation of the line excitation temperaturewhich is sensitive to density, requires an assumption of the un-known beam filling factor. We find kinetic temperatures between21 - 31 K with an average of 25 K, this value is above the High(22.4 K) and Low (21.7 K) groups of Molinari et al. (1996) andsimilar to the values found in a survey of ammonia clumps in theOrion and Cepheus clouds by Harju et al. (1993). An importantcaveat is that these regions are at di ff erent distances to G305 andobserved with higher resolution and depth. We assume that our de-rived column densities are a lower limit due to beam dilution, andthe unknown beam filling factor. The NH (1,1) and (2,2) emissionappears to be biased to cool gas which is extended over the size ofthe Mopra beam and does not reflect the physical conditions of theunresolved dense cores and filaments we expect to find at sub arcminute scales. The presence of hotter gas is indicated by (3,3) emis-sion and so clumps 1 - 5 and 8 are likely to be more evolved thanother, cooler clumps. The clumps that exhibit (3,3) emission arefound to correlate well with PAH emission, with the strongest andtherefore hottest clumps found close to the central cavity boundary.Using an assumed NH to H abundance ratio of 3 × − wecalculate H column densities of > cm − . We assume that theammonia abundance fraction is constant across the entire region,which is not likely to be the case, however with our current lowresolution observations our chosen value gives an appropriate es-timate. The abundance ratio of NH to H has been found to varyfrom 1 × − to 1 × − according to the study of Orion clumps(Harju et al. 1993), thus our measurements of the H column den-sity could be an order of magnitude lower than shown in Table 6and so our results should be taken as an upper limit.The sizes of the NH clumps have been derived using the ge-ometric mean of the linear x and y distances, they range from 2.6to 10.1 pc. The sizes of the (2,2) emission are smaller than those ofthe (1,1) emission in all cases and in the same way (3,3) emissionis smaller than (2,2) emission. The size of these clumps is signif-icantly greater than those detected in a survey carried out by Wuet al. (2006) (0.4-3.1 pc) and similar observation in the Orion andCepheus cloud carried out by Harju et al. (1993) (mean 0.3 pc).However the same di ff ering resolution and sensitivity caveat ap-plies, and due to our large beam we are only able to resolve NH structure of > . ∼ × M ⊙ . It is clear thatthere is su ffi cient dense molecular gas to form multiple massivestellar clusters within the detected clumps, and this is likely to oc-cur around the central region. The virial mass is found to be signifi-cantly less than the calculated mass for all clumps, we would there-fore assume these clumps are gravitationally bound and given thelarge di ff erence between their actual masses and their virial massesit is likely they are in a state of gravitational collapse, however thismay not be an appropriate comparison. The mass of the detectedclumps in G305 is higher than expected for isolated clumps thattend to have mass on the order of a few tens to hundreds of so-lar masses. The simplest explanation for this is that our detectedclumps are in fact comprised of a number of unresolved clumps. c (cid:13) , 1–14 L.Hindson
Table 6. NH Clump parameters including; main line and integrated optical depth, rotational temperature, kinetic temperature H column density, derivedmass and virial mass. Region Clump No τ (1 , , m ) R τ (1 , T T kin N(H ) Mass Virial mass(K) (K) 10 cm − (M ⊙ ) 10 (M ⊙ )NE 1 0.79 6.5 21.3 ± ± ± ±
25 313 1.13 11.3 22.3 ± ± ± ±
22 24NW 4 1.32 15.8 23.5 ± ± ± ±
59 517 0.86 5.7 20.1 ± ± ± ±
20 149 0.29 1.9 19.5 ± ± ± ± ± ± ± ±
27 378 1.03 15.0 24.4 ± ± ± ±
46 546 0.47 3.5 20.6 ± ± ± ±
16 17W 5 1.06 6.8 18.7 ± ± ± ±
25 1110 1.10 8.5 10 - 40 10 - 74 9 - 16 21 - 37 811 1.46 10.2 10 - 40 10 - 74 11 - 20 27 - 48 912 0.36 1.8 18.3 ± ± ± ± ± ± ± ±
27 1414 1.23 10.7 21.2 ± ± ± ±
15 1215 1.51 8.5 10 - 40 10 - 74 9 - 17 11 - 20 4
In previous higher resolution studies of NH clumps Harju et al.(1993) the average size of the clumps is significantly lower thanour resolving limit of 2.6 pc, therefore it is no surprise that the av-erage mass of these smaller clumps in the Cepheus, Orion and Tau-rus clouds are found to be 15 or 140, 15 and 4 M ⊙ significantly lessthan our estimates. Given these average clump masses and sizes andcomparing to our derived values it is apparent that the clumps thatwe have detected in G305 most likely consist of a significant num-ber of unresolved clumps and therefore may be better described as“clouds” rather than “clumps”.As mentioned above we are unable to directly estimate the gasdensity however we can make estimate of the density of the ammo-nia clumps by assuming the clumps are spherically symmetric andour estimation of the excitation temperature is correct. This resultsin a H number density of ∼ × cm − for the largest clumps inthe NE, NE and SE region, significantly lower than the critical den-sity of ammonia of 10 − cm − , we find that the smallest clumpsdetected in the western region have the highest H number densitiesof ∼ . This is further evidence that the large clumps we find inthe NE, NW and SE region are in fact comprised of unresolved substructure which are smoothed by the beam into the single clumpswe detect. It appears that the mass of several unresolved clumpsis spread out over the detected area by the ∼ ′ beam, this drivesdown the estimated density for the largest clumps but has less of ane ff ect on the smaller clumps which are likely to consist of at mosta few cores. This means that the mass we estimate for the largestclumps is most likely an overestimate as the area of emission ap-pears greater than it physically is.There is compelling evidence of substructure within the de-tected clumps, this coupled with the unknown abundance ratio, con-tribute the main sources of error in our calculations. Thus our re-sults should be taken as an upper limit due to the unknown beamfilling factor and may be up to an order of magnitude too high dueto the unknown abundance ratio. A key goal of our observations was to study the relationship be-tween the dense molecular gas in G305, as traced by NH and thedistribution of known sites of recent or ongoing star formation. Tolocate areas of ongoing star formation we make use of the RedSource MSX Survey (RMS) (Urquhart et al. 2008; Hoare et al. 2005) and 6.7 GHz methanol masers drawn from Caswell (2009)and the Methanol Multi Beam (MMB) survey (Green et al. 2009),as well as our H O maser results. In Figures 5 - 8, we present apanoramic view of the ongoing star formation towards G305.The RMS survey aims to locate massive young stellar objects(MYSOs), and so highlight regions of recent star formation. Colourselection in the MSX point source catalogue is used to identify ob-jects with colours the same as well known MYSOs. Many objectshave very similar colours to the very red MYSOs, this includesUC HII regions, compact planetary nebula, low mass YSOs andevolved stars. We have selected objects that are either MYSOs,YSOs, HII or UC HII detections. We do not distinguish betweenthe di ff erent types of selected RMS sources as we will be usingthem only as a general indicator of recent star formation and haveinsu ffi cient numbers to carry out any statistical analysis. We find 26RMS sources in the direction of G305 and overplot these with redboxes in Figures 5 - 8.The Methanol Multi Beam survey Green et al. (2009) is thewidest area survey yet carried out for 6.7 GHz methanol masers,comprising an initial maser search with Parkes and sub-arcsecondpositional followup by the Australia Telescope Compact Arrayand MERLIN. Methanol masers have been recognised as one ofthe brightest signposts to the formation of massive young stars,(Menten 1991) and unlike other strong masers (OH, H O and SiO)have thus far only found close to high mass star forming regions(Minier et al. 2003). The MMB survey has not yet published theircatalogue for the l =
305 region and so we use preliminary maserpositions kindly provided by the MMB team (Fuller & Caswell,priv. comm.) and masers previously documented by Caswell(2009) we identify 13 6.7 GHz methanol masers in the direction ofG305 and overplot their locations with cyan circles in Figures 5 –8. Below we briefly describe the locations of star formation tracerswith respect to the ammonia emission.
NW Region:
In this region (Fig. 5) we note the highest numberof RMS sources (nine) of which six are found within and aroundthe periphery of clump 4. Two RMS sources are seen to the westof clump 7 and one RMS source is seen within clump 9. We findtwo methanol masers associated with clump 4, one is seen justoutside the leading edge of the ammonia emission facing WR48 a,the other methanol maser is seen towards the centre of the clump.Three H O masers are detected, two of which reside within clump c (cid:13) , 1–14 he G305 star forming complex: Wide-Area molecular mapping of NH and H O masers Figure 5. NH (1,1) and (2,2) emission contoured over greyscale GLIMPSE5.4 µ m image in black and red respectively clump numbers are shownwithin circles. Detected H O masers are shown by blue crosses, MMBsources by yellow circles and RMS sources by red boxes. Contours begin at0.15 K and increment by 0.1 K. O masers and methanolmasers all within the bounds of clump 4. as mentioned in section3.1 unlike other regions we see that the clump is coincident withstrong PAH emission towards the centre of clump. High resolutionobservations are required to resolve the internal structure andkinematics but it is likely that there are a number of individualdense cores at di ff erent velocities contributing to the large FWHMline width of clump 4. The RMS and MMB sources appear to liearound the core of the detected clumps and are predominantlyfound on the western side of the clump. The H O masers again arelocated close to the clump cores. Nine 1.2 mm cores are detectedby Hill et al. (2005) six are found within clump 4, three are foundin clump 9. Clump 1 in the NE region is the largest clump ingeometric size but clump 4 is the most massive
NE Region:
In the bounds of this region (Fig. 6) we find four RMSsources, three of which are found along the western edge of theNH emission, one is located close the centre of clump 3. Thereare three methanol masers, two of which (G305A and B) are foundclose the centre of clump 3 and one on the western edge of clump 1coincident with an RMS source and strong PAH emission. The twomethanol masers G305A and B are found to be coincident withCH OH emission, which surrounds the two masers and is elon-gated to the north. We note the position of a HII region (see Clark& Porter (2004)) to the west of the detected NH clumps, seen asa ring in the 5.4 µ m PAH emission, this coincides with the lowdensity region separating clump 1 and 3. Star formation tracers are Figure 6. NH (1,1) and (2,2) emission contoured over greyscale GLIMPSE5.4 µ m image in black and red respectively, clump numbers are shownwithin circles. Detected H O masers are shown by blue crosses, MMBsources by yellow circles and RMS sources by red boxes. Contours begin at0.15 K and increment by 0.1 K. clustered on the western side of the region facing this HII bubble,there are no signs of star formation to the east of the clumps. Wedetect two H O masers towards the centre of two clumps. It appearsthat RMS and MMB star formation tracers are confined within thelower density NH facing the ionising HII region to the west, theH O masers however are found embedded towards the centre of thetwo clumps.This region has been studied in some detail (Walsh & Burton2006; Hill et al. 2005; Longmore et al. 2007). Hill et al. (2005)detect nine 1.2 mm cores within this region concentrated withinclump 1 but o ff set from the peak to the west. Observations inthe mid infrared Walsh et al. (2001), near infrared Walsh et al.(1999) and mm line observations Walsh & Burton (2006) ( CO,HCO + , N H + , CH CN and CH OH) were carried out towardsthe two methanol maser sites designated G305A and G305B(G305.21 + + c (cid:13) , 1–14 L.Hindson
Figure 7. NH (1,1) and (2,2) emission contoured over greyscale GLIMPSE5.4 µ m image in black and red respectively, clump numbers are shownwithin circles. Detected H O masers are shown by blue crosses, MMBsources by yellow circles and RMS sources by red boxes. Contours begin at0.15 K and increment by 0.1 K. evolution, but probably older than G305A. Longmore detect twelveembedded sources with significant IR excess in this region locatedaround the southern edge of clump 3 within the region of lowemission separating clump 1 and 3. An embedded source is foundin the centre of the HII region highlighted by 5.4 µ m PAH emissionto the west of clump 1 and 3. Longmore et al. (2007) concludethat the expanding bubble may be responsible for triggering a thirdgeneration of star formation in clump 1 and 3. SE Region:
This region is shown in Fig. 7, coincident with thisstrip of NH emission we find five RMS sources, four H O masersand two methanol masers. We also note that in the central cavityof G305 there is an RMS source and H O maser coincident withstrong PAH emission but devoid of any NH emission. The FourH O masers within the NH emission are all in close proximityto the cores of the NH emission. A striking feature of the regionis that four of the five RMS sources and both methanol masersare clustered within the low density region of clump 8 and extendtowards the core of the clump, this area is also the location of a HIIregion documented in Clark & Porter (2004), this is clearly a veryactive area of ongoing star formation. All of these star formationtracers are found on the inward side of the ammonia emissionclose to the boundary between the central cavity, again facing theionising central sources. Hill et al. (2005) detect seven 1.2 mmcores, four of these are found in the intense star forming regionbetween clumps 8 and 6. W Region:
This region, shown in Fig. 8 is distinctly di ff erentfrom the other regions we have discussed. We find a total of eightRMS sources, six MMB sources and six H O masers. Whereasthe previous regions all show strong NH emission around theperiphery of the central HII region, this region shows no significant Figure 8. NH (1,1) and (2,2) emission contoured over greyscale GLIMPSE5.4 µ m image in black and red respectively, clump numbers are shownwithin circles. Detected H O masers are shown by blue crosses, MMBsources by purple circles and RMS sources by red boxes. Contours begin at0.15 K and increment by 0.1 K. NH at this boundary. However, in the absence of any significantammonia emission at the boundary we do still see significant signsof star formation coincident with clump 14 close to the cavityboundary, with four RMS sources and a single MMB source.Further to the west of the cavity we see dispersed signs of ongoingstar formation the majority of which are associated with isolatedNH emission. Clump 5 is host to three H O masers, the mostdetections coincident with a clump in the complex. This region isalso host to the most methanol maser detections, 5 of which arefound on the edge of NH clumps. Hill et al. (2005) detect eleven1.2 mm cores, which are confined to clump 14 and clump 13.Unlike the other regions where 1.2 mm emission is correlated withstrong NH (1,1) and (2,2) emission, there is no 1.2 mm emissionassociated with the strongest ammonia emission in this region,in clump 5. It is clear that this region is host to star formation,however there is much less molecular material in this region andthe star formation appears much more dispersed.It is clear that all regions are host to significant signs of on-going star formation. Of all the detected clumps only clump 1 and3 have been studied in any detail by the authors (mentioned in theNE clump section). A striking feature of all these signs of star for-mation is that they are for the most part all coincident with densemolecular material and are located only on the side of the molecu-lar emission facing the PDR. In cases where we appear have a sideon line of site we see star formation only occurring in the low den-sity regions of the ammonia clumps with exception to H O masers,which appear closer to the dense clump cores. This is most apparentin the NE and S regions, where RMS and MMB detections are seenonly in the less dense gas towards the PDR, whereas H O masersare found deeply embedded in the NH emission. This may indicatean age spread in the star formation with H O masers highlighting ayounger age of star formation embedded towards the dense clumpcentres and older YSOs and visible HII regions found on the clumpperiphery in the lower density regions. Warmer clumps, highlightedby NH (3,3) emission, correlate well with regions of intense star c (cid:13) , 1–14 he G305 star forming complex: Wide-Area molecular mapping of NH and H O masers Table 7.
G305 Comparison, column 3 is molecular mass and column 5shows the number of massive OB stars in the complex. Parameters havebeen taken from a literature search, numbers in brackets correspond to thefollowing papers.Region Distance Mass NLyc Starskpc M ⊙ s − OBG305 3.5 - 4 (1) 10 (*) 145.5 (2) >
31 (1)W49A 10.2 - 12.6 (4) 10 (3) 171.7 (2) >
40 (4)NGC3603 5 - 8 (5) 10 (5) 187.7 (2) 20 (6)M17 1.3 - 1.9 (9) 10 (9) 53.7 (2) 16 (9)W3 main 1.9 - 2.1 (10) 10 (10) 8.8 (2) 12 (10)Rosette Nebula 1.4 - 1.6 (11) 10 (11) 9.8 (14) 30 (11)Westerlund 2 5 - 7 (12) 10 (12) 95.9 (2) >
12 (12)Carina 2.3 (13) 10 (13) 21.6 (2) 65 (13)(*) This work (1) Clark & Porter (2004) (2) Smith et al. (1978) (3) Smithet al. (2009) (4) de Pree et al. (1997) (5) N¨urnberger et al. (2002) (6) Mo ff at(1983) (7) Dougherty et al. (2009) (8) Luna et al. (2009) (9) Povich et al.(2009) (10) Tieftrunk et al. (1998) (11) Wang et al. (2008) (12) Dame(2007) (13) Smith & Brooks (2008) (14) Churchwell (1975) formation, indeed all clumps with (3,3) detections are either host toa large number of star formation tracers or strong PAH emission.In terms of the star formation distributed within G305 and itsrelation to the dense molecular material surrounding the centralcavity, we have expanded on the work of Clark & Porter (2004),who suggested that the morphology of the nebula is strongly sug-gestive of star formation on the periphery of the cavities (see Clark& Porter 2004 Section 4.3). We confirm that star formation is tak-ing place on the periphery of the central cavity created by Danks 1& 2, WR48 and also that a potentially younger generation of starformation is taking place around the younger HII regions locatedaround the central cavity, indicating multiple epochs of massivestar formation in G305. The dense molecular material surround-ing Danks 1 & 2 takes the form of several discrete and massiveclumps, with significant “gaps” (particularly towards the westernregion) highlighting areas where the ionising photons from thesemassive stellar clusters can escape. Thus the integrated radio fluxvery likely underestimates the true ionising photon flux from G305.We are still unable to ascertain if the photoionisation driven shocksare responsible for triggering the observed star formation, howeverwe do note a possible age spread in star formation tracers in the re-gions closest to the cavity. The western region may be more likelyto be spontaneous in nature star formation due to its dispersed mor-phology. To put G305 into context it is useful to compare the region to otherwell known GMCs and HII regions in the galaxy. We have per-formed a literature search and present parameters for comparisonin Table 7. It is apparent that G305 is one of the closest and mostmassive GMCs in the galaxy, with a molecular mass approximatelyequal to or exceeding that of other well studied regions. G305 issignificantly closer than the regions with similar masses detailed inTable 7, and regions that are found closer to the sun tend to havea lower molecular mass. With its numerous HII regions, the inte-grated Lyc photon count is on par with the most luminous HII re-gions in the galaxy. However, whereas many of these regions com-prise single HII regions, G305 is host to a number of individual HIIregions surrounding a central cavity. As proposed in Clark & Porter(2004) given the morphology of the G305 complex and in particu- lar the fact that su ffi cient natal material has been dispersed to allowDanks 1 and 2 to become optically visible, there is a likelihood ofsubstantial photon leakage and thus the massive stellar populationof G305 is likely to be an underestimate. This would make G305one of the closest examples of intense massive star formation inthe galaxy, comparable to W49A which lies at ∼ − We present radio observations of the ∼ . ◦ × ◦ region G305 inorder to uncover the dense molecular component of the region.This study includes observations of the NH (1,1), (2,2) and (3,3)spectral lines, as well as the 22 GHz H O maser line. To comple-ment our observations we have made use of mid-IR data from theGLIMPSE survey and traced signs of star formation using the RMSand MMB surveys. Combining these data sets has allowed us to ob-tain a panoramic view of ongoing star formation within G305 andits relation to the dense molecular material.We detect 15 NH (1,1) clumps, 12 NH (2,2), 7 NH (3,3)clumps and 16 H O masers. Our observations reveal that G305is comprised of several massive molecular clumps mostly foundtowards the northern and southern edges of the main cavity. Themolecular material towards the western region of G305 is more dis-persed in nature. Kinetic temperatures of the clumps ranges from ∼ −
31 K, the total mass of dense gas traced by NH is estimatedto be ∼ × M ⊙ with an average clump mass of ∼ × M ⊙ .We detect the majority ( ∼ ffi cient material to form a number of massive clusters and thatthere is a high degree of interaction between the massive stars andthe molecular material. It is possible that such interactions couldbe responsible for triggering the star formation seen around the pe-riphery of the NH clumps.We find 27 RMS sources and 13 MMB masers, as well as16 H O masers. With 56 star formation tracers G305 is obviouslya region of intense star formation. We are able to clearly identifythe PDR and find it correlates with the molecular gas traced byNH ˙Star formation appears to be taking place within and on theperiphery of the detected NH clumps, with >
80% of star forma-tion tracers found to be associated with NH emission. In almostall cases star formation is located on the side of the NH clumpthat faces the ionising radiation highlighted by the PDR, evidenceperhaps of triggered star formation.These low resolution observations are intended as a pathfinderfor future high resolution follow up, which will allow us to un-cover the sub parsec structure of the ammonia emission we predictis present, and thus allow us to comment on the e ff ects massive starformation is having upon the dense molecular material. ACKNOWLEDGMENTS
We would like the thank the Director and sta ff of the Paul WildObservatory for their assistance with a pleasant and productive ob-serving run. We would like to thank the referee, Michael Burton,for a thoroughly constructive and useful report. The Mopra tele-scope is part of the Australia Telescope which is funded by theCommonwealth of Australia for operation as a National Facilitymanaged by CSIRO. The University of New South Wales DigitalFilter Bank used for the observations was provided with supportfrom the Australian Research Council. This research has made use c (cid:13) , 1–14 L.Hindson of the NASA / IPAC Infrared Science Archive, which is operated bythe Jet Propulsion Laboratory, California Institute of Technology,under contract with the National Aeronautics and Space Adminis-tration.
REFERENCES
Benjamin R. A., Churchwell E., Babler B. L., Bania T. M.,Clemens D. P., Cohen M., Dickey J. M., Indebetouw R., Jack-son J. M., Kobulnicky H. A., Lazarian A., Marston A. P., MathisJ. S., Meade M. R., Seager S., Stolovy S. R., Watson C., Whit-ney, B. A., Wol ff , M. J., Wolfire, M. G., 2003, PASP, 115, 953Berry D. S., Reinhold K., Jenness T., Economou F., 2007, inR. A. Shaw, F. Hill, & D. J. Bell ed., Astronomical Data AnalysisSoftware and Systems XVI Vol. 376 of Astronomical Society ofthe Pacific Conference Series, CUPID: A Clump Identificationand Analysis Package. pp 425– + Breen S. L., Ellingsen S. P., Johnston-Hollitt M., Wotherspoon S.,Bains I., Burton M. G., Cunningham M., Lo N., Senkbeil C. E.,Wong T., 2007, MNRAS, 377, 491Caswell J. L., 2009, Publications of the Astronomical Society ofAustralia, 26, 454Churchwell E., 1975, in T. L. Wilson & D. Downes ed., H II re-gions and related topics Vol. 42 of Lecture Notes in Physics,Berlin Springer Verlag, Evolved H II regions. pp 245– + Churchwell E., Babler B. L., Meade M. R., Whitney B. A., Ben-jamin R., Indebetouw R., Cyganowski C., Robitaille T. P., PovichM., Watson C., Bracker S., 2009, pasp, 121, 213Clark J. S., Porter J. M., 2004, aap, 427, 839Dame T. M., 2007, ApJl, 665, L163Danby G., Flower D. R., Valiron P., Schilke P., Walmsley C. M.,1988, MNRAS, 235, 229de Pree C. G., Mehringer D. M., Goss W. M., 1997, ApJ, 482, 307Dougherty S. M., Clark J. S., Negueruela I., Johnson T., ChapmanJ. M., 2009, ArXiv e-printsElmegreen B. G., 2002, in Geisler D. P., Grebel E. K., Minniti D.,eds, Extragalactic Star Clusters Vol. 207 of IAU Symposium,Triggering the Formation of Young Clusters. pp 390– + Elmegreen B. G., Lada C. J., 1977, ApJ, 214, 725Furuya R. S., 2003, in J. M. De Buizer & N. S. van der Blieked., Galactic Star Formation Across the Stellar Mass SpectrumVol. 287 of Astronomical Society of the Pacific Conference Se-ries, H O Masers in Young Stellar Objects. pp 367–372Green J. A., Caswell J. L., Fuller G. A., Avison A., Breen S. L.,Brooks K., Burton M. G., Chrysostomou A., Cox J., DiamondP. J., Ellingsen S. P., Gray M. D., Hoare M. G., MashederM. R. W., McClure-Gri ffi ths N. M., Quinn L., Thompson M. A.,2009, MNRAS, 392, 783Habing H. J., Israel F. P., 1979, araa, 17, 345Harju J., Walmsley C. M., Wouterloot J. G. A., 1993, aaps, 98, 51Hill T., Burton M. G., Minier V., Thompson M. A., Walsh A. J.,Hunt-Cunningham M., Garay G., 2005, MNRAS, 363, 405Ho P. T. P., Townes C. H., 1983, araa, 21, 239Hoare M. G., Lumsden S. L., Oudmaijer R. D., Urquhart J. S.,Busfield A. L., Sheret T. L., Clarke A. J., Moore T. J. T., AllsoppJ., Burton M. G., Purcell C. R., Jiang Z., Wang M., 2005, inR. Cesaroni, M. Felli, E. Churchwell, & M. Walmsley ed., Mas-sive Star Birth: A Crossroads of Astrophysics Vol. 227 of IAUSymposium, The RMS survey: Massive young stars throughoutthe galaxy. pp 370–375Leger A., Puget J. L., 1984, aap, 137, L5 Longmore S. N., Maercker M., Ramstedt S., Burton M. G., 2007,MNRAS, 380, 1497Luna A., Mayya Y. D., Carrasco L., Rodr´ıguez-Merino L. H.,Bronfman L., 2009, in Revista Mexicana de Astronomia y As-trofisica Conference Series Vol. 37 of Revista Mexicana de As-tronomia y Astrofisica Conference Series, The large scale molec-ular environment towards Westerlund 1. pp 32–37Menten K., 1991, in A. D. Haschick & P. T. P. Ho ed., Atoms,Ions and Molecules: New Results in Spectral Line AstrophysicsVol. 16 of Astronomical Society of the Pacific ConferenceSeries, Methanol Masers and Submillimeter WavelengthWaterMasers in Star-Forming Regions. pp 119– + Minier V., Ellingsen S. P., Norris R. P., Booth R. S., 2003, aap,403, 1095Mo ff at A. F. J., 1983, aap, 124, 273Molinari S., Brand J., Cesaroni R., Palla F., 1996, aap, 308, 573N¨urnberger D. E. A., Bronfman L., Yorke H. W., Zinnecker H.,2002, aap, 394, 253Pineda, J. E., Rosolowsky, E. W., & Goodman, A. A., 2009, ApJl,699, L134Povich M. S., Churchwell E., Bieging J. H., Kang M., WhitneyB. A., Brogan C. L., Kulesa C. A., Cohen M., Babler B. L., Inde-betouw R., Meade M. R., Robitaille T. P., 2009, ApJ, 696, 1278Purcell C. R., 2006, PhD thesis, The University of New SouthWalesSmith L. F., Biermann P., Mezger P. G., 1978, aap, 66, 65Smith N., Brooks K. J., 2008, The Carina Nebula: A Laboratoryfor Feedback and Triggered Star Formation. pp 138– + Smith N., Whitney B. A., Conti P. S., de Pree C. G., Jackson J. M.,2009, MNRAS, 399, 952Tieftrunk A. R., Megeath S. T., Wilson T. L., Rayner J. T., 1998,aap, 336, 991Urquhart J. S., Hoare M. G., Lumsden S. L., Oudmaijer R. D.,Moore T. J. T., 2008, in H. Beuther, H. Linz, & T. Henning ed.,Massive Star Formation: Observations Confront Theory Vol. 387of Astronomical Society of the Pacific Conference Series, TheRMS Survey: A Galaxy-wide Sample of Massive Young StellarObjects. pp 381– + Urquhart J. S., White G. J., Pilbratt G. L., Fridlund C. V. M., 2003,aap, 409, 193Urquhart J. S., Hoare M. G., Purcell C. R., Brooks K. J., VoronkovM. A., Indermuehle B. T., Burton M. G., Tothill N. F. H., Ed-wards P. G., 2010, ArXiv e-printsWalmsley C. M., Ungerechts H., 1983, aap, 122, 164Walsh A. J., Bertoldi F., Burton M. G., Nikola T., 2001, MNRAS,326, 36Walsh A. J., Burton M. G., 2006, MNRAS, 365, 321Walsh A. J., Burton M. G., Hyland A. R., Robinson G., 1999,MNRAS, 309, 905Walsh A. J., Lo N., Burton M. G., White G. L., Purcell C. R.,Longmore S. N., Phillips C. J., Brooks K. J., 2008, Publicationsof the Astronomical Society of Australia, 25, 105Wang J., Townsley L. K., Feigelson E. D., Broos P. S., GetmanK. V., Rom´an-Z´u˜niga C. G., Lada E., 2008, ApJ, 675, 464Wilson T. L., Gaume R. A., Johnston K. J., 1993, ApJ, 402, 230Wu Y., Zhang Q., Yu W., Miller M., Mao R., Sun K., Wang Y.,2006, aap, 450, 607 c (cid:13)000