The Millimeter Astronomy Legacy Team 90 GHz (MALT90) Pilot Survey
Jonathan B. Foster, James M. Jackson, Elizabeth Barris, Kate Brooks, Maria Cunningham, Susanna C. Finn, Gary A. Fuller, Steve N. Longmore, Joshua L. Mascoop, Nicholas Peretto, Jill Rathborne, Patricio Sanhueza, Frédéric Schuller, Friedrich Wyrowski
aa r X i v : . [ a s t r o - ph . GA ] A ug Draft version April 17, 2018
Preprint typeset using L A TEX style emulateapj v. 11/10/09
THE MILLIMETER ASTRONOMY LEGACY TEAM 90 GHZ (MALT90) PILOT SURVEY
Jonathan B. Foster , James M. Jackson , Elizabeth Barris , Kate Brooks , Maria Cunningham , Susanna C.Finn , Gary A. Fuller , Steve N. Longmore , Joshua L. Mascoop , Nicolas Peretto , Jill Rathborne ,Patricio Sanhueza , Fr´ed´eric Schuller , Friedrich Wyrowski Draft version April 17, 2018
ABSTRACTWe describe a pilot survey conducted with the Mopra 22-m radio telescope in preparation for theMillimeter Astronomy Legacy Team Survey at 90 GHz (MALT90). We identified 182 candidate densemolecular clumps using six different selection criteria and mapped each source simultaneously in 16different lines near 90 GHz. We present a summary of the data and describe how the results of thepilot survey shaped the design of the larger MALT90 survey. We motivate our selection of targetsources for the main survey based on the pilot detection rates and demonstrate the value of mappingin multiple lines simultaneously at high spectral resolution.
Subject headings:
ISM: molecules — stars: massive — stars: formation — surveys INTRODUCTION
The goal of the Millimeter Astronomy Legacy TeamSurvey at 90 GHz (MALT90) is to characterize the physi-cal and chemical conditions of dense molecular clumps as-sociated with high-mass star formation over a wide rangeof evolutionary states. MALT90 will do this by takingadvantage of the newly upgraded Mopra Spectrometer(MOPS ) and the fast mapping capability of the Mopra22-m radio telescope . The survey will obtain molecu-lar line maps of 3000 candidate dense molecular clumps.The clumps will be selected so as to cover a broad rangeof evolutionary states, from pre-stellar clumps to accret-ing high-mass protostars and on to H II regions. Thesurvey will be conducted at 90 GHz because this fre-quency regime contains numerous molecular lines whichhave typical critical densities for collisional excitation of & cm − and are therefore excellent tracers of densegas. Such data will allow us to study the Galactic dis-tribution of these clumps, their physical properties, andtheir chemical variation and evolution; this basic infor-mation is necessary to constrain theories of high-massstar formation. In addition, MALT90 will provide a valu-able database of dense molecular clumps associated withhigh-mass star formation for future ALMA observations. [email protected] Institute for Astrophysical Research, Boston University,Boston, MA 02215, USA CSIRO Astronomy and Space Science, PO Box 76, Epping,NSW 1710, Australia School of Physics, The University of New South Wales,Sydney 2052, Australia Jodrell Bank Centre for Astrophysics, School of Physics andAstronomy,University of Manchester, Manchester M13 9PL, UK ESO Headquarters, Karl-Schwarzschild-Str. 2, 85748Garching bei M¨unchen, Germany Harvard-Smithsonian Center for Astrophysics, MS 42, 60Garden Street,Cambridge, MA 02138, USA Laboratoire AIM, CEA/DSM-CNRS-Universitˆe ParisDiderot, IRFU/Service d’Astrophysique, C.E. Saclay, Orme deMerisiers, 91191 Gif-sur-Yvette, France Departamento de Astronom´ıa, Universidad de Chile, Chile Max-Plank-Institut f¨ur Radioastronomie, Auf dem H¨ugel69, D-53121 Bonn, Germany
MALT90 will map roughly 3,000 dense molecularclumps, providing an order of magnitude more sourcesthan previous comparable surveys (e.g., Shirley et al.2003; Pirogov et al. 2003; Gibson et al. 2009; Wu et al.2010). A large number of sources will allow us to dividethe sample into sub-samples (based on mass, evolution-ary phase, etc.) yet retain a sufficient number of sourcesin each sub-sample for statistical analysis. Because densemolecular gas occupies only a small solid angle of theGalactic plane and molecular emission at 90 GHz is rel-atively faint, a blind fully-sampled 90 GHz survey of asignificant portion of the Galactic plane is impractical.Instead, we must choose targets based on other methodsfor identifying dense molecular clumps. The main pur-pose of the MALT90 pilot survey described herein is tochoose the best method for identifying dense molecularclumps, with the twin aims of having a high percentageof detections within our sensitivity limits and covering abroad range of evolutionary states.Throughout this paper, we will use the term “densemolecular clump” to refer to our sources. Thechoice of “clump” follows the naming system usedby Williams et al. (2000) and Bergin & Tafalla (2007)which distinguishes between molecular clouds, clumps,and cores. In this scheme, clumps are coherent regionsin position-velocity space with typical masses of 50-500 M ⊙ , typical sizes of 0.3-3 pc and typical mean densitiesof 10 -10 cm − which may contain additional substruc-tures called cores which give rise to individual stars orstellar systems. Our goal in MALT90 is to identify andmap the clumps that give rise to a cluster of stars con-taining one or more high-mass stars.This paper will focus predominantly on the technicalvalidation of the survey and explain the design choicesmotivated by the pilot survey. Data from the pilot sur-vey will be combined with the full MALT90 survey data(where the diverse selection criteria in the pilot surveyare not detrimental to statistical analysis) for the spe-cific scientific projects in MALT90. These analyses willappear in future papers. MOPS was funded in part through a grant instigated by theUniversity of New South Wales (UNSW) under the Australian Re-
Foster
Figure 1.
A proposed evolutionary sequence for high-mass starformation, running from pre-stellar clumps [left], to clumps withembedded accreting protostars [middle], to H II regions [right].These three-color images show IRAC band 1 (3.6 µ m) as blue,IRAC band 4 (8.0 µ m) as green, and MIPS 24 µ m as red. Thesecategories correspond to different Spitzer emission morphologiesand motivated the three different catalogs based on GLIMPSE andMIPSGAL data. Specifically, clumps which are dark at GLIMPSEwavelengths (3.6 - 8 µ m) are identified as pre-stellar, clumps witha MIPSGAL 24 µ m point source are identified as protostellar, andclumps with extended 8 µ m flux are identified as H II regions. TARGET SELECTION
We used six different input catalogues for selectingsources; three were lists that we produced for our pi-lot survey and three were based on pre-existing catalogs.From these lists we chose 20 - 40 sources near integerGalactic longitudes covering the range of longitudes ac-cessible from Mopra. This assured a random selection ofsources from the three pre-compiled lists and allowed usto focus on a limited portion of the sky when developingour own lists, while still covering a broad range of Galac-tic longitudes. The six selection criteria are summarizedin Table 1.The first three lists were produced for our pilot survey.For these lists we chose sources based on the mid-infaredmorphology of candidate dense molecular clumps as re-vealed by two
Spitzer
Space Telescope Legacy surveys:the Galactic Legacy Infrared Mid-Plane Survey Extraor-dinaire (GLIMPSE; Benjamin et al. 2003) and the 24and 70 Micron Survey of the Inner Galactic Disk withMIPS (MIPSGAL; Carey et al. 2009). GLIMPSE coversthe Galactic plane in the Infrared Array Camera (IRAC;Fazio et al. 2004) bands from 3.6 to 8.0 µ m, while MIPS-GAL covers much the same area at 24 and 70 µ m usingthe Multiband Infrared Photometer for Spitzer (MIPS;Rieke et al. 2004).We examined the GLIMPSE and MIPSGAL mosaicsusing three criteria designed to select sources in distinctevolutionary states as shown in Figure 1. A preliminaryversion of the Peretto & Fuller (2009) catalog of InfraredDark Clouds (IRDCs) was used to identify 8 µ m extinc-tion features near integer Galactic longitudes. This cat-alog was then trimmed to remove any sources which con-tained a 24 µ m point source (since this is most likely aprotostar). We refer to this the GLIMPSE-Dark cata-log, and these sources should correspond to the earliestphase of high-mass star formation. We then examinedthe Spitzer mosaics near integer Galactic longitudes byhand to choose candidate dense molecular clumps con-taining either 24 µ m point sources or bright extended8 µ m emission. If a clump of 8 µ m dark extinction was co- search Council Grants scheme for Linkage, Infrastructure, Equip-ment and Facilities (LIEF), and in part by CSIRO Astronomy andSpace Science. The Mopra radio telescope is part of the Australia TelescopeNational Facility which is funded by the Commonwealth of Aus-tralia for operation as a National Facility managed by CSIRO. incident with a 24 µ m point source, we assigned it to theMIPSGAL catalog and classified this candidate clumpas protostellar. In the case of bright extended emissionat 8 µ m we assigned the source to the GLIMPSE-Brightcatalog and classified it as an H II region.The correspondence between the appearance of asource in the Spitzer surveys and its evolutionary stateis clearly imperfect. The projection of unrelated objectsalong a given line of sight, inhomogeneities in the diffuse8 µ m emission, and sensitivity limits (e.g., our abilityto detect 24 µ m point sources will depend on intrinsicluminosity and distance) are three possible sources ofmisidentification. However, this system provides a quickand uniform way to make an initial assessment of a can-didate dense molecular clump’s evolutionary state.The other three lists in our pilot survey were producedusing pre-existing catalogs. The first came from the H OSouthern Galactic Plane Survey (HOPS; Walsh et al.2008) which used Mopra at 1.2 cm to map the Galac-tic plane from − ◦ < l < ◦ in NH and H O. Be-cause the HOPS NH (1,1) and (2,2) lines have a sim-ilar critical density (n ∼ cm − ) as the MALT9090 GHz lines (n ∼ cm − ), bright NH sources inHOPS are likely to be detected by Mopra in the 90GHz lines. The last two catalogs came from mm/submmcontinuum surveys which reveal the location of regionswith high dust column density, typically corresponding todense molecular clumps. The Beltr´an et al. (2006) sur-vey at 1.2 mm made maps around Infrared AstronomicalSatellite (IRAS) point sources using the Swedish-ESOSubMillimeter Telescope (SEST) and the SEST ImagingBolometer Array (SIMBA). Because the Beltr´an et al.(2006) maps were made toward IRAS point sources, thesesources are likely to contain a protostar or H II region,which gives rise to the IRAS emission; we shall refer tothis catalog as the IRAS catalog for convenience. Fi-nally, the APEX (Atacama Pathfinder Experiment) Tele-scope Large Area Survey of the Galaxy (ATLASGAL;Schuller et al. 2009) is a survey of the Galactic plane( ± ◦ in longitude over ± ◦ in latitude and -80 ◦ ≤ l ≤ -60 ◦ with -2 ◦ ≤ b ≤ ◦ ) at 870 µ m. From a pre-liminary compact source catalog we chose ATLASGALsources with peak fluxes above 2 Jy/beam closest to in-teger Galactic longitudes.In two cases (G336.994 − − Spitzer -identified sources and theBeltr´an et al. (2006) catalog of IRAS sources), and twoof which (ATLASGAL and HOPS) we expected to beless biased with respect to evolutionary status. For bothATLASGAL and HOPS, the most luminous sources willtend to be the hottest, more evolved sources. To selecta broad range of evolutionary states from these surveysit is necessary to include some additional information aswe discuss in § DATA
We carried out observations for the MALT90 pilot sur-vey in the austral winter of 2009 from June 15-24. TheALT90 Pilot Survey 3
Table 1
MALT90 Pilot Survey Input CatalogsData Criterion Object Identified ShorthandGLIMPSE 3.6 to 8 µ m Dark Extinction Pre-stellar (IRDC) Clump GLIMPSE-DarkGLIMPSE 8 µ m Extended Emission H II Region GLIMPSE-BrightMIPSGAL 24 µ m Point Source Accreting Protostar MIPSGALHOPS a Source Dense Clump HOPSIRAS + 1.2 mm emission b IRAS + mm Continuum Star-forming Dense Clump IRASATLASGAL c µ m Compact Continuum Dense Clump ATLASGAL a Walsh et al. (2008) b Beltr´an et al. (2006) c Schuller et al. (2009)
On-The-Fly (OTF) mapping mode of Mopra was used.Maps were made with the beam center running on a 3 ′ .4 x3 ′ .4 grid. At typical distances to high-mass star-formingregions (several kpc) this map size is sufficient to coverthe expected spatial extent of a few parsecs for our densemolecular clumps. The scan rate was 3.92 ′′ per second.The map is made with 12 ′′ spacing between the rows,giving 17 rows per map. Since the Mopra beam at 90GHz is 36 ′′ , this row spacing provides redundancy inthe map. OFF positions were chosen at ± ′′ . Typical system temperatures (T sys ) were150 - 250 K and were measured by paddle scans every15 minutes. Weather conditions were variable. Sourcesobserved under poor system temperatures (T sys > sys is presented here.The full 8 GHz bandwidth of MOPS was split into 16zoom bands of 138 MHz each providing a velocity res-olution of ∼ − in each band, easily sufficientto resolve line emission in a high-mass star-forming re-gion. The central frequencies are shown in Table 2, alongwith the line targeted at that frequency and what infor-mation that line primarily provides. The strongest lineswere N H + (1-0), HNC(1-0), HCO + (1-0), and HCN(1-0).These lines are all good tracers of dense gas, but provideslightly different information. N H + is more resistantto freeze-out on grains than the carbon-bearing species(Bergin et al. 2001). HNC is particularly prevalent incold gas (Hirota et al. 1998). HCO + often shows infallsignatures and outflow wings (e.g., Rawlings et al. 2004;Fuller et al. 2005). These strong lines can all be opticallythick. Two isotopologues, H CO + (1-0) and H CN (1-0) were also observed and provide optical depth and lineprofile information. CS (2-1) is another optically thincolumn density tracer by virtue of its low abundance. Wealso include C S (2-1) but this molecule is too rare tobe detected. A number of lines were chosen as tracers of hot corechemistry: CH CN ( J K = 5 − ), HC N ( J = 10 − CCN ( J = 10 − , F = 9 − J K a ,K b =4 , − , ), HNCO ( J K a ,K b = 4 , − , ) (Brown et al.1988). These carbon-bearing species are typically onlyseen in the hot cores around high-mass protostars oncemolecules have been liberated off dust grains by radia-tion or shocks. Three more lines trace particular envi-ronments: the recombination line H41 α traces ionizedgas (Shukla et al. 2004); SiO (2-1) is seen when SiO isformed from shocked dust grains, typically in outflows(Schilke et al. 1997); C H is produced in photodissocia-tion regions (e.g., Lo et al. 2009; Gerin et al. 2011), the N = 1 − , J = 3 / − / , F = 2 − H lines in this spectral win-dow. Henceforth we will refer to these line transitions bythe molecule name where this usage is unambiguous (i.e.HCO + instead of HCO + (1-0)).The maps were reduced using the Livedata and
Gridzilla packages . Livedata performs bandpasscalibration using reference OFF scans and fits a 2nd or-der polynomial to the baseline.
Gridzilla uses this out-put to construct a uniformly gridded cube. Both polar-izations were averaged together. A top-hat smoothingkernel with radius of 30 ′′ was used to determine whichspectra contribute signal to a pixel in the output map,and spectra were weighted by the system temperature.This choice of parameters produces an effective beamsize of FWHM = 72 ′′ . The final cube is over-sampled inspatial frequency (9 ′′ pixels) and is 4 ′ .6 x 4 ′ .6 with theedges having significantly lower coverage, i.e. integrationtime. The data are presented on the antenna tempera-ture scale (T ∗ A ). The beam efficiency of Mopra is between0.49 at 86 GHz and 0.42 at 115 GHz (Ladd et al. 2005)for converting T ∗ A into main-beam brightness tempera-ture (T mb ). All the data are publicly available from theMALT90 website . ANALYSIS
The MALT90 pilot survey data were used to test someof the analysis tools in development for the full survey.Three different methods were used to assess detectionstatistics: (1) “by-hand” examination, (2) automatedGaussian fitting based on the HOPS (Walsh et al. 2008)pipeline and (3) moment maps. Although Gaussian fit-ting is critical for certain measurements (particularly forlines with hyperfine component), moment maps are a fast http://malt90.bu.edu Foster
Table 2
MALT90 Pilot Survey LinesIF Species Transition ν (GHz) Primary Information Provided1 N H + J = 1 − CS J = 2 − α CN J K = 5 − N J = 10 − C S J = 2 − J = 1 − CCN J = 10 − , F = 9 − + J = 1 − J = 1 − J K a ,K b = 4 , − , J K a ,K b = 4 , − , H N = 1 − , J = 3 / − / , F = 2 − J = 2 − CO + J = 1 − CN J = 1 − Note . — Frequencies listed above are the rest frequencies used in converting to velocity scale (FITS keyword RESTFREQ). and relatively robust way to measure basic line proper-ties and this study will focus only on properties well-measured by moment analysis. The results of the “by-hand” examination were used to select the parametersused in generating moment maps.As a first step to making moment maps, we calculatedan error map for each spatial pixel by computing thestandard deviation of the spectra at that position using3 σ iterative rejection. In this way, we remove strong linefeatures from the calculation and derive an estimate ofthe per-channel noise in the spectrum at each point inthe map. Our on-the-fly maps have less integration timeat the edges, so a spatial error map is required in orderto properly assess features near the noisy edges of themap.Zeroth (M ; integrated intensity), first (M ; central ve-locity), and second (M ; line-width) moment maps weremade according to M = Z I ( ν ) dν (1) M = 1 M Z I ( ν ) ν dν (2) M = s M Z I ( ν )( ν − M ) dν (3)where I ( ν ) is the intensity at a given frequency, ν . Theerror on the zeroth moment is simply σ M = σ √ n (4)where σ is the per-channel noise in the spectrum as calcu-lated for our error map and n is the number of spectralchannels used. Errors on the first and second moment( σ M and σ M ) are calculated from propagation of un-certainty on the formulae for M and M above, but areomitted for space.The main choice in making moment maps lies in iden-tifying the region of the cube to use. For the pilot sur-vey automatic line detection was hindered by baselineripples (particularly in worse weather) and noisy edgeson the bandpasses. Improvement to the data processing pipeline are expected to mitigate baseline ripples for thefull MALT90 survey and allow for automatic detection oflines, but these were not available for processing the pilotdata. Therefore we use hand-identified velocities for eachsource to make moment maps in fixed-width windowsaround these velocities. Hand-identified velocities wereestimated by recording the velocity at the center of eachline as estimated by eye and averaging the velocities fromwhichever of the four main lines (N H + , HNC, HCO + ,HCN) were clearly detected above the noise. Where noline could be identified (53 sources), no moment map wasmade.Two different velocity ranges were used for making mo-ment maps, a narrow range for detection and a broaderrange for measuring line properties. A narrow velocityrange ( ± − ) produced the highest signal tonoise measurement for weak, narrow lines by limiting thespectral region considered to the peak of the line. Typ-ical full-width at half-maxima line-widths (∆ V FWHM)for our sources are between 5 and 8 km s − (as mea-sured in HNC; see § ± − ) was also usedto make moment maps. This range typically covers mostof the line down to the noise, and thus comes much closerto estimating the true moments of the line. In addition,it includes the hyperfine components in both the N H + and HCN lines, providing a better measure of the inte-grated intensity for those lines; the trade-off is highernoise. We therefore report detections from the narrowvelocity integration range ( ± − ) and reportmoment information from the broader range ( ± − ).Integrated intensity ( M ) maps for each source de-tected in any line are presented in Figure Set 2. To fa-cilitate inter-comparison, all maps are displayed on thesame intensity scale, with the lowest contour at 1 K kms − which is a typical 5 σ uncertainty in the integratedintensity. These moment maps are all made in a fixedvelocity range around hand-identified central velocities.We show the spectra for our four main lines at theirrespective positions of maximum integrated intensity inALT90 Pilot Survey 5 Figure 2.
Example integrated intensity (zeroth moment; M )map of G263.620 − − which is a typical5 σ m contour for our dataset. Maps taken in worse weather havehigher noise and the edges of the maps have higher noise, so notall emission at this level is necessarily significant. We use spatiallyvarying noise maps to identify genuinely significant emission forthe analysis presented in the text. Figure 1 of 134 in this figureset. Figure 3.
Spectra of our four main lines at their respective posi-tions of maximum integrated intensity (or in the center of the mapfor non-detections) for G263.620 − Figure Set 3. For sources without any detections FigureSet 3 shows the spectra at the center of the map.Basic source properties, including our hand-determined centroid velocities and the per-channelstandard deviation at the center of the map (which isrepresentative of the fully-sampled portion of the map)are summarized in Table 3. In five cases, two distinct and widely separated velocity components were seenin a source. In these cases, each velocity componentwas used to create moment maps. The stronger line islisted first in Table 3 and is the main line used whenconsidering detection statistics. We did not considerthese as separate sources for the detection statistics(because we are interested in knowing if a given catalogwill give us a detection at a given spatial position), butdid consider them as separate sources (giving us a totalof 187 sources) when considering the distributions ofmeasured moments (because they are likely two separatedense molecular clumps at distinct distances as well asvelocities).The positions of maximum integrated intensity withineach map were found by making a signal-to-noise ratiomap from M and σ M , setting the poorly-sampled threeedge pixels to zero, boxcar smoothing by a factor of three(i.e. taking the sliding average of three pixels), and iden-tifying the maximum value. This process produces apotentially different maximum integrated intensity po-sition for each line, but this is desirable as several of oursources exhibit strong spatial variation in line intensityratios. The parameters of the four main lines (N H + ,HNC, HCO + , HCN) at their respective positions of max-imum integrated intensity are listed in Table 4. RESULTS
Detection Statistics and Line Properties
The large size of our data set requires that we set ahigh level of significance when searching for features toavoid many false positives. With 187 sources, each with16 lines and 31 ×
31 pixels in each map we are searchingfor line detections in nearly 3 million spectra. A 5 σ de-tection criteria should produce one false positive per 1.7million measurements (for a perfectly normal distribu-tion). We consider this to be an acceptably small levelof contamination, and refer to a 5 σ detection as a robustdetection. Additional selection criteria combined witha lower detection threshold could be used to search foradditional weak lines. For instance, to improve the com-pleteness of H CO + detections we could adopt a lower σ threshold while constraining the search to locations withsignificant HCO + flux.Our robust detection rates of the four main lines(N H + , HNC, HCO + , and HCN ) were high ( > < Spitzer emission (see Figure 4). Detectionrates were comparable for the HOPS, ATLASGAL andIRAS samples, and similar for all four species. Five ad-ditional species had robust detections: C H, CS, SiO,H CO + and H CN. These detection rates are presentedin Figure 5. C H, in particular, was commonly seen,with detection rates between 10 and 90% for the sixdifferent surveys. Again, the HOPS, ATLASGAL andIRAS catalogs produced more robust detections thanthe catalogs based on the morphology of
Spitzer emis-sion. The low detection rates of the three input cata-logs based on
Spitzer morphology are likely due to thesecatalogs identifying features which are not truly asso-ciated with dense clumps. For instance, 10% to 20%of the IRDC candidates in the Peretto & Fuller (2009)catalog are not detected in the Herschel Hi-GAL survey Foster(Peretto et al. 2010) and only 58% of the IRDC candi-dates in the Simon et al. (2006) catalog are detected inCS (Jackson et al. 2008). These non-detections suggestthat the IRDC catalogs contain sources with a range ofcolumn densities, including sources with low column den-sities that do not have sufficient column density to beobserved with the sensitivity limits of this survey. Ta-ble 5 presents the detections and non-detections of linesfor all the sources.The detection statistics correspond to the brightest in-tegrated emission anywhere in the map, not necessarilyat the center of the map. Each input catalog providesa central position of the source, which was used as thecenter of the map. The positions of maximum integratedintensity tend to be clustered at the center of our images(see Figure 6) with 50% of maximum integrated inten-sity detections for each of the four main lines occurringwithin 40 ′′ of the map center. This suggests that the in-put catalog positions are good choices for the center ofthe map.The observed distributions of the integrated intensitiesand line-widths of the four main species (N H + , HNC,HCO + , and HCN) at the brightest point in each mapare displayed in Figures 7 and 8. Again, the detectioncriteria is M > σ in the narrow ( ± − ) in-tegration range, but the integrated intensities shown inFigures 7 and 8 and reported in Table 4 are based onthe broader range ( ± − ); some lines are nolonger 5 σ measurements when using the broader velocityintegration range.The integrated intensities for the four main species attheir brightest location in each map show broadly simi-lar distributions. All are incomplete below 2 K km s − due to our noise and detection level. Two sources haveextremely bright and broad HCO + lines, with M >
20K km s − , possibly indicating the presence of outflows.HNC has relatively fewer lines which are both broadand bright. Although integrated intensity is a distance-dependent measurement, most sources are detected inall four lines at the same velocity (and thus distance)or in none of these lines. Therefore the similarity of in-tegrated intensity distributions shows that the line lu-minosity distributions for these transitions are similarfor the majority of these sources (see § V FWHM from the secondmoment (M ) with the formula for a Gaussian profile(∆ V FWHM = √ × M ), despite the fact that manylines deviate from a Gaussian profile. ∆ V FWHM are of-ten reported as a proxy for second moment, so we reportthis effective ∆ V FWHM to facilitate comparison withother studies. We report this quantity only for HNCand HCO + . N H + and HCN are excluded because theirhyperfine structure prohibits making a line-width mea-surement solely from our moment maps.We compare the HCO + and HNC line-width distribu-tions in Figure 8 for sources where M /σ M >
3. Webreak down the HCO + distribution based on whetherH CO + is detected for a given source. The detectionof this rare isotope typically indicates an optically thickHCO + line (although a non-detection of H CO + is nota guarantee that HCO + is optically thin). The distri-butions of ∆ V FWHM for HCO + and HNC are broadly H O P S A T L A S G A L I R A S M I P S G A L G L I M P S E - B r i g h t G L I M P S E - D a r k P e r c e n t D e t e c t e d N H + HNCHCO + HCN
Figure 4.
The percentage of robust detections of the four mainpilot survey lines (N H + , HNC, HCO + , and HCN ) as a functionof input catalog. A robust detection is a source with maximumintegrated intensity (M ) > σ M , excluding the poorly samplededge 3 pixels (27 ′′ ). similar. We apply a two-sided Kolmogorov-Smirnov (K-S) test and find that we cannot reject the hypothesisthat the HNC and HCO + line-widths are drawn from thesame population when considering just the HCO + line-widths in sources without a H CO + detection (p-value= 9%) or all the HCO + line-widths (p-value = 32%).The line-width distributions shows in Figure 8 arethe line-widths at the positions of maximum integratedintensity for each molecular line transition. If we re-strict our analysis to sources for which the positionsof maximum integrated intensity for HCO + and HNCare within one 9 ′′ pixel (within 13 ′′ to include diag-onally adjacent pixels) we can compare line-widths atroughly the same position. Figure 9 shows the results ofthis comparison for sources where M /σ M > CO + detection (i.e.where HCO + is likely to be optically thick) the HCO + line-width is typically larger than the HNC line-width( h ∆ V FWHM(
HCO + ) − ∆ V FWHM(
HN C ) i = 1 . CO + detection, the line-width ratios are correlated and centered around unity( h ∆ V FWHM(
HCO + ) − ∆ V FWHM(
HN C ) i = 0 . CO + detectionsdo have optically thick HCO + emission, and that this iswhat produces their larger line-widths. In the pilot sur-vey, we have no rare isotopologue of HNC to study whereHNC might be optically thick, but the main MALT90survey will include HN C instead of H CN (HCN, be-cause of its hyperfine structure, is less likely to be opti-cally thick for similar line intensities).
Choice of Input Catalog
The first goal of the MALT90 pilot survey was to selectan input catalog for the full MALT90 survey. Of the 6catalogs tested, only HOPS, ATLASGAL and IRAS hadsufficiently high ( > H O P S A T L A S G A L I R A S M I P S G A L G L I M P S E - B r i g h t G L I M P S E - D a r k P e r c e n t D e t e c t e d C HSiO CSH CO + H CN Figure 5.
The percentage of robust detections of the frequentlydetected weaker pilot survey lines (C H, CS, SiO, H CO + andH CN ) as a function of input catalog. A robust detection is asource with maximum integrated intensity (M ) > σ M , excludingthe poorly sampled edge 3 pixels (27 ′′ ). N u m b e r N H + N u m b e r HCO + Figure 6.
Radial offset of maximum integrated intensity for eachof the four main MALT90 pilot lines (N H + , HNC, HCO + , andHCN) from the center of the map. Offsets are relative to the tar-geted center of the map, which is determined differently for thedifferent input surveys. Pointing error is estimated to be less than10 ′′ . N u m b e r N H + (cid:0) ]05101520253035 N u m b e r HCO + (cid:1) ]05101520253035 HCN Figure 7.
Maximum integrated intensity histograms of the fourmain species. The temperature scale is T ∗ A . Detections are incom-plete below 2 K km s − at the chosen 5 σ level. The distributionsfor the four lines are broadly similar. Figure 8.
Effective ∆ V FWHM of the two strong single-component lines ([left] HNC and [right] HCO + ) where M / σ M >
3. Light gray bars on the HCO + histogram show points withH CO + detections where the HCO + line is likely optically thickand self-absorbed. The distributions of line-widths are similar forthe two lines. (cid:2) V FWHM (HNC) [km s (cid:3) ]24681012 (cid:4) V F W H M ( H C O + ) [ k m s (cid:5) ] With H CO + DetectionNo H CO + Detection
Figure 9.
Comparison of the effective ∆ V FWHM of the twostrong single-component lines ([left] HNC and [right] HCO + ) wherethe positions of maximum integrated intensity are within 13 ′′ andM / σ M >
3. The dashed line is unity. Crosses indicate sourceswith H CO + detections where the HCO + line is likely opticallythick and self-absorbed; in these sources the HCO + linewidth tendsto have larger than the HNC linewidth. For the sources withoutH CO + detections, the linewidths are on average the same forboth HCO + and HNC. The ATLASGAL catalog provides the optimal source listfor MALT90. There are three major factors in favor ofusing ATLASGAL: (1) catalog size, (2) a broad range ofsource positions and velocities, and (3) a range of evolu-tionary states.ATLASGAL provides a much larger catalog thanHOPS or the Beltr´an et al. (2006) survey of IRASsources. Schuller et al. (2009) report from the initial re-sults of the survey about 6000 sources brighter than 0.25Jy in 95 deg in the Galactic range − ◦ ≤ l ≤ +11 . ◦ and +15 ◦ ≤ l ≤ +21 ◦ with | b | ≤ ◦ . In contrast theBeltr´an et al. (2006) survey contains 235 sources andHOPS (Walsh et al. 2008) is expected to contain a fewhundred bright NH sources. Neither HOPS nor IRAS Fostercontains a sufficient number of sources for the sciencegoals of MALT90.ATLASGAL sources appear to sample many Galac-tic structures. Figure 10 shows the Galactic longitudeand velocity of sources with detections (using hand-determined velocities) plotted on the Dame et al. (2001)CO map. The Dame et al. (2001) CO map is pre-sented as a longitude-velocity diagram integrated over − ◦ < b < ◦ and has units of K arcdeg. The positionsof our sources in this plot all fall within the 0.3 K arcdegCO contour and most cluster in the portions of strongerCO emission, as expected for dense, star-forming gas.The presence of ATLASGAL sources at many positionsin the longitude-velocity diagram indicates that this cat-alog is detecting sources in a range of Galactic locations.ATLASGAL sources cover a range of evolutionarystates. The initial results of ATLASGAL found two-thirds of the sources do not have a mid- or far-infrared counterpart in the Midcourse Space Experi-ment (MSX) and IRAS catalogs. A closer inspectionof IRAS/MSX dark sources in more sensitive Spitzer
GLIMPSE/MIPSGAL images reveals many of them as-sociated with weaker infrared sources but still a con-siderable fraction of the submillimeter emission appeardark in the
Spitzer images (e.g., Fig. 13 and 14 ofSchuller et al. 2009) and work on the first compact sourcerelease catalog demonstrates that ATLASGAL will pro-vide enough sources in each evolutionary state as assessedby the
Spitzer emission morphology scheme shown inFigure 1.We thus choose ATLASGAL as the sole input cata-log for the MALT90 survey because it meets all our re-quirements for a source list. ATLASGAL sources hadhigh detection rates in this pilot survey, include a diver-sity of evolutionary states, and cover a broad range ofGalactic positions. Choosing ATLASGAL as the sole in-put catalog also provides the benefits of having a singleuniform catalog when selecting sources. We can chooseour sources with a single uniform criteria and compareour MALT90 measurements against ATLASGAL catalogproperties such as the flux and extent of 870 µ m emis-sion. MALT90 Survey Strategy
The second goal MALT90 pilot survey was to test theobserving set up and verify that it allows us to achieveour science goals. MALT90 is fundamentally a mappingsurvey; although some science goals (such as determin-ing distances to clumps) could be achieved with a sin-gle pointing, the majority of our science goals rely onmaps. Our configuration allows us to map sources inmultiple lines at high spectral resolution. Maps of multi-ple lines allows us to study the chemical variation withina clump, which is most useful if clumps are typicallyspatially resolved and at least sometimes exhibit strongchemical variation. Mapping at high velocity resolution(0.11 km/s) allows us to study spatial variation in lineprofiles which may indicate changes in the strength ofturbulence, large scale motions (rotation, shear, infallor outflow), or multiple velocity components. The pi-lot survey allowed us to verify that our on-the-fly mapshad sufficient sensitivity and that we could make mapswithout significant artifacts.
Mapping Multiple Dense-gas Tracers to RevealChemistry
We see strong chemical variation in the MALT90 pi-lot sources which validates our decision to map multi-ple lines in these sources. All four of our main lines(N H + HCN HCO + and HNC) are ground state transi-tions of molecules with similar critical densities. As trac-ers of dense gas, the emission from these lines typicallyshow similar morphologies in MALT90 pilot sources, butthis is not always the case. Figure 11 shows two exam-ple sources where N H + varies significantly with respectto the other species. Figure 11 shows one source wherethe maximum N H + integrated intensity is a factor of 4-8times weaker than the other three main lines. Conversely,the other source in Figure 11 has N H + emission whichis twice as strong as that of HCN, HCN or HCO + . Largevariation in the HCO + /N H + integrated intensity ratioin high-mass star-forming regions has been noted before(e.g., Turner & Thaddeus 1977; Walsh & Burton 2006).Thus, the combination of mapping in several lines simul-taneously gives us the most complete picture of the spa-tial distribution of the various molecules, which is crucialto studying variations in the chemistry with the MALT90sources. Mapping Strategy
Mapping artifacts are seen in many of our maps, typi-cally manifesting as stripes in the direction of scans. Oursources were generally mapped with scans of constantGalactic longitude so that each strip in Galactic longi-tude uses the same reference spectrum. Noise or gainvariations in this reference spectrum can therefore pro-duce stripes in the map, and this phenomenon is partic-ularly prevalent in sources observed in bad weather. Wechose to map using scans of constant Galactic longitudefor the pilot survey because most extended structures inthe Galactic plane are parallel to the plane. Thus, noisestripes are easier to identify, since they typically runperpendicular to real features. We mapped two sources(G305.887+00.016 and G308.058 − − C S cube. The striping visiblein the scan direction in both individual maps is signif-icantly reduced in the combined image. We thereforedecided to map using scans of both constant Galacticlongitude and latitude in the full MALT90 survey.
The Value of High-resolution Velocity Information
Figure 13 shows the central portion of the maps forthe source G321.935 − + . TheHCO + line shows self-absorption at the systemic veloc-ity (traced by the HNC which shows little or no non-gaussianity). In the lower-right portion of the map, thisself-absorption shows an asymmetric profile with brighterblue-shifted emission. Such a profile is characteristicof infall of cold gas toward a hot central source (e.g.,Mardones et al. 1997). This characteristic shape is notpresent in the upper-left of the map, where we see ared-shifted profile usually associated with expansion. Itis possible that infall is happening only in part of thissource or that other kinematic complexity is present; theALT90 Pilot Survey 9 Figure 10.
Distribution of MALT90 pilot sources in velocity and Galactic longitude, overlaid on the Galactic CO distribution fromDame et al. (2001) integrated over Galactic latitude. The CO contours are at 0.3, 1, 3, and 10 K arcdeg. Different input catalogs arelabelled as follows: [yellow] = Dark GLIMPSE source, [purple] = Bright GLIMPSE source, [green] = MIPSGAL source, [red] = HOPSsource, [blue] = IRAS source, [orange] = ALTASGAL source. Sources with no detected line emission (and thus no velocity) are omitted.In addition, four IRAS sources with Galactic longitude between -70 and -100 degrees were omitted to display the remaining sources at alarger scale. These four omitted sources all also lie within the CO 0.3 K arcdeg emission contour.
Figure 11.
Two examples of strong chemical variation in N H + . Left: G287.814 − − ) HNC, HCO + ,and HCN lines, but weak (1.65 ± − ) N H + . Right: G322.932+01.393 shows a very strong N H + line (8.93 ± − ),but comparatively weak HNC and HCO + (3.5 - 4.5 K km s − ) and a non-detection of HCN( < .
66 K km s − ). large variance of this complex line shape over the sourcedemonstrates the value of mapping at high velocity res-olution. Studying Different Stages of Evolution
The many lines observed in MALT90 provide infor-mation about the evolutionary state of the clumps ob-served. This information can be combined with
Spitzer morphological classification and dust temperature deter-mination from spectral-energy distribution fitting (e.g.,Rathborne et al. 2010; Peretto et al. 2010) to constrainthe evolutionary state of a clump. As a short example wepresent three sources in different
Spitzer morphologicalstates and show what information can be gained fromthe molecular lines in each case.Figure 14 shows G330.873 − Spitzer
GLIMPSE/MIPSGAL imagesdue to its strong extended 8 and 24 µ m emission. Thepresence of CH CN ( J K = 5 − ) emission identifiesthe hot core associated with a massive protostar sincethis molecule is seen only in warm ( T >
100 K) anddense ( n > cm − ) regions and is often detected in HII regions (e.g., Purcell et al. 2006). The location of thecore is also the position of maximum integrated inten-sity for most of the other molecules (HCN, HNC, SiO, CS), but not N H + , which peaks in the south at theposition of a 24 µ m point source. Although this source isclearly identified as an H II region from the characteristic Spitzer appearance of a 24 µ m inner bubble surroundedby a ring of 8 µ m emission (e.g., Watson et al. 2008), the0 Foster Figure 12.
Zeroth moment (integrated intensity) maps of C Sin G308.058 − C S has very low abundance, so thesemaps show only noise. In panel (a) the map was made with scansof constant Galactic longitude, which is the default mode for theMALT90 pilot survey. In panel (b) the map was made with scansof constant Galactic latitude. In panel (c), the two individual mapswere combined, reducing the striping artifacts visible in maps (a)and (b). This reduction of artificial structure motivates combiningmaps scanned in both directions in the full MALT90 Survey. detection of a hot core molecule identifies the locationof the central exciting source. The chemical difference(N H + /HCO + ratio) between the southern source sug-gests either a different luminosity for the exciting sourceor a different evolutionary state.Figure 15 shows G335.075 − Spitzer
GLIMPSE/MIPSGAL images due tothe presence of 24 µ m point sources within a dark extinc-tion feature without extended 8 or 24 µ m emission. Thespatial coincidence of the 24 µ m point sources and the8 µ m extinction feature suggests that the 24 µ m pointsources are associated with the clump, and the MALT90pilot survey data confirms this association. The N H + integrated intensity contours trace the 8 µ m emission ex-tinction feature. There is SiO emission at the same ve-locity at the position of the brightest 24 µ m point source.Since SiO emission is normally associated with outflowactivity in protostars (e.g., L´opez-Sepulcre et al. 2011), adetection of this line at the same velocity as the clump isstrong evidence that the 24 µ m point source is associatedwith this clump.Figure 16 shows G322.668 − Spitzer
GLIMPSE/MIPSGAL images dueto the lack of 8 or 24 µ m emission inside the 8 µ m ex-tinction feature. As a quiescent clump in the early stagesof evolution, this object displays less complex chemistrythan clumps in more evolved stages with only the fourmain lines (N H + , HNC, HCO + , and HCN) detected.The HNC integrated intensity emission shows two dis-tinct peaks associated with two of the darkest 8 µ m ex-tinction features. The velocity field of HNC shows thatthese two peaks are at very similar velocities (-64 km s − and -65.6 km s − ), strongly suggesting that both peaksare at the same distance and that the entire extinctionfeature is a single physical object. We use the Clemens(1985) rotation curve to calculate a kinematic distancefor this clump; the near distance is 4.27 kpc and thefar distance is 9.25 kpc. Because we see the clump asan extinction feature against the diffuse Galactic back-ground, it is reasonable to assume that the near distanceis correct. The MALT90 map therefore allows us (1) toidentify which extinction features are likely a single phys-ical object versus a chance projection and (2) to assigna distance which is useful for any further study of thisobject. CONCLUSION
We have described the MALT90 pilot survey, carriedout to demonstrate the feasibility of the MALT90 survey,identify the best input catalog for choosing MALT90 tar-gets, and optimize the survey parameters. We choose theATLASGAL (Schuller et al. 2009) catalog as our sourcelist on the basis of its high detection rates for the mainfour survey lines ( >
90% for N H + , HNC, HCO + , andHCN) and the large number of dense molecular clumpsin different evolutionary stages in this catalog. The sur-veys which provided a prior selection for regions of highcolumn density, either from optically-thin dust (ATLAS-GAL at 870 µ m, the Beltr´an et al. (2006) survey at 1.2-mm) or another dense gas tracer (NH from HOPS) pro-duced much higher detection rates than choosing sourcesidentified based on Spitzer emission morphology withoutthis prior.We have briefly summarized the data obtained fromthe MALT90 pilot survey and highlighted some of the sci-ence possible with this survey including studying chem-ical variation, the kinematics of massive dense clumpsand the galactic distribution of dense molecular clumpsassociated with high-mass star formation. We have madethe full data-set publicly available through this publica-tion and the MALT90 website, including reduced data-cubes and uniform moment maps which facilitate easyinspection of the data. This collection is already oneof the largest sets of 90 GHz molecular line maps fordense molecular clumps. The full survey is underway andplans to map a total of 3,000 candidate dense molecularclumps, increasing this sample by an order of magnitudeand providing a valuable database for studying many as-pects of high-mass star formation. ACKNOWLEDGEMENTS
The Mopra telescope is part of the Australia Telescopeand is funded by the Commonwealth of Australia foroperation as National Facility managed by CSIRO. TheUNSW-MOPS Digital Filter Bank used for the observa-tions with the Mopra telescope was provided with sup-port from the Australian Research Council, together withthe University of New South Wales, University of Syd-ney and Monash University. This research has made useof the NASA/ IPAC Infrared Science Archive (for accessto GLIMPSE and MIPSGAL images), which is operatedby the Jet Propulsion Laboratory, California Instituteof Technology, under contract with the National Aero-nautics and Space Administration. This research hasmade use of NASA’s Astrophysics Data System Bibli-ographic Services. JMJ gratefully acknowledges fund-ing support from NSF Grant No. AST-0808001. TheMALT90 project team gratefully acknowledges the useof dense core positions supplied by ATLASGAL. AT-LASGAL is a collaboration between the Max PlanckGesellschaft (MPG: Max Planck Institute for Radioas-tronomy, Bonn and the Max Planck Institute for As-tronomy, Heidelberg), the European Southern Observa-tory (ESO) and the University of Chile. Thanks to AnitaTitmarsh and the duty astronomers and staff at the PaulWild Observatory for their assistance during the obser-vations.ALT90 Pilot Survey 11
Figure 13.
HCO + [thick/black] and HNC [thin/red] line profiles over the central portion of the G321.935 − + line shows a strong dip at the systemic velocity, but the relative strengths of the blue andred wings vary throughout this map. Adjacent spectra are separated by 9 ′′ . Figure 14.
An example H II clump (G330.873 − Spitzer /IRAC band 1 (3.6 µ m) and band4 (8.0 µ m) from GLIMPSE in blue and green and Spitzer /MIPSband 1 (24 µ m) from MIPSGAL in red. The white box shows theextent of the MALT90 pilot survey map. Dashed (red) contoursare HCO + integrated intensity (plotted at SNR of 15, 30, 45, 60; σ ∼ − ). Thin (blue) contours are N H + integratedintensity (plotted at SNR of 10, 20, 30, 40; σ ∼ − ).Thick (black) contours are CH CN integrated intensity (plottedat SNR of 2, 4; σ ∼ − ). The presence of CH CNidentifies the hot core associated with a massive protostar. Thislocation is also the position of maximum integrated intensity formost of the other molecules (HCN, HNC, SiO, CS), but N H + peaks in the south as the position of another 24 µ m point source. Figure 15.
An example protostellar clump (G335.075 − Spitzer /IRAC band 1 (3.6 µ m)and band 4 (8.0 µ m) from GLIMPSE in blue and green and Spitzer /MIPS band 1 (24 µ m) from MIPSGAL in red. The whitebox shows the extent of the MALT90 pilot survey map. Dashed(red) contours are HCO + integrated intensity (plotted at SNR of5, 7; σ ∼ − ). Thin (blue) contours are N H + inte-grated intensity (plotted at SNR of 7, 9, 13, 19, 25; σ ∼ − ). Thick (yellow) contours are SiO integrated intensity (plottedat SNR of 4; σ ∼ − in the small velocity window).This clump was drawn from the HOPS catalog and classified asprotostellar due to the presence of several 24 µ m point sources.The strongest 24 µ m point sources are associated with SiO emis-sion at the same velocity as the clump, indicating the presence ofprotostellar outflows in the clump. ALT90 Pilot Survey 13
Figure 16.
An example quiescent clump (G322.668 − Spitzer /IRAC band 1 (3.6 µ m) andband 4 (8.0 µ m) from GLIMPSE in blue and green and Spitzer /MIPS band 1 (24 µ m) from MIPSGAL in red. The white box shows theextent of the MALT90 pilot survey map. The extinction seen at 8.0 µ m is well traced by the HNC integrated intensity contours (plottedas SNR of 4, 6, 8, 10; σ ∼ − ) which show two distinct peaks. The right panel shows the centroid velocity M in color withthe HNC integrated intensity contours. The two peaks are at very similar velocities (-64 km s − and -65.5 km s − , suggesting that theextinction feature is all at the same distance. REFERENCESBeltr´an, M. T., Brand, J., Cesaroni, R., et al. 2006, A&A, 447,221Benjamin, R. A., Churchwell, E., Babler, B. L., et al. 2003,PASP, 115, 953Bergin, E. A., Ciardi, D. R., Lada, C. J., Alves, J., & Lada, E. A.2001, ApJ, 557, 209Bergin, E. A. & Tafalla, M. 2007, ARA&A, 45, 339Brown, P. D., Charnley, S. B., & Millar, T. J. 1988, MNRAS,231, 409Carey, S. J., Noriega-Crespo, A., Mizuno, D. R., et al. 2009,PASP, 121, 76Clemens, D. P. 1985, ApJ, 295, 422Dame, T. M., Hartmann, D., & Thaddeus, P. 2001, ApJ, 547, 792Fazio, G. G., Hora, J. L., Allen, L. E., et al. 2004, ApJS, 154, 10Fuller, G. A., Williams, S. J., & Sridharan, T. K. 2005, A&A,442, 949Gerin, M., Ka´zmierczak, M., Jastrzebska, M., et al. 2011, A&A,525, A116+Gibson, D., Plume, R., Bergin, E., Ragan, S., & Evans, N. 2009,ApJ, 705, 123Hirota, T., Yamamoto, S., Mikami, H., & Ohishi, M. 1998, ApJ,503, 717Jackson, J. M., Finn, S. C., Rathborne, J. M., Chambers, E. T.,& Simon, R. 2008, ApJ, 680, 349Ladd, N., Purcell, C., Wong, T., & Robertson, S. 2005, PASA, 22,62Lo, N., Cunningham, M. R., Jones, P. A., et al. 2009, MNRAS,395, 1021L´opez-Sepulcre, A., Walmsley, C. M., Cesaroni, R., et al. 2011,A&A, 526, L2+ Mardones, D., Myers, P. C., Tafalla, M., et al. 1997, ApJ, 489, 719Peretto, N. & Fuller, G. A. 2009, A&A, 505, 405Peretto, N., Fuller, G. A., Plume, R., et al. 2010, A&A, 518, L98+Pirogov, L., Zinchenko, I., Caselli, P., Johansson, L. E. B., &Myers, P. C. 2003, A&A, 405, 639Purcell, C. R., Balasubramanyam, R., Burton, M. G., et al. 2006,MNRAS, 367, 553Rathborne, J. M., Jackson, J. M., Chambers, E. T., et al. 2010,ApJ, 715, 310Rawlings, J. M. C., Redman, M. P., Keto, E., & Williams, D. A.2004, MNRAS, 351, 1054Rieke, G. H., Young, E. T., Engelbracht, C. W., et al. 2004,ApJS, 154, 25Schilke, P., Walmsley, C. M., Pineau des Forets, G., & Flower,D. R. 1997, A&A, 321, 293Schuller, F., Menten, K. M., Contreras, Y., et al. 2009, A&A, 504,415Shirley, Y. L., Evans, II, N. J., Young, K. E., Knez, C., & Jaffe,D. T. 2003, ApJS, 149, 375Shukla, H., Yun, M. S., & Scoville, N. Z. 2004, ApJ, 616, 231Simon, R., Jackson, J. M., Rathborne, J. M., & Chambers, E. T.2006, ApJ, 639, 227Turner, B. E. & Thaddeus, P. 1977, ApJ, 211, 755Walsh, A. J. & Burton, M. G. 2006, MNRAS, 365, 321Walsh, A. J., Lo, N., Burton, M. G., et al. 2008, PASA, 25, 105Watson, C., Povich, M. S., Churchwell, E. B., et al. 2008, ApJ,681, 1341Williams, J. P., Blitz, L., & McKee, C. F. 2000, Protostars andPlanets IV, 97Wu, J., Evans, N. J., Shirley, Y. L., & Knez, C. 2010, ApJS, 188,313
ALT90 Pilot Survey 15
Table 3
MALT90 Pilot SourcesName Glon Glat RA J2000 Dec J2000 Catalog Velocity T rms [degrees] [degrees] [HH:MM:SS] [DD:MM:SS] LSR [km s − ] [K]G263.620 − − − − − − − − − − · · · − − · · · − − − − − − − − · · · · · · − − − − − − − − · · · · · · − − · · · − − − − · · · − − − − · · · − − · · · − − · · · · · · − − · · · · · · − − − − · · · − − − − · · · − − · · · − − · · · · · · − − · · · − − · · · − − · · · − − − − · · · − − − − · · · − − · · · · · · − − · · · − − Table 3 — Continued
Name Glon Glat RA J2000 Dec J2000 Catalog Velocity T rms [degrees] [degrees] [HH:MM:SS] [DD:MM:SS] LSR [km s − ] [K]G321.013 − − − − · · · − − − − · · · − − · · · − − − − − − − − · · · · · · − − · · · − − · · · − − − − − − − − · · · − − − − · · · − − − − − − · · · − − − − · · · · · · − − · · · − − − − − − · · · − − − − · · · − − · · · − − · · · − − − − − − − − · · · − − − − · · · − − · · · − − · · · − − − − − − − − · · · − − − − − − − − ALT90 Pilot Survey 17
Table 3 — Continued
Name Glon Glat RA J2000 Dec J2000 Catalog Velocity T rms [degrees] [degrees] [HH:MM:SS] [DD:MM:SS] LSR [km s − ] [K]G333.767 − − − − − − − − − − − − − − · · · − − − − · · · − − − − − − − − · · · − − · · · − − − − − − − − − − − − Note . — Central source velocities are determined by hand-examination of the four main lines for each source. Where two velocity componentsare seen, the stronger is listed first. T rms gives the noise per channel in the spectrum as measured at the central position in the map.
Table 4
Properties of Main Lines at Position of Maximum Integrated IntensityName Maximum Integrated Intensity [K km s − ] Offset from Map Center a [9 ′′ pixels] FWHM [km s − ]N H + HNC HCO + HCN N H + HNC HCO + HCN HNC HCO + G263.620 − ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G300.968+01.145 5.11 ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G302.018 − < ± ± < · · · (-5, -6) (-3, 9) · · · ± ± ± ± ± ± ± ± < ± ± < · · · (-2, -1) (-1, 0) · · · ± ± − < < ± < · · · · · · (-2, 4) · · · · · · ± − < ± ± ± · · · (2, -1) (0, 0) (1, 0) 5.26 ± ± − < < < < · · · · · · · · · · · · · · · · · · G303.992+00.209 < < < < · · · · · · · · · · · · · · · · · · G304.887+00.635 1.79 ± ± ± ± ± ± < < ± ± · · · · · · (-4, 1) (-8, 6) · · · ± < ± ± ± · · · (2, -6) (2, -8) (-5, -3) 6.71 ± ± ± ± ± ± ± ± ± ± ± ± ± ± − ± ± ± ± ± ± ± ± ± ± ± ± − ± ± ± ± ± ± ± ± ± ± ± ± < < ± < · · · · · · (1, -5) · · · · · · ± − < < ± ± · · · · · · (-2, -3) (-3, -1) · · · ± − < < < < · · · · · · · · · · · · · · · · · · G307.017+00.706 < < < < · · · · · · · · · · · · · · · · · · G308.006 − < < ± < · · · · · · (3, 0) · · · · · · ± < < < < · · · · · · · · · · · · · · · · · · G308.058 − ± ± ± ± ± ± ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G309.014+00.208 < ± ± ± · · · (8, 1) (8, 0) (6, 2) 7.01 ± ± − ± ± ± ± ± ± ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G310.013+00.387 3.85 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G311.005 − < < ± < · · · · · · (8, -4) · · · · · · ± < < < < · · · · · · · · · · · · · · · · · · G312.038+00.077 1.09 ± ± ± ± ± ± ± ± < < · · · · · · ± · · · G312.992+00.172 < < < < · · · · · · · · · · · · · · · · · · G313.015 − < < < < · · · · · · · · · · · · · · · · · · G313.976+00.139 < < < < · · · · · · · · · · · · · · · · · · G313.994 − < ± ± ± · · · (2, -2) (2, -2) (2, -2) 8.51 ± ± ± ± ± ± ± ± ± ± < < · · · · · · ± · · · G314.995 − < < < < · · · · · · · · · · · · · · · · · · G315.981 − < ± ± ± · · · (8, 3) (0, 1) (-1, 1) 6.79 ± ± − < < < < · · · · · · · · · · · · · · · · · · G316.139 − ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± < < < < · · · · · · · · · · · · · · · · · · G317.106 − < < < < · · · · · · · · · · · · · · · · · · G317.242+00.017 < < < < · · · · · · · · · · · · · · · · · · G317.977 − < < < < · · · · · · · · · · · · · · · · · · G318.049+00.086 4.43 ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G318.145 − < < < < · · · · · · · · · · · · · · · · · · G318.776 − ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G318.916 − < ± ± < · · · (2, -3) (-4, 3) · · · ± ± − ± ± ± ± ± ± < < < < · · · · · · · · · · · · · · · · · · G319.033 − < < < < · · · · · · · · · · · · · · · · · · G319.990+00.368 < < < < · · · · · · · · · · · · · · · · · · G320.019 − < < < < · · · · · · · · · · · · · · · · · · G320.168+00.824 8.55 ± ± ± ± ± ± ALT90 Pilot Survey 19
Table 4 — Continued
Name Maximum Integrated Intensity [K km s − ] Offset from Map Center a [9 ′′ pixels] FWHM [km s − ]N H + HNC HCO + HCN N H + HNC HCO + HCN HNC HCO + G320.270+00.293 < ± ± < · · · (1, -1) (6, -7) · · · ± ± − < ± ± < · · · (-2, -1) (1, -11) · · · ± ± − < ± ± ± · · · (-1, -3) (0, -4) (-4, -7) 7.20 ± ± − < < < < · · · · · · · · · · · · · · · · · · G321.030 − ± ± ± ± ± ± ± ± ± ± ± ± − ± ± ± ± ± ± < < < < · · · · · · · · · · · · · · · · · · G322.032 − < < < < · · · · · · · · · · · · · · · · · · G322.668+00.038 1.19 ± ± ± ± ± ± ± ± ± < · · · ± ± − ± ± ± ± ± ± ± ± ± < · · · ± ± − < ± ± ± · · · (-2, 3) (0, 1) (-2, 0) 5.73 ± ± − ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G325.030+00.051 < < < < · · · · · · · · · · · · · · · · · · G325.127+00.029 1.75 ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G326.410+00.596 < < ± ± · · · · · · (-5, 0) (0, -3) · · · ± − < < < < · · · · · · · · · · · · · · · · · · G326.794+00.386 3.48 ± ± ± ± ± ± ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G327.054+00.037 2.52 ± ± ± ± ± ± − ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G327.397 − ± ± ± ± ± ± − ± ± ± ± ± ± − < ± ± ± · · · (2, 2) (-2, 0) (1, 1) 1.97 ± ± ± ± ± ± ± ± < < < < · · · · · · · · · · · · · · · · · · G327.981 − < < ± < · · · · · · (2, 2) · · · · · · ± < ± < < · · · (4, 3) · · · · · · ± · · · G328.255 − ± ± ± ± ± ± < < < < · · · · · · · · · · · · · · · · · · G328.971+00.019 < < < < · · · · · · · · · · · · · · · · · · G329.034 − < < < < · · · · · · · · · · · · · · · · · · G329.036 − ± ± ± ± ± ± − ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G329.457+00.506 7.92 ± ± ± ± ± ± − < ± ± ± · · · (-6, 3) (-9, 6) (5, -9) 7.90 ± ± ± ± ± ± ± ± ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G330.042+01.059 5.05 ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G330.620 − < < < < · · · · · · · · · · · · · · · · · · G330.778+00.256 1.79 ± ± ± < · · · ± ± − ± ± ± ± ± ± − < < ± ± · · · · · · (7, 2) (-4, 2) · · · ± ± ± ± ± ± ± − < ± ± ± · · · (-7, -1) (4, 3) (-6, -4) 8.69 ± ± − ± ± ± ± ± ± < ± ± < · · · (8, -1) (2, -2) · · · ± ± < < < < · · · · · · · · · · · · · · · · · · G331.230 − ± ± ± < · · · ± ± − ± ± ± ± ± ± ± ± ± ± ± ± < < < < · · · · · · · · · · · · · · · · · · G331.914 − < < < < · · · · · · · · · · · · · · · · · · G332.003 − < < < < · · · · · · · · · · · · · · · · · · G332.070+00.503 1.52 ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G332.963+00.773 6.50 ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± Table 4 — Continued
Name Maximum Integrated Intensity [K km s − ] Offset from Map Center a [9 ′′ pixels] FWHM [km s − ]N H + HNC HCO + HCN N H + HNC HCO + HCN HNC HCO + G333.067 − ± ± ± ± ± ± − ± ± ± ± ± ± ± ± ± ± ± ± − ± ± ± ± ± ± ± ± ± ± ± ± − ± ± < < · · · · · · ± · · · G334.975 − < ± ± ± · · · (8, -10) (-3, 0) (-1, 0) 7.32 ± ± − ± ± ± < · · · ± ± − ± ± ± ± ± ± − ± ± ± < · · · ± ± ± ± ± ± ± ± − < < < < · · · · · · · · · · · · · · · · · · G336.022 − ± ± ± ± ± ± − ± ± ± ± ± ± < < < < · · · · · · · · · · · · · · · · · · G336.994 − · · · b ± ± ± · · · (4, 2) (2, 3) (-1, -4) 7.60 ± ± ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − < ± ± ± · · · (4, -3) (0, -4) (-1, -5) 6.00 ± ± · · · b ± ± ± · · · (0, 1) (1, 1) (0, 2) 2.14 ± ± ± ± < < · · · · · · ± · · · G338.927+00.635 9.70 ± ± ± ± ± ± < ± ± ± · · · (10, -5) (0, -2) (0, -2) 9.21 ± ± < < < < · · · · · · · · · · · · · · · · · · G339.968 − ± ± ± ± ± ± < < < < · · · · · · · · · · · · · · · · · · G342.975+02.673 < ± ± ± · · · (-8, 5) (-5, -1) (-5, -1) 9.93 ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± < ± ± ± · · · (1, 5) (-8, 6) (0, 0) 9.67 ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± ± ± ± ± ± ± Note . — Results are listed for lines with 5 σ detections in the narrow ( ± .
25 km s − ) moment maps, but these measurements report themoments calculated in a range of ( ± .
25 km s − ). Not all integrated intensity measurements at 5 σ results in this broader velocity range.a Offsets are (x,y) offsets in units of pixels, where each pixel is 9 ′′ . Offsets are relative to the targeted center of the map, which is determineddifferently for the different input surveys.b N H + is outside the spectral coverage for sources with V LSR < − − ALT90 Pilot Survey 21
Table 5
Robust (5 σ ) Detections of LinesName N H + 13 CS HNC HCO + HCN C H SiO H CO + H CNG263.620 − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − Table 5 — Continued
Name N H + 13 CS HNC HCO + HCN C H SiO H CO + H CNG321.013 − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − ALT90 Pilot Survey 23
Table 5 — Continued
Name N H + 13 CS HNC HCO + HCN C H SiO H CO + H CNG334.458 − − − − − − − − − − − − − − − − − − − − − Note . — Robust detections correspond to 5 σ integrated intensity detections excluding the 3 pixels (27 ′′ ) on the edge of each map. Detectionstatistics are not shown for H41 α , CH CN, HC N, C S, HC CCN, HNCO 4 , or HNCO 4 ,4