The Bolocam Galactic Plane Survey IV: 1.1 and 0.35 mm Dust Continuum Emission in the Galactic Center Region
John Bally, James Aguirre, Cara Battersby, Eric Todd Bradley, Claudia Cyganowski, Darren Dowell, Meredith Drosback, Miranda K Dunham, Neal J. Evans II, Adam Ginsburg, Jason Glenn, Paul Harvey, Elisabeth Mills, Manuel Merello, Erik Rosolowsky, Wayne Schlingman, Yancy L. Shirley, Guy S. Stringfellow, Josh Walawender, Jonathan Williams
aa r X i v : . [ a s t r o - ph . GA ] N ov DRAFT: September 20, 2018
The Bolocam Galactic Plane Survey: λ = 1.1 and 0.35 mm DustContinuum Emission in the Galactic Center Region John Bally , James Aguirre , Cara Battersby , Eric Todd Bradley , Claudia Cyganowski ,Darren Dowell , Meredith Drosback , Miranda K Dunham , Neal J. Evans II , AdamGinsburg , Jason Glenn , Paul Harvey , Elisabeth Mills , Manuel Merello ErikRosolowsky , Wayne Schlingman , Yancy L. Shirley , Guy S. Stringfellow , JoshWalawender , and Jonathan Williams CASA, University of Colorado, UCB 389, Boulder, CO 80309,
[email protected] Department of Physics and Astronomy, University of Pennsylvania, Philadelphia, PA , [email protected] CASA, University of Colorado, UCB 389, Boulder, CO 80309,
[email protected] Department of Physics and Astronomy, University of Central Florida, , [email protected] Department of Astronomy, University of Wisconsin, Madison, WI 53706 , [email protected] Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Dr. ,Pasadena, CA 91104, [email protected] Department of Astronomy, University of Virginia, P. O. Box 400325, Charlotesville, VA22904, [email protected] Department of Astronomy, University of Texas, 1 University Station C1400, Austin, TX78712, [email protected] Department of Astronomy, University of Texas, 1 University Station C1400, Austin, TX78712, [email protected] CASA, University of Colorado, UCB 389 Boulder, CO 80309,
[email protected] CASA, University of Colorado, UCB 389, Boulder, CO 80309,
[email protected] CASA, University of Colorado, UCB 389, Boulder, CO 80309, [email protected] Department of Physics Astronomy, University of California, Los Angeles, Los Angeles,CA 90095, [email protected] Department of Astronomy, University of Texas, 1 University Station C1400, Austin, TX78712, [email protected] Department of Physics and Astronomy, University of British Columbia, Okanagan, 3333University Way, Kelowna BC V1V 1V7 Canada , [email protected] Steward Observatory, University of Arizona, 933 North Cherry Ave., Tucson, AZ 85721 , [email protected] Steward Observatory, University of Arizona, 933 North Cherry Ave., Tucson, AZ 85721 , [email protected] CASA, University of Colorado, UCB 389, Boulder, CO 80309,
[email protected] Institute for Astronomy (IfA), University of Hawaii, 640 N. Aohoku Pl., Hilo, HI 96720, [email protected] Institute for Astronomy (IfA), University of Hawaii, 2680 Woodlawn Dr., Honolulu, HI96822, [email protected]
ABSTRACT
The Bolocam Galactic Plane Survey (BGPS) data for a six square degreeregion of the Galactic plane containing the Galactic center is analyzed and com-pared to infrared and radio continuum data. The BGPS 1.1 mm emission consistsof clumps interconnected by a network of fainter filaments surrounding cavities,a few of which are filled with diffuse near-IR emission indicating the presence ofwarm dust or with radio continuum characteristic of HII regions or supernovaremnants. New 350 µ m images of the environments of the two brightest regions,Sgr A and B, are presented. Sgr B2 is the brightest mm-emitting clump in theCentral Molecular Zone and may be forming the closest analog to a super starcluster in the Galaxy. The Central Molecular Zone (CMZ) contains the highestconcentration of mm and sub-mm emitting dense clumps in the Galaxy. Most1.1 mm features at positive longitudes are seen in silhouette against the 3.6 to 24 µ m background observed by the Spitzer Space Telescope. However, only a fewclumps at negative longitudes are seen in absorption, confirming the hypothesisthat positive longitude clumps in the CMZ tend to be on the near-side of theGalactic center, consistent with the suspected orientation of the central bar inour Galaxy. Some 1.1 mm cloud surfaces are seen in emission at 8 µ m, pre-sumably due to polycyclic aromatic hydrocarbons (PAHs). A ∼ ◦ ( ∼
30 pc)diameter cavity and infrared bubble between l ≈ ◦ and 0.2 ◦ surrounds theArches and Quintuplet clusters and Sgr A. The bubble contains several clumpydust filaments that point toward Sgr A ∗ ; its potential role in their formation isexplored.Bania’s Clump 2, a feature near l = 3 ◦ to 3.5 ◦ which exhibits extremelybroad molecular emission lines (∆ V >
150 km s − ), contains dozens of 1.1 mmclumps. These clumps are deficient in near- and mid-infrared emission in theSpitzer images when compared to both the inner Galactic plane and the CentralMolecular Zone. Thus, Bania’s Clump 2 is either inefficient in forming stars oris in a pre-stellar phase of clump evolution.The Bolocat catalog of 1.1 mm clumps contains 1428 entries in the Galacticcenter between l = 358.5 ◦ to l = 4.5 ◦ of which about 80% are likely to be withinabout 500 pc of the center. The mass-spectrum above about 80 M ⊙ can be 4 –described by a power-law ∆ N/ ∆ M = N M − . . , − . . The power-law indexis somewhat sensitive to systematic grain temperature variations, may be highlybiased by source confusion, and is very sensitive to the spatial filtering inherentin the data acquisition and reduction. Subject headings:
Galaxy: center (ISM): clouds (ISM): dust, extinction stars:formation surveys-
1. Introduction
The formation of massive stars and star clusters remains one of the outstanding prob-lems in star formation research (Stahler et al. 2000; Bally & Zinnecker 2005). The CentralMolecular Zone (CMZ) of our Galaxy, located within about 500 pc of the nucleus, containsabout 5 to 10% of the molecular gas mass and the largest concentration of massive starsand star clusters in the Milky Way (Morris & Serabyn 1996; Ferri`ere et al. 2007). The CMZhosts two of the most massive, densest, and remarkable star forming complexes in the MilkyWay, Sgr A and Sgr B2 (Yusef-Zadeh et al. 2008, 2009). Giant molecular clouds (GMCs) inthe CMZ are one to two orders of magnitude denser and have over an order of magnitudelarger line-widths than typical GMCs in the Galactic plane beyond 3 kpc from the nucleus(Bally et al. 1987, 1988; Oka et al. 1998; Tsuboi et al. 1999; Oka et al. 2001b). The study ofthe interstellar medium in the CMZ and the immediate vicinity of the central black hole isan essential step towards understanding star and cluster formation in extreme environmentssuch as starburst galaxies and active galactic nuclei. The CMZ may be a Galactic analogof the nuclear star-forming rings observed in the centers of star forming barred galaxies(Kormendy & Kennicutt 2004; Kormendy & Cornell 2004; Liszt 2009).The Galactic plane just outside the CMZ between l = 1.3 ◦ to 5 ◦ contains several remark-able, localized cloud complexes with unusually large velocity extents. Longitude-velocitydiagrams in species such as HI, CO, and high-density gas tracers such as CS and HCO + show that near l = 355 ◦ , 1.3 ◦ , 3 ◦ , and 5 ◦ , these cloud complexes have velocity extents of 100to 200 km s − in regions less than 0.5 ◦ ( ∼
75 pc) in diameter (Dame et al. 2001; Liszt 2006,2008). These remarkable cloud complexes may trace the locations where gas is entering thedust lanes at the leading edges of the bar in the center of our Galaxy and passing through aseries of shocks (Liszt 2006), or dust lanes along the bar’s leading edge seen nearly end-on.A crucial step in the observational study of massive star and cluster formation is theidentification and characterization of clumps that will soon form or are actively formingstars, without biases introduced by targeting sources with already known signposts of star 5 –formation such as masers, IR sources, or HII regions. Massive stars and clusters form fromcool high-density clumps with very large column densities and extinctions that can sometimesexceed A V = 100 magnitudes. Such clumps are best investigated at millimeter and sub-millimeter wavelengths. Spectral lines provide excellent diagnostics of line-of-sight motionsin a cloud. However, the interpretation of various gas tracers that produce emission lines inthis portion of the spectrum can be very difficult. Variations in tracer abundances causedby depletions and complex chemical processing, uncertainties in excitation conditions, andthe impacts of radiation fields and shocks make the derivation of column densities, masses,and other physical properties of clumps highly uncertain.The continuum emission from warm dust provides a somewhat more reliable tracer ofthe column density and clump masses. At 1.1 mm, hν/k ≈
13 K, and most sources areexpected to be optically thin. Thus, the column density is roughly proportional to thetemperature and optical depth of the emitting medium. Shorter wavelength observationscan be more difficult to obtain due to lower atmospheric transmission and greater variationsin sky emission (“atmospheric” or “sky noise”). Furthermore, the stronger dependenceof column density on temperature when hν/k > T dust makes the derivation of physicalparameters more difficult.Contamination of the dust continuum emission in the Bolocam 1.1 mm filter by spectrallines and free-free emission is likely to be small, less than a few percent in almost all cases.As discussed in detail in Aguirre et al. (2010), the Bolocam 1.1 mm filter excludes the bright1.3 mm CO lines. In the worst case, that of the hot cores in Sgr B2, contamination byother spectral lines can account for at most 20% of the detected 1.1 mm flux. Such spectralline emission is much weaker in the other clumps. Contamination by free-free emissionwould only occur from ultra-compact HII regions. However, such objects tend to be only afew arc-seconds in diameter, and will be beam diluted and completely overwhelmed by thelarger-angular scale dust continuum. Comparison of the 20 cm radio continuum emission(see discussion of Figure 11 below) shows that 20 cm radio and 1.1 mm dust continuumemission are not correlated, supporting the argument that there is little contamination ofthe 1.1 mm fluxes by free-free emission.This paper presents two new data sets obtained with the 10.4 meter Caltech Submil-limeter Observatory; 1.1 mm continuum observations obtained with Bolocam, and 350 µ mdata obtained with SHARC-II. These data are compared with a variety of previously pub-lished infrared and radio continuum data. The new data presented here are part of a 1.1 mmcontinuum Bolocam Galactic Plane Survey (BGPS) (Aguirre et al. 2010; Rosolowsky et al.2009; Dunham et al. 2009) that has mapped about 170 square degrees of the northern Galac-tic Plane. The BGPS data were released into the public domain on 24 June 2009 through the 6 –Infrared Processing and Analysis Center (IPAC) at the California Institute for Technology(http://irsa.ipac.caltech.edu/data/BOLOCAM GPS/).This paper presents an analysis of BGPS clumps in the region between l = 358.5 and+4.5 ◦ which contains the Galactic nucleus, the Central Molecular Zone (CMZ) of dense, tur-bulent, and warm molecular clouds (Morris & Serabyn 1996), and the most prominent high-velocity-dispersion cloud complex, Bania’s Clump 2 (Stark & Bania 1986) located at l = 3.2 ◦ .Additionally, 350 µ m images are presented of the two most luminous star forming complexesin the CMZ, Sgr A and its neighboring cloud complexes, and the Sgr B2 region. Previous(sub)millimeter studies of the Galactic center (Lis et al. 1991, 1994; Lis & Carlstrom 1994;Lis & Menten 1998; Lis et al. 2001) tend to be based on maps of small fields. The wide-fieldBGPS data are presented and compared to images obtained with the Spitzer Space Telescopeat 3.6 to 8.0 µ m (Arendt et al. 2008) and Spitzer 24 µ m images now in the public domain(Yusef-Zadeh et al. 2009). BGPS is also compared with the SCUBA / JCMT 450 and 850 µ m dust continuum maps (Pierce-Price et al. 2000) from an approximately 2 ◦ by 0.5 ◦ regioncontaining the CMZ. Finally, the 1.1 mm data are compared with the 20 cm radio continuum(Yusef-Zadeh et al. 2004) to constrain recent and on-going massive star formation activityin the CMZ and Bania’s Clump 2.This paper is organized as follows: Section 2 describes the observations and analysismethods, including a discussion of mass estimation (Section 2.1), and the determination ofdust properties for the two fields where 350 µ m observations are presented (Section 2.2).Section 3 presents the observational results. Section 3.1 presents a discussion of clump massspectra extracted from the Bolocat catalog (Rosolowsky et al. 2009). Section 3.2 describesthe emission from the Central Molecular Zone containing the Sgr A, B1, B2, and C complexesand its relationship to objects detected at other wavelengths. Section 3.3 presents a moredetailed discussion of the environment of Sgr A. Section 3.4 presents a discussion of theemission along the inner Galactic plane and Bania’s Clump 2. Section 3.5 presents a shortdiscussion of features likely to be located in front of or behind the center of the Galaxy in theGalactic plane. Section 4.1 explores the energetics of the cavity in the Sgr A region. Section4.2 discusses the clumps that may be fueling star formation in Sgr A. Section 4.3 discussesthe low star formation rate in Bania’s Clump 2. Conclusions are presented in Section 5.Detailed descriptions of some of the brighter or otherwise more noteworthy sources are givenin an Appendix. A tabulation of Bolocat clumps located within the field of view coveredby Figure 1 and 2 along with their derived properties is given in Table 3 in the electronicversion of this paper. 7 –
2. Observations
The observations reported here were obtained with Bolocam between June 2005 andJuly 2007 on the Caltech Submillimeter Observatory (CSO) 10-meter diameter telescope.SHARC-II observations were obtained during the best weather on these observing runs andin April 2008. The observations used in this paper were obtained on the dates listed inTable 1.Bolocam (Glenn et al. 2003) is a 144-element bolometer array with between 110 and115 detectors working during the observations reported here. The instrument consists of amonolithic wafer of silicon nitride micromesh AC-biased bolometers cooled to 260 mK. Alldata were obtained with a 45 GHz bandwidth filter centered at 268 GHz ( λ = 1.1 mm)which excludes the bright 230 GHz J=2-1 CO line. However, as discussed in Aguirre et al.(2010), the effective central frequency is 271.1 GHz. The individual bolometers are arrangedon a uniform hexagonal grid. Each bolometer has an effective Gaussian beam with a FWHMdiameter of 31 ′′ , but the maps presented here have an effective resolution of 33 ′′ due to beam-smearing by the scanning strategy, data sampling rate, and data reduction as discussedby Aguirre et al. (2010). At an assumed distance of 8.5 kpc to the Galactic center, the33 ′′ diameter Bolocam effective resolution corresponds to a length-scale of 1.36 pc and 1 ◦ corresponds to about 148.7 pc. The instantaneous field-of-view (FOV) of the focal-planearray is 7 ′ .5 but the effective filter function limits spatial frequency sensitivity to about 3 ′ to5 ′ . Bolocam observations were only obtained when atmospheric conditions were clear with a225 GHz zenith opacity ranging from 0.06 to 0.15. During better observing conditions, 350 µ m SHARC-II observations were obtained as discussed below.Data were obtained by raster-scanning Bolocam on the sky at a rate of 120 ′′ per second.Each ‘observation’ consisted of three pairs of orthogonal raster-scans each covering a fullsquare-degree or a 3 ◦ by 1 ◦ field in Galactic coordinates with scan-lines separated by 162 ′′ on the sky. The second and third pairs of orthogonal raster-scans were offset from the firstpair by ± ′′ . Four to over a dozen ‘observations’ taken on different days were averaged toobtain the maps presented here.The Bolocam raw data were reduced using the methods described in detail in Aguirre et al.(2010) based on the earlier techniques described in Enoch et al. (2006) and Laurent et al.(2005). The new reduction pipeline handles the large dynamic range and complex structurecharacteristic of Galactic star forming regions better. In summary, real-time pointing datawere merged with the time-series data from each bolometer whose exact location in the > few Jansky) sources and crowdingrequired the removal of 13 PCA components. Following removal of signals attributed to theatmosphere, the maps are then “iteratively mapped” to restore astronomical structure onangular scales comparable to the size of the Bolocam array (Enoch et al. 2006; Aguirre et al.2010). The final data were sampled onto a uniform 7 ′′ .2 grid. The data were flux calibratedusing observations of the planets Mars, Uranus, and Neptune. In the Galactic center fieldspresented here, the 3 σ r.m.s. noise is about 40 mJy beam − .The V1.0 data released at the IPCA site use 7 ′′ .2 pixels. The data value in each pixelcorresponds to the flux measured at that location on the sky by the 33 ′′ effective beam andis in units of Jy/beam. The area under a 33 ′′ gaussian beam is the same as the area undera top-hat beam with a radius of r B = F W HM/ (4 ln / = 19.8 ′′ or an aperture with adiameter of ≈ ′′ . There are πr B / . = 23.8 pixels in such an aperture. Thus, to convertthe data in the V1.0 images from Jy/beam to Jy/pixel where each pixel is a 7 ′′ .2 square,divide the data by 23.8. To measure the total flux over some aperture, sum the pixels valuesin the aperture (after converting to Jy/pixel units).Only flux produced by structure smaller than about one-half of the Bolocam array-sizeof approximately 7.5 ′ can be measured reliably. Larger-scale emission smoothly distributedin 2 dimensions is under-represented. However, long filaments or structure in which at leastone dimension is compact compared to the Bolocam array size can be recovered. A moredetailed discussion of the spatial transfer-function of the BGPS is given in Aguirre et al.(2010).Comparison between Bolocam and 1.2 mm data acquired with two other instruments,MAMBO on the IRAM 30 meter (Motte et al. 2003, 2007) and SIMBA on the SEST (Matthews et al.2009) 15 meter, are presented in Aguirre et al. (2010). (Aguirre et al. 2010) find that theBGPS Version 1.0 Bolocat and image data release nead a mean correction factor of 1.5 to beapplied to align BGPS fluxes with these previous measurements and to yield dust spectralindices when compared to data at other wavelengths.During the very best weather conditions at the CSO when the zenith sky opacity at 225GHz was less than about 0.06, observations of selected fields were obtained at 350 µ m withthe 384 element SHARC-II focal plane array (Dowell et al. 2003). SHARC-II consists of a 12by 32 pixel array of “pop-up” bolometers providing an effective resolution (beam diameter) of 9 –about 9 ′′ on the CSO. The SHARC-II 350 µ m filter is matched to the atmospheric window inthis part of the spectrum and has an effective passband of ∆ λ/λ = 0.13 (111 GHz) centeredat λ µ m.Observations of five slightly overlapping fields in the Sgr A region were obtained in June2005 using 10 ′ by 10 ′ BOX SCANs. A second set of observations were obtained on 6 April2008 using a 30 ′ by 30 ′ BOX SCAN covering the fields near Sgr A using a scan rate of 60 ′′ per second for a total integration time of 40 minutes. The Sgr B2 observations were obtainedon 31 March 2004 with a 13 ′ × ′ BOX SCAN with a scan-rate of 30 ′′ per second. Thetotal integration time was 20 minutes. While the June 2005 data were reduced with thefacility software package, CRUSH (Kov´acs 2008), the 350 µ m images shown in the figureswere processed with the code SHARCSOLVE written by co-I Dowell. Minimal filtering wasused; many iterations fit only an offset and gain to each bolometer time-stream, a spatiallyflat atmospheric signal at each time, ending with a few iterations of fitting an atmosphericgradient over the detector array.Due to their better uniformity, only the March 2004 and April 2008 images are used inthe figures. The June 2005 observations were used as an independent check on flux calibrationand image fidelity. Although there are slight differences in the background intensities, theoverall structure of the emission during the two epochs agrees extremely well. The fluxcalibration on compact objects such as G359.984,-0.0088] agrees to within 10% (29 vs 32Jy/beam) once an annular background is subtracted. The 3 σ sensitivity of these observationsis about 1Jy/beam at 350 µ m. Estimation of the dust column densities and masses from the 1.1 mm emission requiresknowledge of the dust properties such as emissivity, spectral index, and temperature, whichrequires input from additional data. Column densities, densities, and masses are estimatedfor a subset of the brightest clumps based on dust parameters appropriate for the Solarneighborhood (Table 2). Masses, column densities, and densities estimated for all Bolocatentries are tabulated in Table 3 in the electronic version of this paper. For all mass estimationgiven in this paper, the images and fluxes reported in the V1.0 data release have beenmultiplied by an empirically determined scaling factor of 1.5 as discussed above.The mass within the 33 ′′ BGPS effective beam (or equivalently, in a 40 ′′ diameter circularaperture) is given by M = 1 . × − S . ( J y ) D κ . B . ( T ) 10 –where B . ( T ) is the Planck function for dust temperature T , and κ . is the dust opacityat λ = 1.1 mm in units of cm g − . As a function of wavelength, κ λ = κ ( λ /λ ) β where β is the dust emissivity. Hildebrand (1983) calibrated κ at λ = 0.25 mm, giving a value κ . = 0.1 cm g − for interstellar dust and a gas-to-dust ratio of 100. For β = 1.5, theHildebrand opacity implies κ . = 1 . × − cm g − . Ossenkopf & Henning (1994) (OH94)calculated the opacity of grains with ice mantles that have coagulated for 10 years at a gasdensity of 10 cm − . Enoch et al. (2006) interpolated logarithmically the OH94 opacities toobtain κ . = 1 . × − cm g − , close to the estimate based on the Hildebrand opacity.As discussed in Aguirre et al. (2010) the effective central frequency of the Bolocam 1.1 mmfilter is 271.1 GHz. Log-interpolating the OH94 opacity at 1.00 mm to this frequency implies κ . = 1 . × − cm g − . Adopting this value gives M = 14 . e . /T ( K ) − S ( J y ) D kpc ( M ⊙ )where D kpc is in units of 1.0 kpc and T is in Kelvin.The greater metallicity of the Galactic center may result in a larger dust opacity, andpossibly a different dust emissivity law. This effect has not been taken into account in thepresent analysis. The spatial filtering inherent in BGPS, and large uncertainties inherent inthe estimation of masses using spectral lines, it is not at present possible to determine clumpmasses with sufficient precision to detect a difference in the dust properties. Use of a largerdust opacity would decrease the estimated masses and column densities given here.The dust temperature in the Central Molecular Zone has been studied by Sodroski et al.(1994) who used COBE/DIRBE 140 to 240 µ m data to conclude that 15 to 30% of theemission arises from molecular clouds with temperatures of about 19 K, 70 to 75% arisesfrom the HI phase with a dust temperature of 17 to 22 K and less than 10% comes from ∼
29K dust in the extended HII phase. Rodr´ıguez-Fern´andez et al. (2004) used the Infrared SpaceObservatory (ISO) to study a representative sample of 18 molecular clouds not associatedwith prominent thermal radio continuum emission. They found typical dust temperatures,measured from λ = 40 to 190 µ m, required a combination of a cold, 15 K componentand a warm component with a temperature ranging from 27 to 42 K. Thus, in the analysispresented here, the dust column densities and masses are parameterized in terms of a fiducialdust temperature of 20 K. For T = 20 K, M = 13 . S ( J y ) D kpc ( M ⊙ )which at the 8.5 kpc distance to the Galactic center gives M = 944 S ( J y ) M ⊙ . This masscan be converted to a column density averaged over the beam, N ( H ) = Mπµ H m H r B ( cm − ) 11 –where M is in grams, µ H = 2.8 is the mean molecular weight per hydrogen molecule(Kauffmann et al. 2008), r B = D GC × F W HM/ (205265 × √ ln
2) is the effective beam ra-dius in centimeters at the Galactic center, and D GC = 8.5 kpc. Following Kauffmann et al.(2008), the mean molecular weight per particle, µ p ≈ .
37 should be used when estimat-ing quantities such as collision rate or pressure. For BGPS,
F W HM ≈ ′′ , giving effectivebeam radius or R B = 19 . ′′ , which at a distance of 1 kpc corresponds to r B ( kpc ) = 2 . × cm and at the distance to the Galactic center is r B = 2 . × cm. The resulting columndensity per beam is N ( H ) = 2 . × [ e . /T ( K ) − S ( J y ) ≈ . × S ( J y ) ( cm − )where the right side was evaluated for T = 20 K. This column density corresponds to avisual extinction of A V ≈ . N ( H ) = 1 . × cm − per visual magnitude in the standard Johnson / Cousins Vfilter centered at 0.55 µ m . These column densities can be converted into surface densities;Σ( g cm − ) = 4 . × − N ( H ) or Σ( M ⊙ pc − ) = 0 . A V .For extended sources, masses are given by M = µm H N ( H ) A S where A S is the areasubtended by the aperture used to estimate the flux. At the Galactic center, the mass isgiven by M tot = 944 h S . ( J y ) i [Ω S / Ω B ] M ⊙ where < S . ( J y ) > is the flux averaged over thesolid angle Ω S subtended by the measurement aperture and Ω B = πr B ≈ . × − is theBGPS beam solid angle in steradians. A temperature of T = 20 K is assumed to evaluatethe numerical coefficient.An estimate of the volume density can be made by assuming that each clump is sphericaland uniformly filled with emitting matter. Table 2, columns 9 and 10 give the densitiesestimated from the fluxes measured in 40 ′′ and 300 ′′ diameter apertures under the assumptionthat each clump is either 40 ′′ or 300 ′′ in diameter. Interferometric measurement of clumps inthe Solar vicinity show that much of the millimeter continuum emission arises from regionshaving diameters of order 10 AU or less. Thus, BGPS clumps are likely to consist of clustersof unresolved objects much smaller than the beam. The BGPS-based density estimates arevolume averaged quantities. Inside cores, the densities are likely to be much higher whilein-between cores, densities may be lower.Table 2 lists the locations, fluxes, masses and densities for a sample of regions, manyof which are marked in Figure 9. S gives the flux measured with the FWHM = 33 ′′ Bolocam effective resolution at the location of the brightest emission in each source. Theuncertainties in the listed quantity range from about 20% to over 30% for bright sourcesdue to a combination of calibration errors, and incomplete flux recovery in the iterativemapping phase of data reduction. The 1.1 mm images have also been smoothed with a“tophat” function having a 20 pixel radius to provide an estimate of the total flux produced 12 –by extended objects in a 300 ′′ diameter, circular region. At the distance of the Galacticcenter (8.5 kpc), these masses are given by M (300) = 5 . × S ( J y ) M ⊙ for 20 Kdust where S ( J y ) is the average flux density in the 300 ′′ aperture. (Note that for severalhigh-latitude regions suspected to be in the foreground, a distance of 3.9 kpc is assumed inestimating masses as indicated in the comments column.) These mass estimates are lowerbounds because they only refer to the flux in small-scale structure to which Bolocam ismost sensitive. Because BGPS data starts to lose sensitivity to structure on scales rangingfrom 3 ′ to 6 ′ , 300 ′′ is the largest aperture for which reliable photometry can be obtained.As discussed below, comparison of mass determinations in the 300 ′′ aperture to previouslypublished measurements for regions such as Sgr A and Sgr B2 agree to within the estimateduncertainties. µ m and 1.1 mm data Two fields around Sgr A and Sgr B2 were observed at 350 µ m. The 350 µ m to 1100 µ mflux ratio can be used to constrain dust temperatures or emissivities. Three corrections mustbe applied to the SHARC-II data prior to comparisons with BGPS; matching of angularresolution, correction for signals-picked up in the sidelobes, and matching of the spatialfrequency response functions. No correction has been made for contamination by spectrallines.The SHARC-II 350 µ m images were obtained with a 9 ′′ FWHM beam. Thus, they haveto be convolved with a gaussian function with σ = 13.5 ′′ to match the 33 ′′ effective BGPSbeam. Because the data values in both the 350 and 1100 µ m images are in units of Janskysper beam, the pixel values in the convolved SHARC-II images must be multiplied by theratio of beam areas (13.4 when scaling from 9 ′′ to a 33 ′′ beam).Inspection of the 350 µ m beam delivered by the CSO on bright point sources such asplanets shows that in addition to the main 9 ′′ beam, there is a low-level error pattern severalarc-minute diameter with a roughly hexagonal shape. The main-beam efficiency of the CSOat 350 µ m is only about 0 . ± .
15. In a complex field such as the Galactic center withan abundance of bright, extended emission, fluxes will be over-estaimated when comparedwith a point source calibrator in the main beam due to the signal entering the sidelobes.Because bright emission near both Sgr A and Sgr B2 is extended on scales of many arc-minutes, we assume that about 30% of the signal at a typical location in the maps arisesfrom signal picked up in the sidelobes. As an approximate correction for this effect, we scalethe convolved 350 µ m images by a factor 0 . µ m maps of the Sgr A region discussed above shows thatthe rapidly-scanned SHARC-II images recover flux on larger angular scales than the 1.1 mmBGPS images. Thus, the SHARC-II images are spatially filtered with an ”unsharp mask”.The mask in constructed by convolving the 350 µ m with a gaussian function having a kernel σ k , and subtracting a fraction, f of the mask from the original image. The optimal valuesof σ k and f were chosen by finding the closest match between the spatial structure of theresulting ”unsharp-masked” SHARC-II image with the corresponding BGPS image. For theSgr A field, the best choice is σ k = 90 ′′ (FWHM = 212 ′′ ) and f = 0.8; for Sgr B2, σ k = 68 ′′ (FWHM = 160 ′′ ) and f = 0.5. The beam-matched, spatially filtered 350 µ m images weredivided by the corresponding 1.1 mm BGPS images to form flux-ratio maps.The observed intensity of continuum emission at each location in a map is given by S ν = Ω beam [1 − e − τ νd ] B ν ( T ) ( J y ) where Ω beam ≈ . × − sr is the beam solid anglecorresponding to the 33 ′′ effective resolution of the 1.1 mm data and the convolved, beam-matched, 350 µ m observations, τ ν is the optical depth of the emitting dust, and B ν ( T ) is thePlanck function at each frequency. The optical depth in the sub-mm to mm regime scaleswith frequency as ν β where β is the emissivity power-law index. For small grains, β is close to2 in the mm to sub-mm regime but can be smaller if the grains have experienced substantialgrowth or are coated by ice mantles (Ossenkopf & Henning 1994). In the optically thin limit,the flux-ratio at each point in a ratio map is given by R = S µm /S . mm = ν β [ exp ( hν . /kT d ) − ν β . [ exp ( hν /kT d ) − T d . A look-up table is generated for each choiceof β and T d and used to analyze points in the ratio maps. For hot β =2.0 dust where both350 µ m and 1100 µ m are in the Rayleigh-Jeans limit, the observed intensity ratio should bearound R = 100. Larger ratios require β >
2. Where the pixels have good signal at both350 µ m and 1.1mm, flux ratios have values R <
70, corresponding to dust temperatures lessthan 70 K for β = 2 dust. The interpretation of these results will be discussed below.
3. Results
The inner two degrees of the Galactic plane (Figures 1 and 2) contain the brightestmillimeter-wave continuum emission in the sky. The Central Molecular Zone (CMZ, whichconsists of the high-surface brightness region between l ≈ ◦ to 1.8 ◦ ) has a radius of about200 to 300 pc and contains the highest concentration of dense gas and dust in the Galaxy(Morris & Serabyn 1996). 14 –Figure 3 shows an image of the peak antenna temperature of the CO J = 1 – 0 linetaken from the data set presented by Bally et al. (1987, 1988). Comparison of Figures 1through 3 and maps of dense gas tracers such as CS (Miyazaki & Tsuboi 2000) shows thatthe brightest 1.1 mm continuum emission is closely associated with the brightest CO andCS emission.Figure 4 shows the spatial-velocity behavior of CO-emitting gas at positive latitudesintegrated from b = 0.20 to 0.48 ◦ . Emission from the Galactic disk is confined to velocitiesbetween V LSR = – 70 and +10 km s − and can be distinguished from the Galactic center gasby its relatively narrow line-widths of order ∆ V ∼ few km sec − . Foreground CO emissionis mostly associated with the nearby Sagittarius Arm (near V LSR = 0 km s − ), the Scutum-Centaurus Arm (near V LSR = -35 km s − ), and the 3 kilo-parsec Arm (near V LSR = -60km s − ). The systematic progression in (negative) radial velocity as one approaches theGalactic center provides evidence for the presence of a central bar (Binney et al. 1991). Thehigh velocity CO feature (near V LSR = 100 to 200 km s − ) between l = –1.5 ◦ and +2 ◦ hasbeen called the “expanding molecular ring” (EMR) which in the barred models of the Galaxyis interpreted as gas in x orbits located mostly along the leading edge of the central bar(labeled as “leading edge of the bar” in Figure 4) whose positive longitude side is thoughtto be closer to us (Morris & Serabyn 1996). The major axis of the bar is thought to have anangle with respect to our line-of-sight between 20 ◦ to 45 ◦ (Binney et al. 1991; Bissantz et al.2003; Ferri`ere et al. 2007; Pohl et al. 2008).Figures 3 and 4 show the remarkable molecular feature known as Bania’s Clump 2(Stark & Bania 1986). This feature is localized to a roughly 1/2 degree diameter ( ∼
70 pc)region in the spatial direction. But, as illustrated by Figure 4, it extends over a velocity rangeof order 200 km s − . Figures 1 and 2 show that Bania’s Clump 2 is prominent at 1.1 mm,consistent with the previous detection of emission by high density gas tracers. Throughoutthe Galactic plane, bright dust continuum sources generally indicate the presence of embed-ded massive stars or clusters. But, as discussed below, there is virtually no massive staror cluster formation in this region as indicated by the lack of infrared and radio continuumemission.The 1.1 mm data presented here are qualitatively similar to the recent ATLASGALsurvey of the Galactic plane conducted with the LABOCA array on the APEX 12 meterdiameter telescope in Chile at a wavelength of 870 µ m (Schuller et al. 2009). Althoughthe angular resolution of ATLASGAL is nearly a factor of two better than BGPS, theimages presented in Schuller et al. (2009) show similar structures to what is seen in BGPS.Specifically, the CMZ and Bania’s Clump 2 dominate the emission in the region between l =358.5 ◦ and 4.5 ◦ . A quantitative comparison of fluxes in ATLASGAL, BGPS, and eventually 15 –Herschel Space Observatory images, will enable the measurement of dust temperatures andemissivities. The Bolocat clump catalog contains the highest concentration of clumps in the skytowards the inner few degrees of the Galaxy, and the 1.1 mm emission is about 5 timesbrighter than adjacent parts of Galactic plane. The region between l = 0 ◦ to 1.5 ◦ contains542 Bolocat entries, or 361 objects per square degree. In contrast, the Galactic disk awayfrom the Galactic center between l = 5.5 ◦ and 11.5 ◦ contains 511 entries in 6 square degreesor 85 objects per square degree on average. This field is chosen for comparison since it likelyhas a similar surface density of foreground BGPS clumps as the Galactic center region. Thesymmetrically located field in the 4-th quadrant was not observed to comparable depth dueto its lower elevation as seen from Mauna Kea. The surface density of clumps is somewhatlarger around l = 20 ◦ to 35 ◦ where the line-of-sight passes through the tangent points inthe Molecular Ring. Thus, the surface density of clumps in the CMZ is about 4.3 timeshigher than in the plane. When Bolocat entries located at higher latitudes above and belowthe CMZ are excluded, the surface density of Bolocat clumps located close to the Galacticcenter is more than a factor of 5 higher. This is a lower limit since confusion is likely to makeBolocat have lower fidelity in the Galactic center. Thus, at least 80% of Bolocat entries inthe inner few degrees of the Galaxy are likely to be within 500 pc of the nucleus, located atapproximately 8.5 kpc from the Sun. Less than 20% are likely to be located in the molecularring at distances of 2 to 6 or 11 to 14 kpc from the Sun.The concentration of clumps in the CMZ permits the generation of a clump mass spec-trum for a large sample of objects located at a common distance between about 8 and 9 kpc.The formulae in section 2.1 are used to estimate individual clump masses from the peak fluxper beam in a 40 ′′ diameter (1.64 pc at D = 8.5 kpc) aperture tabulated in column 13 inBolocat. Figure 5 shows the mass spectrum using the assumption that all clumps have thesame temperature of T dust = 20 K. For masses above about 100 M ⊙ , the mass spectrum canbe fit by a power-law, dN/dM = kM α with an index α ≈ –2.14 ± M ⊙ , the slope is slightly steeper with α ≈ –2.58 ± M ⊙ , the slope of the mass function is α =–1.59 ± l − b and l − V maps; it is far less ‘lumpy’ than the BGPS mapswhich are dominated by small-scale strucutre. Thus, it is not surprising that the overlap inmass between the CS and BGPS mass-spectra is small or that the majority of CS featuresare considerably more massive than the BGPS clumps. The slope of the 1.1 mm BGPS massspectrum above about 100 M ⊙ is similar to the mass spectrum of low-mass cloud cores in theSolar vicinity which has a slope of around –2.3, similar to the stellar initial mass function(Motte et al. 1998, 2001; Andr´e et al. 2007; Enoch et al. 2008); it is considerably steeperthan found for CO-emitting GMCs in the Galactic plane.Because at the distance of the Galactic center, most BGPS clumps will form clustersrather than individual (or multiple) stars, it may be more appropriate to compare the BGPSmass spectra with the mass spectra of star clusters. Observations of Galactic star clusters(Battinelli et al. 1994) and OB associations (McKee & Williams 1997) reveal steep massspectra scaling as M − . Massive young star clusters in the merging galaxy pair known asthe ‘Antennae’ also have dN/dM = N M − (Zhang & Fall 1999) in the mass range 10 . to 10 . M ⊙ . Similar results have been found for clusters in M51 (Bik et al. 2003) andolder clusters in the Large Magellanic Cloud (de Grijs & Anders 2006). Nearby low-massstar-forming cores, more distant and more massive BGPS clumps in the Galactic center,and Galactic and extra-galactic star clusters share the common trait of having steep massspectra. That all three classes of object are gravitationally bound may have something todo with this trait.The BGPS data acquisition and reduction pipeline results in images with highly atten-uated response to extended structure, effectively acting as a high-pass filter that attenuateslow-spatial frequencies (Aguirre et al. 2010). Furthermore, the Bolocat watershed algorithm 17 –(Rosolowsky et al. 2009) tends to subdivide sources with complex structure into multipleclumps. The Sgr B2 complex provides a good example. While single dish maps in tracerssuch as CO or other high-dipole moment molecules such as CS show that this region consistsof a single cloud complex about 0.2 ◦ by 0.3 ◦ in extent (e.g. Figure 3), the BGPS imagessuch as Figures 9 highlight local maxima and interconnecting ridges. Bolocat subdividesthe Sgr B2 complex into several dozen clumps. Attenuation of low spatial frequencies com-bined with the watershed algorithm of Bolocat may remove objects from the high-mass,large-extent portion of the mass-spectrum and re-distribute the flux into a larger numberof lower-mass entries. This has the effect of steepening the power-law index of the derivedmass-spectrum. Detailed simulations to quantify this effect are needed and will be the sub-ject of a future paper. In complex fields such as the Galactic plane, line-of-sight blending ofphysically unrelated sources may lead to further confusion which can alter the slope of themass function. We urge the reader to exercise great caution in the interpretation of massspectra of diffuse objects.Bolocat (Rosolowsky et al. 2009) tabulates fluxes in four different ways. In additionto the flux in a 40 ′′ aperture, fluxes are also measured in 80 ′′ and 120 ′′ apertures. Thecatalog also tabulates the total flux in the area assigned to each clump along with a beam-deconvolved “effective” radius of the clump (the radius of a circle that has the same area asthe clump). Mass-spectra based on fluxes measured in larger apertures have shallower slopesby about 0.1 to 0.2.The mass-spectrum derived from the Bolocat fluxes measured in a 120 ′′ diameter aper-ture using a constant dust temperature of 20 K has a power-law index α = –2.04 (Figure 6),slightly shallower than the index based on measurements in a 40 ′′ aperture. This differencemay be due to increased source confusion within the larger aperture. The relatively smallslope change may suggest that source blending and confusion already affects the slope basedon fluxes in the smaller aperture. The slope also depends slightly on the widths of the binsused to generate the histograms.A curious puzzle about the Galactic center is that the dust temperature appears to belower than the gas temperature despite the high average density. Both the COBE and ISOsatellites determined that the average dust temperature in the CMZ ranges from 15 to 22K (Reach et al. 1995; Lis et al. 2001). Thus, the use of T dust = 20 K in mass estimationis justified. However, the dense gas associated with this dust is considerably warmer with80% of the ammonia having temperatures between 20 and 80 K and 18% warmer than 80 K(Nagayama et al. 2007).Is the slope of the derived mass spectrum sensitive to variations in T dust ? Two types ofdust temperature variations are considered to answer this question: First, the dust temper- 18 –ature of a clump is assumed to decrease with increasing projected distance from Sgr A as apower law. Second, the dust temperature of each clump is assigned a value depending onthe observed flux; a 100 Jy source such as Sgr B2 is assumed to have dust with T = 50 Kwhile a 1 Jy source is assumed to have T = 20 K. This corresponds to a dust temperaturevarying with flux, F , as F . .Given a source of luminosity L located at a distance r from a grain, the mean graintemperature depends on the luminosity L and distance r as T ( r ) = T L /γ g r − /γ g (Scoville & Kwan 1976) where T is a normalization constant. For blackbodies larger thanthe emitted and absorbed wavelengths, γ g = 4. For interstellar grains, γ g ranges between 5and 6 for grain emissivity power-law indices ranging from 1 to 2. Assume that the heatingluminosity L is generated by stars in Galactic disk and bulge, that the luminosity-to-massratio is constant, and grains at a given distance r from the Galactic center are heated mostlyby the stars within a sphere of radius r centered on the Galactic center. Thus, the enclosedluminosity, L ( r ), is a function of r which can be estimated by assuming that the densityof stars decreases with increasing distance from the Galactic center as ρ ( r ′ ) = ρ r ′− δ . For δ = 2 (the singular isothermal sphere), integrating the mass distribution from r ′ = 0 to r implies that the mass enclosed within a radius r is proportional to r . Thus, the luminosity L ( r ) is also proportional to r . Inserting this in the above equation for T ( r ) gives a simpleapproximation for the radial variation of the grain temperature T ( r ) = T r − /γ g where T is normalized to a temperature at some radius. This estimate applies to grainswhich are located mostly at cloud surfaces or otherwise are exposed to the visual and near-IR component of starlight. As a concrete example, choose β = 1 and γ g = 5 so that T ( d ) = T r − . because it yields reasonable cloud surface temperatures over a wide range ofGalactocentric distances. Taking T = 50 K at r = 10 pc implies T = 27.5 K at r = 200 pc,15.5 K at r = 3 kpc, and 12.5 K at 8.5 kpc. This parametrization provides reasonable basetemperatures at the surfaces of clouds lacking internal or close-by heating sources. It is usedto test the sensitivity of the mass spectrum to a systematic decrease of grain temperaturewith increasing galacto-centric distance.Systematic temperature variations do not have a large impact on the derived slopeof the mass spectrum. However, the elevated temperature does decrease the estimates oftotal mass. The mass-spectrum for γ g = 5 corresponding to a radial temperature gradientexpected for a grain emissivity that decreases with wavelength as a power-law with index β = 1 has a slope of α = –2.44 (Figure 7). 19 –Finally, it is possible that internal heating by massive stars results in an average dusttemperature that is a function of 1.1 mm flux. To determine the consequences of such a modelon the derived mass-spectrum, we assume that the dust temperature varies as T = T F . . where the normalization constant T is chosen to give a dust temperature of 20 K for a 1.1mm flux density per beam of 1 Jy. This model results in a steeper mass spectrum shown inFigure 8 with a slope of α = -2.64.These results show that the mass spectrum of mm clumps having masses above about70 M ⊙ is represented by a steep power-law comparable to low-mass star forming cores withmasses below about 10 M ⊙ (Motte et al. 1998, 2001; Andr´e et al. 2007). The derived power-law index is relatively insensitive to assumptions about the dust temperature. In summary,the power-law index of Bolocat clumps in the Galactic center is 2.14 (+0.4,–0.1). The mass-spectrum in the Galactic center is steeper than GMC mass-spectra in the inner galaxy andLMC and comparable to the outer Galaxy and M33 (Rosolowsky 2005). However, greatcaution must be exercised in the interpretation of these mass spectra owing to the dataacquisition and reduction methodology.The total H mass of the 1428 Bolocat clumps between l = –1.5 ◦ to 4.5 ◦ , measured in a40 ′′ aperture, assuming that all are located at a distance of 8.5 kpc, have a dust temperatureof 20 K, and multiplied by the empirically determined scaling factor of 1.5 needed to bringBGPS fluxes into agreement with MAMBO (Motte et al. 2003, 2007) and SIMBA (Matthewset al. 2009) gives M ( H ) ∼ . ± × M ⊙ . Assuming a temperature gradient thatdecreases with distance from Sgr A as proposed above gives the lowest total mass of order ∼ . ± . × M ⊙ . Assuming a dust temperature that scales with detected 1.1 mmflux as described above gives a total mass of ∼ . ± . × M ⊙ in a 40 ′′ aperture.In the 120 ′′ aperture, (assuming a constant dust temperature of 20 K) this mass grows to ∼ . ± . × M ⊙ . Using the total integrated flux (Bolocat column 19) within the effectiveradius (Bolocat column 12) results in a total mass ∼ . ± . × M ⊙ .Summing all BGPS flux in this field above 100 mJy gives a mass of about ∼ . ± . × M ⊙ . The total mass of molecular gas within 500 pc of the center has been estimated tobe between 3 to 8 × M ⊙ (Sodroski et al. 1994; Dahmen et al. 1998; Tsuboi et al. 1999).Thus, the total 1.1 mm emission from clumps in a 40 ′′ aperture represents about 0.7% to2.2% of the total mass in molecular gas while the mass in a 120 ′′ aperture is about 7% to18% of the total. Summing the flux above 100 mJy correponds to 8% to 21% of the totalmolecular gas mass. This is about a factor of several larger than what was found in theGem OB1 association toward the Galactic anticenter (Dunham et al. 2009), indicating thata larger fraction of the molecular gas near the Galactic center is in compact clumps.It is important to recall that BGPS filters out extended emission on scales larger than 20 –about 5 ′ and that the 1.1 mm data is only sensitive to dust in compact structure. Suchspatial-filtering is not present present in the single-dish spectral line data which thereforecan trace gas in extended regions. Thus, the small mass fraction detected by BGPS maybe a consequence of the spatial filtering or diffuse, extended emission. Spatial filtering ofthe spectral line data to match BGPS could be used to constrain the relative abundances ofdust and the tracers producing the lines. Figure 1 shows an image of the 1.1 mm emission from the Galactic center region. Figure 2shows this image as contours superimposed on the Spitzer Space Telescope IRAC 8 µ m image.As at most long IR, sub-mm, and radio wavelengths, the area-integrated 1.1 mm emissionis about 5 times brighter than in adjacent portions of the inner Galactic plane. Whilemost of the Galactic plane mm-wave continuum emission is dominated by discrete clumps(Aguirre et al. 2010; Rosolowsky et al. 2009; Dunham et al. 2009), the CMZ contains brightarcminute-scale extended structures interconnected by a lacy network of filaments. Figure9 shows the inner 2 square degrees of the Galactic plane at 1.1 mm. Figure 10 shows thesame field at 8.0 µ m as observed by the Spitzer Space Telescope (Arendt et al. 2008) withcontours of 1.1 mm emission superimposed.The chain of bright 1.1 mm clumps extending from Sgr B2 (G0.68,–0.03) to G0.26+0.03and the surrounding lower level diffuse emission is seen in silhouette against the brightbackground in the infrared image. The 20 ′ halo of 1.1 mm emission surrounding Sgr B2 canalso be seen in silhouette against the infrared background. Thus, most of BGPS emissionbetween l = 1.0 ◦ and 0.25 ◦ must originate in front of most of the stars and mid-infrared lightin the central bulge of the Galaxy. This is consistent with the suspected orientation of thecentral bar whose positive longitude part is closer.Several features of the dust continuum emission from the CMZ are apparent. First,most of the 1.1 mm emission with a flux density greater than about 1 Jy / beam originatesfrom a few dozen clumps (in the Bolocat clump catalog, many of these objects are subdividedfurther). The two brightest peaks correspond to Sgr B2 and the Sgr A*. At lower levels, theCMZ can be decomposed into hundreds of individual clumps and filaments. Bolocat contains1428 entries in the field covered in Figure 1. Second, most of the 1.1 mm emission in theCMZ closely follows the distribution of dense molecular gas traced by high-dipole momentmolecules and the large line-width CO emission (Bally et al. 1987, 1988; Oka et al. 1998,2001b). In the inner square degree, both the 1.1 mm dust continuum and the moleculartracers of the CMZ exhibit a smaller scale-height than emission from the Galactic disk 21 –measured beyond | l | > ◦ . Near Sgr A ( l ≈ ◦ ), the vertical extent (orthogonal to theGalactic plane) of the brightest BGPS emission and broad-line molecular gas is only about2 ′ to 3 ′ (5 to 8 pc); near Sgr B2 around l ≈ ◦ this layer has a thickness of about 10 ′ to15 ′ (25 to 40 pc). Beyond | l | ∼ ◦ , the scale-height of 1.1 mm clumps associated with theCMZ increases dramatically. Around l = 1.3 ◦ and 3.2 ◦ the gas layer is nearly 60 ′ (150 pc)thick. Away from the Galactic center, between l = 5 ◦ and 35 ◦ the dust layer has an averagethickness of about 0.8 ◦ or 125 pc. The spatially averaged 1.1 mm flux from the Galacticcenter features is about 3 to 5 times higher than the average flux from the roughly ± ◦ scale-height clump population of the Molecular Ring which is at least 3 kpc in front of (orbehind) the Galactic center. Third, many 1.1 mm features can be seen in silhouette in the8 µ m IRAC and 24 µ m MIPS images. Fourth, at levels up to several hundred mJy/beam, acomplex network of filaments forms a frothy background of emission which surrounds voids.The filaments may either trace the walls of stellar-wind bubbles or SNR, the warm edges ofmolecular clouds, cavities produced by HII regions, or “fossil” cavities whose energy sourcesare extinct.Comparison with the 20 cm radio maps (Figure 11) shows that only a small subset ofthese 1.1 mm cavities are associated with radio continuum emission. Figure 11 shows the20 cm emission imaged with 30 ′′ resolution (Yusef-Zadeh et al. 2004) superimposed on the1.1 mm image. The Sgr A, B1, B2, C, and D regions that contain young massive stars andstar clusters stand out in both the radio continuum and at 1.1 mm. The non-thermal radiocontinuum filaments tend to be located adjacent to 1.1 mm clumps. For example, the largestand brightest group of filaments near l = 0.18 ◦ crosses the Galactic plane at the positivelatitude end of the ridge of clumps associated with the 50 km s − cloud that stretches from l = 0 ◦ to 0.17 ◦ . The fainter non-thermal filament located directly above (in latitude) Sgr Aknown as N5 (Yusef-Zadeh et al. 2004) passes through the cluster of clumps associated withthe HII region G359.96,+0.17. The brightest part of filament N8 abuts a diffuse 200 mJyBGPS clump at [359.78,0.17] (not marked).Some of the diffuse 1.1 mm emission between l = 0.24 ◦ and –359.6 ◦ located at positivelatitudes is visible in emission at 8 µ m. At positive longitudes, this region contains thethermal “arched filaments” near the Arches cluster. The 1.1 mm emission may either tracewarm dust or free-free emission from dense, photo-ionized plasma illuminated by the ArchesCluster. At negative longitudes, 1.1 mm clumps associated with the high-latitude, positivevelocity portion of the “expanding molecular ring” ( b = 0.0 ◦ to about 0.2 ◦ ; V LSR = 100to 200 km s − ) are also associated with 8 µ m emission. At negative longitudes, the 20km s − cloud located south of Sgr A* is visible in silhouette and must therefore be in frontof the central cluster. A pair of oppositely facing comet-shaped clouds near the right edge ofFigures 9 through 11 (associated with Sgr C and G359.62–0.24) are both seen prominently 22 –in extinction. Except for these clouds, few of the 1.1 mm sources at negative longitudes arevisible in extinction in the IRAC data.The fainter 1.1 mm continuum is structured and organized into a filamentary networkof arcs and circular features. This is most evident in left half of Figure 9. Some rings andpartial rings surround radio continuum features and therefore probably mark dust heated inPDRs and the edges of cavities created by massive stars and star clusters. However, mostrings, arcs, and filaments are not associated with known HII regions or supernova remnants.Massive stars that have shed their natal cocoons are hard to separate from the multitudeof intrinsically fainter foreground stars, because in the IR distant massive stars and closerlower-mass stars have similar colors and spectral shapes. These cavities may also be theremnants of older generations of stars and clusters whose most massive members have eitherexploded as supernovae more than a million years ago so that all radio continuum has longsince faded, or whose stars were ejected by dynamical processes that create run-away, high-velocity stars. Run-away stars are most common among spectral type O (Gies & Bolton1986; Gies 1987); many of these stars were ejected by dynamic interactions in dense clusters(Gualandris et al. 2004). The shells and filaments may be tracers of fossil star and clusterformation in the CMZ. However, it is also possible that randomly superimposed filamentscreate the impression of cavities.The large kidney-bean-shaped clump at [0.26,0.03] (see Figures 9 and 12) is remarkablefor being one of the brightest features at 1.1 mm and for being the most prominent infrareddark cloud (IRDC) in the entire Galactic center region (Liszt 2009). This feature is seen inabsorption at 24 µ (Figure 14). Although Bolocat lists two entries for this object (Nos. 96and 99), dense gas tracers such as CO, CS, and HCN indicate that the entire structure isat approximately the same radial velocity (V LSR ≈
40 km s − with a line-width extendingfrom 20 to 60 km s − ). Liszt (2009) use the MSX data to set a lower bound on the columndensity of N(H ) > . × cm − (A V >
53 magnitudes), consistent with the 1.1 mm-based column density estimate in a 40 ′′ aperture. The higher resolution 350 µ m observations(Figure 12) indicate that it has considerable sub-structure with consequently larger columndensities.Figure 13 shows the ratio map formed from the 350 µ m and 1.1 mm data in the vicinityof Sgr A. In regions with significant emission at both wavelengths, there is a general trendthat the clump centers where the fluxes peak have the smallest flux ratios. The flux ratiostend to increase towards the clump edges. Ratios at a number of locations in Figure 13 areindicated by the blue arrows. In Figure 13, the flux in the 1.1 mm BGPS maps have beenscaled-up by a factor of 1.5. Without this scaling, a large fraction of the pixels have ratioslarger than 100. Figure 13 shows that clumps in the CMZ have warmer dust than either 23 –foreground or background clumps in the Molecular Ring.Sgr B2 is the brightest clump in the entire BGPS survey. Figure 15 shows the 350 µ mimage. As shown in Table 2, it has the largest total mass ( > × M ⊙ ), highest volume-averaged density ( n ( H ) > cm − ), and greatest column density ( N ( H ) > × cm − in 33 ′′ beams centered on Sgr B2 Main and North) of any BGPS clump. Thus, theaverage extinction towards Sgr B2 M and N corresponds to at least 1,000 magnitudes atvisual wavelengths. Molecular line measurements indicate that the Sgr B2 complex has amass of at least 5 × M ⊙ , at least a factor of 10 greater than implied by the total 1.1mm flux detected by BGPS and listed in Table 2 (Jones et al. 2008). Over 50 compactHII regions and powerful masers indicate that a major burst of star formation is occurringhere (Gaume et al. 1995; de Pree et al. 1998; De Pree et al. 2005). The molecular line dataprovide evidence that the intense burst of star formation occurring in Sgr B2 may havebeen caused by the supersonic collisions between two giant Galactic center molecular clouds(Jones et al. 2008; Liszt 2009). Sgr B2 may be located near the high-longitude end of ‘x2’family of orbits in a bared potential (Contopoulos & Papayannopoulos 1980) where it mayencounter gas moving along the ‘x1’ family of orbits. Sgr B2 is located at the high-longitudeend of a chain of 1.1 mm clouds, mid-IR extinction features, embedded 24 µ m sources, andHII regions starting near the less-extincted and older Sgr B1 complex (Yusef-Zadeh et al.2009). Sgr B2 may represent the most recent event in a chain of sequential star formationthat started near Sgr B1. The Sgr B2 complex may be giving birth to a massive star cluster oreven a super star cluster comparable to the somewhat older Arches and Quintuplet clusterslocated in the CMZ.Figures 17 and 18 show the BGPS 1.1 mm contours superimposed on the Spitzer 8 and24 µ m images respectively. These images clearly show that most of the dust associated withthe Sgr B2 region is in front of the bulk of the 8 and 24 µ m emission. Furthermore, severalof the roughly circular cavities in the BGPS images are translucent in the IR data. Thelarge Sgr B1 complex of HII regions and infrared sources has created a giant cavity on thelow-longitude and low-latitude side of Sgr B2. Highly obscured nebulosity and IR sourcesabut the the Sgr B2 sub-mm/mm peaks on the low-latitude side. However, no IR sourcescan be seen at the locations of Sgr B2N and Sgr B2M. More discussion of the Sgr B2 regionis given in the Appendix.Figure 16 shows the 350 µ m / 1.1 mm flux ratio map derived from the 350 µ m imagematched in spatial frequency response to BGPS divided by a 1.1 mm map scaled-up by afactor of 1.5. No correction for line contamination has been applied. Sgr B2 is likely to bemost affected by such contamination; Lis & Goldsmith (1991) found about 30% of the 350GHz flux is due to lines and contamination. The two brightest peaks, corresponding to Sgr 24 –B2 North and Sgr B2 Middle (or Main) have the smallest flux ratios (25 and 34, respectively)which for β = 2 dust, implies temperatures of 14 – 17 K. It is likely that the dust is internallyheated by forming massive stars to at least 20 K, and possibly 40 K (Lis et al. 1991). Asthese authors proposed, it is possible that the low flux ratio is the result of an unusually low β ; for β = 1.5, ratios of 25 and 34 imply 21 K and 33 K. ∗ and its Environment The second most prominent peak in the 1.1 mm data (Sgr B2 is the first) is centeredon Sgr A* which is thought to mark the location of the 4 . ± . × M ⊙ black hole atthe Galactic center (Genzel et al. 2003; Ghez et al. 2008). The extended radio source knownas Sgr A consists of Sgr A East, a several arc-minute diameter oval of non-thermal radiocontinuum emission thought to trace a supernova remnant, and Sgr A West, an HII regionwhich consists of a “mini-spiral” of dense, thermal plasma with the central black hole, SgrA* at its center. The “mini-spiral” is surrounded by the ∼ µ m image shown in Figure 12.Emission from Sgr A* peaks near a wavelength of 1 mm (Serabyn et al. 1997). Be-tween 1993 and 1996 the peak flux near this wavelength was around 3.5 Jy (Figure 2 inSerabyn et al. (1997)). In July 2005, Sgr A* was about 1.6 times brighter at 1.1 mm witha flux density of 5.7 Jy. This may be an underestimate because the BGPS pipeline maynot fully recover the flux of bright sources or those surrounded by an extended envelope(Aguirre et al. 2010; Rosolowsky et al. 2009). Diffuse emission produced by dust or hotplasma along the line of sight contributes at most 1 Jy per beam. Thus, the increasedbrightness of Sgr A* in the 1.1 mm maps is most likely due to variability which is consistentwith a non-thermal origin for most of the 1.1 mm continuum emission from that source.Figure 12 shows the 350 µ m SHARC-II image of the 30 ′ field surrounding Sgr A. Figure14 shows contours of 350 µ m emission superimposed on the Spitzer 24 µ m image. In contrastto 1.1 mm, there is very little 350 µ m emission from Sgr A* itself. At 350 µ m, Sgr A*has a flux density between 0.5 and 1.5 Jy depending on how the large background flux issubtracted. While emission from the CND and Sgr A* are blended at 1.1 mm by the 33 ′′ effective resolution, in the 350 µ m image, the CND is resolved (Figure 12) and appearsas a nearly complete oval. Several dozen discrete clumps and filaments can be identifiedin the region surrounding the ring in addition to the prominent 20 and 50 km s − clouds.Interferometric observations of molecules such as HCN resolve the CND into a collection ofdense clumps (Christopher et al. 2005). The detection of maser emission and hyper-compact 25 –sources of radio continuum emission indicates on-going massive star formation in the CND(Yusef-Zadeh et al. 2008).The central 30 ′ region contains several chains of 1.1 mm clumps and filaments elongatedtowards Sgr A (turquoise lines in Figures 9 through 11 red lines in Figures 12 and 14). Thesechains are narrower at the end facing Sgr A and may therefore be cometary clouds sculptedby this source. The two most prominent chains are located about 1 ′ south of the Archescluster and 2 ′ south of the Quintuplet cluster and Pistol Star; these chains are each about6 ′ to 8 ′ long. A shorter, fainter chain is located between the Arches and Quintuplet clusters,and a nearly 10 ′ long chain is located about 3 ′ north of the Arches cluster. These featuresare also seen in the 350 µ m image (Figures 12 and 14) where they break-up into complicatedsubstructures with an overall elongation towards or away from Sgr A.About a half dozen compact clumps located between the IRDC at [0.26,0.03] and theQuintuplet cluster near l ∼ ◦ are also elongated along the Galactic plane and maybe smaller cometary clouds less than 1 ′ to 2 ′ in length. These are best seen in Figure 9.Figure 11, shows that these clumps are located at the high-longitude side of the prominentnon-thermal filaments that cross the Galactic plane at l ≈ ◦ .Mid-infrared images from the MSX and ISO satellites reveal a very bright, limb-brightennedbubble of infrared emission surrounding the Quintuplet and Arches clusters (Moneti et al.2001). Inspection of the 24 µ m images obtained by the Spitzer Space Telescope (programP20414; PI - F. Yusef-Zadeh) also shows the prominent elliptical bubble with a major axisbounded towards positive latitudes by the Arched Filaments around [ l, b ] ≈ [+0.11 ◦ ,+0.10 ◦ ]and toward negative latitudes by a rim at [ l, b ] ≈ [+0.12 ◦ ,–0.18 ◦ ] and extending from l ≈ µ m ovoidis indicated by the large ellipse in Figures 9 through 14.The chains facing Sgr A are in the projected interior of the infrared “bubble” surroundingthe Arches and Quintuplet clusters. It is likely that this bubble traces the walls of a cavitythat extends from the non-thermal filaments near l ≈ ◦ to at least l ≈ .0.0 ◦ near the 50km s − cloud and contains both the Arches and Quintuplet clusters and possibly the Sgr Acomplex. Figure 11 shows that this region is filled with diffuse 20 cm continuum emission.Figure 12 shows this region at 350 µ m.The dashed line in Figure 2 outlines the boundary of a large region relatively devoidof clouds as traced by 8 or 24 µ m emission. This feature may trace an extension of thebubble associated with the Arches and Quintuplet clusters that opens-up above and belowthe Galactic plane. While this feature can be traced to the edge of the observed field aboveSgr A, below the Galactic plane it becomes confused with the foreground bubble centered 26 –on [ l, b ] = [+0.41 ◦ ,–0.50 ◦ ]. An extended bubble blowing-out of the Galactic plane may havebeen created by energy release from Sgr A, Arches, and the Quintuplet regions. The Galactic center contains several remarkable cloud complexes which exhibit line-widths ∆
V >
100 km s − over a region typically about 0.5 ◦ in diameter (Bania et al.1986; Stark & Bania 1986; Oka et al. 1998; Liszt 2006, 2008). Although Bania’s Clump 1 at[ l, b ] ≈ [–355 ◦ ,+0.4 ◦ ] and Clump 2 [ l, b ] ≈ [+3.2 ◦ ,+0.2 ◦ ] were the first to be noted, similarbut less distinct, large line-width complexes exist near [+1.3 ◦ ,+0.2 ◦ ] and near [+5.0 ◦ ,+0.2 ◦ ](Oka et al. 1998, 2001a; Dame et al. 2001; Bally et al. 1987, 1988). Bania’s Clump 2, locatedat a projected distance of about 450 pc from the Galactic center, is the most prominent ofthese features with a line width of ∆ V ≈
200 km s − . The feature extends from the mid-planeto about b = +0.8 ◦ and consists of about 16 CS-emitting clumps each with masses of order5 × M ⊙ and H densities in excess of 2 × cm − (Stark & Bania 1986) surrounded byan envelope of CO emitting material. Such dense and massive clumps are usually associatedwith on-going massive star formation.Figure 19 shows the BGPS 1.1 mm map of the region around l = 3 ◦ . Some BGPS datawere obtained up to b = ± ◦ that shows several additional features associated with Bania’sClump 2. These data are noisier than the BGPS data along the plane and are not shownhere. The 1.1 mm clumps coincide with the CS and CO clouds in Bania’s Clump 2. Figure20 shows the Spitzer Space Telescope 8 µ m image from GLIMPSE (Benjamin et al. 2005)with superimposed 1.1 mm contours. The most striking result is the absence of extendedinfrared sources or bubbles associated with most Clump 2 features. In contrast to Clump2, 1.1 mm clumps located along the Galactic plane tend to contain infrared sources at 4.5,and 5.8 µ m, extended 8 µ m emission, and prominent bubbles, indicating massive star andstar cluster formation. Inspection of the 24 µ m MIPS scan-maps available from the SpitzerScience Center archives also shows a general absence of mid-IR sources in Bania’s Clump 2.This result is consistent with the lack of 20 cm radio continuum emission (Yusef-Zadeh et al.2004).Comparison of the 1.1 mm and Spitzer images shows that there is a slight depression inthe spatially-averaged 8 and 24 µ m emission from the bulge of the Galaxy at the location ofthe 1.1 mm features in Clump 2. The Clump 2 dust features are weak infrared dark clouds(IRDCs) indicating that most of the dust in this region must be somewhat in the foregroundof the bulge. This is consistent with the suspected location of the Clump 2 material in thenear-side of the central bar. 27 –Two clumps located in Bania’s Clumps to appear to be foreground objects. The 1.1mm clump located at [ l, b ] = [–3.405,+0.877] is associated with IRAS 17470-2533 at thehigh-latitude end of Clump 2. Located several arcminutes southwest of the clump, there isan arc-minute diameter bubble of 8 µ m emission which contains a small cluster. The BGPSclump is seen as an IRDC projected in front of the background star field and diffuse emission.Its high latitude and appearance as an IRDC is consistent with this clump and adjacent HIIregion being located in the foreground in the Galactic disk. The bright BGPS clump at [ l, b ]= [+3.34,+0.44] is associated with IRAS 17484-2550 and the clump CO003.34+00.44 in theGalactic center survey of Oka et al. (1998, 2001b). The radial velocity ranges from 0 to 40km s − . A low-contrast IRDC is associated with this feature. Several prominent 1.1 mm clumps are located at relatively high Galactic latitudes | b | > ◦ . Some of these clumps are associated with prominent HII regions or bubbles visible inthe IRAC images. The Spitzer 8.0 µ m images (Figure 10) reveal a prominent bubble centeredon the 1.1 mm clump located at [ l, b ] = [0.41 ◦ ,–.50 ◦ ] (Figure 9). In the 8.0 µ m Spitzer image,the bubble appears as a nearly circular ring of emission with a radius of about 7 ′ ( ∼
17 pc)with a dark region centered on the 1.1 mm clump. The 1.1 mm clump is located at theapex of a cometary cloud seen at 8 µ m that projects from more negative latitude toward thecenter of the bubble (Figures 9 and 10). The bubble is rimmed by several 1.1 mm clumps,the brightest of which is located at [ l, b ] = [0.28 ◦ , –0.48 ◦ ]. Several additional 1.1 mm clumpsare located at larger projected radii from the center of the bubble along its rim. Figure 11shows that the bubble interior is lit-up with faint 20 cm radio continuum emission. Thisregion is reminiscent of the IC 1396 bubble in Cepheus with its intruding tongue of moleculargas. Several other relatively low-latitude 1.1 mm clumps such as [ l, b ] = [359.71,–0.37], [ l, b ] =[359.91,–0.31], and [ l, b ] = [359.97,–0.46] form a chain below the bulk of dust associated withthe Galactic center. This chain of clouds and those associated with the Spitzer ring havebeen recently noted by Nagayama et al. (2009) who identified them in near-IR extinctionmaps as well as in CO data. These authors estimate distances using the cumulative numbersof stars and reddening to determine distances and column densities. The distance to thischain of clouds located about 0.2 ◦ to 0.4 ◦ below the Galactic plane ranges from 3.6 to 4.2kpc. K-band extinctions range from 0.4 to 1.0 magnitudes. In Table 2, we assume a uniformdistance of 3.9 kpc for mass estimation.Another high-latitude region is associated with the compact cluster of 1.1 mm clumps 28 –listed in Table 1 and in the Figures as [ l, b ] = [359.96,+0.17]. The Spitzer image (Figure 10)shows bright 8.0 µ m emission at the high-longitude (northeast) side of the cluster of BGPSclumps. This region may be a foreground HII region complex exhibiting a “champagne flow”towards high Galactic longitudes from a molecular cloud containing several dense clumps.The clumps at [ l, b ] = [0.53,+0.18] and [ l, b ] = [0.84,+0.18] are located well away fromthe dense-gas layer of the CMZ which suggests that they may be foreground clumps in theGalactic plane far from the Galactic center region. Thus, for these objects, as well as thecluster of clumps near [ l, b ] = [359.96,+0.17], we assume a distance of 3.9 kpc.
4. Discussion4.1. A Cavity Extending 30 pc from Sgr A?
Millimeter and sub-mm imaging of the CMZ reveals a network of dust filaments, shells,and dense clumps. The ∼
30 pc diameter infrared bubble between Sgr A and l ≈ ◦ is associated with a 1.1 mm and 350 µ m cavity that contains the Quintuplet and Archesclusters and Sgr A. Several chains of clumps in the bubble interior point towards Sgr A.The high combined luminosity of the Arches and Quintuplet clusters and of circum-nuclearcluster surrounding Sgr A* can plausibly produce the cavity. Below, the physical parametersof this feature are estimated.The central few hundred parsecs of the Galaxy contain at least 4 different types of gasreservoir; cold (T < K) and dense molecular clouds (n(H ) > cm − ) that comprisethe CMZ (Morris & Serabyn 1996; Oka et al. 1998), a warm inter-cloud medium (T ≈ to10 K) traced by species such as H +3 in absorption (Oka et al. 1998, 2005; Goto et al. 2008)and diffuse CO and CI in emission (Martin et al. 2004), a warm ionized component (T ≈ K) traced by recombination and fine structure lines, radio scattering (Lazio & Cordes1998), and free-free emission and absorption, and an ultra-hot component (T > K)traced by X-rays (Wang et al. 2002).The volume filling factor of the dense, gravitationally bound cold molecular phase is inthe range 0.01 (Cotera et al. 2000) to 0.1(Morris & Serabyn 1996). The remaining volumeis either filled with gravitationally unbound (to individual clouds or star clusters) warmneutral atomic and molecular, photo-ionized, or X-ray emitting gas (Martin et al. 2004;Oka et al. 2005; Goto et al. 2008). Abundant 3 to 4 µ m H +3 absorption towards massivestars in the central 0.2 ◦ of the Galaxy indicates that warm, but mostly neutral gas unboundto any particular cloud may occupy a large fraction of the volume in the CMZ. However, thelarge angular-diameters of background extra-galactic radio sources produced by scattering 29 –of their radio continuum emission by a foreground screen of electrons (Lazio & Cordes 1998)observed towards the CMZ indicates the presence of relatively dense HII regions towardsmost lines-of-sight through the CMZ. Yusef-Zadeh et al. (2007) show that 6.4 keV Fe K α emission fills the projected area of the cavity delineated by the Spitzer 24 µ m bubble andthe 1.1 mm cometary clouds (see their Figure 13). However, the strongest 6.4 keV emissionappears to be associated with the molecular clouds traced by the 1.1 mm continuum. Theseauthors suggest that the X-ray emission is produced by the interaction of low-energy cosmicrays with dense clouds.We consider a scenario in which UV radiation and winds emerging from the Arches,Quintuplet, and Sgr A* clusters and the central black hole are responsible for carving outthe cavity. The massive Quintuplet and Arches clusters may have contributed to formationof the individual cells in which these clusters are embedded. Towards positive longitudes theouter-boundary of the cavity is near l ≈ ◦ at a projected distance of about 30 pc fromSgr A* (Figure 9). As shown in Figure 12, the high longitude end of the cavity is marked bya chain of small dust clouds located around l = 0.22 ◦ (northeast of the Quintuplet clustermarked with a “Q” and below the bright 1.1 mm cloud at [0.26 ◦ ,+0.03 ◦ ]). The cell associatedwith the Quintuplet cluster may extend to l = 0.28 ◦ , or about 40 pc from Sgr A. Below (tothe right of) Sgr A, the intense emission from the 20 and 50 km s − clouds masks the cavityif it extends this far. The 20 and 50 km s − clouds are physically close to Sgr A; they mayblock the penetration of UV irradiation and winds. The Spitzer 24 µ m image shows thatthe low-longitude end of the cavity is near l = 0 ◦ .An order-of-magnitude estimate of the parameters of an ionized cavity can be obtainedfrom its size and the 20 cm radio continuum flux, which imply that the average electrondensity in the cavity can not be much above 100 cm − . Assuming that the cavity interior isionized and has a uniform hydrogen density, the Lyman continuum luminosity (in units ofthe number of ionizing photons emitted per second) required to ionize the cavity is L ( LyC ) =3 × n r where n is the average ionized hydrogen density in units of 100 cm − and r is the mean radius of the cavity in units of 10 pc. The above parameters imply an ionizedmass of about M HII ≈ . × n r Solar masses around the Arches, Quintuplet, andSgr A.The Lyman continuum luminosity of the Sgr A region has been estimated from the fluxof free-free radio continuum to be around L ( LyC ) ∼ f ew × ionizing photons per second.The entire CMZ has a Lyman continuum luminosity of around 1 − × ionizing photonsper second (Morris & Serabyn 1996). This radiation field is produced mostly by youngmassive stars with a possible contribution from the central black hole in the immediatevicinity of Sgr A. Dust within the cavity and the CMZ may absorb a substantial fraction of 30 –the Lyman continuum radiation. However, even if nearly 90% is absorbed, the UV radiationfield is sufficient to maintain the central cavity in an ionized state for a mean density of order10 − cm − .The orbital time scale in the center of the Galaxy sets a constraint on the age of thecavity and its cometary clouds. Using a typical orbit velocity V o = 120 km s − , a test particlein a circular orbit would have an orbital period of t o = 2 πr/V o ∼ r V − Myr where V is the orbit velocity in units of 120 km s − and r is the distance from the Galactic centerin units of 10 pc. Portions of clouds at different radii that are gravitationally unbound fromeach other would be sheared on a time-scale comparable to the orbit time. The presenceof recognizable cometary clouds implies that they have been ionized on a time-scale shorterthan the orbit time. On the other hand, dense clouds are expected to orbit on largely radialtrajectories. Thus, the observed cometary features may be objects that have only recentlyfallen into the interior of cavity with a speed comparable to the orbit speed. In this case, theelongation of these features away from Sgr A may be a result of the survival of dense gas anddust in the region shielded by the dense head facing Sgr A. Alternatively, this cavity may beonly recently ionized by the Arches and Quintuplet clusters. If these clusters have high spacevelocities, they may have completed several orbits around the nucleus during their lifetimes(Stolte et al. 2008) and the cometary clouds in their vicinity may be a consequence of thismotion. The formation mechanism of massive stars within the central parsec of the Galaxy hasbeen a long-standing mystery. Wardle & Yusef-Zadeh (2008) present a model in which a gi-ant molecular cloud passes through the center. Gravitational focusing of trajectories by thecentral black hole and the central stellar cluster of older stars causes gas to collide and dissi-pate angular momentum and kinetic energy in the wake. This results in the trapping of somegas in eccentric, circum-nuclear orbits, forming a disk of compressed material. Gravitationalinstabilities in the disk can result in the formation of massive stars.It has been long suspected that the two most prominent molecular clouds near Sgr A,the 20 and 50 km s − clouds, are located within the central 30 pc (Morris & Serabyn 1996;Coil & Ho 1999; Wright et al. 2001; McGary et al. 2001; Christopher et al. 2005; Lee et al.2008). High-dipole moment molecules such as HCN and CS show that they are among thedensest and warmest clouds in the CMZ (Miyazaki & Tsuboi 2000). Together, these twoclouds give the CMZ its apparently small (few arcminute) scale-height near Sgr A. 31 –The 50 km s − cloud is located about 10 pc in projection from the CND towards higherlongitudes. Several 350 µ m filaments connect the 50 km s − cloud to the CND (Figure 12);these features correspond to ammonia streamers seen by McGary et al. (2001). Lee et al.(2008) present a 3D model based on near-infrared observations of H and argue that the 2pc radius CND was recently fueled by the passage of the 50 km s − cloud within a few pc ofthe nucleus. Assuming that the 50 km s − cloud is on an x2 orbit, the projected separationof about 10 pc from Sgr A* implies that for a proper motion of 100 km s − , it passed bythe nucleus about 10 years ago. The passage of the 50 km s − cloud may have formed theCND, leading to the current episode of massive star formation activity within its clumps.The 20 km s − cloud is the most prominent object at 350 µ m in the inner square degreeof the Galactic center (Figures 12 and 14); it is located about 7 to 20 pc from the nucleusin projection. It is elongated with a length-to-width ratio of about 5 to 10:1 with a majoraxis that points about to 2 ′ below the CND (Figure 12). Table 2 shows that for an assumedgrain temperature of 20 K, the 20 km s − cloud is the second most massive and densest cloudin the CMZ (Sgr B2 is the most massive and dense). A bright filament of 350 µ m and 1.1mm dust emission connects the 20 km s − cloud to the CND; this feature is likely to be thesame as the one detected in ammonia by Coil & Ho (1999). This structure provides supportfor the model in which CND continues to be actively fueled, currently by flow from the 20km s − cloud (Coil & Ho 1999).If the center of the 20 km s − cloud is currently located about 20 pc from the nucleus,and on a near collision course with the nuclear cluster, it may provide fuel for furtherstar formation in the nucleus in the near future. Assuming that this cloud is on an x2orbit (Binney et al. 1991; Marshall et al. 2008; Rodriguez-Fernandez & Combes 2008) whosevelocity is mostly orthogonal to our line-of-sight, and has speed of about 100 km s − , it willcome closest to the nuclear region in a few hundred thousand years.The IRS 16 cluster of massive stars surrounding Sgr A* could not have formed from anycurrently visible cloud. The IRS 16 cluster has an age of at least several Myr; Tamblyn & Rieke(1993) give an age of 7 to 8 Myr. The formation of a previous circum-nuclear ring wouldhave required the passage of a cloud about 5 to 10 Myr ago. Any such cloud would havemoved too far from the nucleus to have a clear connection to it and may have been disruptedby a combination of tidal forces, UV radiation and winds. 32 – Figure 19 shows the 1.1 mm image of the region containing Bania’s Clump 2. Figure20 shows this image superimposed on the Spitzer Space Telescope IRAC image. Theseimages show that despite the presence of dozens of clumps emitting brightly at 1.1 mm,Clump 2 lacks significant mid-infrared emission from either extended or point sources. Itis not forming stars at a rate comparable to similar clumps elsewhere in the Galaxy. Inthe Galactic plane, HII regions and nebulae excited by young, massive stars are frequentlyseen at 3.6 to 5.8 µ m. The 8 µ m Spitzer images trace O, B, and A stars by the presenceof bubbles rimmed by bright emission from UV-excited PAH molecules and nano-particles.Few of these signatures of massive stars are observed towards Bania’s Clump 2. The longerwavelength Spitzer images (e.g. 24 µ m) trace highly embedded young stars and clusters.Inspection of the 24 µ m Spitzer images shows that Bania’s Clump 2 also lacks a substantialpopulation of embedded sources visible at 24 µ m. Perhaps the clumps are embedded ina medium having an unusually high pressure or large non-thermal motions which suppressgravitational instabilities on the mass-scale of stars or even clusters. Many of the Clump 2features are seen in silhouette against background star light and diffuse emission from theISM in the 3.6 to 8 µ m Spitzer images, indicating that they are in front of much of thecentral bulge.The absence of emission in the Spitzer images is not the result of high extinction.Regions having similar amounts of 1.1 mm emission closer to the mid-plane of the Galaxyshow a factor of 2 to 4 times more emission between 6 and 24 µ m. Furthermore, thebrightest 1.1 mm emission maxima towards Bania’s Clumps 2 imply localized extinctionsof A V <
10 magnitudes averaged over a 40 ′′ aperture. However, the low-spatial frequencyemission resolved-out by BPGS may amount to as much as A V = 30 magnitudes based onthe line-of-sight column density of CO. But, this would be insufficient to completely hide allSpitzer mid-IR emission.The exceptionally large line-widths and velocity extent of this complex may be causedby three processes: First, the abrupt change in the orientation of the orbital velocity where aspur or inner dust lane encounters the ridge of dense gas at the leading edge of the Galacticbar. Second, at the ends of the so-called “x1” orbits in a tri-axial potential where they becomeself-intersecting (Contopoulos & Papayannopoulos 1980; Binney et al. 1991; Marshall et al.2008; Rodriguez-Fernandez & Combes 2008). In either mechanism cloud-cloud and cloud-ISM collisions result in high ram-pressures, shocks, and turbulence that can produce largeline-widths and non-thermal motions. Third, broad-line features such as Bania’s Clump 2may consist of unrelated clouds along an x1 orbit which has a leg aligned with our line-of- 33 –sight. In this scenario, Bania’s Clump 2 traces gas in a dust lane at the leading edge of thebar which is elongated along our line-of-sight.Figure 21 shows a cartoon of the possible configuration of the central regions of the MilkyWay as seen from the north Galactic pole. The x1 orbits are elongated along the major axisof the Galactic bar in a frame of reference rotating with the bar. As a family of orbits withdecreasing semi-major axes (corresponding to lower angular momenta about the nucleus)are considered, the ends first become more “pointed”, then they become self-intersecting.Clouds whose orbits decay eventually enter the regions of self-intersection where they mayencounter other clouds or the lower density ISM on the same trajectory with supersonicvelocities. The resulting shock waves and consequent dissipation of orbital energy causes theclouds to rapidly migrate onto the smaller “x2” orbits whose major axes are orthogonal tothe bar. Most of the observed dense gas in the CMZ is thought to occupy the near-side ofthe X2 orbits at positive longitudes where they are seen as IRDCs.Models of gas flows in a tri-axial potential can reproduce the major features observed inthe longitude-velocity diagrams of molecular tracers of the inner Galaxy (Bissantz et al. 2003;Rodriguez-Fernandez et al. 2006; Rodriguez-Fernandez & Combes 2008; Rodriguez-Fernandez2009; Pohl et al. 2008; Liszt 2009). It is suspected that the major axis of the central bar iscurrently oriented within 15 ◦ to 45 ◦ of our line-of-sight. The conventional interpretation ofthe Galactic center molecular line emission places the high-velocity rhombus evident in l -Vdiagrams in the “180 pc molecular ring” (sometimes called the “Expanding Molecular Ring”or EMR - see Figure 4 in Morris and Serabyn 1996). This feature is tilted with respect tothe Galactic plane so that the positive latitude part is seen at positive velocities. In Figure4, this feature is labeled as the “Leading Edge of the Bar” since in most models, the gasthat is transiting from x1 to x2 orbits is found in such a dust lane (see Figure 5 in Morrisand Serabyn 1996).Bania’s Clump 2 is located at higher longitudes than the 180-pc molecular ring. Thus, itis either located where the x1 orbits become self-intersecting, or where a spur encounters thedust lane at the leading-edge of the bar, or traces gas in the dust lane which is oriented nearlyparallel to our line-of-sight. Rodriguez-Fernandez et al. (2006) shows the likely location ofBania’s Clump 2 in a hypothetical face-on view of the Galaxy (see their Figure 2). Thesuperposition along the line-of sight of clouds moving near the apex of the innermost, self-intersecting x1 orbits can exhibit a large range of radial velocities within a very restrictedspatial region. This phenomenon is caused by the near co-location of clouds moving towardsand away from the Galactic center. Some of these clouds will be colliding, resulting incompression, heating, and turbulence generation. While low-velocity cloud-cloud collisionscan trigger star formation, high-velocity collisions are likely to be disruptive; sufficiently fast 34 –collisions dissociate molecules and raise post-shock temperatures.The presence of 1.1 mm clumps and associated gas seen in high-dipole moment moleculessuch as CS and HCN suggests that the smallest mass that can become gravitationally un-stable is comparable to the observed clump mass of 10 to 10 M ⊙ . The low-rate of starformation in these clumps implies that fragmentation is suppressed in this environment,possibly by shear or large-fluxes of energy and momentum from large-scales to small.If cloud collisions and supersonic internal motions suppress star formation, the dissi-pation of even a small fraction of this energy must emerge as radiation, resulting in theexcitation of fine structure cooling lines such as the 63 µ m [OI], 157 µ m [CII], and 205 µ m[NII], or high-J lines of CO and OH. A substantial portion of the internal energy may alsogo into heating grains which would emerge as sub-mm continuum radiation. The HerschelSpace Observatory may be able detect the resulting fluxes.While such shocks may promote star formation in other environments such as in SgrB2, they have failed to do so here. Perhaps the mm sources in Bania’s Clump 2 are in anearlier, pre-stellar stage of evolution. Clumps could evolve to denser configurations capableof star formation in a few crossing times. Using a typical Galactic center cloud line-widthof ∆ V = 20 km s − for each clump, a clump with a radius of r = 1 ′ ( ∼ t ∼ R/ ∆ V f ∼ . × f − years where f (= 0.1 to 1) is the area fillingfactor of gas in the clump. Thus, it is possible that we are seeing the Clump 2 objects in anearly state of evolution and that they will become active star formers in the next few millionyears.The Clump 2 complex has a diameter of about 0.5 ◦ (70 pc), implying R ∼ . × cm (25 pc), and a velocity extent of about 2 V = 200 km s − . The time-scale on which theentire Clump 2 region should evolve is about t ∼ R / ∆ V f ∼ . × f − years. Figure19 implies an area filling factor of about f ∼ ∼
15 Myr orbitaltime-scale at the projected speration between Sgr A ∗ and Clump 2.
5. Conclusions
The first wide-field λ = 1.1 mm continuum map of the inner six degrees of the Galaxyis presented. This emission traces dozens of bright star forming clumps and an extended,nearly continuous filamentary network. Filaments tend to outline a lattice of voids. Somecontain bright, diffuse nebulosity in infrared images that may indicate the presence of HII 35 –regions. A few may contain supernova remnants. However, many cavities do not containobvious energy sources. They may trace fossil cavities carved in the Galactic center ISM bythe action of massive stars that have died in the recent past. The large velocity dispersionof gas in the Galactic center region combined with the shear of differential rotation wouldtend to erase unsupported cavities on a relatively short time-scale. Thus, the cavities requireenergy injection within the last few Myrs. The four main results of this study are:First, the 1.1 mm Bolocam Galactic Plane Survey (BGPS) reveals over a thousandindividual clumps within a few degrees of the Galactic center that are associated with theCentral Molecular Zone (CMZ). Their small scale height compared to the rest of the Galacticplane and association with the dense molecular gas in the CMZ indicates that over 80% ofthese clumps are likely to be within the central few hundred parsecs of the Galaxy. The 20and 50 km s − clouds near Sgr A, and the Sgr B2 complex, host the brightest 1.1 mm sourcesin the field of study. Comparison of the 1.1 mm emission with 350 µ m maps of the regionsaround Sgr A and Sgr B2 is used to constrain the dust emissivity, grain temperatures,and optical depths. These measurements indicate peak 1.1 mm optical depths of around0.1, and grain temperatures around 20 to 70 K for an emissivity power-law index of 2.0.Sgr B2 appears to have an atypically low emissivity index for reasonable dust temperatures,indicating that substantial grain growth has occurred. Both the 350 µ m and 1.1 mm emissionprovide evidence for about a half dozen elongated, cometary dust clouds with dense headspointing towards the Sgr A complex. Some of these features delineate the walls of a multi-celled cavity having a mean radius of at least 30 pc. The most prominent cometary dustclouds delineate cells located near the massive Arches and Quintuplet clusters. These dustfeatures point towards Sgr A, and indicate that UV radiation from the central cluster ofmassive stars has sculpted the Galactic center environment within the last 1 Myr. Themassive 20 and 50 km s − clouds appear to be located in the interior of this cavity and mayshield the nuclear ISM at low longitudes and latitudes from ionization.Second, the collection of clouds located between l = 3 ◦ and 3.5 ◦ and V LSR = 0 to 200km s − known as “Bania’s Clump 2” are highly deficient in active star formation given their1.1 mm dust continuum fluxes. It is possible that the clumps in Clump 2 are in a pre- starforming state. This observation provides evidence for models in which Clump 2 consists ofgas and dust that has recently been shocked and rendered highly turbulent. This feature mayeither trace gas that has recently fallen in from the innermost x1 orbit (elongated along thebar) onto the outermost x2 orbit (elongated orthogonal to the bar), or gas that is enteringthe shock on the leading edge of the bar from a spur, or a dust lane at the leading-edge ofthe bar that is elongated along our line-of-sight.Third, under the assumption of a constant dust temperature in the CMZ, the mass 36 –spectrum of clumps M >
70 M ⊙ in the Galactic center is characterized by a steep power-law dN/dM = kM − . − . , +0 . . This index is significantly steeper than found from CSobservations of the Galactic center and similar to the index found for star forming cores inthe Solar vicinity. While such a steep index may be a real feature of the Galactic clumppopulation, it may in part be an observational artifact produced by the attenuation oflow spatial-frequency flux, source confusion, and the watershed algorithm used to constructBolocat which tends to subdivide large objects into clusters of smaller ones.Fourth, comparison of the 350 and 1100 µ m images places new constrains on dustproperties. Values of β between 1.5 and 2.0 give the most reasonable dust temperatureswith values ranging from 15 K to as high as 80K. However, Sgr B2 is anomalous. The lowflux ratios here require lower values of β , providing evidence for substantial grain growth.The BGPS project is Supported in part by the National Science Foundation throughNSF grant AST-0708403. J.A. was supported by a Jansky Fellowship from the NationalRadio Astronomy Observatory (NRAO). The first observing runs for BGPS were supportedby travel funds provided by NRAO. Support for the development of Bolocam was provided byNSF grants AST-9980846 and AST-0206158. NJE and MKD were supported by NSF grantAST-0607793. This research was performed at the Caltech Submillimeter Observatory 9CS),supported by NSF grants AST-05-40882 and AST-0838261. CB and CC were supported bya National Science Foundation Graduate Research Fellowship. We thank Farhad Yusef-Zadeh for providing his 20 cm radio image. We thank an anonymous referee for very helpfulcomments that improved the manuscript. 37 –
6. Appendix I: Comments on Individual Regions
In this section, individual regions Listed in Table 2 are discussed in order of increasingGalactic longitude. Fluxes and mass estimates for the marked features (based on visualinspection of the images rather than Bolocat) are given in Table 2.
The Great Annihilator (1E1749.7-2942) ; [ l, b ] = 359.116,–0.106): Long suspected to bea major source of 511 keV emission from decaying positronium, this hard X-ray and γ -ray source was initially suspected to be stellar-mass black hole accreting from a molecularcloud (Bally & Leventhal 1991; Mirabel et al. 1991). However, recent observations indicatethat this object may be a low-mass X-ray binary (Main et al. 1999). The molecular cloudlocated adjacent to 1E1749.7-2942 was recently observed at millimeter wavelengths with theCARMA interferometer by Hodges-Kluck et al. (2009) who provides some mass estimates.The molecular cloud is detected in our 1.1 mm image (Figure 22). Thus, we can comparethe dust-based mass estimate with the gas-based estimate. Interestingly, while the clumpnear 1E1749.7-2942 is clearly seen, the brightest HCN and HCO + clump in the maps ofHodges-Kluck et al. (2009) is not evident in the 1.1 mm dust continuum. The 1.1 mm fluxat the location of their HCO + and HCN peak 2b is only about 30 mJy, comparable to thenoise in our maps.The masses estimated for the cloud adjacent to the Great Annihilator from the 1.1mm emission in 40 ′′ and 300 ′′ apertures are 2 . × and 7 . × M ⊙ (for fluxes scales-up by a factor of 1.5). The mass in the smaller aperture was measured at the locationof the HCO + peak found by Hodges-Kluck et al. (2009). On the other hand, the mass inthe larger aperture was determined in an aperture centered on the 1.1 mm peak at [ l, b ] =359.135, –0.100. Hodges-Kluck et al. (2009) use the Virial theorem to estimate the massfrom the clump-size and line-width for their clump 1 (at the location shown by the smallcircle immediately next the the Great Annihilator in Figure 22). They find M = 3 . × M ⊙ , in between the two BGPS estimates in 40 ′′ and 300 ′′ apertures. Sgr C ([ l, b ] = 359.47,–0.11): There are relatively few major star forming complexesat negative Galactic longitudes. The exception is the Sgr C complex of HII regions. TheBGPS images reveal a prominent comet shaped cloud clump complex facing west in the SgrC region.[ l, b ] = 359.47,–0.03: This feature is a 1.1 mm clump located at the low-longitude endof a chain of clumps that can be trace back to just above Sgr A. This chain is part of theso-called “Expanding Molecular Ring” that may mark gas and dust at the leading edge ofthe Galactic center bar. In CO spatial-velocity diagrams, this feature is part of the positivelatitude, positive velocity feature that defines the “rhombus” thought to mark the innermost 38 –x1 orbit (marked in Figure 4). A 20 cm non-thermal filament crosses the Galactic plane onthe low-longitude side of the marked clump in Figure 11.[ l, b ] = 359.62,–0.24: Located 15 ′ east of Sgr C, there is a second, east-facing cometarycloud in the 1.1 mm continuum image. The dense head of this cometary feature abutsthe HII region G359.62-0.25 which consists of a series of 20 cm filaments east of the 1.1 mmclump. Extended emission east of the clump is evident in the GLIMPSE 8 µ m image (Figure10). The 1.1 mm clump is the brightest portion of a roughly east-west ridge that extendswest towards Sgr C. Both the Sgr C and G359.62-0.24 cometary clouds are clearly seen insilhouette against background 8 µ m emission.[ l, b ] = 359.71,–0.37: A compact 1.1 mm knot associated with a cometary feature in theGLIMPSE images. This is one of the clumps associated with the band of foreground cloudslocated at about 3.9 kpc from the Sun (Nagayama et al. 2009).[ l, b ] = 359.91,–0.31: This is another of the clumps associated with the band of fore-ground clouds located at about 3.9 kpc from the Sun (Nagayama et al. 2009). It is a cometarycloud facing the Galactic plane that is visible in silhouette against the 8 µ m background inthe Spitzer GLIMPSE images. The 20 km s − cloud. ([ l, b ] = 359.88,–0.08): The central twenty parsecs of the Galaxycontains the Sgr A radio complex and the 20 and 50 km s − clouds; both clouds are seen insilhouette in the Spitzer 8 images (Figure 10). The 50 km s − cloud located east-northeastof Sgr A* contains bright MSX point sources while the 20 km s − cloud lacks bright MSXcounterparts. Thus, the former is more evolved and is actively forming stars while the latteris more similar to IRDCs and likely to be either in a pre-star forming stage of evolution, orits star formation may be inhibited by its proximity to the Galactic nucleus.The 20 km s − cloud is the third brightest source of 1.1 mm emission in the CentralMolecular Zone (next only to Sgr B2 and and Sgr A*) and has a mass of at least 1 . × M ⊙ .The cloud is elongated along a direction pointing towards Sgr A*. The 350 µ m image (Figure12) shows complex structure consisting of several bright arcs of dust protruding above andbelow the cloud. Although there is abundant extended 20 cm continuum emission in thegeneral vicinity, no radio features can be directly associated with the cavities outlined bythese arcs. Filaments and clumps of dust connect this cloud to the Circum-Nuclear Disk(CND) surrounding Sgr A* and to the 50 km s − cloud discussed below. As discussed inSection 4.2 this cloud may be plunging towards the central parsecs of the Galaxy on an x2orbit. If it is on a radial trajectory towards Sgr A*, it is the most likely cloud to inject newgas into the central few parsecs of the Galaxy. The G359.94,+0.17 Complex : Located 0.2 ◦ directly above (in Galactic latitude) the 39 –Galactic nucleus, this complex is associated with several compact 8 µ m bubbles. It containsIRAS 17417-2851, a compact cluster of at least four prominent 1.1 mm clumps adjacent toan HII region complex that wraps around the clumps with PDRs and ionization fronts on thenorth, east and south sides. The non-thermal filament, N5 in Figure 11 (Yusef-Zadeh et al.2004) is located along the line-of-sight to this clump. However, because of its high-latitudelocation and visibility in the Spitzer IRAC data as both an IRDC and as a bright region ofemission at 8 µ m, we assume that it is a foreground complex at an approximate distance ofabout 3.9 kpc (Table 2). The 50 km s − cloud. ([ l, b ] = 359.98,–0.08): Located only a few arc-minutes east ofSgr A*, the passage of this cloud near the Galactic center may have fueled the injection ofgas into the circum-nuclear ring about 1 to 2 × years ago. The cloud contains severalcompact radio sources (Goss et al. 1985; Yusef-Zadeh et al. 2004, 2008) and evidence for starformation. Tendrils of dust continuum emission appear to link this cloud to both the CNDand to the 20 km s − cloud discussed above.[ l, b ] = 359.97,–0.46: An isolated cometary cloud associated with an 8 µ m bubble,presumably an HII region. This clump appears to be part of band of foreground cloudslocated at about 3.9 kpc from the Sun (Nagayama et al. 2009).[ l, b ] = 0.05,–0.21: A 1.1 mm clump located directly below Sgr A* and the 50 km s − cloud with a prominent Spitzer nebulosity on its low-longitude side. The Quintuplet and Arches Clusters ([ l, b ] = 0.163,–0.060 and [ l, b ] = 0.121,+0.018): Oneof the most striking features of the 20 cm image are series of concentric shells centered at l, b = [0.15,–0.06] near the location of the “Quintuplet” cluster of massive stars (Figer et al.1999). The innermost ring may trace the ionization front and PDR associated with thiscluster; the northeast part of this ring is known as the “Sickle” due to its radio morphology(Lang et al. 2005). The outer, northwest portion of these concentric rings comprise the“Arched” filaments thought to be illuminated by the massive Arches cluster (Figer et al.2002; Stolte et al. 2005).While there is no 1.1 mm clump at the location of the Quintuplet cluster, an extensivenetwork of clumps is located adjacent to and to the northeast. The Quintuplet appears tobe in a cavity which is consistent with its relatively evolved status. The 4 to 5 Myr oldQuintuplet cluster contains one of the most luminous stars yet discovered: the Pistol Star.The VLA 20 cm continuum map shows a prominent ionization front north and east of thiscluster that is associated with a ridge of 1.1 mm dust continuum emission along the Galacticplane.The Arches Cluster is another of the most massive young star clusters in the Galaxy 40 –(Figer et al. 2002; Stolte et al. 2005) and is associated with a diffuse 1.1 mm source. Thethermal radio continuum filaments known as the ‘Arches’ trace ionization fronts and photon-dominated regions (PDRs) illuminated by this cluster. Several clumps of 1.1 mm emissionare embedded within the ridges of radio continuum emission. Fainter filaments of 1.1 mmcontinuum follow the Arches about 10 ′′ to 30 ′′ due east of the radio continuum ridges. Asshown by Serabyn et al. (1999), the radio emission traces the western edges of a chain ofmolecular clouds seen in CS. The 1.1 mm emission is located between the radio filamentsand the CS clouds, indicating a layer of warm dust on the side of these clouds facing theArches cluster. This layer is most likely associated with a PDR. (Yusef-Zadeh et al. 2003).Some of the 1.1 mm emission may be free-free emission from the plasma. Wang et al. (2006)proposed that the Arches cluster is colliding with the molecular clouds in this region. [ l, b ] = 0.26,+0.03 : The “lima bean’” shaped molecular cloud extending from [ l, b ] =0.23,+0.01 to 0.26,+0.03 is remarkable for being a bright 1.1 mm source not associatedwith radio continuum sources or any other indicators of on-going star formation such asbright sources in MSX or Spitzer images. In the 2MASS, Spitzer, and MSX images, GCM0.25+0.01 is seen as the most prominent absorption feature in the CMZ (Figure 10). Thecloud qualifies as an infrared-dark-cloud (IRDC) which does not yet contain evidence ofstar formation. In the SHARC II 350 µ m images GCM 0.25+0.01 is resolved into a dozenindividual clumps which could be precursors to star-forming clumps. Lis & Menten (1998)find broad molecular line profiles typical of Galactic center Giant Molecular Clouds (GMCs)and a line center velocity of 20 to 40 km s − similar to the GMCs known to be interactingwith the non-thermal filaments that cross the Galactic plane at l = 0.18 near the Pistol andQuintuplet Clusters. GCM 0.25+0.01 may be the best example of a high column densityGMC in a pre-star forming state in the inner molecular zone of the Galaxy.[ l, b ] = 0.28,–0.48: This is the brightest clump in a cluster of clumps associated with theband of foreground clouds located at about 3.9 kpc from the Sun (Nagayama et al. 2009).This clump is located on the rim of the large Spitzer 8 µ m bubble centered on the clump at[ l, b ] = 0.41,–0.50. [ l, b ] = 0.28,–0.48 is associated with its own bright Spitzer nebulosity whichmay indicate the presence of a compact HII region whose birth may have been triggered bythe expansion of the Spitzer bubble.[ l, b ] = 0.32,–0.20: A clump of 1.1 mm emission associated with a bright but compactcluster of extended Spitzer nebulosity. The clump lies at the low-longitude end of fan-shapedgroup of at least three filaments that converge on this object and which can be traced for atleast 0.1 ◦ towards higher longitudes in both 1.1 mm emission and 8 µ m extinction.[ l, b ] = 0.41,–0.50: The easternmost 1.1 mm clump associated with the band of fore-ground clouds located at about 3.9 kpc from the Sun (Nagayama et al. 2009). This particular 41 –clump is located at the high-latitude tip of a large pillar of dust and is centered on the 0.15 ◦ (10 pc) radius ring of 8 µ m Spitzer emission.[ l, b ] = 0.41,+0.05: The low-longitude end of the chain of clouds that extend towardsand includes Sgr B2.[ l, b ] = 0.48,–0.00: A clump in this chain located directly above the Sgr B1 complex ofHII regions.. Sgr B1 ([ l, b ] = 0.506,–0.055 - not marked on the images): Sgr B1 is a giant HII regionto the southwest of Sgr B2 and associated with extensive recent star formation traced byradio continuum emission, masers, and IR sources. In the BGPS map, it appears as a largecavity, marked only by a few faint filaments and bright clumps located to the northwest.Thus, this region must be more evolved than Sgr B2 because gas and dust has been expelled.[ l, b ] = 0.53,+0.18: The brightest member of a cluster of 1.1 mm clumps located abovethe Galactic plane and Sgr B1. At least three extended 8 µ m Spitzer emission nebulae areassociated with this complex. Sgr B2 ([ l, b ] = 0.687,–0.030): The Sgr B2 complex consists of a 5 ′ long north-to-southchain of bright cloud clumps, surrounded by a ‘halo’ of fainter ones, filaments, and shellsextending over nearly 20 ′ in R.A. and 10 ′ in Declination. Sgr B2 clump is the hottest andcurrently most active star forming region in a roughly 25 by 50 parsec complex of emission.Sgr B2 is associated with molecular gas between V LSR = 30 to 80 km s − as seen in tracerssuch as CS, HCN, and HCO + (Bally et al. 1987, 1988; Jackson et al. 1996). Figure 15 showsthe Sgr B2 region at 0.35 mm; Figure 17 shows the BGPS 1.1 mm contours superimposedon the Spitzer 8 µ m image; Figure 18 shows the BGPS 1.1 mm contours superimposed onthe Spitzer 24 µ m image.The brightest emission at 1.1 mm originates from the northern component of the SgrB2 complex (Sgr B2N) which has S . ∼
100 Jy / beam and S . ∼ ,
303 Jy / beam. SgrB2M is a bit fainter at 1.1 mm with S . ∼
80 Jy / beam. However, it is the dominant sourceat 350 µ m with S . ∼ ,
375 Jy / beam. (Note that the 1.1 mm beam has a diameter of33 ′′ while the 0.35 mm beam has a diameter of about 9 ′′ ).Figures 17 and 18 show that the large cavities located at low longitudes and latitudeswith respect to Sgr B2 are filled with bright IR emission. Furthermore, there is a closecorrelation between the mm and sub-mm emission and dark clouds seen in silhouette againstbackground stars and diffuse infrared light.Sgr B2 exhibits a rich millimeter-wave spectrum consisting of thousands of individualspectral lines tracing transitions of over 100 molecular species. Combined with the 10 to 42 –50 km s − Doppler widths of individual lines, this results in a nearly continuous forest ofspectral features. In Sgr B2, the combined flux between 218 to 236 GHz from mostly heavyorganic molecules provides a frequency variable background spectrum with peak temper-atures ranging from 0.1 to over 10 K (Nummelin et al. 1998) that results in a pass-bandaveraged flux in this spectral range of about 1 K. Nummelin et al. (1998) argue that about22% of the total flux from Sgr B2N, and 14% of the flux from the clump known as M isproduced by spectral lines. Thus, in the densest clumps, up to tens of percent of the fluxin the 1.1 mm BGPS pass-band is produced by spectral lines rather than dust. Away fromthese few bright regions, the contribution of spectral lines is likely to be smaller than 10%and less than the flux calibration error.The average 1.1 mm flux of the Sgr B2 complex in a 150 ′′ radius circle centered on SgrB2N is 10.7 Jy. Assuming a dust temperature of 20 K, the mass enclosed in this regionis about 8 . × M ⊙ This region is surrounded by a 0.2 ◦ by 0.25 ◦ diameter envelope offilamentary emission with an average 1.1 mm flux of about 0.9 Jy that implies an additionalmass of at least 1 . × M ⊙ . Thus, the total mass of the Sgr B2 complex is at least 1 . × M ⊙ . This estimate is a severe lower bound because the spatial frequency transfer functionof the BGPS observations and pipeline attenuates emission that is smooth on scales largerthan about 3 ′ and completely filters out emission uniform on scales larger than 7.5 ′ .The extended Sgr B2 complex is the densest and most massive giant molecular cloudcomplex in the Galaxy Lis et al. (1991). About a dozen cavities are apparent in the 1.1mm image. The large 0.15 ◦ by 0.08 ◦ cavity at located at low Galactic longitudes (centeredroughly near l = 0.50 ◦ , b = –0.06 ◦ (Figures 10 and 18) contains bright extended 8 and 24 µ memission in the Spitzer IRAC images that consists of diffuse emission, filaments, and compactsources. This region apparently contains mature HII regions and massive stars formed in aprevious episode of star formation. A similar but smaller cavity, centered at l, b = 0.680 ◦ ,–0.600 ◦ is also filled with 8 µ m emission. These cavities contain an earlier generations ofmassive stars and mature HII regions.[ l, b ] = 0.84,+0.18: An isolated clump associated with extended Spitzer emission. Sgr D ([ l, b ] = 1.12,–0.11): The brightest 1.1 mm clump within a 0.5 ◦ region. It isassociated with bright Spitzer emission, and two bubbles of 20 cm continuum located northand south of the clump (Figures 9 through 11).[ l, b ] = 1.33,+0.16: A typical 1.1 mm clump in this portion of the survey. However, thisone is associated with a blister HII region that extends to the south of the mm clump.[ l, b ] = 1.60,+0.02: One of the brightest clumps in the l = 1.3 ◦ complex of molecularclouds. The surrounding complex consists of several prominent CO-emitting filaments that 43 –extend vertically out of the Galactic plane for at least 0.5 ◦ ( ∼
75 pc). This one degreediameter complex has the largest scale-height in the Central Molecular Zone. The radialvelocities of CO emitting clouds extend over at least 120 km s − . This region may be similarin nature to Bania’s Clump 2. IRAS 17481-2738 ([ l, b ] = 1.74,–1.41): Located at the low-latitude end of the l = 1.5 ◦ complex of molecular clouds, this is the second brightest clump in the region. IRAS 17469-2649 ([ l, b ] = 2.30,+0.26): A typical example of one of the faint, diffuseclumps in this portion of the Galactic plane. It was singled out because it is associated withbright and extended Spitzer emission as well as a 20 cm radio continuum complex.[ l, b ] = 2.62,+0.13: A prominent 1.1 mm clump associated with IRDCs, a complex ofSpitzer nebulosities, and a bright, compact cometary 20 cm continuum feature facing east.[ l, b ] = 2.89,+0.03: The low-latitude end of a complex of 1.1 mm clumps which containsa compact Spitzer source and filament of bright emission. A small and bright 20 cm radiosource may trace a compact HII region. A low-surface brightness and extended bubble of20 cm continuum emission extends about 0.1 ◦ towards low-longitudes. This is one of thefew 1.1 mm clumps in the Bania’s Clump 2 region that exhibits signs of recent massive starformation. The mm clump is seen as a dim IRDC. It may be located in the foregroundportion of the Galactic plane. IRAS 17496-2624 ([ l, b ] = 2.96,–0.06): This 1.1 mm clump is associated with the bright-est 20 cm radio continuum source in the Bania’s Clump 2 region. Spitzer shows a compactlimb-brightened bubble opening towards high latitudes. This feature may trace a compactHII region less than 1 ′ in diameter.[ l, b ] = 3.09,+0.16: One of the brightest 1.1 mm clumps in Bania’s Clump 2. It is notassociated with either a 20 cm or a Spitzer nebula. Located at the high-longitude end of acomplex of clumps that may be faintly seen in silhouette at 8 µ m.[ l, b ] = 3.14,+0.41: A relatively bright 1.1 mm clump in Bania’s Clump 2. It is notassociated with either a 20 cm or a Spitzer nebula.[ l, b ] = 3.31,–0.40: A bright 1.1 mm clump associated with a compact IRDC and withno bright Spitzer sources.[ l, b ] = 3.34,+0.42: The largest and brightest 1.1 mm clump in Bania’s Clump 2. Itdoes not contain any Spitzer sources. However, it is seen faintly in silhouette at 8 µ m andmay be associated with very dim, extended 20 cm continuum (0.04 Jy/beam).[ l, b ] = 3.35,–0.08: A very bright 1.1 mm and Spitzer source which is associated with a 44 –bright cometary 20 cm continuum source located at the high-longitude end of a large HIIregion complex.[ l, b ] = 3.44,–0.35: The brightest 1.1 clump in this part of the survey which is associatedwith a compact and prominent IRDC. A dim 8 µ m stellar object is located near the centerof the cloud. IRAS 17470-2533 ([ l, b ] = 3.41,+0.88): The highest latitude object near Bania’s Clump2, this 1.1 mm clump is surrounded by two tails, one extending towards the Galactic plane,and another towards higher longitudes. The clump contains a bright Spitzer and IRASsource and extended emission . It may be a foreground complex. IRAS 17504-2519 ([ l, b ] = 4.00,+0.34): An isolated 1.1 mm clump seen as an IRDC thatcontains a bright and compact 8 µ m Spitzer source, likely to be a compact HII region.[ l, b ] = 3.39,+0.08: A collection of 1.1 mm clumps associated with a complex of Spitzeremission nebulae.[ l, b ] = 4.43,+0.13: The brightest 1.1 mm clump associated with the above complex.The Spitzer 8 µ m emission shows that this clump is a IRDC superimposed on the extendedbright emission. The IRDC consists of a compact arc of absorption no more than 8 ′′ wide,indicating that the feature is much thinner in one dimension than the Bolocam beam. 45 –
7. Appendix II: Bolocat Clump Masses listed in Table 3
The electronic version of this paper contains a table of masses, column densities, anddensities computed using the methods outlined in Section 2.1. These masses are based onthe Bolocam V1.0 fluxes scaled-up by the empirically determined factor of 1.5. The variousentries in Table 3 are described here.Column 1: The clump number in Bolocat V1.0 released in June via the IPAC website athttp://irsa.ipac.caltech.edu/data/BOLOCAM GPS/ .Column 2: Galactic longitude in degrees.Column 3: Galactic latitude in degrees.Column 4: J(2000) Right Ascension in degrees.Column 5: J(2000) Declination in degrees.Column 6: Mass estimate for each clump measured in a 40 ′′ aperture assuming T = 20 K asdiscussed in Section 2.1.Column 7: H column density estimate for each clump measured in a 40 ′′ aperture assumingT = 20 K as discussed in Section 2.1.Column 8: H volume number density estimate for each clump measured in a 40 ′′ apertureassuming T = 20 K as discussed in Section 2.1.Column 9: The mass derived from the Bolocat flux in the effective beam-deconvolved area ofeach clump tabulated in column 19 of the Bolocat catalog available from the IPAC website.Column 10: The Bolocat effective clump radius tabulated in column 12 of the Bolocat catalogavailable from the IPAC website.Column 11: H column density estimate for each clump measured in the beam-deconvolvedeffective area and based on the mass in column 10 assuming T = 20 K as discussed in Section2.1.Column 12: H volume number density estimate for each clump measured in the beam-deconvolved effective area and based on the mass in column 10 assuming T = 20 K asdiscussed in Section 2.1.Column 13: Mass estimate for each clump measured in a 120 ′′ aperture assuming T = 20 Kas discussed in Section 2.1.Column 14: Mass estimate for each clump measured in a 40 ′′ aperture assuming a tempera- 46 –ture gradient that declines as a power-law with distance from Sgr A*. The power-law has aslope γ g = 0.2 and is normalized to have T = 50 K at a distance of 10 pc from Sgr A*. Thispower-law implies T = 27.5 K at d = 200 pc from Sgr A*, 15.5 K at d = 3 kpc, and 12.5 Kat d = 8.5 kpc. See discussion in Section 3.1. 47 – REFERENCES
Aguirre, J. E. et al. 2010, ApJ, submittedAndr´e, P., Belloche, A., Motte, F., & Peretto, N. 2007, A&A, 472, 519Arendt, R. G. et al. 2008, ApJ, 682, 384Bally, J. & Leventhal, M. 1991, Nature, 353, 234Bally, J., Stark, A. A., Wilson, R. W., & Henkel, C. 1987, ApJS, 65, 13—. 1988, ApJ, 324, 223Bally, J. & Zinnecker, H. 2005, AJ, 129, 2281Bania, T. M., Stark, A. A., & Heiligman, G. M. 1986, ApJ, 307, 350Battinelli, P., Brandimarti, A., & Capuzzo-Dolcetta, R. 1994, A&AS, 104, 379Benjamin, R. A. et al. 2005, ApJ, 630, L149Bik, A., Lamers, H. J. G. L. M., Bastian, N., Panagia, N., & Romaniello, M. 2003, A&A,397, 473Binney, J., Gerhard, O. E., Stark, A. A., Bally, J., & Uchida, K. I. 1991, MNRAS, 252, 210Bissantz, N., Englmaier, P., & Gerhard, O. 2003, MNRAS, 340, 949Christopher, M. H., Scoville, N. Z., Stolovy, S. R., & Yun, M. S. 2005, ApJ, 622, 346Coil, A. L. & Ho, P. T. P. 1999, ApJ, 513, 752Contopoulos, G. & Papayannopoulos, T. 1980, A&A, 92, 33Cotera, A. S., Simpson, J. P., Erickson, E. F., Colgan, S. W. J., Burton, M. G., & Allen,D. A. 2000, ApJS, 129, 123Dahmen, G., Huttemeister, S., Wilson, T. L., & Mauersberger, R. 1998, A&A, 331, 959Dame, T. M., Hartmann, D., & Thaddeus, P. 2001, ApJ, 547, 792de Grijs, R. & Anders, P. 2006, MNRAS, 366, 295de Pree, C. G., Goss, W. M., & Gaume, R. A. 1998, ApJ, 500, 847 48 –De Pree, C. G., Wilner, D. J., Deblasio, J., Mercer, A. J., & Davis, L. E. 2005, ApJ, 624,L101Dowell, C. D. et al. 2003, in Presented at the Society of Photo-Optical InstrumentationEngineers (SPIE) Conference, Vol. 4855, Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series, ed. T. G. Phillips & J. Zmuidzinas, 73–87Dunham, M. K. et al. 2009, ApJ, in preparationEnoch, M. L., Evans, II, N. J., Sargent, A. I., Glenn, J., Rosolowsky, E., & Myers, P. 2008,ApJ, 684, 1240Enoch, M. L. et al. 2006, ApJ, 638, 293Ferri`ere, K., Gillard, W., & Jean, P. 2007, A&A, 467, 611Figer, D. F., McLean, I. S., & Morris, M. 1999, ApJ, 514, 202Figer, D. F. et al. 2002, ApJ, 581, 258Gaume, R. A., Claussen, M. J., de Pree, C. G., Goss, W. M., & Mehringer, D. M. 1995,ApJ, 449, 663Genzel, R. et al. 2003, ApJ, 594, 812Ghez, A. M. et al. 2008, ApJ, 689, 1044Gies, D. R. 1987, ApJS, 64, 545Gies, D. R. & Bolton, C. T. 1986, ApJS, 61, 419Glenn, J. et al. 2003, in Proc. SPIE, Vol. 4855, Millimeter and Submillimeter Detectors forAstronomy, 30Goss, W. M., Schwarz, U. J., van Gorkom, J. H., & Ekers, R. D. 1985, MNRAS, 215, 69PGoto, M. et al. 2008, ApJ, 688, 306Gualandris, A., Portegies Zwart, S., & Eggleton, P. P. 2004, MNRAS, 350, 615Hildebrand, R. H. 1983, QJRAS, 24, 267Hodges-Kluck, E. J., Pound, M. W., Harris, A. I., Lamb, J. W., & Hodges, M. W. 2009,ArXiv e-prints 49 –Jackson, J. M., Heyer, M. H., Paglione, T. A. D., & Bolatto, A. D. 1996, ApJ, 456, L91+Jones, P. A. et al. 2008, MNRAS, 386, 117Kauffmann, J., Bertoldi, F., Bourke, T. L., Evans, II, N. J., & Lee, C. W. 2008, A&A, 487,993Kormendy, J. & Cornell, M. E. 2004, in Astrophysics and Space Science Library, Vol. 319,Penetrating Bars Through Masks of Cosmic Dust, ed. D. L. Block, I. Puerari, K. C.Freeman, R. Groess, & E. K. Block, 261–+Kormendy, J. & Kennicutt, Jr., R. C. 2004, ARA&A, 42, 603Kov´acs, A. 2008, in Presented at the Society of Photo-Optical Instrumentation Engineers(SPIE) Conference, Vol. 7020, Society of Photo-Optical Instrumentation Engineers(SPIE) Conference SeriesLang, C. C., Johnson, K. E., Goss, W. M., & Rodr´ıguez, L. F. 2005, AJ, 130, 2185Laurent, G. T. et al. 2005, ApJ, 623, 742Lazio, T. J. W. & Cordes, J. M. 1998, ApJ, 505, 715Lee, S. et al. 2008, ApJ, 674, 247Lis, D. C. & Carlstrom, J. E. 1994, ApJ, 424, 189Lis, D. C., Carlstrom, J. E., & Keene, J. 1991, ApJ, 380, 429Lis, D. C. & Goldsmith, P. F. 1991, ApJ, 369, 157Lis, D. C. & Menten, K. M. 1998, ApJ, 507, 794Lis, D. C., Menten, K. M., Serabyn, E., & Zylka, R. 1994, ApJ, 423, L39Lis, D. C., Serabyn, E., Zylka, R., & Li, Y. 2001, ApJ, 550, 761Liszt, H. S. 2006, A&A, 447, 533—. 2008, A&A, 486, 467—. 2009, ArXiv e-printsMain, D. S., Smith, D. M., Heindl, W. A., Swank, J., Leventhal, M., Mirabel, I. F., &Rodr´ıguez, L. F. 1999, ApJ, 525, 901 50 –Marshall, D. J., Fux, R., Robin, A. C., & Reyl´e, C. 2008, A&A, 477, L21Martin, C. L., Walsh, W. M., Xiao, K., Lane, A. P., Walker, C. K., & Stark, A. A. 2004,ApJS, 150, 239Matthews, H. et al. 2009, AJ, 138, 1380McGary, R. S., Coil, A. L., & Ho, P. T. P. 2001, ApJ, 559, 326McKee, C. F. & Williams, J. P. 1997, ApJ, 476, 144Mirabel, I. F., Paul, J., Cordier, B., Morris, M., & Wink, J. 1991, A&A, 251, L43Miyazaki, A. & Tsuboi, M. 2000, ApJ, 536, 357Moneti, A., Stolovy, S., Blommaert, J. A. D. L., Figer, D. F., & Najarro, F. 2001, A&A,366, 106Morris, M. & Serabyn, E. 1996, ARA&A, 34, 645Motte, F., Andre, P., & Neri, R. 1998, A&A, 336, 150Motte, F., Andr´e, P., Ward-Thompson, D., & Bontemps, S. 2001, A&A, 372, L41Motte, F., Bontemps, S., Schilke, P., Schneider, N., Menten, K. M., & Brogui`ere, D. 2007,A&A, 476, 1243Motte, F., Schilke, P., & Lis, D. C. 2003, ApJ, 582, 277Nagayama, T., Omodaka, T., Handa, T., Iahak, H. B. H., Sawada, T., Miyaji, T., & Koyama,Y. 2007, PASJ, 59, 869Nagayama, T. et al. 2009, PASJ, 61, 283Nummelin, A., Bergman, P., Hjalmarson, A., Friberg, P., Irvine, W. M., Millar, T. J., Ohishi,M., & Saito, S. 1998, ApJS, 117, 427Oka, T., Geballe, T. R., Goto, M., Usuda, T., & McCall, B. J. 2005, ApJ, 632, 882Oka, T., Hasegawa, T., Sato, F., Tsuboi, M., & Miyazaki, A. 1998, ApJS, 118, 455—. 2001a, PASJ, 53, 779Oka, T., Hasegawa, T., Sato, F., Tsuboi, M., Miyazaki, A., & Sugimoto, M. 2001b, ApJ,562, 348 51 –Ossenkopf, V. & Henning, T. 1994, A&A, 291, 943Pierce-Price, D. et al. 2000, ApJ, 545, L121Pohl, M., Englmaier, P., & Bissantz, N. 2008, ApJ, 677, 283Reach, W. T. et al. 1995, ApJ, 451, 188Rodriguez-Fernandez, N. J. 2009, ArXiv e-printsRodriguez-Fernandez, N. J. & Combes, F. 2008, A&A, 489, 115Rodriguez-Fernandez, N. J., Combes, F., Martin-Pintado, J., Wilson, T. L., & Apponi, A.2006, A&A, 455, 963Rodr´ıguez-Fern´andez, N. J., Mart´ın-Pintado, J., Fuente, A., & Wilson, T. L. 2004, A&A,427, 217Rosolowsky, E. 2005, PASP, 117, 1403Rosolowsky, E. et al. 2009, ArXiv e-printsSchuller, F. et al. 2009, A&A, 504, 415Scoville, N. Z. & Kwan, J. 1976, ApJ, 206, 718Serabyn, E., Carlstrom, J., Lay, O., Lis, D. C., Hunter, T. R., & Lacy, J. H. 1997, ApJ, 490,L77+Serabyn, E., Shupe, D., & Figer, D. F. 1999, in ASP Conf. Ser. 186: The Central Parsecs ofthe Galaxy, ed. H. Falcke, A. Cotera, W. J. Duschl, F. Melia, & M. J. Rieke, 320–+Sodroski, T. J. et al. 1994, ApJ, 428, 638Stahler, S. W., Palla, F., & Ho, P. T. P. 2000, Protostars and Planets IV, 327Stark, A. A. & Bania, T. M. 1986, ApJ, 306, L17Stolte, A., Brandner, W., Grebel, E. K., Lenzen, R., & Lagrange, A.-M. 2005, ApJ, 628,L113Stolte, A., Ghez, A. M., Morris, M., Lu, J. R., Brandner, W., & Matthews, K. 2008, ApJ,675, 1278Tamblyn, P. & Rieke, G. H. 1993, ApJ, 414, 573 52 –Tsuboi, M., Handa, T., & Ukita, N. 1999, ApJS, 120, 1Wang, Q. D., Dong, H., & Lang, C. 2006, MNRAS, 371, 38Wang, Q. D., Gotthelf, E. V., & Lang, C. C. 2002, Nature, 415, 148Wardle, M. & Yusef-Zadeh, F. 2008, ApJ, 683, L37Wright, M. C. H., Coil, A. L., McGary, R. S., Ho, P. T. P., & Harris, A. I. 2001, ApJ, 551,254Yusef-Zadeh, F., Braatz, J., Wardle, M., & Roberts, D. 2008, ApJ, 683, L147Yusef-Zadeh, F. et al. 2009, ApJ, 702, 178Yusef-Zadeh, F., Hewitt, J. W., & Cotton, W. 2004, ApJS, 155, 421Yusef-Zadeh, F., Muno, M., Wardle, M., & Lis, D. C. 2007, ApJ, 656, 847Yusef-Zadeh, F., Nord, M., Wardle, M., Law, C., Lang, C., & Lazio, T. J. W. 2003, ApJ,590, L103Zhang, Q. & Fall, S. M. 1999, ApJ, 527, L81
This preprint was prepared with the AAS L A TEX macros v5.2.
Table 1. Observing Epochs for the Galactic center dataBegin End Instrument2 Jul 2005 9 Jul 2005 Bolocam2 Jun 2006 29 Jun 2006 Bolocam + SHARC-II7 Jul 2007 14 Jul 2007 Bolocam + SHARC-II13 Apr 2008 14 Apr 2008 SHARC-II 53 –Table 2. 1.1 mm Peak Fluxes and Masses of Selected Regions l b S S M M N N n n Commentsdeg deg Jy Jy 10 M ⊙ M ⊙ cm − cm − cm − cm − − cloud359.91 –0.31 1.5 0.32 3 33 30 6 195 566 D=3.9 kpc359.94 –0.04 11.7 2.90 – 1425 234 57 701 2348 Sgr A*, 50 km s − cloud359.96 +0.17 3.3 0.77 7 81 65 15 431 1374 IRAS 17417-2851, D=3.9 kpc359.97 –0.46 3.3 0.41 7 42 66 8 431 728 D=3.9 kpc0.05 –0.21 1.8 0.41 17 195 36 8 108 3350.12 +0.02 1.4 0.78 13 390 27 15 81 642 Near Arches cluster0.16 –0.06 0.3 0.83 3 413 6 17 18 680 Near Quintuplet cluster0.26 –0.03 5.7 1.50 54 750 114 30 341 1235 Lima bean0.28 –0.48 2.3 0.80 5 84 40 17 396 1428 for D=3.9 kpc0.32 –0.20 3.9 0.59 38 293 78 12 233 482 IRAS 17439-28450.41 –0.05 4.2 1.2 39 600 84 24 251 9890.41 –0.50 1.8 0.33 4 35 36 6 236 593 for D=3.9 kpc0.48 –0.00 7.7 2.0 72 975 153 39 458 16070.53 +0.18 2.4 0.50 23 240 48 11 144 4080.68 –0.03 140 16.1 1320 8010 2796 323 8351 13224 Sgr B20.84 +0.18 2.0 0.32 18 158 39 6 117 2601.12 -0.11 5.1 0.92 48 458 68 12 204 503 Sgr D1.33 +0.16 0.6 0.36 6 180 12 8 36 297 IRAS 17450-27421.60 +0.02 2.1 0.48 20 240 42 9 126 3961.74 –0.41 2.3 0.47 21 233 45 9 135 383 IRAS 17481=27382.30 +0.26 0.5 0.05 4 23 9 2 27 38 IRAS 17469-26492.62 +0.13 1.5 0.12 14 60 30 3 90 992.89 +0.03 0.9 0.30 9 150 18 6 54 2482.96 –0.06 0.6 0.14 6 68 12 3 36 111 IRAS 17496-26243.09 +0.16 0.9 0.15 9 75 18 3 54 1233.14 +0.41 0.9 0.15 9 75 18 3 54 1233.31 –0.40 1.5 0.17 14 83 30 3 90 1373.34 +0.42 1.2 0.30 11 150 24 6 72 2483.35 –0.08 2.3 0.12 14 60 45 3 135 993.41 +0.88 0.8 0.06 7 30 15 2 45 50 IRAS 17470-25333.44 –0.35 4.8 0.26 45 128 96 5 288 2104.00 +0.34 0.9 0.06 9 30 18 3 54 50 IRAS 17504-25194.39 +0.08 0.8 0.15 7 75 15 3 45 1234.43 +0.13 2.1 0.23 20 113 42 5 126 186 Notes: [1] Flux densities: S is the flux (in Jy per beam) measured in a 33 ′′ diameter effective CSO beam (equivalent to a 40 ′′ diameter top-hat beam with a radius of 19.82 ′′ ). S is the average flux measured (in Jy per beam) in a solid anglecorresponding to a 300 ′′ diameter ”top-hat” beam. [2] Masses: M and M are the masses estimated for an assumed dusttemperature of 20 K and a distance of D = 8.5 kpc (unless otherwise noted) in 40 ′′ and 280 ′′ apertures. The tabulated valuesare in Solar masses ( M ⊙ ) divided by 10 . [3]: Column densities: N and N are the equivalent average column densitiesestimated in the 40 ′′ and 300 ′′ apertures, assuming that the emission is uniformly spread over the aperture. These are lowerbounds on the actual column densities as discussed in the text. The units are in cm − divided by 1 . × which is about A V ≈ [4] H volume densities: n and n are the equivalent average volume densities of H estimated inthe 40 ′′ and 300 ′′ apertures under the assumption that the clumps are spherical with a radius equal to the projected radius ofthe measurement aperture in which the emitting material is spread uniformly. These are a lower bounds on the actualdensities as discussed on the text. The units for n are in cm − divided by 10 . The units for n are in cm − .
54 –Fig. 1.— A 1.1 mm BGPS continuum image showing the central parts of the Galactic Planefrom l = 358.5 ◦ to 4.5 ◦ . This image has been processed with 13 PCA components and 50iterations of the iterative mapper. While this level of processing restores much of the faintstructure, it removes flux from bright sources ( > about 5 Jy) and creates significant negativebowls around them. The display presented here uses a logarithmic scale from -0.5 to 100Jy/beam. 55 –Fig. 2.— BGPS 1.1 mm contours superimposed on the Spitzer Space Telescope 8 µ m imageshown on a logarithmic intensity scale and at the same spatial scale as Figure 1. Variousregions discussed in the text are marked. Bania’s Clump 2 is shown on the left and the l =1.3 ◦ complex is located in the middle. The oval to the left of Sgr A shows the approximateoutline of the infrared cavity that contains the Arches and Quintuplet clusters and is rimmedby mid-IR and radio continuum emission. The dashed line-segments illustrate the apparentwalls of a larger cavity bounded by Spitzer 8 and 24 µ m emission. The interior of this regionis filled with 20 cm radio continuum. See text for discussion. The 1.1 mm contour levels areat 0.10, 0.20, 0.37, 0.73, 1.4, 2.7, 5.3, 10.3, and 20 Jy/beam. 56 –Fig. 3.— The Central Molecular Zone in the the Galactic center as traced by the peakantenna temperature, T ∗ A of the J =1–0 CO transition. The location of Bania’s Clump 2is indicated along with some of the other prominent Galactic center features. Taken fromthe AT&T Bell Laboratories data presented by Bally et al. (1987, 1988). 57 –Fig. 4.— A longitude-velocity diagram showing the Central Molecular Zone in the Galacticcenter as traced by the peak antenna temperature, T ∗ A of the J =1–0 CO transition averagedover the Galactic latitudes from b = 0.20 ◦ to 0.48 ◦ . The location of Bania’s Clump 2 isindicated along with some of the other prominent Galactic center features. Taken from theAT&T Bell Laboratories data presented by Bally et al. (1987, 1988). 58 –Fig. 5.— The mass spectrum of Bolocat clumps derived from the fluxes measured in a 40 ′′ diameter aperture under the assumption that all Bolocat clumps are located at D = 8.5 kpcand have a temperature of T = 20 K. The top panel shows the cumulative distribution alongwith a line representing a differential mass distribution with the indicated slope. The bottompanel shows the differential mass distribution with counts in bins whose width increasedlogarithmically along with a best fit slope to the differential mass distribution for constantwidth mass-bins. 59 –Fig. 6.— Same as Figure 5 but using fluxes measured in a 120 ′′ diameter aperture. 60 –Fig. 7.— Same as Figure 5 using fluxes measured in a 40 ′′ diameter aperture but assumingthat the dust temperature declines with increasing distance, d , from Sgr A as a power-lawwith T = T d − . . See text for details. 61 –Fig. 8.— Same as Figure 5 using fluxes measured in a 40 ′′ diameter aperture but assumingthat the dust temperature increases with increasing 1.1 mm flux, F , as a power-law with T = T F . . See text for details. 62 –Fig. 9.— The 1.1 mm image of the Central Molecular Zone. The red lines show the ori-entations and locations of cometary clouds and chains of clumps that face Sgr A. Thesefeatures are located in the interior of the cavity surrounding Sgr A that is discussed in thetext and form the walls of its multiple chambers. The features marked in red are likely tobe foreground clouds at an assumed distance of 3.9 kpc as indicated in Table 1. 63 –Fig. 10.— Contours of 1.1 mm dust continuum emission from the CMZ superimposed on theSpitzer IRAC 8 µ m image (Arendt et al. 2008) show with a logarithmic greyscale presentedat the same spatial scale as Figure 9. The 1.1 mm contour levels are 0.3, 0.45, 0.75, 1.2, 2.0,3.0, 5.0, 7.5, 12.0, 20, and 30 Jy/beam. 64 –Fig. 11.— Contours of 1.1 mm dust continuum emission from the CMZ superimposed onthe 20 cm radio continuum image from Yusef-Zadeh et al. (2004) shown on a logarithmicgreyscale and presented at the same spatial scale as Figure 9. The 1.1 mm contour levelsare 0.3, 0.45, 0.75, 1.2, 2.0, 3.0, 5.0, 7.5, 12.0, 20, and 30 Jy/beam. 65 –Fig. 12.— A SHARC-II 350 µ m image showing warm dust continuum emission from thevicinity of Sgr A in the center of the Galaxy with 9 ′′ resolution. This image has been”unsharp-masked” to suppress large scale structure and to match the spatial frequency re-sponse of the Bolocam 1.1 mm images. 66 –Fig. 13.— A map showing the ratio of the surface brightness of 350 µ m emission dividedby 1.1 mm emission in the vicinity of Sgr A. The display range goes from 0 (white) to 100(black). Regions with ratios larger than 100, which corresponds to the Rayleigh-Jeans limitfor β = 2.0 dust have been masked and are shown as white. The ratios at the edge of theSHARC-II field are spurious. As discussed in the text, the 350 µ m map has been convolvedwith a gaussian kernel to have the same effective beam size as the 1.1 mm map (33 ′′ effectivebeam). The italic numbers indicate the flux ratio at the locations indicated by the associatedarrows. The numbers in parentheses give the derived dust temperatures for an emissivitypower-law index of β = 2.0. Emission from the central black hole dominates the 1.1 mm fluxoriginating in the center of the CND. Because the source of emission is unlikely to dust, nodust temperature is given. 67 –Fig. 14.— Contours of SHARC-II 350 µ m emission superimposed on a greyscale rendition ofthe 24 µ m Spitzer image showing the vicinity of Sgr A (Yusef-Zadeh et al. 2009) displayedon a logarithmic intensity scale. The sharp-edged white regions indicate locations where theSpitzer image is saturated. Contour levels are: 4.5, 7, 9.0, 13, 18, 26, 37, 53, 75, 106, and150 Jy/beam. 68 –Fig. 15.— A SHARC-II 0.35 mm image showing the Sgr B2 complex on a logarithmicintensity scale. The bright emission is contoured with levels shown at 100, 130, 170, 220,285, 371, 482, 627, 814, 1060, and 1375 Jy/beam. The two peaks in the center correspondto Sgr B2N and SgrB2 Main. 69 –Fig. 16.— A map showing the ratio of the surface brightness of 350 µ m emission dividedby 1.1 mm emission in the vicinity of Sgr B2. The display range goes from 0 (white) to100 (black). Regions with ratios larger than 100, which corresponds to the Rayleigh-Jeanslimit for β = 2.0 dust have been masked and are shown as white. The ratios at the edgeof the SHARC-II field are spurious. As discussed in the text, the 350 µ m map has beenconvolved with a gaussian kernel to have the same effective beam size as the 1.1 mm map(33 ′′ effective beam). The italic numbers indicate the flux ratio at the locations indicated bythe associated arrows. The numbers in parentheses give the derived dust temperatures foran emissivity power-law index of β = 2. 70 –Fig. 17.— The Bolocam 1.1 mm contours superimposed on a Spitzer 8 µ m image displayedin logarithmic intenstiy. Contour levels are at 0.30, 0.54, 1.0, 1.8, 3.3, 6.0, 11, 20, 36, 66,and 120.0 Jy/beam. 71 –Fig. 18.— The Bolocam 1.1 mm contours superimposed on a Spitzer 24 µ m image displayedin logarithmic intenstiy. Contour levels are at 0.30, 0.54, 1.0, 1.8, 3.3, 6.0, 11, 20, 36, 66,and 120.0 Jy/beam. 72 –Fig. 19.— A 1.1 mm image centered at l = 3 ◦ , the region that contains Bania’s Clump 2with carious clumps marked. 73 –Fig. 20.— Contours of 1.1 mm continuum emission centered at l = 3 ◦ , the region thatcontains Bania’s Clump 2 (red), superimposed on the GLIMPSE2 8 µ m image. Contourlevels are at 0.15, 0.24, 0.38, 0.6, 1.0, 1.5, 2.4, 3.8, 6.0, 10, and 15 Jy/beam. 74 –Fig. 21.— A cartoon showing a face-on view of the central 500 pc region of the Milky Wayas viewed from the northern Galactic pole. The Sun is located below the figure and positivelongitudes to the left. The diffuse grey-scale shows the current orientation of the stellar bar,thought to have its major axis tilted between 20 to 45 ◦ with respect to our line-of sight.The large oval show a non intersecting x1 orbit; Lower angular momentum x1 orbits becomeself-intersecting. The ”rhombus” of molecular emission in l - V diagrams probably occupiesthe receding and approaching portions of the innermost x1 orbits. The 3 smaller inscribedellipses show x2 orbits. Bania’s Clump 2 is thought to be located on the receding portionthe last stable x1-orbit where it is self-intersecting. Approaching gas presumably enters asan atomic phase; shock-copresison and subsequent cooling results in its conversion to themolecular phase. The receding gas above Bania’s clump 2 traces the leading edge of the barand populates extended network of positive velocity clouds in l - V diagrams such as Figure4. The CMZ emission is associated with gas mostly located on the near-side of the Galacticcenter on the x2 orbits. The possible locations of the Sgr C, 20 km − , 50 km − , and SgrB2 complexes are indicated. 75 –Fig. 22.— A 1.1 mm image showing the field of view surrounding the ”Great Annihilator”(GA) thought to be a major source of 511 keV positronium emission. The large white circleis 40 ′′ in diameter and shows the location of the X-ray source 1E1740.7-2942. The smallcircles mark the position centroids of two HCO + clumps detected by Hodges-Kluck et al.(2009). The circle next to the GA is their clump 1; the circle above is their clump 2b whichdoes not appear to be associated with 1.1 mm emission. The display is linear intensity. 76 –Table 3. 1.1 mm Masses, Column Densities, and Densities for 1428 Bolocat Clumps BGPS l b R.A. Dec.
M(40 ′′ ) N ( H ) n ( H ) M(r) r) N r ( H ) n r ( H ) M(120 ′′ ) M(gradient T)( ⊙ cm − cm − M ⊙ ( ′′ ) cm − cm − M ⊙ M ⊙ [1] Assuming a dust temperature of 20 K. [2][2]