1: 

2: \documentclass[preprint2,longabstract]{aastex}

3: 

4: \newcommand{\vdag}{(v)^\dagger}

5: \newcommand{\myemail}{olmi@arcetri.astro.it}

6: \newcommand{\beq}{\begin{equation}}

7: \newcommand{\eeq}{\end{equation}}

8: \newcommand{\bdi}{\begin{displaymath}}

9: \newcommand{\edi}{\end{displaymath}}

10: \newcommand{\snr}{S/N}

11: \newcommand{\kms}{km\,s$^{-1}$}

12: 

13: 

14: \slugcomment{To appear in the Astrophysical Journal}

15: 

16: \shorttitle{BLAST Survey in Vulpecula}

17: \shortauthors{Chapin, E.~L.~et al.}

18: 

19: 

20: \begin{document}

21: 

22: \title{ The Balloon-borne Large Aperture Submillimeter Telescope (BLAST) 2005:

23:   A 4~deg$^2$ Galactic Plane Survey in Vulpecula ($\ell=59^\circ$)

24: }

25: 

26: \author{E.~L.~Chapin,\altaffilmark{1,\dag}

27:         P.~A.~R.~Ade,\altaffilmark{2}

28:         J.~J.~Bock,\altaffilmark{3,4}

29:         C.~Brunt,\altaffilmark{5}

30:         M.~J.~Devlin,\altaffilmark{6}

31:         S.~Dicker,\altaffilmark{6}

32:         M.~Griffin,\altaffilmark{2}

33:         J.~O.~Gundersen,\altaffilmark{7}

34:         M.~Halpern,\altaffilmark{1}

35:         P.~C.~Hargrave,\altaffilmark{2}

36:         D.~H.~Hughes,\altaffilmark{8}

37:         J.~Klein,\altaffilmark{6}

38:         G.~Marsden,\altaffilmark{1}

39:         P.~G.~Martin,\altaffilmark{9,10}

40:         P.~Mauskopf,\altaffilmark{2}

41:         C.~B.~Netterfield,\altaffilmark{10,11}

42:         L.~Olmi,\altaffilmark{12,13}

43:         E.~Pascale,\altaffilmark{11}

44:         G.~Patanchon,\altaffilmark{1,14}

45:         M.~Rex,\altaffilmark{6}

46:         D.~Scott,\altaffilmark{1}

47:         C.~Semisch,\altaffilmark{6}

48:         M.~D.~P.~Truch,\altaffilmark{15}

49:         C.~Tucker,\altaffilmark{2}

50:         G.~S.~Tucker,\altaffilmark{15}

51:         M.~P.~Viero,\altaffilmark{10}

52:         D.~V.~Wiebe\altaffilmark{11}}

53: 

54: 

55: \altaffiltext{1}{Department of Physics \& Astronomy, University of

56: British Columbia, 6224 Agricultural Road, Vancouver, BC V6T~1Z1,

57: Canada}

58: 

59: \altaffiltext{2}{Department of Physics \& Astronomy, Cardiff University, 5 The Parade, Cardiff, CF24~3AA, UK}

60: 

61: \altaffiltext{3}{Jet Propulsion Laboratory, Pasadena, CA 91109-8099}

62: 

63: \altaffiltext{4}{Observational Cosmology, MS 59-33, California Institute of Technology, Pasadena, CA 91125}

64: 

65: \altaffiltext{5}{School of Physics, University of Exeter, Stocker

66: Road, Exeter, EX4~4QL, UK}

67: 

68: \altaffiltext{6}{Department of Physics and Astronomy, University of Pennsylvania, 209 South 33rd Street, Philadelphia, PA 19104}

69: 

70: \altaffiltext{7}{Department of Physics, University of Miami, 1320 Campo Sano Drive, Carol Gables, FL 33146}

71: 

72: \altaffiltext{8}{Instituto Nacional de Astrof\'isica \'Optica y Electr\'onica (INAOE), Aptdo. Postal 51 y 72000 Puebla, Mexico}

73: 

74: \altaffiltext{9}{Canadian Institute for Theoretical Astrophysics, University of Toronto, 60 St. George Street, Toronto, ON M5S~3H8, Canada}

75: 

76: \altaffiltext{10}{Department of Astronomy \& Astrophysics, University of Toronto, 50 St. George Street, Toronto, ON  M5S~3H4, Canada}

77: 

78: \altaffiltext{11}{Department of Physics, University of Toronto, 60 St. George Street, Toronto, ON M5S~1A7, Canada}

79: 

80: \altaffiltext{12}{Istituto di Radioastronomia, Largo E. Fermi 5, I-50125, Firenze, Italy}

81: 

82: \altaffiltext{13}{University of Puerto Rico, Rio Piedras Campus, Physics Dept., Box 23343, UPR station, San Juan, Puerto Rico}

83: 

84: \altaffiltext{14}{Laboratoire APC, 10, rue Alice Domon et L{\'e}onie Duquet 75205 Paris, France}

85: 

86: \altaffiltext{15}{Department of Physics, Brown University, 182 Hope Street, Providence, RI 02912}

87: 

88: \altaffiltext{\dag}{\url{echapin@phas.ubc.ca}}

89: 

90: \begin{abstract}

91:   We present the first results from a new 250, 350, and 500\,\micron\

92:   Galactic Plane survey taken with the Balloon-borne Large-Aperture

93:   Submillimeter Telescope (BLAST) in 2005. This survey's primary goal

94:   is to identify and characterize high-mass proto-stellar objects

95:   (HMPOs).  The region studied here covers 4\,deg$^2$ near the open

96:   cluster NGC~6823 in the constellation Vulpecula ($\ell=59^\circ$).

97:   We find 60 compact sources ($<60''$ diameter) detected

98:   simultaneously in all three bands.  Their spectral energy

99:   distributions (SEDs) are constrained through BLAST, {\it IRAS}, {\it

100:     Spitzer} MIPS, and {\it MSX} photometry, with inferred dust

101:   temperatures spanning $\sim 12$--40\,K assuming a dust emissivity

102:   index $\beta=1.5$.  The luminosity-to-mass ratio, a

103:   distance-independent quantity, spans

104:   $\sim0.2$--130\,L$_\odot$\,M$_\odot^{-1}$. Distances are estimated

105:   from coincident $^{13}$CO$(1 \rightarrow 0)$ velocities combined

106:   with a variety of other velocity and morphological data in the

107:   literature. In total, 49 sources are associated with a molecular

108:   cloud complex encompassing NGC~6823 (distance $\sim 2.3$\,kpc), 10

109:   objects with the Perseus Arm ($\sim 8.5$\,kpc) and one object is

110:   probably in the outer Galaxy ($\sim 14$\,kpc).  Near NGC~6823, the

111:   inferred luminosities and masses of BLAST sources span $\sim

112:   40$--$10^4$\,L$_\odot$, and $\sim15$--$700$\,M$_\odot,$

113:   respectively.  The mass spectrum is compatible with molecular gas

114:   masses in other high-mass star forming regions.  Several luminous

115:   sources appear to be Ultra Compact \ion{H}{2} regions powered by

116:   early B stars. However, many of the objects are cool, massive

117:   gravitationally-bound clumps with no obvious internal radiation from

118:   a protostar, and hence excellent HMPO candidates.

119: \end{abstract}

120: 

121: \keywords{submillimeter --- stars: formation --- ISM: clouds ---

122:   balloons}

123: 

124: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

125: \section{INTRODUCTION}     

126: \label{sec:intro}

127: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

128: 

129: The Balloon-borne Large Aperture Submillimeter Telescope (BLAST) is a

130: 2-m stratospheric balloon telescope that observes simultaneously at

131: 250, 350, and 500\,\micron\ using bolometric imaging arrays

132: \citep{pascale2007}.  During the first BLAST science flight, a 100~hr

133: Arctic flight from Sweden to Canada in June 2005 (BLAST05), BLAST

134: conducted the first sensitive large-scale Galactic Plane surveys at

135: these wavelengths.  The focus of these surveys is the earliest stages

136: of massive star formation. As noted in the recent review by

137: \citet{zinn2007} our understanding of this evolutionary phase is still

138: limited.  Over the past 15 years submillimeter observations at longer

139: wavelengths (350--1200\,\micron), e.g. with SCUBA on the 15-m James

140: Clerk Maxwell Telescope \citep{holland1999}, or MAMBO on the IRAM 30-m

141: telescope \citep{kreysa1998,motte2007}, have opened up studies of the

142: earliest evolutionary stages of molecular core collapse and

143: proto-stellar formation.  Such cores lack an internal source of

144: radiation and are very cold ($\la 25$\,K), so that they emit the bulk

145: of their radiation at submillimeter wavelengths. Since this emission

146: is optically thin in the submillimeter band, observed flux densities

147: are proportional to column density and mass.  There is a variety of

148: terminology used for these elusive early stages.  For example,

149: \citet{zinn2007} distinguish cores (size 0.1\,pc) embedded within

150: clumps (size 0.5\,pc).  \citet{motte2007} refer to molecular cloud

151: fragments which could be high-luminosity infra-red protostars,

152: infra-red quiet protostars, or high-mass pre-stellar cores. Along the

153: lines of the latter, here we refer to high-mass proto-stellar objects

154: (HMPOs) as compact sources residing in dense molecular clouds, that

155: have the potential to form (one or more) massive OB stars, having

156: luminosities in the range $\sim10^2$--$10^5$\,L$_{\odot}$, but without

157: associated radio continuum emission.  This latter qualification

158: distinguishes them from massive young stellar objects, a later stage

159: in which a hot star has formed (perhaps still accreting), providing

160: the ionizing radiation necessary to form a high emission measure Ultra

161: Compact (UC) \ion{H}{2} region (and hence radio emission).  Note that

162: at the distances of high-mass star forming regions surveying

163: instruments detect quite massive clumps, possibly harboring groups of

164: stars or their precursors, and not resolving the substructure

165: associated with individual nodes of collapse; however, once star

166: formation is underway the most massive objects dominate the luminosity

167: and ionization.

168: 

169: BLAST is presently unique in its ability to detect and characterize

170: cold dust emission from a range of pre- and proto-stellar sources,

171: constraining the temperatures of objects with $T \la 25$\,K

172: ($\beta=1.5$) using its three-band photometry near the peak of the

173: spectrum.  An earlier, less sensitive balloon-borne submillimeter

174: telescope called ProNaOS \citep{dupac2001} had similar goals.

175: %

176: Previous studies which do not sample the spectral peak of the thermal

177: emission \citep[e.g.,][]{johnstone2000, pierceprice2000, kirk2005,

178: reid2005, thompson2005, enoch2006, hill2006, schneider2006,

179: thompson2006, young2006, enoch2007, moore2007} have been limited by

180: their relative inability to measure the temperature, producing large

181: uncertainties in the derived luminosities and masses.  Recent surveys

182: with {\it Spitzer} MIPS can constrain the temperatures of warmer

183: objects \citep{car05}, but the youngest and coldest objects are

184: potentially not detected even in the long-wavelength {\it Spitzer}

185: bands.

186: 

187: BLAST is also very efficient at mapping, and so conducted a series of

188: surveys spread across the entire portion of the Galactic Plane that

189: was available during the flight.  These maps encompass (primarily)

190: known high-mass star forming regions: a 10\,deg$^2$ map of Cygnus X; a

191: 4\,deg$^2$ map in Vulpecula (described in detail in this paper); a

192: 3\,deg$^2$ map in Sagitta; and finally a 6\,deg$^2$ map in Aquila

193: towards the Galactic Ring Survey molecular cloud GRSMC~45.60+0.30

194: \citep{rathborne2004}. The locations and sizes of these maps are shown

195: in Figure~\ref{fig:coverage}. The Vulpecula map was studied first, as

196: it is the deepest of the four regions, and consists of 6.5 hours of

197: data. The analysis of the other three fields, and those from the 2006

198: Antarctic flight, including a diffraction-limited 50\,deg$^2$ map of

199: the Vela molecular ridge, will be presented in forthcoming papers.

200: 

201: %\clearpage

202: \begin{figure*}

203: \centering

204: \includegraphics[width=\linewidth]{f1.eps}

205: \caption{Locations of fields for the BLAST05 Galactic Plane surveys.

206:   Results from the Vulpecula field are discussed in detail in this

207:   paper. The background intensity image is 100\,\micron\ emission from

208:   \citet{schlegel1998}.}

209: \label{fig:coverage}

210: \end{figure*}

211: %\clearpage

212: 

213: To put these surveys in context, \citet{motte2007} state that ``large

214: fast-mapping images with MAMBO-2 are currently the best tool to study

215: density structure of molecular clouds with high spatial resolution.''

216: In about 33 hours of observations (and 30 more with the less efficient

217: MAMBO) they mapped 3\,deg$^2$ in Cygnus~X; not the entire region but

218: targeted areas of high extinction.  They catalogued 129 compact

219: sources (down to a scale of 15\arcsec) to a peak flux level of 80\,mJy

220: ($5\,\sigma$) and proposed an additional 40 somewhat more extended

221: structures (2\arcmin).  Larger structures (10\arcmin) were filtered

222: out.

223: %

224: In the survey reported here, we mapped a contiguous (no threshold for

225: extinction) 4\,deg$^2$ in 6.5 hours.  Although the very largest scales

226: (including the DC level) are absent from our maps, our images maintain

227: power well beyond 10\arcmin; the images reveal interesting large-scale

228: structures crossing the entire region.  As reported below, we catalog

229: 60 sources to a flux level of 10\,Jy at 250\,\micron.  Extrapolating

230: this detection limit using a typical spectral shape ($T = 20$\,K and

231: $\beta = 1.5$) this corresponds to 170\,mJy at 1.2\,mm.  We also find

232: fewer sources per square degree compared to the \citet{motte2007}

233: survey of Cygnus~X because this region is slightly more distant, not

234: as active, and our survey did not target areas of high extinction.

235: There are exciting large sensitive surveys to look forward to at these

236: same wavelengths with the SPIRE instrument for {\it Herschel}

237: \citep{griffin2003}.

238: 

239: The region chosen for this study was centered near NGC~6823 in the

240: constellation Vulpecula ($\ell=59^\circ$), and covers approximately

241: 4\,deg$^2$. This region of the Galactic Plane is prominent in images

242: of thermal dust emission (e.g., {\it IRAS} in

243: Figure~\ref{fig:coverage}), as well as in the radio and the optical.

244: In the radio, the identifiable extended \ion{H}{2} region excited by

245: the massive stars is called Sh2-86 (nominal position indicated in

246: Figure~\ref{fig:blastmaps}).  The stellar HR diagram for NGC~6823 has

247: been examined by \citet{massey1995}.  They find an age of 5--7\,Myr

248: for the bulk of the stars.  There are several evolved OB supergiants

249: of mass $\sim 25$\,M$_\odot$.  The most massive star, O7 V ((f)), with

250: mass $\sim40$\,M$_\odot$, appears younger (2\,Myr) than the rest.  In

251: the optical this massive star is responsible for the illumination of

252: many ``elephant trunks'' or ``pillars'' in the eastern portion of the

253: \ion{H}{2} region.

254: 

255: Very high-mass stars evolve quickly into supernovae.  We note that a

256: supernova remnant (SNR) discovered by \citet{taylor1992} is within

257: 20\arcmin\ of the O7 star in projection ($< 12$\,pc laterally,

258: assuming the same distance), and even closer to one B supergiant ---

259: nominal positions are indicated in Figure~\ref{fig:blastmaps}.

260: \citet{massey1995} find several pre-main sequence stars of mass

261: 5--7\,M$_\odot$, which pass through this stage in less than 2\,Myr.

262: Earlier stages of evolution are revealed by the bright Infra Red

263: Astronomy Satellite ({\it IRAS}) sources clustered in this region.  At

264: least seven of these objects have been observed in molecular line

265: studies \citep{beu02,zha05,bel06}

266: 

267: This paper describes a technique developed for deconvolving the

268: out-of-focus beams common to all data from BLAST05

269: \citep[see][]{truch2007} and for detecting compact objects in the

270: BLAST maps. We report a robust list of 60 submillimeter sources with

271: sizes $\la60''$ diameter that are detected simultaneously in the three

272: BLAST bands (\S\ref{sec:obs}).  Infra-red photometry for BLAST sources

273: is obtained from comparisons with {\it IRAS} and {\it Spitzer} MIPS

274: maps, and the Midcourse Space Experiment ({\it MSX}) point source

275: catalog (\S\ref{sec:firsed}).  The BLAST and infra-red data are

276: combined to constrain dust temperatures and integrated far infra-red

277: (FIR) fluxes using isothermal modified blackbody SED fits

278: (\S\ref{sec:coldsed}).  Our analysis has benefited from a $^{13}$CO$(1

279: \rightarrow 0)$ data cube obtained as a part of a Galactic Plane

280: survey at FCRAO \citep{brunt2007}. Noting coincidences between compact

281: structures in the $^{13}$CO emission, and comparing velocities and

282: morphologies with the VLA Galactic Plane Survey, {\it Spitzer} GLIMPSE

283: \citep{whitney2005} and Digitized Sky Survey images, distance

284: estimates are obtained for all of the sources in \S\ref{sec:dist}. Of

285: 60 BLAST sources, 49 objects are likely associated with the molecular

286: clouds surrounding NGC~6823 at $\sim2.3$\,kpc, 10 appear to be in the

287: more distant Perseus arm ($\sim8.5$\,kpc), and a single object is

288: thought to lie in the outer Galaxy ($\sim14$\,kpc). These distances

289: are used to calculate source luminosities and masses in

290: \S\ref{sec:mass}.  In this section we also present an estimate of the

291: mass function for sources associated with NGC~6823.  In this survey

292: high-mass clumps are seen in a range of evolutionary stages: those

293: with embedded UC \ion{H}{2} regions; luminous objects without

294: substantial radio emission (HMPOs); and cold gravitationally-bound

295: clumps (or cores) with low luminosity-to-mass ratio and no evidence of

296: star formation (\S\ref{sec:disc}).

297: 

298: 

299: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

300: \section{BLAST OBSERVATIONS TOWARDS NGC~6823} 

301: \label{sec:obs}

302: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

303: 

304: %****************************************

305: \subsection{Observing Strategy}

306: %****************************************

307: 

308: BLAST05 is a 2-m Cassegrain telescope, whose under-illuminated primary

309: mirror is designed to produce diffraction-limited beams with FWHM

310: 40\arcsec, 58\arcsec, and 75\arcsec at 250, 350, and 500\,\micron\

311: respectively.  The camera consists of three silicon-nitride ``spider

312: web'' bolometer arrays \citep{turner2001} almost identical to those

313: for SPIRE on {\it Herschel} \citep{griffin2003}, with 149, 88, and 43

314: detectors at 250, 350, and 500\,\micron, organized in a hexagonal

315: close-packed pattern.  Radiation is coupled to the bolometers using

316: 2$f\lambda$ spaced conical feed-horns, so that the $14' \times 7'$

317: field-of-view (simultaneously imaged by all three arrays) is

318: instantaneously under-sampled. The telescope must therefore scan in

319: order to produce fully-sampled images.  Scanning is also required to

320: modulate the bolometer signals to remove low-frequency detector drift.

321: A full description of the BLAST telescope and detectors is given in

322: \citet{pascale2007}. The 2005 Sweden flight performance is described

323: in \citet{truch2007}.

324: 

325: The optimal detector noise was obtained by scanning the telescope in

326: azimuth at 0.1\,deg\,s$^{-1}$ while drifting slowly in elevation such

327: that the scan lines are spaced 65\arcsec\ apart. In addition,

328: observations of Vulpecula were split between times when it was rising

329: and setting, resulting in cross-linked scans at angles of $\sim

330: 45^\circ$. This cross-linking greatly reduces large-scale $1/f$ noise

331: in the map.  Details of the BLAST scanning technique are given in

332: \citet[][]{pascale2007}.

333: 

334: 

335: %****************************************

336: \subsection{Data Reduction} \label{sec:obsandred}

337: %****************************************

338: 

339: The raw BLAST05 data are reduced using a common pipeline detailed in

340: \citet{pascale2007} and \citet{truch2007}.  First, the 100\,Hz sampled

341: bolometer and gondola pointing data are de-spiked and the digital

342: filter responses are deconvolved.  Time-varying bolometer

343: responsivities are tracked using an internal calibration lamp. The

344: absolute gain of the instrument (including antenna efficiency),

345: determined from regular observations of Arp 220, is measured with

346: absolute uncertainties of 8\%, 10\%, and 12\% at 500, 350, and

347: 250\,\micron\ respectively.  The relative pointing of the telescope

348: within a single map is determined to $<5''$ rms by integrating rate

349: gyroscopes with star trackers providing an absolute reference.

350: Residual pointing offsets between different observations of Vulpecula

351: are removed by direct rebinning of the data into maps, and aligning

352: the peaks of the seven brightest compact sources from each pass.

353: Finally, maps are made using a new algorithm called Signal And Noise

354: Estimation Procedure Including Correlations (SANEPIC). This technique

355: has evolved from strategies used to produce maps of the Cosmic

356: Microwave Background. A model is first developed for the raw bolometer

357: time-stream data in which the astronomical signal is produced by

358: projecting the estimate of the map into the time domain using the

359: known positions of each detector over time, and assuming that the

360: noise is Gaussian, stationary, and correlated in time and between

361: detectors.  The brightness of each map pixel is considered a free

362: parameter in this model. A maximum likelihood solution for the map is

363: then found using a direct inversion technique. This calculation

364: includes the full noise power spectra of the bolometers, and

365: propagates knowledge of correlations in the bolometer data to pixels

366: in the map.  Low-frequency noise (predominantly slowly-varying sky

367: emission) is naturally removed by this process.  In particular, to

368: reconstruct large spatial scales in the map, this method gives more

369: weight to data taken during a single visit, during which the bolometer

370: zero-point drift is minimal.  The DC level of the map is

371: unconstrained, and set to 0 by applying a weak high-pass filter to the

372: bolometer data before map-making.  The details of SANEPIC are given in

373: \citet{patanchon2007}.

374: 

375: 

376: %****************************************

377: \subsection{Overview of the Maps} \label{sec:mapoverview}

378: %****************************************

379: %\clearpage

380: \begin{figure} 

381: \centering

382: \includegraphics[width=0.95\linewidth]{f2.eps}

383: \caption{Raw BLAST05 maps of Vulpecula produced with SANEPIC

384:   \citep{patanchon2007}. The similarity between the bands is due to

385:   the filters sampling primarily the Rayleigh-Jeans spectrum of

386:   thermal dust emission.  Previously identified {\it IRAS} sources

387:   associated with high-mass star forming regions are indicated on the

388:   350\,$\mu$m map (see \S\ref{sec:mapoverview}). It is assumed

389:   that most of the submillimeter emission originates in a molecular

390:   cloud complex at the distance of the open cluster NGC~6823

391:   ($\sim2.3$\,kpc away) indicated in the 500\,\micron\ map (the two circles

392:   give the cluster core size, and maximum extents, respectively).  The

393:   locations of several OB supergiants, an O7 V ((f)) star, and a

394:   supernova remnant, all associated with NGC~6823, are also shown.

395:   Sh2-86 is an extended \ion{H}{2} region excited by these massive

396:   stars.}

397: \label{fig:blastmaps}

398: \end{figure}

399: %\clearpage

400: The BLAST05 maps of Vulpecula are shown in Figure~\ref{fig:blastmaps}.

401: The similarity of the three maps is striking, both in the diffuse

402: emission and the peaks, demonstrating that the bulk of the

403: measurements fall on the Rayleigh-Jeans tail of the thermal emission.

404: In this case, they are a good measure of dust column density in the

405: molecular cloud and the mass of the compact sources.

406: %\clearpage

407: \begin{deluxetable}{ccrrr}

408: \tablewidth{0pt}

409: \tablecaption{Vulpecula map sensitivities.  \label{tab:mapsens}}

410: \tablehead{

411: \colhead{Band} &

412: \colhead{FWHP} &

413: \colhead{$\sigma_\mathrm{b}$} &

414: \colhead{$\sigma_\mathrm{p}$} &

415: \colhead{$\sigma_\mathrm{s}$} \\

416: \colhead{(\micron)}  &

417: \colhead{(\arcsec)} &

418: \colhead{(\arcsec)} &

419: \colhead{(mJy)} &

420: \colhead{(Jy)}

421: }

422: \startdata

423: \multicolumn{5}{c}{\it Raw maps} \\

424: 250 & 207 & 88.2 &  18.0 & 0.32 \\

425: 350 & 206 & 87.8 &   9.9 & 0.17 \\

426: 500 & 212 & 90.4 &   9.9 & 0.18\vspace{1 ex}\\

427: \multicolumn{5}{c}{\it Deconvolved maps} \\

428: 250 &  40 & 17.0 & 350.0 & 1.17 \\

429: 350 &  50 & 21.3 &  99.0 & 0.41 \\

430: 500 &  60 & 25.5 & 122.0 & 0.61 \\

431: \enddata

432: \tablecomments{Full-Width Half-Power (FWHP) is the diameter in which

433:   50\% of the beam power is contained (equivalent to Full-Width

434:   Half-Maximum for Gaussian beams).  The standard deviation (size) of

435:   a Gaussian beam with the equivalent FWHP is given by

436:   $\sigma_\mathrm{b}$. The rms noise in a single 18\arcsec\ map pixel

437:   is given by $\sigma_\mathrm{p}$. The rms noise in the equivalent

438:   FWHP Gaussian beam is given by $\sigma_\mathrm{s}$.  Values are

439:   presented both for ``Raw maps'' produced with SANEPIC

440:   (\S\ref{sec:mapoverview}) and for ``Deconvolved maps''

441:   (\S\ref{sec:decon}).}

442: \end{deluxetable}

443: %\clearpage

444: 

445: The far-infrared (60 and 100\,\micron) emission in this part of

446: Vulpecula is dominated by several luminous high-mass star forming

447: regions.  Three of these have been studied recently by \cite{bel06}

448: and \citet{zha05}, associated with the following IRAS sources (for

449: later cross-reference we append the BLAST name from

450: Table~\ref{tab:iras}): 19368+2239 (V03); 19374+2352 (V05); and

451: 19388+2357 (V08).  A further four regions have been studied

452: extensively by \cite{beu02}, using CS multi-line, multi-isotopolog

453: observations and the 1.2\,mm dust continuum: 19403+2258 (V18);

454: 19410+2336 (V30); 19411+2306 (V32); and 19413+2332 (V40).

455: %

456: Note that these do not exhaust the list of bright sources, either

457: at submillimeter or far-infrared wavelengths.

458: 

459: In the submillimeter, the brightest high-mass star forming region near

460: IRAS~19410+2336 is visible in the north-east part of the BLAST maps

461: (Figure~\ref{fig:blastmaps}). The other bright regions are located on

462: a roughly ``C''-shaped arc readily apparent in the {\it IRAS} maps.

463: As with {\it IRAS}, the BLAST emission is in most cases centrally

464: peaked, representing a single massive clump at the observed resolution

465: of BLAST.  There is also diffuse emission from extended clouds of

466: dust, as well as blending of (clustered) point sources convolved with

467: the beam.

468: 

469: In the 2005 flight, the point spread function (PSF) was not

470: diffraction-limited (40\arcsec, 58\arcsec, and 75\arcsec\ at 250, 350,

471: and 500\,\micron\ respectively) as originally designed. The resulting

472: full-width half-powers (FWHP, equivalent to the full-width

473: half-maximum for a Gaussian) in all three bands are approximately

474: $3\farcm5$ and exhibit complex shapes that vary as a function of

475: wavelength and position in the focal plane \citep{truch2007}.

476: 

477: The beam shape is clearly recognizable at many locations across the

478: entire field, suggesting the presence of point-like objects with

479: angular scales smaller than the beam. However, the overlapping beam

480: patterns also indicate that the maps are highly source-confused

481: (Figure~\ref{fig:blastmaps}).  Even though only 3--8\% of the beam

482: power is found in the central diffraction-limited peak at each

483: wavelength, the signal-to-noise (\snr) of the maps are high enough

484: (compare sensitivity for ``Raw maps'' in Table~\ref{tab:mapsens} with

485: typical source flux densities in later tables) that we are able to

486: deconvolve a significant fraction of the beam to recover angular

487: scales close to the diffraction limit.  We can then use these

488: deconvolved maps to search for compact sources.

489: 

490: %****************************************

491: \subsection{Image Deconvolution}

492: \label{sec:decon}

493: %****************************************

494: 

495: Deconvolution in astronomy has a long history \citep[see the review

496: by][]{starck2002}. The convolution problem can be stated, using the

497: notation of \citet{starck2002}, as

498: 

499: \begin{equation}

500: \label{eqn:convprob}

501: I(x,y) = (O \ast P)(x,y) + N(x,y),

502: \end{equation}

503: 

504: \noindent where $I$ is the observed map, $O$ is the true image, $P$ is

505: the instrument's PSF, $N$ is measurement noise, and ``$\ast$'' is the

506: convolution operator. Invoking the convolution theorem, this

507: expression can be written in Fourier space as

508: 

509: \begin{equation}

510: \hat{I}(u,v) = \hat{O}(u,v) \hat{P}(u,v) + \hat{N}(u,v) ,

511: \end{equation}

512: 

513: \noindent where $\hat{X}$ is the Fourier transform of $X$. A naive

514: solution to the convolution problem is found by simple division in

515: Fourier space,

516: 

517: \begin{equation}

518: \label{eqn:convsol}

519: \hat{\tilde{O}}(u,v) = \frac{\hat{I}(u,v)}{\hat{P}(u,v)} = 

520:   \hat{O}(u,v) + \frac{\hat{N}(u,v)}{\hat{P}(u,v)}.

521: \end{equation}

522: 

523: However, in practice, this direct inversion method amplifies the noise

524: at high spatial frequencies. Many methods have been developed to solve

525: Equation~\ref{eqn:convprob} iteratively, including least squares,

526: maximum likelihood and wavelet-based methods.  These algorithms

527: typically require high-precision knowledge of the PSF, which we do not

528: have for the BLAST05 maps. The effective PSFs vary significantly

529: across the maps due to asymmetries in the beam-pattern (which itself

530: is well-measured).  Over time the orientation of this pattern on the

531: sky changes, such that the effective shape of a point source depends

532: on the amount of time spent observing at different parallactic angles,

533: and instrumental noise variations. Iterative methods for deconvolution

534: were attempted, but the variable PSFs proved to be too problematic. We

535: concluded that strong artifacts due to convolution are inevitable and

536: we decided to use the direct inversion method, even though it is

537: probably non-optimal.

538: 

539: 

540: Equation~\ref{eqn:convsol} can be rewritten in real space as

541: $\tilde{O} = I \ast K$, where $K$, the deconvolution kernel, is the

542: inverse transform of $\hat{P}^{-1}$. In order to suppress the

543: amplification of noise at high frequencies we effectively re-convolve

544: the map by a Gaussian, $G$, with width approximately equal to the

545: designed diffraction limit of the telescope. This can be combined with

546: the deconvolution formulation by writing $\hat{K} = \hat{G} /

547: \hat{P}$. Here $\hat{K}$ is also smoothed at large spatial frequencies

548: to further suppress noise spikes as $\hat{P}$ goes to zero.

549: Additionally, spikes in $\hat{K}$ due to zero-crossings in $\hat{P}$

550: are clipped.

551: 

552: The PSFs are estimated by two methods. In the first case, a synthetic

553: PSF is constructed based on measurements of point sources throughout

554: the flight. The synthetic PSF attempts to account for the range of

555: parallactic angles over which the field was observed by averaging

556: together appropriately-weighted beam rotations. The second method uses

557: a point source directly from the map. An isolated source from the

558: north-west corner of the map is used (V05 in Figure~\ref{fig:decon}),

559: this having relatively low surrounding diffuse emission. A low-order

560: polynomial is fit to the background and subtracted from the source to

561: remove the diffuse emission.  The deconvolutions are performed using

562: both types of PSF.  The synthetic PSFs provide the best results at 350

563: and 500\,\micron, while the isolated source is better for the

564: 250\,\micron\ map, based on the amplitude of the ripples seen in the

565: resulting deconvolved maps.

566: 

567: In order to reduce edge effects caused by the Fourier transforms, both

568: the maps, $I$, and the PSFs, $P$, are apodized and zero-padded prior

569: to deconvolution. The map is apodized over a scale of 5\arcmin\ along

570: a rectangle bordering the map and the PSF is apodized over 2\arcmin\

571: along a circle of radius 4\arcmin.

572: 

573: Finally, the variance map produced by SANEPIC is propagated through

574: the deconvolution filter, providing a noise map for use in

575: source-finding and fitting. The noise at each wavelength both before

576: and after deconvolution are given in Table~\ref{tab:mapsens} under

577: ``Raw maps'' and ``Deconvolved maps'' respectively.

578: 

579: %\clearpage

580: \begin{figure} 

581: \centering

582: \includegraphics[width=\linewidth]{f3.eps}

583: \caption{The grayscale image is a resolution enhanced BLAST

584:   350\,\micron\ map, with the locations of compact sources

585:   (Table~\ref{tab:src}) indicated: 49 objects associated with the

586:   clouds surrounding NGC~6823 at $\sim$2.3\,kpc (circles); 10 objects

587:   in the Perseus Arm at $\sim$8.5\,kpc (triangles); and a single

588:   outer-galaxy object at $\sim$14\,kpc (diamond) --- see

589:   \S\ref{sec:dist} for a discussion of distance estimates.  The

590:   resolution enhancement was accomplished with direct Fourier

591:   deconvolution of the PSF (see \S\ref{sec:decon}).  Residual

592:   ringing is an artifact of uncertainties in the beam shape which

593:   varies slightly across the map.}

594: \label{fig:decon}

595: \end{figure}

596: %\clearpage

597: 

598: %****************************************

599: \subsection{Compact Source Identification}

600: \label{sec:sourceid}

601: %****************************************

602: 

603: We search for individual compact sources in the deconvolved maps using

604: source-finding analysis based on the use of a ``compensated PSF'',

605: also called the Mexican Hat Wavelet (MHW) technique \citep[see

606: e.g.,][]{barnard2004}, which identifies objects in confused images by

607: subtracting a local background. We apply the MHW with a characteristic

608: width equal to the nominal resolution (the width of $G$ in the

609: deconvolution), to the BLAST deconvolved maps. The peaks of all $\sim

610: 6$-$\sigma$ clumps in each waveband are identified as potential

611: sources. Due to the excess noise produced by the deconvolution

612: process, many false detections are found near bright sources.  No

613: robust method to automatically reject these false peaks was found,

614: since the ripples are not uniform across the image. We visually reject

615: any peak that appears to be associated purely with the noise ripples.

616: In addition, 11 of the recovered sources that lie slightly further

617: away from the brightest objects remain clearly affected by these

618: residual ripples. Since the ripples are different in each band the net

619: effect is to contaminate the observed BLAST colors.  In total,

620: $\sim50$\% of the rejected sources have formal statistical

621: significances below $9$-$\sigma$, although there is a tail of rejected

622: sources extending to $\sim25$-$\sigma$ that have relatively higher

623: systematic errors since they lie on the largest ripples next to the

624: brightest submillimeter objects.  The positions of BLAST sources

625: derived from the deconvolved maps are shown in Figure~\ref{fig:decon}.

626: 

627: The list of robust detections in each waveband are combined and a

628: circular region around each peak in the three maps is then

629: simultaneously fit with a single Gaussian. In addition nearby sources

630: are fit using multiple Gaussians. In each region, a fourth order

631: polynomial baseline is fit to the surrounding regions and subtracted.

632: The amplitudes in each waveband, position, and width are all free

633: parameters in the fit. The fit is calculated using a non-linear

634: least-squares minimization routine. Constraints on source position and

635: width are included to reduce the possibility of divergence in the

636: fitting procedure. Leaving the source width as a free parameter in the

637: fit biases the resulting flux densities high, but is necessary since

638: the beam varies significantly across the field.  This bias is

639: investigated in the next section.  Given the \snr\ of these sources,

640: however, even after deconvolution, the flux density uncertainty is

641: dominated by the calibration error.  The catalog of source positions

642: and BLAST flux densities are given in Table~\ref{tab:src}.

643: 

644: %****************************************

645: \subsection{Monte Carlo Simulations}

646: \label{sec:montecarlo}

647: %****************************************

648: 

649: In order to estimate the effectiveness of our compact source

650: extraction process, a series of Monte Carlo simulations are performed.

651: Flux densities, biases, completeness estimates, and positional error

652: distributions are investigated by inserting a point source, convolved

653: with the estimate of the beam, into the actual BLAST maps produced

654: with SANEPIC. The deconvolution and source extraction processes are

655: applied and the extracted flux densities compared with the inputs. The

656: source extraction routine is modified slightly from that described in

657: \S\ref{sec:sourceid} in order to avoid manually rejecting unreliable

658: sources in residual ripples around bright sources --- a simple

659: automatic routine was implemented that is able to reproduce the manual

660: rejection procedure for the original source list to an accuracy of

661: $\sim 10$\%.

662: 

663: In order to extract the flux density bias and errors, 500 iterations

664: are performed at each of several input flux densities spanning the

665: range of source brightnesses found in the map.  At each iteration, a

666: 250\,\micron\ flux density is chosen and the 350 and 500\,\micron\

667: flux densities are assigned based on the median source colors found in

668: the field. A simulated source is considered detected if an additional

669: peak is found compared to the number of sources originally found in

670: the map. In some cases, the simulated source will by chance land on

671: top of a real source at the same position. In such cases, the

672: detection of the simulated source is only counted if the flux density

673: exceeds that of the real source that it obscures. The resulting output

674: flux density distributions are roughly Gaussian with a positive tail.

675: The Bayesian 68\% confidence limits are found and the distributions

676: are fit by the Gaussian that passes through these points. We find that

677: the measured flux density is biased high by a factor of 1--1.2,

678: independent of input flux density (Figure~\ref{fig:blastfluxsims}).

679: Based on these results, we can describe the spread in flux density

680: with a 2-component model: the first is independent of flux density,

681: reflecting noise in the map, and the other is linear in flux density,

682: due to uncertainty in the fitting process.  Positional errors are

683: calculated using the same simulations. These errors are described by

684: circular Gaussians with $\sigma$ ranging from $\sim 7$\arcsec\ at the

685: bright end to $\sim 70$\arcsec\ at the faint end.  

686: %\clearpage

687: \begin{figure}%[t] 

688: \centering

689: \includegraphics[width=.8\linewidth]{f4.eps}

690: \caption{The flux density bias, determined from 500 Monte Carlo

691:   simulations, is shown for each waveband. The bias is modeled as a

692:   constant function of input source flux density and the errors are

693:   modeled with two components (described in

694:   \S\ref{sec:montecarlo}).}

695: \label{fig:blastfluxsims}

696: \end{figure}

697: %\clearpage

698: Completeness is estimated using the same set of simulations, except

699: that the flux density fits are not performed. At each of the input

700: flux densities, 1000 simulations are performed and the number of times

701: that the input is detected is counted. The results are shown in

702: Figure~\ref{fig:blastcomplsims}.

703: %\clearpage

704: \begin{figure}%[t] 

705:   \centering

706:   \includegraphics[angle=90,width=\linewidth]{f5.eps}

707:   \caption{The completeness as a function of 250\,\micron\ flux

708:     density determined from 1000 Monte Carlo simulations (described in

709:     \S\ref{sec:montecarlo}).}  

710:   \label{fig:blastcomplsims}

711: \end{figure} 

712: %\clearpage

713: 

714: In total, 60 compact sources are detected in the Vulpecula maps. Flux

715: densities and their uncertainties are given in Table~\ref{tab:src}.

716: Note that the tabulated values include a color correction based on SED

717: fits; see \S\ref{sec:colorcorrect} for a description of this

718: procedure.  Figure~\ref{fig:decon} indicates the positions of the

719: detected sources.

720: \section{INFRA-RED PHOTOMETRY}

721: \label{sec:firsed} 

722: 

723: The BLAST data alone cannot accurately constrain the SED corresponding

724: to thermal emission in many of the sources from this survey.  At

725: temperatures $T\ga25$\,K (and $\beta=1.5$) all of the BLAST filters

726: sample the Rayleigh-Jeans tail. To measure the luminosity and dust

727: mass, additional photometry at slightly shorter wavelengths

728: ($\sim$50--100\,\micron) is required to identify the emission peak.

729: FIR measurements and upper limits are estimated for all of the sources

730: using {\it IRAS} 60 and 100\,\micron\ photometry and the new MIPS

731: 70\,\micron\ map from MIPSGAL \citep{car05}.  Although most of the

732: luminosity is produced at submillimeter and FIR wavelengths, a

733: non-negligible portion is also observed in the mid-infrared (MIR)

734: ($\lambda\sim$10--50\,\micron). This radiation is emitted from hotter

735: dust that lies close to the heat source.  Protostars produce

736: significant amounts of radiation in this band, since the material in

737: the vicinity of the object is optically thick. This material is

738: irradiated with UV light which is then re-emitted at longer

739: wavelengths.  To probe the MIR SEDs of BLAST sources we used {\it

740:   IRAS}\/ 12 and 25\,\micron\ measurements and the {\it MSX}

741: \footnote{\url{http://www.ipac.caltech.edu/ipac/msx/msx.html}} catalog

742: at 8, 12, 14, and 21\,\micron.

743: 

744: %****************************************

745: \subsection{{\it IRAS} 12, 25, 60, and 100\,\micron\ Flux Densities}

746: \label{sec:iras}

747: %****************************************

748: 

749: A significant fraction of the BLAST sources have clear counterparts in

750: the {\it IRAS} Point Source Catalog version 2.0

751: \citep[PSC,][]{helou1988}. We add the BLAST positional uncertainties

752: in quadrature with the semi-major axes of the PSC error ellipse to

753: determine search radii for each source. The BLAST uncertainties are

754: taken from the simulations described in \S\ref{sec:montecarlo},

755: but we conservatively set the error at the bright end to 30\arcsec.

756: Identifications from the {\it IRAS} PSC were found for 23 of the 60

757: BLAST sources, as summarized in Table~\ref{tab:iras}.

758: 

759: For sources that lack PSC counterparts or measurements in any of the

760: {\it IRAS} bands, we produce measurements or upper limits directly

761: from {\it IRAS} maps. We use the {\it IRAS} Galaxy Atlas

762: \citep[IGA,][at 60 and 100\,\micron]{cao1997} and the Mid-Infrared

763: Galaxy Atlas \citep[MIGA,][at 12 and 25\,\micron]{kerton2000}. These

764: maps are produced using the resolution-enhancing algorithm HIRES

765: \citep{aumann1990}. The PSFs vary across these maps, showing strong

766: elongation along the scan direction. However, these HIRES maps resolve

767: sources that are also detected with BLAST (see

768: Figure~\ref{fig:irascompare}) yet appear confused in {\it IRAS} maps

769: that have not undergone resolution enhancement \citep[e.g., the IRIS

770: maps of][]{miv2005}.

771: 

772: We use aperture photometry to measure flux densities in the IGA and

773: MIGA maps. A circular aperture radius of 2\farcm4 is used at

774: 100\,\micron.  For the remaining bands an elliptical aperture aligned

775: with the scan direction is used, with semi-axes $1\farcm8 \times

776: 1\farcm2$. These apertures are centred over the BLAST coordinates and

777: the maps are integrated.  Baseline pixel values are estimated as the

778: median in an annulus between the outer-edge of the photometry aperture

779: and a second circle (or ellipse) that is larger by a factor of 1.3.

780: This size was chosen through trial-and-error as a compromise between

781: smaller sizes which encompass fewer pixels (baseline estimates with

782: larger statistical errors), and larger sizes which suffer greater

783: contamination from large-scale extended structure or adjacent sources

784: (baseline estimates with larger systematic errors).  Since many of the

785: sources remain confused, the flux density and baseline measurements

786: are both checked visually. In cases where a small amount of emission

787: is detected in the baseline annulus, the confused fraction is excised

788: in the estimates.  Furthermore, if additional sources are seen near

789: the edge of the measurement aperture, or if it appears to lie on a

790: bright gradient of background emission, the measurement is flagged as

791: an upper-limit.  Finally, if no source is visible in the aperture, it

792: is flagged as a non-detection. The uncertainty in all of the

793: measurements is estimated as the rms in the non-detections. The

794: results of this photometry procedure are summarized in

795: Table~\ref{tab:mapphot}.

796: 

797: To verify our technique, we compare these aperture measurements with

798: flux densities from the {\it IRAS} PSC and find no significant bias,

799: with scatters of 12\% at 60\,\micron\ and 25\% at 100\,\micron.

800: 

801: %****************************************

802: \subsection{MIPS 70\,\micron\ Flux Densities}

803: \label{sec:mips70}

804: %****************************************

805: 

806: After examining the {\it IRAS} PSC, and measuring flux densities in the

807: IGA maps, 32 of the 60 BLAST sources lacked a clear detection at

808: either 60 or 100\,\micron. To constrain the FIR SEDs in these cases,

809: we use mosaics of the 70\,\micron\ MIPSGAL \citep{car05} images

810: downloaded from the {\it Spitzer} public web site using the {\it

811:   Spitzer} Pride

812: software.\footnote{\url{http://ssc.spitzer.caltech.edu/propkit/spot/}}

813: The 8\,$\mu$m IRAC image (Figure~\ref{fig:iracblast}) is also produced

814: using this software.

815: 

816: Similar to the flux densities obtained from {\it IRAS} maps in

817: \S\ref{sec:iras}, aperture photometry with baseline correction is used

818: for the 70\,\micron\ map. The radius of the measurement aperture is

819: 1\arcmin\ and the baseline aperture is an annulus with inner and outer

820: radii of 1\arcmin\ and 1\farcm3. It was noticed that these

821: measurements yield flux densities systematically lower than the

822: 60\,\micron\ observations, despite the fact that the SED for every

823: source should be brighter at 70\,\micron, since these wavelengths fall

824: on the Wien tail of the thermal emission in all cases. This problem

825: arises because saturation and non-linearities affect MIPS 70\,\micron\

826: sources with flux densities $\ga 20$\,Jy (A.  Noriega-Crespo, private

827: communication).  Fortunately only the fainter BLAST sources require

828: 70\,\micron\ photometry to constrain their FIR SEDs, and adopting a

829: 20\,Jy cut in the catalog of 70\,\micron\ measurements yields useful

830: data for 10 BLAST sources with no {\it IRAS} detections. The

831: uncertainty in this photometry was estimated as 3\,Jy from the rms of

832: observations that were visually flagged as non-detections.  A

833: comparison between MIPS, {\it IRAS}, and BLAST is shown in

834: Figure~\ref{fig:irascompare}, and the MIPS photometry is summarized in

835: Table~\ref{tab:mapphot}.

836: 

837: %****************************************

838: \subsection{500--8\,\micron\ Map Comparison}

839: \label{sec:mapcompare}

840: %****************************************

841: 

842: %\clearpage

843: \begin{figure}%[t] 

844: \centering

845: \includegraphics[width=\linewidth]{f6.eps}

846: \caption{Comparison between BLAST05 deconvolved maps, {\it IRAS} maps

847:   produced with HIRES, and the {\it Spitzer} MIPS (70\,\micron) and

848:   IRAC (8\,\micron) maps. Each thumbnail has a size of $0\fdg3 \times

849:   0\fdg3$. Circles on the BLAST maps and IRAC map indicate source

850:   positions with arbitrary diameters 2\farcm5 and 2\arcmin\

851:   respectively.  The circles and ellipses indicated on the {\it IRAS}

852:   and MIPS maps are the apertures that were used for photometry. The

853:   SED of the coldest object, V11, is shown in

854:   Figure~\ref{fig:mergedSED2}.}

855: 

856: \label{fig:irascompare}

857: \end{figure}

858: %\clearpage

859: 

860: Figure~\ref{fig:irascompare} shows an expanded view of a $0\fdg3

861: \times 0\fdg3$ region of the maps described in the previous sections,

862: encompassing the BLAST sources V11, V12, V13, V15, and V16.  It is

863: clear that the relative brightnesses of sources vary greatly over the

864: range 500--8\,\micron, due to their different intrinsic SEDs.  None of

865: the BLAST sources in this region of the map have counterparts in the

866: {\it IRAS} PSC\@.  Source confusion is also an issue; for example,

867: although V12 is not in the {\it IRAS} PSC, a counterpart is seen in

868: the higher resolution IGA 60\,\micron\ and the MIPS 70\,\micron\

869: images (and MIGA as well).  V16 has a faint counterpart at

870: 70\,\micron.  In the same images, {\it IRAS} PSC sources not seen by

871: BLAST appear: 19395+2313 between V11 and V13, 19397+2309 to the NE of

872: V12, and 19399+2312 W of V16; the source SW of V15 is not in the {\it

873:   IRAS} PSC.  Based on the brightness and color of 19397+2309 at 60

874: and 100\,\micron\, (23 and 80\,Jy respectively), we would expect to

875: have detected this source with BLAST.  Comparison of the 250\,\micron\

876: image with the 70\,\micron\ image (where the source is resolved into

877: two), suggests a detection by BLAST, but it did not meet our

878: compactness criteria to be tabulated as a point source.  At even

879: shorter wavelengths, additional sources and nebulosities appear, for

880: example 19397+2315 E of V11.

881: 

882: 

883: %****************************************

884: \subsection{{\it MSX} 8, 12, 14, and 21\,\micron\ Flux Densities}

885: \label{sec:msx}

886: %****************************************

887: 

888: {\it MSX} flux densities for the BLAST sources are obtained from a

889: cross-correlation with the {\it MSX} Point Source Catalog version 2.3

890: \citep{egan2003}.  As with the {\it IRAS} PSC

891: (\S\ref{sec:iras}), {\it MSX} counterparts are identified within

892: a search radius of the BLAST positions that varies as a function of

893: the source brightness. The positional uncertainties in the {\it MSX}

894: catalog are negligible by comparison (rms $\sim4$--5\arcsec).  We find

895: potential {\it MSX} counterparts for 40 objects in the BLAST catalog.

896: Inspection of the {\it MSX} maps shows that some of these sources are

897: found in regions with diffuse emission, and, given the angular

898: resolution of the BLAST catalog, we cannot presently determine whether

899: all of the proposed {\it MSX} associations are BLAST objects, or other

900: protostars in the same star forming region at an older evolutionary

901: stage. In cases where a single {\it MSX} counterpart is identified we

902: assume that it is in fact the same source. In cases with multiple

903: proposed {\it MSX} counterparts, however, the sum of the flux

904: densities is taken as an upper limit. Presently no attempt has been

905: made to use the shape of the {\it MSX} SEDs to justify their

906: associations with the BLAST objects. All candidates are given in

907: Table~\ref{tab:msx}.

908: 

909: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

910: \section{SUBMILLIMETER--MIR SEDS} 

911: \label{sec:coldsed}

912: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

913: 

914: With submillimeter--MIR photometry, we fit the temperature and

915: bolometric flux produced by the coldest dust, as well as the

916: bolometric flux produced in the MIR.  We note that all of our sources

917: are likely composed of regions at different temperatures, typically a

918: warmer core embedded in a colder and less dense medium.  Our goal is

919: to use a simple SED model as an interpolation function for the

920: sparsely sampled photometry to estimate the total luminosity of each

921: source, and to infer an approximate temperature for the dominant

922: emission from cold dust.

923: 

924: We assume optically-thin emission from an isothermal modified

925: blackbody,

926: %

927: \begin{equation}

928:   S_{\nu} = A

929:   \left(\frac{\nu}{\nu_0}\right)^{\beta} B_{\nu}(T),

930: \label{eq:sed}

931: \end{equation}

932: %

933: where $A$ is the amplitude of the SED, $B_{\nu}(T)$ is the Planck

934: function, $\beta$ is the dust emissivity index, and the emissivity

935: factor (brackets) is normalized at a fixed frequency $\nu_0$.  Note

936: that the amplitude can be expressed in terms of a total clump mass,

937: $M_{\rm c}$, the dust mass absorption coefficient $\kappa_0$

938: (evaluated at $\nu_0$), and the distance to the object, $d$,

939: %

940: \begin{equation}

941: A = \frac{M_{\rm c} \kappa_0}{Rd^2}.

942: \label{eq:mass}

943: \end{equation}

944: %

945: Since $\kappa_0$ refers to a dust mass, the gas-to-dust mass ratio,

946: $R$, is required in the denominator to infer total masses.  We adopt

947: $\kappa_{0} = 10$\,cm$^2$\,g$^{-1}$, evaluated at

948: $\nu_0=c/250$\,\micron, from \citet{hildebrand1983}. A factor of 100

949: is assumed for $R$. We note that the combined mass uncertainty due to

950: $\kappa_{0}$ and $R$ is at least as large as a factor $\sim2$

951: \citep[e.g.,][]{hildebrand1983,ossenkopf1994,kerton2001}, depending on

952: assumptions about the environment, age, chemical composition, and

953: shape of the dust grains.  Note that our adopted value, along with

954: $\beta = 1.5$, is equivalent to the 1.2\,mm opacity adopted by

955: \citet{motte2007}.

956: 

957: Equation~\ref{eq:sed} is fit to all of the available photometry from

958: 500--60\,\micron\ using $\chi^2$ optimization (except for some cases

959: where the FIR peak is very broad, in which case it is only fit from

960: 500--100\,\micron). As in \citet{truch2007} the band-averaged flux

961: density of the model SED is calculated with knowledge of the BLAST

962: filter passbands before comparing them with measurements. We also

963: include the correlated calibration uncertainties with the statistical

964: uncertainties estimated in \S\ref{sec:montecarlo}.  The data

965: covariance matrix, $C$, is constructed by placing the estimated

966: variance for each data point along the diagonal. For the BLAST data

967: points, the diagonals are calculated as the quadrature sum of the

968: statistical uncertainties with the calibration uncertainties listed in

969: Table~1 in \citet{truch2007}. The off-diagonal cross-correlation terms

970: are estimated directly from the Pearson correlation coefficients in

971: the same table.  $\beta$ is not well constrained and is therefore

972: fixed to $\beta=1.5$ (consistent with Figure~\ref{fig:blastcolcol}) so

973: that only $A$ and $T$ are allowed to vary. For the 11 cases where the

974: BLAST colors are unreliable (\S\ref{sec:sourceid}; marked in

975: Table~\ref{tab:src}), a temperature of 20\,K has been adopted.

976: 

977: 

978: %****************************************

979: \subsection{Including Upper Limits in $\chi^2$}

980: \label{sec:upperlim}

981: %****************************************

982: 

983: Many of the fainter sources have only BLAST detections. In order to

984: make the best use of FIR upper-limits, ``survival analysis'' is used

985: to include the upper-limits in the calculation of $\chi^2$.  Given a

986: measurement that is deemed an upper limit, with flux density,

987: $s_{\mathrm{l}}$, and uncertainty, $\sigma_{\mathrm{l}}$, the likelihood

988: of the model flux density, $\tilde{s}$, is obtained from the integral of

989: the appropriate tail of the likelihood of a detection,

990: %

991: \begin{equation}

992: \label{eqn:survival}

993:   p(\tilde{s}|s_{\mathrm{l}},\sigma_{\mathrm{l}}) = 

994:   \frac{1}{\sqrt{2\pi}}\int^\infty_{(\tilde{s}-s_{\mathrm{l}})/\sigma_{\mathrm{l}}} e^{-z^2/2} dz,

995: \end{equation}

996: %

997: where we have assumed Gaussian noise.  Equation~\ref{eqn:survival} is

998: the ``survival function'' for censored data \citep{isobe1986}. The

999: behavior of this function is easily seen with a few examples.  If the

1000: model flux density is much smaller than the measured limit, then the

1001: likelihood goes to 1. If the model is equal to the measured limit, the

1002: likelihood is 0.5. The likelihood of models more than a few $\sigma$

1003: above the limit drops to $\sim0$.

1004: 

1005: Noting that $\chi^2$ is the negative log-likelihood function, and

1006: assuming the likelihood of the upper limits is independent of the

1007: detections, then

1008: %

1009: \begin{equation}

1010: \label{eqn:chisq}

1011:   \chi^2 = (\tilde{s}-s_{\mathrm{d}})C^{-1}(\tilde{s}-s_{\mathrm{d}})^T - 

1012:   \ln p(\tilde{s}|s_{\mathrm{l}},\sigma_{\mathrm{l}}).

1013: \end{equation}

1014: %

1015: Here $s_{\mathrm{d}}$ are the measured detections associated with the

1016: data covariance matrix, $C$. Equation~\ref{eqn:chisq} gives the

1017: expression that we minimize in order to fit the region of the data

1018: consistent with the isothermal SED model ($\ga 60$\,\micron).

1019: 

1020: %****************************************

1021: \subsection{SED Fits}

1022: \label{sec:sedfits}

1023: %****************************************

1024: 

1025: Uncertainties for the model parameters $A$ and $T$, and non-linearly

1026: dependent quantities such as the FIR integrated flux, $M_\mathrm{c}$,

1027: and their distance-independent ratio $L_{\mathrm{FIR}}/M_\mathrm{c}$,

1028: are obtained from Monte Carlo simulations.  Mock data sets are

1029: generated from realizations of Gaussian noise, including both

1030: correlated and uncorrelated errors, as described by the covariance

1031: matrix. The $\chi^2$ minimization process is repeated for each data

1032: set, and the resulting parameters and dependent quantities placed in

1033: histograms.  Means and 68\% Bayesian confidence intervals measured

1034: from the relevant histograms are given in Table~\ref{tab:sed}.  Note

1035: that errors in $T$ have opposite effects on $L_{\mathrm{FIR}}$ and

1036: $M_\mathrm{c}$, which exaggerates the uncertainty in

1037: $L_{\mathrm{FIR}}/M_\mathrm{c}$.

1038: 

1039: 

1040: Figure~\ref{fig:mergedSED} shows an example SED for V30, the brightest

1041: BLAST object in the sample. The solid gray lines indicate the 68\%

1042: confidence envelope of modified blackbodies that fit the BLAST and

1043: {\it IRAS} 100\,\micron\ data. The black line consists of the best-fit

1044: modified blackbody at wavelengths $> 100$\,\micron, and a series of

1045: piecewise-continuous power-laws through the {\it IRAS} and {\it MSX}

1046: data points at wavelengths $\le 100$\,\micron. The best fit

1047: temperature to the submillimeter data is 26\,K, considerably smaller

1048: than 46\,K estimated from MIR and mm data away from the peak

1049: \citep{Sridharan02}.  The luminosity fraction given in column 5 of

1050: Table~\ref{tab:sed} is the ratio (as a percentage) of the FIR

1051: integrated flux from the best-fit modified blackbody (solid gray

1052: lines) to the total bolometric flux, estimated from the integral of

1053: the black line across the wavelength range 2--5000\,\micron.

1054: Similarly, Figure~\ref{fig:mergedSED2} shows the SED of V11, the

1055: coldest object in the sample ($T=12.3$\,K). As seen in this example,

1056: observations across the BLAST wavelength range clearly reveal a

1057: turnover in the FIR SED for the coldest objects. The lack of

1058: counterparts in the shorter-wavelength bands is emphasized in

1059: Figure~\ref{fig:irascompare}.

1060: 

1061: The range of modified blackbody models from the Monte Carlo

1062: simulations are generally very tight (e.g.~for V30 $\Delta T =

1063: 1.0$\,K, and for V11 $\Delta T = 1.6$\,K). The narrowness of this

1064: range is largely due to the prior constraint that the dust emissivity

1065: index is precisely $\beta=1.5$. There is a strong degeneracy between

1066: the values of $\beta$ and $T$; for example, at larger values of

1067: $\beta$ the steepness of the Rayleigh-Jeans spectrum increases,

1068: requiring lower values of $T$ to compensate for that steepness, by

1069: bringing the FIR peak to longer wavelengths and hence closer to the

1070: shape of the data reported in the BLAST bands. For the two examples

1071: discussed here, we re-fit the $\lambda \ge 100$\,\micron\ region of

1072: the SED, letting $A$, $T$, and $\beta$ vary independently in 100 Monte

1073: Carlo simulations.  For V30 the resulting range of temperatures and

1074: dust emissivities are $T=27.6\pm2.8$\,K, and $\beta=1.4\pm0.3$

1075: (compared with $T=25.9\pm1.0$\,K when $\beta=1.5$).  For V11 the

1076: ranges are $T=9.4\pm3.4$ and $\beta=2.9\pm1.5$ (compared with

1077: $T=12.3\pm1.6$ when $\beta=1.5$). Particular realizations from these

1078: Monte Carlo simulations have been selected to illustrate the

1079: approximate 1-$\sigma$ range of $T$ as dashed gray lines in

1080: Figures~\ref{fig:mergedSED} and \ref{fig:mergedSED2}. Many of the

1081: BLAST sources have poorly constrained values of $T$ and $\beta$ as in

1082: the example of V11. We have therefore chosen to fix $\beta=1.5$ for

1083: the remainder of the analysis in this paper, which reduces inferred

1084: errors in $T$ by factors of $\sim2$--3. It also emphasizes the range

1085: of SEDs uncovered by BLAST, since under this constraint $T$ uniquely

1086: determines the wavelength of the FIR peak.

1087: 

1088: In addition to the effect of the prior on $\beta$, the treatment of

1089: correlated uncertainties between the BLAST measurements lead to

1090: inferred temperature ranges that are perhaps smaller than what one

1091: might expect given the sizes of the error bars in

1092: Figures~\ref{fig:mergedSED} and \ref{fig:mergedSED2}.  For example,

1093: V30 has photometric errors that are only $\sim2$\% in all three bands

1094: (Table~\ref{tab:src}). However, the error bars shown in these figures

1095: include approximately $\sim10$\% calibration uncertainties that are

1096: nearly 100\% correlated across the three bands.  This type of

1097: uncertainty only affects the absolute {\em scale} of the SED, $A$, but

1098: not the {\em shape} which is encoded in $T$ and $\beta$ (for which

1099: only the independent 2\% photometric errors are relevant). For this

1100: reason the temperature uncertainty in V30 (under the model

1101: assumptions) is only $\Delta T = 1.0$\,K, demonstrating the superior

1102: ability of BLAST to constrain the shape of the FIR SED of cold sources

1103: when systematic uncertainties are accounted for appropriately.

1104: 

1105: 

1106: %\clearpage

1107: \begin{figure}%[t]

1108: \includegraphics[width=\linewidth]{f7.eps}

1109: \caption{Example SED for V30, the brightest BLAST object

1110:   ($T=25.9$\,K).  BLAST error bars include correlated calibration

1111:   uncertainties ($\sim$10\%). Gray lines show the 68\% confidence

1112:   envelope of modified blackbody models from Monte Carlo simulations

1113:   in which the dust emissivity index is fixed at $\beta=1.5$. Dashed

1114:   gray lines are particular realizations from a second Monte Carlo

1115:   simulation in which $\beta$ is also left as a free parameter, here

1116:   indicating the 1-$\sigma$ envelope of $T$ (\S\ref{sec:sedfits}).

1117:   The black line shows the best-fit modified blackbody connected

1118:   continuously at 100\,$\mu$m to a series of piecewise-continuous

1119:   power-laws between the shorter-wavelength data.  }

1120: \label{fig:mergedSED}

1121: \end{figure}

1122: 

1123: \begin{figure}%[t]

1124: \includegraphics[width=\linewidth]{f8.eps}

1125: \caption{Example SED for V11, the coldest BLAST source ($T=12.3$\,K).

1126:   Symbols have the same meaning as in Figure~\ref{fig:mergedSED}.  The

1127:   FIR spectrum clearly turns over in the BLAST bands, and there are no

1128:   detections at the {\it IRAS}, MIPS or {\it MSX} wavelengths (see

1129:   Figure~\ref{fig:irascompare}). The bolometric flux for this source

1130:   can only be crudely estimated from the integral of the modified

1131:   blackbody.}

1132: \label{fig:mergedSED2}

1133: \end{figure}

1134: %\clearpage

1135: 

1136: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1137: \subsection{Color-corrected BLAST Flux Densities}

1138: \label{sec:colorcorrect}

1139: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1140: 

1141: BLAST filters have broad spectral widths that are $\sim 30$\% of the

1142: central frequency \citep{pascale2007}. Since colors sampled by the

1143: filters are a strong function of temperature for cooler objects

1144: ($T\la25$\,K), a color correction is required to enable direct

1145: comparison with SEDs.  Once the SED (Equation~\ref{eq:sed}) has been

1146: fit to data by minimizing $\chi^2$ (Equation~\ref{eqn:chisq}), we have

1147: the choice of either calculating different effective wavelengths for

1148: each measurement, or correcting each flux density at fixed

1149: wavelengths. We choose the latter, and correct the flux densities to

1150: precisely 250, 350, and 500\,\micron\ \citep[this procedure is

1151: described in][]{truch2007}. These corrected flux densities are given

1152: in Table~\ref{tab:src}, and a color-color plot for all 60 sources is

1153: shown in Figure~\ref{fig:blastcolcol}. This plot illustrates the

1154: variation in temperatures probed by the BLAST wavelengths, and

1155: confirms that the choice of $\beta=1.5$ for the SED fits is close to

1156: the center of the distribution. We have also tested the

1157: color-correction method using $\beta=1$ and $\beta=2$ and find that

1158: the effect on the measured flux densities is small compared with the

1159: photometric uncertainties.

1160: 

1161: %\clearpage

1162: \begin{figure}%[t] 

1163: \centering

1164: \includegraphics[width=\linewidth]{f9.eps}

1165: \caption{Color-color plot for the compact sources detected by BLAST05.

1166:   We have omitted several sources with unreliable colors, identified

1167:   in Table~\ref{tab:src}, and two sources with highly unconstrained

1168:   colors due to faint flux densities.  Modified blackbody models with

1169:   constant $\beta$, equal to 0, 1, 2, and 3, increasing towards the

1170:   bottom, and temperatures ranging from 5 to 200\,K are overplotted as

1171:   solid lines. The dashed line represents the same model with a

1172:   constant temperature of 20\,K and with $\beta$ ranging from $-1$ to

1173:   3. The 1-$\sigma$ statistical error bars are shaded so that the more

1174:   significant detections are darker. The flux densities plotted have

1175:   been band-corrected as discussed in \S\ref{sec:colorcorrect}.}

1176: \label{fig:blastcolcol} 

1177: \end{figure}

1178: %\clearpage

1179: 

1180: 

1181: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1182: \subsection{Distance-independent Clump Properties}

1183: \label{sec:dustproperties}

1184: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1185: 

1186: The range of temperatures inferred from the SED fits to BLAST data

1187: with reliable colors is shown in Figure~\ref{fig:temphist}. The survey

1188: has convincingly detected sources with a range of FIR peak

1189: wavelengths, since the temperature span, $10\,\mathrm{K}\la T\la

1190: 40\,\mathrm{K}$, is much larger than the individual uncertainties that

1191: are typically $\sim1$--2\,K (Table~\ref{tab:sed}). It should be noted

1192: that these latter uncertainties characterize the range of plausible

1193: temperatures {\it under the assumption of the simple SED model}

1194: (Equation~\ref{eq:sed}). There is also a systematic dependence on

1195: $\beta$; for example, increasing $\beta$ from 1.5 to 2.0 shifts most

1196: temperatures lower by about 5\,K, and derived masses higher by a

1197: factor of $\sim$2. Further details on correlations are given in

1198: \S\ref{sec:sedfits}.

1199: 

1200: %\clearpage

1201: \begin{figure}[t]

1202: \includegraphics[width=\linewidth]{f10.eps}

1203: \caption{Histogram of inferred dust temperatures, assuming a dust

1204:   emissivity index $\beta=1.5$, and excluding 11 objects for which the

1205:   BLAST colors are considered unreliable. The entire distribution

1206:   shifts downward by $\sim5$\,K if instead $\beta=2.0$ is adopted

1207:   (uncertainties are discussed in \S\ref{sec:sedfits}).}

1208: \label{fig:temphist}

1209: \end{figure}

1210: %\clearpage

1211: 

1212: As described above, the luminosity-to-mass ratio can also be

1213: calculated from the SED fits without any knowledge of the distance.

1214: For the simple SED model adopted, $L_{\mathrm{FIR}}/M_\mathrm{c}

1215: \propto T^{4+\beta}$ and so we find a broad range in

1216: $L_{\mathrm{FIR}}/M_\mathrm{c}

1217: \sim0.2$--$130$\,L$_\odot$\,M$_\odot^{-1}$ for the objects with $T$ in

1218: the range 12 -- 40\,K (Figure~\ref{fig:temphist}).

1219: 

1220: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1221: \section{DISTANCES}

1222: \label{sec:dist}

1223: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1224: 

1225: To deduce luminosities and masses one must have an estimate of the

1226: distance to each source. There is, however, little information in the

1227: literature regarding distances to known objects in this field.

1228: Cross-correlation with the {\it IRAS} PSC identifies 23 of the 60

1229: BLAST sources with known objects (Table~\ref{tab:iras}).  While some

1230: of these sources have been studied extensively \citep{beu02,bel06},

1231: few have unambiguous distance information.

1232: %

1233: Important clues are provided by the morphology and velocity of the

1234: interstellar material and their inter-relationship.

1235: %

1236: We have carried out a multi-wavelength assessment comparing the BLAST

1237: images and point sources with {\it IRAS} \citep{cao1997,kerton2000};

1238: {\it MSX}; IRAC (\S\ref{sec:mips70}); the STScI Digitized Sky

1239: Survey\footnote{\url{http://archive.stsci.edu/dss/}} (DSS); 21\,cm

1240: radio continuum imaging from the VLA Galactic Plane Survey

1241: \citep[VGPS,][]{stil2006}; and spectral line imaging in

1242: $^{13}$CO\,(1$\rightarrow$0) \citep[FCRAO:][]{brunt2007} and

1243: \ion{H}{1} (VGPS).

1244: 

1245: We argue below that 49 of the 60 BLAST sources in this field are

1246: associated with the molecular cloud complex within which the open

1247: cluster NGC~6823 has already formed (Figure~\ref{fig:blastmaps}).  We

1248: are then able to adopt the photometric distance measured for the

1249: stars.  We use 2.3\,kpc \citep{massey1995}, though we note that

1250: distances in the range 1.5 to 3.2\,kpc have been reported

1251: \citep{hoyle2003,Guetter92,pena2003}.  

1252: %%

1253: In a program measuring parallaxes using methanol masers

1254: \citep{menten2007} a distance of $2.20 \pm 0.01$\,kpc is found for the

1255: masers associated with V30 \citep[IRAS 19410+2336; see][]{szym2000},

1256: about $0\fdg5$ from the cluster center.

1257: %%

1258: NGC~6823 is placed in a Galactic context in Figures

1259: \ref{fig:spiralarms} and \ref{fig:kinematic}.  In neither position nor

1260: velocity \citep[see below and][Figure~2]{lockman1989} is it part of

1261: the main Sagittarius arm.  In addition ten of the sources appear to be

1262: in the more distant Perseus arm ($\sim8.5$\,kpc), and a single object

1263: (V07) is thought to lie beyond that in the outer galaxy

1264: ($\sim14$\,kpc). The sources that are not associated with the NGC~6823

1265: molecular complex are indicated thus in Table~\ref{tab:src}.

1266: 

1267: %\clearpage

1268: \begin{figure}%[t] 

1269: \centering

1270: \includegraphics[width=\linewidth]{f11.eps}

1271: \caption{Galactic spiral arm model of \citet{Taylor93}.  The dashed

1272:   line indicates the line of sight towards 60$^{\circ}$ longitude.

1273:   The star indicates the location of the open cluster NGC~6823 at the

1274:   adopted distance of 2.3\,kpc \citep{massey1995}.  We associate 49 of

1275:   the 60 BLAST sources with the molecular complex within which

1276:   NGC~6823 has already formed.  Another 10 ``Perseus arm sources'' are

1277:   approximately 8.5\,kpc distant (triangle) and one object (V07)

1278:   appears to be in the outer Galaxy (plus).  The square indicates the

1279:   tangent point. }

1280: \label{fig:spiralarms}

1281: \end{figure}

1282: %\clearpage

1283: 

1284: The {\it IRAS} images (see Figure~\ref{fig:coverage}) show that the

1285: bright sources and diffuse emission represent a distinctive

1286: enhancement of size $\sim 2^\circ$, which as we discuss below is also

1287: the scale of the coherent molecular complex.  However, within the

1288: complex there is sub-structure, both in space and velocity.

1289: %

1290: The angular diameter of the NGC~6823 cluster is $\sim 0\fdg5$

1291: \citep{Kharchenko05}, and this is a part of the overall complex.  The

1292: diffuse radio and optical emission from the \ion{H}{2} region Sh2-86

1293: (LBN~135), the portion of the complex that is being ionized, is also

1294: of this size.  Within this region there are ionization fronts, clearly

1295: associated with the ionizing stars in NGC~6823, sculpting the parent

1296: molecular material.

1297: %

1298: A particularly striking example that can be seen in the red DSS image

1299: is a silhouetted ``pillar'' pointing at the stars, with an ionization

1300: front at its end and V39 immediately behind it. V43 and V49 are other

1301: similar examples.

1302: 

1303: At 2.3\,kpc, $\sim 2^\circ$ corresponds to a physical size of 80\,pc,

1304: so that all of the parts of the complex are at essentially the same

1305: distance to within a few percent (there could be a systematic error in

1306: the average distance, as noted), with variations indicated by slightly

1307: different velocities or a silhouette indicating relative position

1308: along the line of sight.

1309: 

1310: To introduce the velocity scale, the radio recombination line velocity

1311: of Sh2-86 measured by \citet{lockman1989} in a 9\arcmin\ beam is $29.4

1312: \pm 2.1$\,\kms\ with FWHM $24.2 \pm 2.9$\,\kms.  This velocity width,

1313: greater than a typical thermal width, and typical of widths found in

1314: that survey, could reflect accelerated gas motions in the ionized

1315: material and/or be intrinsic to the original neutral components

1316: comprising this star forming complex. It is certainly unrelated to

1317: differential Galactic rotation.  For a distance of 2\,kpc, the

1318: gradient in the rotation curve (Figure~\ref{fig:kinematic}) is $\sim

1319: 125$\,pc\,km$^{-1}$\,s, and so the velocity spread would in that

1320: interpretation imply a cloud of extreme elongation along the line of

1321: sight.  Note that this velocity and distance combination does not fall

1322: on the mean rotation curve.  There are substantial systematic motions

1323: in this direction, including considerable emission of both $^{13}$CO

1324: and \ion{H}{1} beyond the nominal tangent-point velocity ($\sim

1325: 34$\,\kms) and so kinematic distances from the rotation curve in this

1326: range are unreliable.

1327: 

1328: %\clearpage

1329: \begin{figure}%[t]

1330: \begin{center}

1331: \includegraphics[width=\linewidth]{f12.eps}

1332: \end{center}

1333: \caption{The solid curve shows the expected kinematic

1334:   distance-velocity relationship derived from the Galactic rotation

1335:   curve of \citet{bra93} projected along $\ell = 60^{\circ}$.  At the

1336:   adopted distance of 2.3\,kpc for NGC~6823 we show the average radio

1337:   recombination line velocity of Sh2-86 as a star.  The dark shaded

1338:   region indicates the spread of $^{13}$CO velocities measured for the

1339:   BLAST sources associated with the main cloud complex.  The uppermost

1340:   dotted line indicates the velocity of V06, which is higher than the

1341:   main spread but still consistent with systematic motions around the

1342:   mean. The light shaded region indicates the ``Perseus arm sources''

1343:   and the lowest line the source V07 in the outer Galaxy.  }

1344: \label{fig:kinematic}

1345: \end{figure}

1346: %\clearpage

1347: 

1348: 

1349: %****************************************

1350: %\subsection{$^{13}$CO Velocities and Morphology}

1351: \subsection{Velocities}

1352: \label{subsec:velocity}

1353: %****************************************

1354: 

1355: Using the accurate BLAST source positions, we have examined

1356: $^{13}$CO\,(1$\rightarrow$0) spectra of Vulpecula, having 46\arcsec\

1357: spatial and 1\,\kms\ velocity resolution (binned from a native

1358: resolution of 0.13\,\kms), respectively. Generally a single compact

1359: spatial/spectral coincident feature is identified (see the example of

1360: source V12 in Figure~\ref{fig:COfig}).  We therefore fit Gaussian

1361: profiles to the velocity component at the position of each source.

1362: The lines at BLAST positions have a typical FWHM of 2.5\,\kms.  Ten of

1363: the BLAST/{\it IRAS} counterparts have previously-measured velocities

1364: in other spectral lines, specifically CS \citep{beu02,bro96} and

1365: NH$_3$ \citep{mol96,Sridharan02,zinchenko1997}.  Our measured

1366: velocities for these objects are on average within 1\,\kms\ of the

1367: published velocities, the maximum deviation being 2\,\kms.  The

1368: velocities are listed in Table~\ref{tab:src} and are presented in

1369: Figure~\ref{fig:velhist}.  The broad peak from 21 to 36\,\kms\ is

1370: similar to that seen in the average spectrum of this region, which

1371: also includes more diffuse gas.  The $^{13}$CO velocity of V06, at

1372: 43\,\kms, still falls well within this range.

1373: 

1374: %\clearpage

1375: \begin{figure}

1376: \centering

1377: \includegraphics[width=3in]{f13a.eps}

1378: \includegraphics[width=3in]{f13b.eps}

1379: \caption{({\it a}) Channel image of $^{13}$CO at a velocity of

1380:   29\,\kms\ for the region shown in Fig.~6.  ({\it b}) Plot of

1381:   $^{13}$CO spectrum averaged over 9 pixels centered on V12 (solid

1382:   line) and average over whole region in panel a (dashed).  The bright

1383:   clump is compact (has a high contrast) both spatially and in

1384:   velocity, making it reasonable to identify it with the BLAST source

1385:   (Table~2).  }

1386: \label{fig:COfig}

1387: \end{figure}

1388: %\clearpage

1389: 

1390: \begin{figure}%[t] 

1391: \begin{center}

1392: \includegraphics[width=\linewidth]{f14.eps}

1393: \end{center}

1394: \caption{Histogram of the $^{13}$CO radial velocities associated with

1395:   BLAST sources.  The majority of these velocities are very similar to

1396:   the radio recombination line velocity of the \ion{H}{2} region

1397:   Sh2-86 measured by \citet{lockman1989}, shown as the Gaussian (with

1398:   arbitrary normalization).  The dashed line indicates the average

1399:   spectrum of the entire $^{13}$CO map, including more diffuse gas.

1400:   The peak at $10$\,\kms\ is the fairly structureless local gas, and

1401:   that at $-5$\,\kms\ is the Perseus arm in which we identify 10

1402:   sources.

1403: %

1404: }

1405: \label{fig:velhist}

1406: \end{figure}

1407: %\clearpage

1408: 

1409: Several BLAST sources are in molecular clumps that clearly affect the

1410: radio emission from the extended \ion{H}{2} region Sh2-86 seen in

1411: Figure~\ref{fig:radio}. For example, V32 is a clear local minimum

1412: surrounded by an ionization front. The pair V22 and V26 are also at a

1413: minimum.  Other objects apparently shaping or influencing the

1414: steepness of the radio contours are V47, V49, and V52.  All of these

1415: sources have CO clump velocities in the range 26--31\,km\,s$^{-1}$,

1416: like the radio recombination line velocity.

1417: 

1418: %\clearpage

1419: \begin{figure} 

1420: \centering

1421: \includegraphics[width=\linewidth]{f15.eps}

1422: \caption{The grayscale shows the VGPS 21-cm radio continuum emission,

1423:   which dominated by a structured \ion{H}{2} region Sh2-86 and a shell

1424:   supernova remnant \citep[diameter $\sim15'$,][]{taylor1992}.

1425:   Overplotted are the BLAST 350\,\micron\ contours, which reveal the

1426:   interfaces between thermal dust emission (neutral material) and the

1427:   ionized gas.  Circles (sources associated with NGC~6823) and

1428:   triangles (Perseus arm sources) identify the compact BLAST sources.

1429:   The squares indicate the most massive stars in the open cluster

1430:   NGC~6823. Labeled BLAST sources (V22, V26, V32, V47, V49 and V52)

1431:   lie in molecular clumps that clearly affect the radio emission from

1432:   Sh2-86. }

1433: \label{fig:radio}

1434: \end{figure}

1435: %\clearpage

1436: 

1437: V23 (with V19 close by) is coincident with a bright compact \ion{H}{2}

1438: region and has the weakest CO signature of all the BLAST sources.

1439: Additional velocity information for V23 was obtained using \ion{H}{1}

1440: absorption against the radio continuum source using VGPS data.

1441: Velocity components are identified at $-5$, 7.5, 21 (very strong), 27,

1442: and 31\,\kms.  The radio recombination line velocity in a 3\arcmin\

1443: beam \citep{lockman1989} is $-2.8 \pm 1.8$\,\kms\ with FWHM $15.2 \pm

1444: 2.6$\,\kms (the unusually low line width suggests an electron

1445: temperature less than 5000\,K).  The spatial position of V23 with

1446: respect to the main CO complex and other {\it IRAS} and BLAST sources

1447: suggest it is at the same distance, but the velocities at $-5$ and

1448: $-2.8$\,\kms\ indicate a much larger distance $\sim9$\,kpc just beyond

1449: the solar circle in the Perseus arm.

1450: %

1451: Nearby, V20 and V24 have similarly low velocities, appearing along a

1452: clumpy arc of $^{13}$CO emission at $-3$ to +2\,\kms.  V17 and V21,

1453: which have $^{13}$CO components near $\sim0$\,\kms\ that are

1454: comparable in strength to weak components in the nominal NGC~6823

1455: molecular cloud velocity range, are probably in this distant cloud

1456: too.  We shall call these the ``Perseus arm sources.''

1457: 

1458: The only other BLAST source for which we can measure the \ion{H}{1}

1459: absorption directly is V02.  It has strong sharp absorption at 31 and

1460: 35\,\kms\ and weaker components at 10\,\kms, and possibly $-6$ and

1461: +43\,\kms as well.  The $^{13}$CO velocity of this clump is 32\,\kms,

1462: but there is considerable gas near $\sim$40\,\kms\ in the surrounding

1463: cloud on this side of the map.

1464: 

1465: There is a weak radio source that is coincident with V07 and a

1466: molecular clump at $-54$\,\kms. This source is too faint in the radio

1467: for any possibility of measuring \ion{H}{1} absorption.  We use the

1468: rotation curve (Figure~\ref{fig:kinematic}) to find a kinematic

1469: distance of 14\,kpc, making this an intrinsically luminous star

1470: forming region in the outer Galaxy.

1471: 

1472: On the low side of the velocity distribution, there are a few other

1473: sources with $^{13}$CO components near $\sim0$\,\kms\ that are

1474: comparable in strength to weak components in the nominal NGC~6823

1475: molecular cloud velocity range, namely V51, V54, V58, and V59.

1476: %

1477: Using C$^{18}$O spectra \citep{brunt2007}, which probe even denser

1478: gas, the low velocity component is chosen for V51 and V58 (these join

1479: the Perseus arm sources).

1480: %

1481: V54 and V59 are seen in projection near the lower outskirts of the

1482: NGC~6823 cloud, as depicted in the zeroth-moment image

1483: (Figure~\ref{fig:co0blast}). Since there is no compelling

1484: morphological evidence to relate them to the higher-velocity gas, we

1485: also consider them to be Perseus arm sources.

1486: 

1487: 

1488: Finally, we note a puzzling result.  Using the \ion{H}{1} absorption

1489: technique to break the distance ambiguity, and Arecibo observations

1490: (with a 4\arcmin\ beam) at the Sh2-86 position \citep{lockman1989},

1491: \citet{kuchar1994} find absorption in the spectrum out to 50\,\kms and

1492: so adopt the far distance solution 6.3~kpc.  Certainly, there is no

1493: question about the velocity of the \ion{H}{2} region and hence the

1494: associated molecular cloud and BLAST sources.  But as we have

1495: mentioned above, the ionization fronts seen on the DSS protruding into

1496: the molecular cloud are clearly caused by the bright OB stars in

1497: NCG~6823 identified by \citet{massey1995}, and these have a

1498: photometric distance much closer to the near solution.  We have looked

1499: at the VGPS \ion{H}{1} data (also with continuum, and 1\arcmin\

1500: spatial resolution) across the face of the radio nebula.  We find that

1501: there are large changes in the emission and absorption spectrum, with

1502: some absorption to 50\, \kms (in line wings) even at the fainter

1503: western edge, where the continuum emission blends into the (non-zero)

1504: background.  This suggests that there is variable distant background

1505: continuum emission that is making it difficult to apply the technique

1506: on this extended source.

1507: 

1508: 

1509: %****************************************

1510: \subsection{Morphology}

1511: \label{subsec:morphology}

1512: %****************************************

1513: 

1514: A map of the $^{13}$CO emission integrated over the velocity interval

1515: 21--36\,\kms\ is shown in Figure~\ref{fig:co0blast}.  

1516: %

1517: Figure~\ref{fig:iracblast} shows the IRAC 8\,\micron\ image for

1518: comparison.  These both trace neutral material, but the latter,

1519: heavily influenced by PAH emission, requires illumination by

1520: sub-ionizing ultraviolet radiation (rather than simply tracing column

1521: density).  Also, 8\,\micron\ is a short enough wavelength for there to

1522: be extinction by large column densities, producing the infra-red dark

1523: cloud (IRDC) phenomenon \citep{ega98,simon2006}.

1524: %

1525: A portion of the molecular cloud associated with V03 and V04 produces

1526: a local minimum in the diffuse 8\,\micron\ emission, suggesting it is

1527: in the foreground.  On a smaller scale, V13 and V56 appear to exhibit

1528: similar behavior.

1529: 

1530: %\clearpage

1531: \begin{figure} 

1532: \centering

1533: \includegraphics[width=\linewidth]{f16.eps}

1534: \caption{The grayscale shows the $^{13}$CO emission integrated over 21

1535:   to 36\,\kms.  Overplotted are the BLAST 350\,\micron\ contours which

1536:   demonstrate a close correspondence between thermal dust emission and

1537:   molecular gas in this velocity range.  Circles (sources associated

1538:   with NGC~6823) and triangles (Perseus arm sources) identify the

1539:   compact BLAST sources.  The squares indicate the most massive stars

1540:   in the open cluster NGC~6823.  Also indicated is the position of the

1541:   supernova remnant, and the extended \ion{H}{2} region Sh2-86.  }

1542: \label{fig:co0blast}

1543: \end{figure}

1544: 

1545: \begin{figure}

1546: \centering

1547: \includegraphics[width=\linewidth]{f17.eps}

1548: \caption{The grayscale shows IRAC 8\,\micron\ emission, with the same

1549:   annotations as Figure~\ref{fig:co0blast}.  We note that the labeled

1550:   BLAST sources (V03, V04, V13, and V56) occur at minima in the

1551:   diffuse 8\,\micron\ emission, possibly indicating that the

1552:   submillimeter objects are (associated with) infra-red dark clouds

1553:   somewhat in the foreground. A prominent filament which appears as 

1554:   a dust lane in the red DSS image is also indicated (probably related to

1555:   the BLAST sources V13, V15, V16, V38, and V41). Optical absorption is

1556:   also associated with V47. These features are 

1557:   discussed in \S\ref{subsec:morphology}.

1558:   \label{fig:iracblast}}

1559: \end{figure}

1560: %\clearpage

1561: 

1562: Overlaid on the figures are the BLAST 350\,\micron\ contours, which

1563: reveal excellent correspondence between this measure of the neutral

1564: column density and the other tracers.  Also indicated are the BLAST

1565: point sources and the ionizing stars in NGC~6823 \citep{massey1995}.

1566: 

1567: A particularly striking filament, which is a prominent dust lane in

1568: the red DSS image, is seen as a ridge in the BLAST, $^{13}$CO, and

1569: IRAC maps (indicated in Figure~\ref{fig:iracblast}), running from V30

1570: to the south west (parallel to the Galactic Plane) toward V03.  BLAST

1571: sources V13, V15, V38, V41, and possibly V16 appear to be related to

1572: this optically-dark filament.  Other optical absorption is associated

1573: with BLAST source regions, such as V47 projected on the \ion{H}{2}

1574: region.

1575: %

1576: 

1577: An interesting phenomenon seen in the \ion{H}{1} data is

1578: self-absorption (HISA), caused by colder foreground material; it is

1579: most convincingly found using the criteria of narrow lines and

1580: fine-scale spatial structure \citep{Gibson2000} but can also be more

1581: widespread.  HISA appears at 24\,\kms\ in this filamentary structure.

1582: Near 33\,\kms\ there is striking filamentary HISA which extends into

1583: the V22--V26--V28 portion of the molecular complex, where the lack of

1584: \ion{H}{1} emission no doubt also reflects a real deficit. At

1585: 36\,\kms\ there is more HISA which traces the diffuse submillimeter

1586: emission along the ridge ending at V06.  This ridge is evident in

1587: $^{13}$CO at these velocities, but is not so clearly seen with IRAC.

1588: 

1589: Examination of the sequence of channel maps in the $^{13}$CO data

1590: cube, or the first moment (average velocity) map reveals a velocity

1591: gradient from the north-east to the south-west (higher to lower

1592: Galactic longitudes).  Along the long filament the velocity changes

1593: quite systematically, indicating a large scale connectivity between

1594: the different portions of the molecular complex.

1595: 

1596: Also revealed in these data are large areas of low emission.

1597: To the west of the stellar cluster is a distinctive void, cleared of

1598: molecular gas and dust.

1599: %

1600: As expected, because the gas has become molecular, there is a general

1601: deficit of \ion{H}{1} across this entire region. This can be seen

1602: extending to adjacent longitudes in the longitude-velocity diagram of

1603: \citet{stil2006}.  The deficit is exaggerated adjacent to the

1604: \ion{H}{2} region (near 33\,\kms\ where the gas is molecular).  This

1605: can appear more localized too, e.g., near V03 and in the V08--V09

1606: cloud at the same velocity as the $^{13}$CO.

1607: 

1608: Another void coincides with a supernova remnant found by

1609: \citet{taylor1992} and is seen in finer detail in the VGPS radio

1610: continuum image (Figure~\ref{fig:radio}).  The void is seen in

1611: \ion{H}{1} and in $^{13}$CO as well near 22\,\kms.  The remnant is not

1612: noticeably interacting with the surrounding dense material (e.g., no

1613: steepening of the $^{13}$CO contours or brightening or deformation of

1614: the radio shell), and so perhaps the supernova occured in a

1615: pre-existing cavity, rather than creating it.  The diameter of the

1616: remnant is only 0\fdg2 (8\,pc) and so it could be at the average

1617: distance of this cloud complex (as the velocity suggests) but not yet

1618: have encountered one of its clumpy components.

1619: 

1620: %****************************************

1621: \section{DUST MASSES AND LUMINOSITIES}

1622: \label{sec:mass}

1623: %****************************************

1624: 

1625: In \S\ref{sec:dist} we determined distances for all of the objects: 49

1626: are associated with the open cluster NGC~6823 at $\sim2.3$\,kpc, 10

1627: with the Perseus arm at $\sim8.5$\,kpc, and 1 object is in the outer

1628: galaxy at $\sim14$\,kpc (summarized in Table~\ref{tab:src}). Here we

1629: use these distance estimates to convert the integrated fluxes from

1630: \S\ref{sec:coldsed} into bolometric luminosities, $L$, and to derive

1631: the mass, $M_\mathrm{c}$, for each source.

1632: 

1633: The clump mass $M_\mathrm{c}$ given in column 3 of Table~\ref{tab:sed}

1634: has been obtained from Equation~\ref{eq:mass}.  Note that although a

1635: mass absorption coefficient at 250\,\micron\ appears in this

1636: expression, $A$ is derived from the fit to the entire submillimeter

1637: and FIR data, rather than just the single color-corrected BLAST

1638: 250\,\micron\ data point alone.

1639: 

1640: The masses of the clumps discovered range over

1641: $\sim15$--700\,M$_\odot$ in NGC~6823, extending to somewhat higher

1642: masses in the Perseus arm. These sources are by definition compact in

1643: the submillimeter (and they are compact in CO as well), with

1644: characteristic sizes $D < 0.44$\,pc at 2.3\,kpc (taking the

1645: de-convolved 250\,\micron\ map beam size of 40\arcsec\ from

1646: Table~\ref{tab:mapsens} as a limit). If observed at higher resolution

1647: many of these ``clumps'' would probably qualify as ``cores'' in the

1648: terminology of \citet{zinn2007}.  The column density for a typical

1649: 100\,M$_\odot$ clump, assuming a uniform sphere limiting diameter $D$,

1650: is of order $10^{23}(0.44\,\mathrm{pc}/D)^2$\,cm$^{-2}$ (visual

1651: extinction to the cloud center $A_V \sim 22(0.44\mathrm{pc}/D)^2$ and

1652: density of order $10^5(0.44\,\mathrm{pc}/D)^3$\,cm$^{-3}$).  The

1653: Bonner-Ebert critical mass for a gravitationally-bound clump, above

1654: which collapse of an unstable configuration occurs, is given by

1655: $M_{\mathrm{crit}}/M_\odot \approx (D/\mathrm{pc})

1656: (T_{\mathrm{gas}}/\mathrm{K})$ \citep[see

1657: e.g.,][Equation~4]{kerton2001}.  In these dense clumps,

1658: $T_{\mathrm{gas}}$ is probably close to the dust temperature and so

1659: for the derived $T$ and $M_{\mathrm{c}}$ most of the clumps that we

1660: have identified would be unstable and are likely supported by

1661: additional pressure from turbulence or magnetic fields.  On release of

1662: this pressure, they have the potential to form a massive star or even

1663: a small cluster.  Interestingly, several of the lowest mass clumps

1664: would be just stable if they were as extended as 1\,pc.

1665: 

1666: The range of luminosities does indeed extend to $10^4$\,L$_\odot$ in

1667: NGC~6823, which corresponds to the luminosity of a B0.5 zero age main

1668: sequence star \citep{panagia1973}.  But as we shall discuss further,

1669: at the low end, $\sim40$\,L$_\odot$, there is by contrast no

1670: indication of significant star formation.

1671: 

1672: %\clearpage

1673: \begin{figure}

1674: \centering

1675: \includegraphics[width=\linewidth]{f18.eps}

1676: \caption{Luminosity vs.\ mass for the 49 objects believed to be

1677:   associated with NGC~6823.  Uncertainties in mass and luminosity are

1678:   estimated from the range of temperatures consistent with the BLAST

1679:   and FIR data under the assumption of the simple SED model of

1680:   Equation~\ref{eq:sed}.  Not included are additional sources of error

1681:   involving the distances to the objects and adopted values of

1682:   $\beta$, $\kappa_0$, and gas-to-dust mass ratio.  The dashed lines

1683:   are loci at fixed temperatures (and $L_{\mathrm{FIR}}/M_\mathrm{c}$)

1684:   of 10--40\,K, assuming a modified blackbody SED with $\beta=1.5$.

1685:   Orthogonal to these are loci (dotted lines) of constant 250\,$\mu$m

1686:   flux density, ranging from 10 to 1000\,Jy, using the same model.

1687: %

1688:   Sources lacking {\it IRAS} PSC or {\it MSX} counterparts are marked

1689:   with circles and squares respectively. Ultra Compact \ion{H}{2}

1690:   region and Hyper Compact \ion{H}{2} region candidates are indicated

1691:   with single and double triangles, respectively.}

1692: \label{fig:lumvsmass}

1693: \end{figure}

1694: 

1695: %\clearpage

1696: \begin{figure}[h]

1697: \centering

1698: \includegraphics[width=\linewidth]{f19.eps}

1699: \caption{Same annotations as Figure~\ref{fig:lumvsmass}, but for the

1700:   10 objects that are associated with the Perseus arm.  }

1701: \label{fig:lumvsmassPerseus}

1702: \end{figure}

1703: %\clearpage

1704: 

1705: Luminosity as a function of mass is shown in

1706: Figures~\ref{fig:lumvsmass} and \ref{fig:lumvsmassPerseus}. This is

1707: essentially a reparameterization of the BLAST flux density vs.\

1708: temperature plane, in terms of intrinsic source parameters.  Because

1709: of the different distances for the two main groups, the Perseus arm

1710: sources and those associated with NGC~6823, the scaling of the axes is

1711: different, requiring similar but separate diagrams.

1712: %

1713: Recall that the derived quantities are based on a SED fit assuming a

1714: modified blackbody SED with $\beta=1.5$.  Using this same model, the

1715: dotted lines are loci (with varying $T$) of constant 250\,$\mu$m flux

1716: densities ranging from 10 to 1000\,Jy.

1717: %

1718: Orthogonal to these, the dashed lines are loci (with varying observed

1719: flux density) at constant $T=10$, 20, 30, and 40\,K.  With our adopted

1720: model, these correspond to constant $L_{\mathrm{FIR}}/M_\mathrm{c}$,

1721: values of 0.07, 3.2, 30, and 140\,L$_\odot$\,M$_\odot^{-1}$,

1722: respectively.

1723: %

1724: The 11 objects with unreliable BLAST colors

1725: (\S\ref{sec:sourceid}, marked in Table~\ref{tab:src}) have been

1726: assigned a temperature of 20\,K and so are seen distinctly along that

1727: locus.  If these sources have different temperatures (probably lower

1728: given the usual lack of {\it IRAS} PSC counterparts), they would move

1729: along the constant flux density locus (down and right if cooler).

1730: %

1731: These Figures can be used to investigate which early stages of star

1732: formation might be present.  To this end, sources have been marked

1733: which lack {\it IRAS} PSC or {\it MSX} counterparts, or are UC

1734: \ion{H}{2} regions

1735: 

1736: %****************************************

1737: \subsection{Mass Spectrum} \label{sec:spec}

1738: %****************************************

1739: 

1740: The clumps that we have detected are much more massive than stars and

1741: therefore represent an early stage in star formation that is not

1742: easily, even empirically, related to the stellar initial mass

1743: function.  Fragmentation and efficiency affect the transfer function

1744: in unknown ways.  Nevertheless, it is interesting to investigate the

1745: mass spectrum, which we have done for the 49 BLAST sources associated

1746: with the NGC~6823 cloud complex.  The masses of individual sources

1747: were placed in logarithmically spaced bins (2 per decade); a lower

1748: limit on the error is then estimated from the Poisson uncertainty for

1749: each bin.  The resulting mass function is shown in

1750: Figure~\ref{fig:mfunc}.

1751: 

1752: %\clearpage

1753: \begin{figure}%[t]

1754: \includegraphics[width=\linewidth]{f20.eps}

1755: \caption{Mass function of the 49 objects associated with

1756:   NGC~6823. Poisson error bars are shown, although there are

1757:   additional sources of uncertainties for the parameters of

1758:   Equation~\ref{eq:mass} for each object.  For reference, the dashed

1759:   line is a power-law, $150(M/50\,\mathrm{M}_\odot)^{-1.7}$.

1760: %

1761:   The two lowest bins suggest a turnover, but this shape is probably

1762:   due to incompleteness.  To illustrate this the 250\,\micron\ flux

1763:   density completeness (Figure~\ref{fig:blastcomplsims}) has been

1764:   converted to a mass completeness using Equation~\ref{eq:mass}

1765:   assuming various temperatures $T=10$, 15, and 20\,K.  These are

1766:   shown as dotted lines, arbitrarily scaled such that 100\%

1767:   completeness is 40 objects per bin.}

1768: \label{fig:mfunc}

1769: \end{figure}

1770: %\clearpage

1771: 

1772: At masses $\ga100$\,M$_\odot$, the distribution is consistent with the

1773: mass functions of molecular clouds. For example, \citet[and references

1774: therein]{kra98} found a power-law index $\alpha \simeq 1.7$ in a mass

1775: range which encompasses the mass interval considered here.  For

1776: reference, we have shown their power-law scaled vertically such that

1777: it passes through the high-mass bins in Figure~\ref{fig:mfunc}: $dN/dM

1778: = 150(M/50\,\mathrm{M}_\odot)^{-1.7}$.  Note that adopting a different

1779: value of dust emissivity affects the inferred dust mass, since it is

1780: highly correlated with the fitted temperature (\S\ref{sec:sedfits}),

1781: and hence mass. For example, choosing $\beta=2.0$ decreases

1782: temperatures by $\sim5$\,K, and increases cloud masses by a factor

1783: $\sim2$. This effect shifts the mass function to the right in this

1784: figure, but does not affect the power-law index.

1785: 

1786: We note that the mass function in our sample appears to turn over at

1787: masses below $\sim100$\,M$_\odot$. However, the completeness of the

1788: survey also drops at these masses, with uncertainties that are

1789: difficult to quantify given the range of object temperatures sampled

1790: (Figure~\ref{fig:lumvsmass}).

1791: %

1792: To attempt to assess the mass completeness, we converted the $x$-axis

1793: of the flux density completeness function

1794: (Figure~\ref{fig:blastcomplsims}) into mass using

1795: Equation~\ref{eq:mass}, assuming consistently a distance of 2.3\,kpc

1796: and $\kappa_0 = 10$\,cm$^2$g$^{-1}$.  Some results are shown in

1797: Figure~\ref{fig:mfunc}.  If we assumed that the sources potentially

1798: missed had a temperature of $20$\,K, then completeness has a

1799: negligible effect, but for lower temperatures, which seem more

1800: probable, the effect is significant, because lower temperatures

1801: correspond to higher masses for a given flux density.  See

1802: \citet{johnstone2000} and \citet{enoch2006} for further discussions

1803: regarding the completeness of mass spectrum estimates from

1804: submillimeter surveys.

1805: %

1806: If, however, the turnover were real at $\sim100$\,M$_\odot$, our

1807: survey may be probing an interesting effect discussed by other authors

1808: \citep[see][and references therein]{bel06}.  They claim that the point

1809: at which the mass distribution flattens is proportional to the mass

1810: range of clumps in star forming clouds.

1811: 

1812: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1813: \section{DISCUSSION}  \label{sec:disc}

1814: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1815: 

1816: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1817: \subsection{Stages of Pre-stellar Evolution}  

1818: \label{sec:prestellar}

1819: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1820: 

1821: This particular survey was not unbiased, since it targeted a region

1822: which has already formed massive stars and which contains bright {\it

1823:   IRAS} sources which are protostar candidates.  But it does fulfill

1824: one of the main goals of BLAST, to explore the sites of formation of

1825: high-mass stars and to examine the pre-stellar evolutionary sequence.

1826: In particular, with the BLAST wavebands we are sensitive to the

1827: earliest phases of massive star evolution, which would be molecular

1828: clumps lacking an internal energy source and thus cool with low

1829: $L_{\mathrm{FIR}}/M_\mathrm{c}$.  In Figure~\ref{fig:lumvsmass} there

1830: are indeed many such low luminosity-to-mass ratio sources that are too

1831: faint and cool to have {\it IRAS} and {\it MSX} counterparts.

1832: 

1833: An interesting perspective is that the luminosity-to-mass ratio for

1834: diffuse cirrus in the local interstellar radiation field is of order

1835: 1\,L$_\odot$\,M$_\odot^{-1}$.  The low ratio (and lower temperature)

1836: found in these molecular clumps, even in the presence of a stronger

1837: impinging radiation field, is a result of the tremendous opacity to

1838: this external radiation, uncompensated by an internal source.  In

1839: absolute terms the luminosity of a cool clump can be quite impressive,

1840: because of the large mass, and yet not be a reliable indicator of

1841: there being star formation within.

1842: 

1843: Once a high-mass pre- or proto-stellar object begins to form, there is

1844: considerable luminosity to be reprocessed, resulting in higher

1845: $L_{\mathrm{FIR}}/M_\mathrm{c}$ and warmer dust.  BLAST can of course

1846: see such objects as well, even those without MIR associations.  There

1847: is not yet a unique definition or set of physical properties

1848: associated with HMPOs, but there is general agreement that they are

1849: heavily obscured objects (even making detection difficult in the MIR)

1850: that have not yet reached the stage at which the emergent Lyman

1851: continuum is sufficient to ionize the surrounding medium to form an

1852: \ion{H}{2} region, and so they are undetected in the radio continuum

1853: as well.  With spectral energy distributions peaking at wavelengths

1854: $\ga 200$\,\micron\ they can be extremely weak and undetectable by

1855: {\it IRAS} at 25\,\micron, thus excluding them from previous surveys

1856: \citep[e.g.,][]{Sridharan02}.  Therefore the most unique BLAST

1857: objects, which are good HMPO candidates, are those with no {\it IRAS}

1858: or {\it MSX} counterparts, and no radio emission.

1859: 

1860: In the subsequent phase, when the massive protostar has become hotter,

1861: near the zero age main sequence, the ionizing capability is enhanced

1862: and a small region in the dense surroundings is ionized.

1863: Historically, such ``ultra compact'' (UC) \ion{H}{2} regions have been

1864: the primary means to investigate the early phases of OB stellar

1865: evolution.  This is a productive approach, since it deals with

1866: intrinsically luminous objects which are readily detectable, and the

1867: radio emission provides unequivocal evidence for an ionizing star.

1868: The more evolved, and therefore warmer, sources are more likely to

1869: have a MIR counterpart and be detectable by {\it MSX} and/or IRAC,

1870: than the colder and/or more deeply embedded sources.  We do see such

1871: objects, as identified in Figures~\ref{fig:lumvsmass} and

1872: \ref{fig:lumvsmassPerseus} (see \S\ref{sec:uch2}).

1873: 

1874: In the past decade it has become clear that ``hot-cores'' represent a

1875: stage in the massive star formation process, probably somewhat earlier

1876: than UC \ion{H}{2} regions \citep[see, e.g.,][and references

1877: therein]{kurtz2000}.  The precise details of the transition are still

1878: lacking, and our understanding is complicated by the co-existence of

1879: objects at different evolutionary phases within the same cluster.

1880: Also, the distinction between hot-cores and UC \ion{H}{2} regions is

1881: often dependent on the sensitivity of the available observations, as

1882: several hot-cores are found to emit radio continuum emission if

1883: observed with enough sensitivity and angular resolution (opacity can

1884: also be a factor for dense objects).  Likewise, although some authors

1885: list hot-cores as a subset of HMPOs \citep[e.g.,][]{Sridharan02},

1886: hot-cores are generally warmer, with $T_{\rm k} \ga 100$\,K, whereas

1887: objects called HMPOs usually have temperatures much lower than 100\,K

1888: \citep[e.g.,][]{mol96}.  Also as noted by \cite{Sridharan02}, among

1889: higher luminosity objects, a low luminosity-to-mass ratio relative to

1890: that for UC \ion{H}{2} regions might indicate a HMPO, though

1891: \citet{bel06} found no evidence to support this difference between

1892: their {\it ``Low''} and {\it ``High''} (less and more evolved)

1893: sources.

1894: 

1895: None of the objects that we detect has an inferred temperature greater

1896: than 40\,K (assuming $\beta=1.5$), even those that are UC \ion{H}{2}

1897: regions.  The temperature that we derive from the SED characterizes

1898: the bulk of the dust emission (more than 50~\%) but of course is a

1899: simplification.  When a protostar forms, the dust nearby is hotter

1900: than in the parent molecular clump, and also appears as a more compact

1901: source.  Thus what is measured and interpreted as mass and luminosity

1902: depends on the radiative transfer and temperature gradient as well as

1903: the angular resolution of the measurements. While $L$ reprocessed by

1904: dust (and the amount of ionization gauged from the radio continuum

1905: emission) are calorimeters for the embedded stellar population that is

1906: forming, the mass is that of the entire parent clump, depleted by

1907: accretion onto stars and later cloud dispersal.  The

1908: luminosity-to-mass ratio can therefore be different, even for the same

1909: stellar population, being maximized for optical depth of order unity

1910: in a compact (and therefore warm) circumstellar region.

1911: 

1912: A plot like Figure~\ref{fig:lumvsmass} could be used to examine the

1913: evolutionary sequence, whether at constant $M_\mathrm{c}$ or

1914: otherwise, determining the lifetimes for each stage from the relative

1915: numbers of objects.  However, this is complicated in practice by

1916: effects like clustering and detection bias and the fact that star

1917: formation does not proceed at precisely the same rate everywhere in a

1918: molecular clump.  Thus without adequate resolution it would be

1919: difficult to discriminate between material associated with an UC

1920: \ion{H}{2} region and another nearby {\it cold} HMPO, for example.

1921: This problem can be remedied by high-resolution and high dynamic range

1922: observations (as anticipated with {\it Herschel} and ALMA); indeed in

1923: some cases interferometric observations have led to the detection of

1924: HMPOs in the same fields as the UC \ion{H}{2} regions \citep[see,

1925: e.g.,][ and references therein]{olmi2003}.

1926: %

1927: Likewise, objects in Figure~\ref{fig:lumvsmass} with candidate {\it

1928: IRAS} and/or {\it MSX} counterparts, might in fact harbor a HMPO that

1929: is associated, but not coincident, with a MIR source at a later

1930: evolutionary stage in the same cluster.

1931: 

1932: % ****************************************

1933: \subsection{Ultra Compact \ion{H}{2} Regions}

1934: \label{sec:uch2}

1935: %****************************************

1936: 

1937: We have detected several {\it bona fide} UC \ion{H}{2} regions, by

1938: which we mean compact objects whose bolometric luminosities {\it and}

1939: radio continuum emission are consistent with an embedded massive star

1940: near the zero age main sequence.  Although within our beam there might

1941: be stars forming over a range of masses, the luminosity (and even more

1942: so the ionizing radiation), is dominated by the most massive and

1943: hottest stars with the earliest spectral types.  Therefore, while not

1944: precise, it is still useful to characterize the region using a single

1945: spectral type.  We use luminosities and ionizing rates from

1946: \citet{panagia1973}, which to the accuracy required here are close to

1947: more modern values \citep{schaerer1997}.

1948: 

1949: These sources can be been identified as UC \ion{H}{2} {\it candidates}

1950: using color criteria based on {\it IRAS} measurements as in the

1951: \citet{wood1989}, \citet{kurtz1994}, \citet{bro96} and

1952: \citet{Watson03} catalogs (qualitatively, these sources are relatively

1953: weak at 12\,\micron). Using the BLAST data we are able to find very

1954: accurate temperatures.

1955: %

1956: Radio continuum measurements in the survey region at 1.4\,GHz are

1957: available from the NRAO VLA Sky Survey

1958: \citep[NVSS\footnote{http://www.cv.nrao.edu/nvss/},][]{condon1998} and

1959: the VGPS \citep{stil2006}. In addition, we searched the second and

1960: third MIT-Green Bank 5\,GHz ``MG'' surveys \citep{langston1990,

1961:   griffith1990} which, though much shallower than NVSS, can provide a

1962: second flux density to estimate the spectral index between 1.4 and

1963: 5\,GHz.  Some additional targeted observations at 4.9\,GHz using

1964: Arecibo are described in \citet{Watson03}.

1965: 

1966: We will discuss these sources briefly in order of decreasing distance.

1967: V07 is the source in the outer Galaxy, which though fairly luminous

1968: ($3 \times 10^4$\,L$_\odot$), is not among the more conspicuous

1969: sources in the maps. It is consistent with excitation (both luminosity

1970: and ionization) by a B0 star (only 20\,M$_\odot$ as compared to

1971: 600\,M$_\odot$ for the clump). In the VGPS radio image this source is

1972: extended and in the NVSS it is double.  The second component, to lower

1973: longitude by about 1\arcmin, coincides with a second {\it IRAS} source which

1974: appears at 12 and 25\,\micron.

1975: 

1976: V23, in the Perseus arm, is the most luminous source in this BLAST

1977: survey ($7\times10^4$\,L$_\odot$; Figure~\ref{fig:lumvsmassPerseus}),

1978: and a prominent radio, submillimeter, and infra-red source, despite

1979: its distance.  Its excitation is consistent with a star as early as

1980: O7--8 (30\,M$_\odot$, as compared to 500\,M$_\odot$ for the clump).

1981: The emission is extended, overlapping a second BLAST source V19, and

1982: is no doubt more complex than this simple description.

1983: 

1984: V02 and V05 are among the sources to which we have assigned the

1985: NGC~6823 distance (Figure~\ref{fig:lumvsmass}).  The luminosity

1986: implies a B2 zero age main sequence star, but the ionization (50\,mJy

1987: radio continuum) requires B0.5 (for these relatively cool stars the

1988: ionization is extremely sensitive to the temperature).  A further note

1989: is that the radio source, while being aligned with the {\it IRAS}

1990: source, is offset by 30\,\arcsec\ from V02.

1991: 

1992: V30, the brightest BLAST source and the most luminous source at this

1993: distance, is by contrast not an NVSS/VGPS source at 1.4\,GHz. However,

1994: using the VLA B array at 8.2\,GHz, \citet{Sridharan02} found weak

1995: emission at the level of 1\,mJy. Lower resolution Arecibo measurements

1996: at 4.9\,GHz \citep{Watson03} give 3\,mJy. Relative to V02 and V05, V30

1997: is under-luminous in the radio continuum by an order of magnitude.

1998: This could be because the central object, though luminous, has not

1999: become hot enough on approach to the zero age main sequence (it is a

2000: HMPO), or because of opacity if the surroundings of the protostar are

2001: very dense.  We note that about 10\% of the sources observed by

2002: \citet{wood1989} are optically thick at 2\,cm, and a higher percentage

2003: may be applicable for the BLAST sources, since they probe the earliest

2004: stages of massive star formation. V30 might be what is called a Hyper

2005: Compact (HC) \ion{H}{2} region \citep[e.g.][]{keto2007}, and has been

2006: marked accordingly in Figure~\ref{fig:lumvsmass}.

2007: 

2008: The situation for the other comparably-luminous sources V08, V18, V32,

2009: and V52 is less clear.  All but V18 meet the color criteria for being

2010: an UC \ion{H}{2} region. However, none are NVSS sources.  The

2011: detection of the V18 and V30 in targeted Arecibo observations

2012: \citep{Watson03} appears to be contradicted by the lower upper limits

2013: of 1\,mJy with the VLA B array \citep{Sridharan02}.  This might

2014: suggest contamination of the Arecibo beam by diffuse continuum

2015: emission; as noted in \S\ref{subsec:velocity}, V32 is projected

2016: against the bright patchy Sh2-86 emission.  On the other hand, V18 is

2017: coincident with a small (20\arcsec) cometary optical nebula and the

2018: Arecibo flux of 8\,mJy might be the more relevant if the VLA resolved

2019: the extended emission. V18 may be another example of a HC \ion{H}{2}

2020: region, and it is marked as a candidate in Figure~\ref{fig:lumvsmass}.

2021: For V08 and V52 there are no sufficiently sensitive observations to

2022: isolate any weak emission, or alternatively, to qualify them as HMPOs.

2023: 

2024: Most of the radio sources in this survey region are not in fact

2025: associated with BLAST and {\it IRAS} sources (and vice versa).  One

2026: must therefore be cautious of chance coincidences \citep[for reference

2027: the surface density at the $\sim$1\,mJy level is $<

2028: 0.007$\,arcmin$^{-1}$,][]{fomalont1991}. There is a VGPS/NVSS source

2029: offset by 75\arcsec\ from V40 and its {\it IRAS} counterpart, also

2030: detected faintly at Arecibo \citep{Watson03}.  The level of radio

2031: emission is at least an order of magnitude greater than would be

2032: expected for the relatively low luminosity B2 star

2033: \citep{panagia1973}, so that it is probably not associated at all.

2034: Using the VLA B array at 8.2\,GHz, \citet{Sridharan02} placed a more

2035: definitive upper limit of 1\,mJy.  Note that such a HMPO will never

2036: become a significant \ion{H}{2} region unless the stellar mass is

2037: increased through further accretion.

2038: % 

2039: There are NVSS sources near a few other even lower luminosity BLAST

2040: objects, V35, V39, V45, and V49 (and a marginal detection near V34).

2041: They are not detected in the shallower MG surveys and so no radio

2042: spectral index could be estimated.

2043: %

2044: However, V35 has been detected in the Westerbork Synthesis Radio

2045: Telescope (WSRT) survey of \citet{taylor1996}. Its spectral index

2046: between 327\,MHz and 1.4\,GHz is $-1.6$, and is therefore very

2047: unlikely to be associated with thermal emission.  All of these sources

2048: are seen projected on Sh2-86, and so the radio ``sources'' could be

2049: just parts of the patchy ionization structure therein.  As mentioned

2050: in \S\ref{sec:dist}, V39 is embedded at the end of a pillar

2051: adjacent to an ionization front caused by the exciting stars of

2052: NGC~6823, not V39.  Inspection of the radio images suggests that V45

2053: and V49 might be similarly confused.

2054: 

2055: Clearly, more sensitive radio observations, with spatial and spectral

2056: index information, are desirable to further elucidate the true nature

2057: of all of these (luminous) sources.

2058: But a firm conclusion that should be emphasized is that none of these

2059: BLAST sources associated with NGC~6823 reveal a protostar or HMPO more

2060: massive than a B0 to B1 star or about 20\,M$_\odot$.  There is thus no

2061: evidence for a new (or lagging) generation of stars that are as

2062: massive as the existing powerhouses of the NGC~6823 cluster, which

2063: reach about 40\,M$_\odot$ \citep[O7~V and O9.5~Ia;][]{schaerer1997}.

2064: 

2065: %****************************************

2066: \subsection{Extended Submillimeter Emission}

2067: \label{sec:extended}

2068: %****************************************

2069: 

2070: In this paper we have focused on the compact sources in the BLAST05

2071: maps of Vulpecula. More diffuse structure could also be studied using

2072: algorithms such as CLUMPFIND \citep{williams1994}. We leave such

2073: quantitative analysis for future work. However, it is worth

2074: illustrating a few features in the context of the discussion from

2075: \S\ref{subsec:morphology} and the cool low luminosity-to-mass ratio

2076: clumps already identified.

2077: 

2078: Between sources V02 and V03 there is a pair of extended BLAST

2079: features, neither sufficiently compact to be included in our source

2080: catalog. 

2081: %

2082: The first, to the west of the line, is a strong 100\micron\ source,

2083: and even more prominent at 60\micron; it is {\it IRAS}~19364+2252.

2084: This region exhibits a striking fan-shaped nebulosity in the IRAC

2085: 8\,\micron\ map (Figure~\ref{fig:iracblast}).  Since the object is

2086: also a radio source, it is probably an evolved object. It has weak

2087: emission in the BLAST maps and is anti-coincident with $^{13}$CO.

2088: %

2089: The second source, right on the line, is possibly related to the {\it

2090:   IRAS} source 19367+2251, which IGA/MIGA show to be much weaker as a

2091: compact source than the first.  This extended BLAST source coincides

2092: with a peak in $^{13}$CO and becomes brighter relative to the first

2093: source at increasing BLAST wavelengths.  This behavior implies that it

2094: is a very cold object.

2095: 

2096: Another example is associated with the pair of {\it IRAS} sources

2097: 19385+2245 and 19389+2233.  They can be seen in all of the {\it IRAS}

2098: bands, but they are both diffuse in the BLAST maps.  The first is weak

2099: in $^{13}$CO and has complex structure in the 8\,\micron\ map. The

2100: second is elongated in BLAST and has a strong elongated CO cloud as

2101: well.  It is also extended at 8\,\micron, with a slight offset from the

2102: submillimeter source.

2103: 

2104: A future analysis of such cold clouds in BLAST maps in conjunction

2105: with MIPS and IRAC source catalogs may be used to determine whether

2106: they represent an even earlier phase before condensations form, or if

2107: they are simply distant cold envelopes of established young stellar

2108: objects.

2109: 

2110: 

2111: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

2112: \section{CONCLUSIONS}  \label{sec:concl}

2113: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

2114: 

2115: In this paper we present the first maps of the Galactic Plane observed

2116: with BLAST during its 2005 LDB flight from Sweden to Canada,

2117: specifically the survey in the direction of star forming clouds in

2118: Vulpecula.

2119: 

2120: Fourier deconvolution is used to improve the angular resolution of the

2121: maps from $\sim3\farcm5$ to $\sim1$\arcmin\ full-width half-power (the

2122: theoretical 500\,\micron\ diffraction limit). In these maps, 60

2123: compact submillimeter sources are detected simultaneously at 250, 350,

2124: and 500\,\micron.  Complementary {\it IRAS}, MIPSGAL, and {\it MSX}

2125: photometry are used to constrain the submillimeter-MIR SEDs of these

2126: objects, and hence infer their cold dust temperatures and bolometric

2127: fluxes. Our sample has convincingly revealed objects with a range of

2128: dust temperatures, from $\sim12$--40\,K under the assumption of an

2129: isothermal modified blackbody with dust emissivity index $\beta=1.5$.

2130: With these SED fits, we derive luminosity-to-mass ratios (independent

2131: of distance) in the range $0.2$--130\,L$_\odot$\,M$_\odot^{-1}$.

2132: Those with low values are highly shielded and cool, with no evidence

2133: for star formation, while those with high values have embedded

2134: high-mass star formation.

2135: 

2136: Using a $^{13}$CO$(1 \rightarrow 0)$ data cube, the VLA Galactic Plane

2137: Survey \ion{H}{1} cube, {\it Spitzer} IRAC and optical DSS images, we

2138: argue that 49 of the 60 sources lie in a molecular cloud complex

2139: associated with the open cluster NGC~6823 at $\sim2.3$\,kpc, 10

2140: objects are associated with the Perseus arm at $\sim8.5$\,kpc and 1

2141: object is in the outer Galaxy at $\sim14$\,kpc.  With these distance

2142: estimates we calculate bolometric luminosities and cloud masses

2143: associated with the thermal emission from cold dust. The most luminous

2144: object ($7\times10^4$\,L$_\odot$) is in the Perseus arm.  Near

2145: NGC~6823, the ranges are $\sim 40$--$10^4$\,L$_\odot$, and $\sim

2146: 15$--$700$\,M$_\odot$.

2147: 

2148: A mass function is constructed for the 49 objects associated with

2149: NGC~6823. It is compatible with the spectrum of molecular gas masses

2150: in other high-mass star forming regions, with a power-law index of

2151: $-1.7$ for the $\ga100$\,M$_\odot$ sources. A flattening at lower

2152: masses might be present, but is affected by detection completeness of

2153: cool sources in this mass range.

2154: 

2155: The luminosity-mass distribution we find is broadly consistent with an

2156: evolutionary sequence, from cool high-mass, low-luminosity clumps

2157: (most are not detected with {\it IRAS} or {\it MSX} and there is no

2158: evidence for any star formation), to more evolved HMPOs and UC

2159: \ion{H}{2} regions.  But interestingly, among the embedded objects in

2160: this molecular complex near NGC~6823 -- the next generation -- there

2161: are none quite as massive as the 40\,M$_\odot$ stars currently

2162: powering the nebula.

2163: 

2164: \acknowledgments

2165: 

2166: The BLAST collaboration acknowledges the support of NASA through grant

2167: numbers NAG5-12785, NAG5-13301, and NNGO-6GI11G, the Canadian Space

2168: Agency (CSA), the UK Particle Physics \& Astronomy Research Council

2169: (PPARC), and Canada's Natural Sciences and Engineering Research

2170: Council (NSERC). We would also like to thank the Columbia Scientific

2171: Balloon Facility (CSBF) staff for their outstanding work.  We thank

2172: Andy Gibb for helpful discussions. We also thank the anonymous referee

2173: for useful comments. LO acknowledges partial support by the Puerto

2174: Rico Space Grant Consortium and by the Fondo Institucional para la

2175: Investigacion of the University of Puerto Rico, and also thanks

2176: students Carlos M.  Poventud and Jorge L.  Morales for assistance with

2177: the analysis. CBN acknowledges support from the Canadian Institute for

2178: Advanced Research. DHH acknowledges the support of Consejo Nacional de

2179: Ciencia y Technolog\'ia (CONACYt) grant 39953-F. This research has

2180: been enabled by the use of WestGrid computing resources.

2181: 

2182: 

2183: \bibliographystyle{apj}

2184: \bibliography{apj-jour,refs}

2185: 

2186: \begin{deluxetable}{ccrrrrrrr}

2187: \tablewidth{0pt} 

2188: \small 

2189: \tablecaption{BLAST Sources \label{tab:src}} 

2190: \tablehead{ 

2191: \colhead{BLAST} & 

2192: \colhead{Source name} &

2193: \colhead{$F_{250}$} & 

2194: \colhead{$\sigma_{250}$} & 

2195: \colhead{$F_{350}$} &

2196: \colhead{$\sigma_{350}$} & 

2197: \colhead{$F_{500}$} &

2198: \colhead{$\sigma_{500}$} & 

2199: \colhead{$V_{\mathrm{lsr}}$} \\

2200: \colhead{ID} & 

2201: \colhead{} & 

2202: \colhead{(Jy)} & 

2203: \colhead{(Jy)} &

2204: \colhead{(Jy)} & 

2205: \colhead{(Jy)} & 

2206: \colhead{(Jy)} & 

2207: \colhead{(Jy)} &

2208: \colhead{km s$^{-1}$} } 

2209: \startdata 

2210: %\input{blasttable.dat} 

2211:   V01 &BLAST J193837+230541 & 59.9 &  4.2 & 27.1 &  2.8 & 11.0 &  1.2 &  30.8\\

2212:   V02 &BLAST J193849+230839 &258.0 &  5.1 & 97.8 &  3.4 & 36.9 &  1.3 &  32.3\\

2213:   V03 &BLAST J193858+224637 &197.9 &  4.7 &101.9 &  3.5 & 44.5 &  1.3 &  36.5\\

2214:   V04\tablenotemark{a} &BLAST J193921+224401 & 23.6 &  4.1 &  8.9 &  2.7 &  7.3 &  1.2 &  27.3\\

2215:   V05 &BLAST J193935+235947 &325.7 &  5.6 &124.1 &  3.8 & 43.6 &  1.3 &  37.0\\

2216:   V06 &BLAST J193943+232601 & 20.4 &  4.1 & 13.1 &  2.7 &  5.8 &  1.2 &  43.2\\

2217:   V07\tablenotemark{b} &BLAST J194046+241007 & 38.7 &  4.1 & 12.2 &  2.7 &  5.2 &  1.1 & $-$53.5\\

2218:   V08 &BLAST J194059+240439 &225.6 &  4.9 & 81.6 &  3.2 & 30.7 &  1.2 &  34.8\\

2219:   V09 &BLAST J194103+240111 &194.8 &  4.7 &102.1 &  3.5 & 41.2 &  1.3 &  34.4\\

2220:   V10 &BLAST J194106+235513 & 21.5 &  4.1 & 12.6 &  2.7 &  6.7 &  1.2 &  34.0\\

2221:   V11 &BLAST J194136+232325 & 18.3 &  4.0 & 18.9 &  2.7 & 10.1 &  1.2 &  27.8\\

2222:   V12 &BLAST J194144+231410 & 33.2 &  4.1 & 16.9 &  2.7 &  7.1 &  1.1 &  29.4\\

2223:   V13 &BLAST J194144+231804 & 26.6 &  4.1 & 16.4 &  2.7 &  7.5 &  1.2 &  23.7\\

2224:   V14 &BLAST J194146+234914 & 19.0 &  4.1 & 10.3 &  2.7 &  3.3 &  1.1 &  34.6\\

2225:   V15 &BLAST J194213+232305 & 41.9 &  4.1 & 21.8 &  2.7 & 10.5 &  1.2 &  24.6\\

2226:   V16 &BLAST J194217+231958 & 23.5 &  4.1 & 10.8 &  2.7 &  5.5 &  1.2 &  33.0\\

2227:   V17\tablenotemark{c} &BLAST J194226+234532 & 30.4 &  4.1 & 10.8 &  2.7 &  5.6 &  1.2 &  $-$2.1\\

2228:   V18 &BLAST J194228+230458 &259.9 &  5.1 &114.3 &  3.6 & 42.8 &  1.3 &  26.6\\

2229:   V19\tablenotemark{ac} &BLAST J194232+234913 & 15.8 &  4.1 &  5.6 &  2.7 &  2.2 &  1.1 &  -3.0\\

2230:   V20\tablenotemark{c} &BLAST J194232+234253 & 27.6 &  4.1 & 17.9 &  2.7 &  9.3 &  1.2 &   0.8\\

2231:   V21\tablenotemark{c} &BLAST J194234+234412 & 20.9 &  4.1 & 15.6 &  2.7 &  6.6 &  1.2 &  $-$0.3\\

2232:   V22 &BLAST J194234+232321 & 60.9 &  4.1 & 34.8 &  2.8 & 15.1 &  1.2 &  30.5\\

2233:   V23\tablenotemark{c} &BLAST J194236+235003 &129.6 &  4.4 & 45.8 &  2.9 & 15.7 &  1.2 &  -3.0\\

2234:   V24\tablenotemark{c} &BLAST J194244+234021 & 62.5 &  4.2 & 37.1 &  2.8 & 12.3 &  1.2 &   0.3\\

2235:   V25 &BLAST J194246+225536 & 30.1 &  4.1 & 17.7 &  2.7 & 10.3 &  1.2 &  26.9\\

2236:   V26 &BLAST J194254+232408 & 73.2 &  4.2 & 42.4 &  2.8 & 18.8 &  1.2 &  31.4\\

2237:   V27 &BLAST J194306+230125 & 19.8 &  4.1 & 14.1 &  2.7 &  7.0 &  1.2 &  27.0\\

2238:   V28 &BLAST J194308+232457 & 21.0 &  4.1 & 15.0 &  2.7 &  6.5 &  1.2 &  33.1\\

2239:   V29\tablenotemark{a} &BLAST J194311+232010 & 15.7 &  4.1 &  2.3 &  2.7 &  5.5 &  1.2 &  24.0\\

2240:   V30 &BLAST J194311+234405 &852.7 & 11.0 &378.5 &  8.5 &153.4 &  2.7 &  22.8\\

2241:   V31 &BLAST J194311+232237 & 13.3 &  4.1 & 19.1 &  2.7 &  2.3 &  1.1 &  33.4\\

2242:   V32 &BLAST J194317+231352 &221.2 &  4.9 &110.7 &  3.6 & 43.3 &  1.3 &  29.3\\

2243:   V33 &BLAST J194319+232639 & 17.5 &  4.0 & 17.7 &  2.7 &  6.7 &  1.2 &  22.5\\

2244:   V34\tablenotemark{a} &BLAST J194319+233906 & 27.5 &  4.1 & 24.1 &  2.7 &  7.4 &  1.2 &  22.9\\

2245:   V35 &BLAST J194323+231512 & 22.0 &  4.1 &  7.9 &  2.7 &  1.4 &  1.1 &  30.4\\

2246:   V36\tablenotemark{a} &BLAST J194325+234503 & 86.6 &  4.2 & 68.3 &  3.1 & 18.9 &  1.2 &  21.3\\

2247:   V37 &BLAST J194326+240618 & 24.8 &  4.1 &  9.3 &  2.7 &  3.5 &  1.1 &  20.7\\

2248:   V38\tablenotemark{a} &BLAST J194327+235638 & 64.4 &  4.2 & 43.8 &  2.9 & 22.6 &  1.2 &  22.4\tablenotemark{d}\\

2249:   V39 &BLAST J194328+232032 & 57.8 &  4.1 & 31.1 &  2.8 & 12.6 &  1.2 &  25.7\\

2250:   V40 &BLAST J194329+234013 &189.7 &  4.7 & 86.9 &  3.3 & 35.4 &  1.3 &  20.5\\

2251:   V41 &BLAST J194331+235728 & 90.8 &  4.2 & 42.7 &  2.8 & 16.3 &  1.2 &  22.8\\

2252:   V42\tablenotemark{a} &BLAST J194333+231445 & 23.8 &  4.1 & 24.0 &  2.7 &  7.5 &  1.2 &  26.9\\

2253:   V43\tablenotemark{a} &BLAST J194334+232737 & 41.8 &  4.1 & 24.7 &  2.8 &  3.6 &  1.1 &  26.9\\

2254:   V44 &BLAST J194335+235332 & 28.2 &  4.1 & 14.7 &  2.7 &  5.5 &  1.2 &  21.4\\

2255:   V45\tablenotemark{a} &BLAST J194338+231725 & 10.6 &  4.1 &  7.0 &  2.7 &  0.4 &  1.1 &  26.6\\

2256:   V46 &BLAST J194340+235506 & 28.8 &  4.1 & 12.5 &  2.7 &  5.3 &  1.1 &  22.5\\

2257:   V47 &BLAST J194342+232006 &129.9 &  4.4 & 70.2 &  3.1 & 33.8 &  1.3 &  26.7\\

2258:   V48 &BLAST J194344+240005 & 23.5 &  4.1 & 11.2 &  2.7 &  3.3 &  1.1 &  22.2\\

2259:   V49\tablenotemark{a} &BLAST J194346+232532 & 27.0 &  4.1 & 17.9 &  2.7 & 11.5 &  1.2 &  26.1\\

2260:   V50 &BLAST J194347+233505 & 18.6 &  4.1 & 10.1 &  2.7 &  3.0 &  1.1 &  22.1\\

2261:   V51\tablenotemark{c} &BLAST J194349+230840 & 22.4 &  4.1 & 16.1 &  2.7 &  6.8 &  1.2 &  -7.0\\

2262:   V52 &BLAST J194350+232839 &486.4 &  7.1 &214.5 &  5.3 & 84.2 &  1.8 &  27.6\\

2263:   V53 &BLAST J194350+231846 & 40.6 &  4.1 & 20.0 &  2.7 &  9.1 &  1.2 &  27.7\\

2264:   V54\tablenotemark{c} &BLAST J194351+234625 & 21.8 &  4.1 &  7.0 &  2.7 &  2.4 &  1.1 &  3.2\tablenotemark{e}\\

2265:   V55\tablenotemark{a} &BLAST J194402+232409 & 30.6 &  4.1 & 21.9 &  2.7 &  5.0 &  1.2 &  29.3\\

2266:   V56 &BLAST J194404+234418 & 53.5 &  4.1 & 37.3 &  2.8 & 15.6 &  1.2 &  21.7\\

2267:   V57 &BLAST J194405+232658 & 40.7 &  4.1 & 19.6 &  2.7 &  7.1 &  1.2 &  27.2\\

2268:   V58\tablenotemark{c} &BLAST J194409+230437 & 15.4 &  4.1 & 10.7 &  2.7 &  5.2 &  1.2 &  -12.7\\

2269:   V59\tablenotemark{c} &BLAST J194424+234239 & 23.8 &  4.1 & 12.9 &  2.7 &  4.1 &  1.1 &  -10.7\\

2270:   V60 &BLAST J194430+233426 & 14.9 &  4.1 & 14.1 &  2.7 &  4.8 &  1.2 &  26.2\\

2271: \enddata

2272: \tablecomments{Flux densities for BLAST sources are quoted at

2273:   precisely 250, 350, and 500\,\micron\ using SED fits to obtain

2274:   color-corrections for the band-averaged flux densities

2275:   (\S\ref{sec:coldsed}). The quoted statistical uncertainties

2276:   are determined from Monte Carlo simulations

2277:   (\S\ref{sec:montecarlo}), and do not include calibration

2278:   uncertainties.}

2279: \tablenotetext{a}{These sources are located on ripples in the

2280:   deconvolved map and the BLAST colors are considered unreliable as a

2281:   result.}  \tablenotetext{b}{V07 is believed to lie in the outer

2282:   galaxy.}  \tablenotetext{c}{These sources are associated with a

2283:   molecular cloud in the Perseus arm.}  \tablenotetext{d}{Also has a

2284:   comparable component at $-$15\,km\,s$^{-1}$.}

2285: \tablenotetext{e}{Also has a comparable component at

2286:   23\,km\,s$^{-1}$.}

2287: \end{deluxetable}

2288: 

2289: % PSC counterparts

2290: \begin{deluxetable}{llrrrrrr}

2291: \tablewidth{0pt}

2292: \small

2293: \tablecaption{{\it IRAS\/} PSC Counterparts \label{tab:iras}}

2294: \tablehead{

2295: \colhead{BLAST} &

2296: \colhead{IRAS} &

2297: \colhead{$\Delta \alpha$\tablenotemark{a}} &

2298: \colhead{$\Delta \delta$\tablenotemark{a}} &

2299: \colhead{$F_{12}$} &

2300: \colhead{$F_{25}$} &

2301: \colhead{$F_{60}$} &

2302: \colhead{$F_{100}$} 

2303: \\

2304: \colhead{ID} &

2305: \colhead{ID} &

2306: \colhead{(\arcsec)} &

2307: \colhead{(\arcsec)} &

2308: \colhead{(Jy)} &

2309: \colhead{(Jy)} &

2310: \colhead{(Jy)} &

2311: \colhead{(Jy)} 

2312: }

2313: \startdata

2314: %\input{psctable.dat}

2315:   V01 & 19364+2258 &$-$19.4 & $-$3.3 &    0.8 &    8.3 &  389.0\tablenotemark{b} &  698.0\tablenotemark{b}\\

2316:   V02 & 19366+2301 &$-$21.0 & $-$1.0 &    8.8 &   62.0 &  389.0 &  698.0\\

2317:   V03 & 19368+2239 & $-$4.5 & $-$5.4 &    2.3 &   11.7 &  112.0 &  231.0\\

2318:   V05 & 19374+2352 &$-$27.0 &  7.3 &    4.6 &   24.3 &  422.0 &  768.0\\

2319:   V07 & 19386+2403 & $-$5.9 &  2.1 &    1.8 &    4.7 &   71.2 &   63.8\\

2320:   V08 & 19388+2357 & $-$3.0 & $-$0.1 &    1.3 &   10.9 &  254.0 &  433.0\\

2321:   V09 & 19389+2354 &  1.5 & $-$0.1 &    1.8 &    1.2\tablenotemark{b} &   58.2 &  433.0\tablenotemark{b}\\

2322:   V14 & 19396+2342 &  3.1 &  9.6 &    1.5\tablenotemark{b} &    0.8\tablenotemark{b} &    3.1 &   16.7\\

2323:   V18 & 19403+2258 &$-$21.0 & 13.9 &   16.2 &  118.0 &  562.0 &  640.0\\

2324:   V20 & 19404+2335 & $-$6.0 &  0.2 &    1.9\tablenotemark{b} &    1.0 &    4.9\tablenotemark{b} &   53.0\tablenotemark{b}\\

2325:   V23\tablenotemark{c} & 19404+2342 &$-$24.0 & $-$7.9 & 9.7 & 35.3 &

2326:   439.0 & 980.0\\

2327:   V24 & 19406+2333 &  0.0 &  8.1 &    0.7 &    5.1 &   35.0 &   89.8\\

2328:   V26 & 19407+2316 & $-$4.5 & $-$6.8 &    4.1\tablenotemark{b} &    4.2 &   23.2 &  165.0\tablenotemark{b}\\

2329:   V30 & 19410+2336 &  4.5 &  0.9 &   14.4 &  109.0 &  982.0 & 1630.0\\

2330:   V32 & 19411+2306 & 12.0 &  6.2 &    5.3 &   41.2 &  159.0 &  557.0\\

2331:   V37 & 19413+2358 & $-$3.0 & $-$8.7 &   15.4\tablenotemark{b} &    0.8\tablenotemark{b} &    4.5 &   37.0\tablenotemark{b}\\

2332:   V38\tablenotemark{d} & 19413+2349 & 10.5 & 20.1 &    0.7 &    4.6 &   42.5 &  132.0\\

2333:   V40 & 19413+2332 & $-$4.5 & $-$9.3 &    4.2 &   38.5 &  217.0 & 1630.0\tablenotemark{b}\\

2334:   V51 & 19416+2301 &  0.0 &$-$12.9 &    1.0\tablenotemark{b} &    1.0 &    4.7\tablenotemark{b} &   35.8\tablenotemark{b}\\

2335:   V52 & 19416+2321 &$-$18.0 & 20.5 &    5.0\tablenotemark{b} &   24.7\tablenotemark{b} &  268.0 &  624.0\\

2336:   V53\tablenotemark{d} & 19416+2312 &$-$39.0 & 39.3 &    0.6 &   10.7\tablenotemark{b} &   13.6\tablenotemark{b} &  250.0\\

2337:   V57 & 19419+2319 &  4.5 &$-$15.0 &    2.8 &    2.6 &   36.1\tablenotemark{b} &  624.0\tablenotemark{b}\\

2338:   V59 & 19422+2335 & $-$1.5 &  2.1 &    1.4 &    2.3 &   25.6 &   46.6\\

2339: \enddata

2340: \tablecomments{{\it IRAS} point sources associated

2341:   with BLAST objects. The search radius is a variable function of both

2342:   BLAST and {\it IRAS} positional uncertainties (see

2343:   \S\ref{sec:iras}).}

2344: \tablenotetext{a}{BLAST source tangent plane offsets (E and N)

2345:   compared to {\it IRAS} source position.}

2346: \tablenotetext{b}{These {\it IRAS} measurements are flagged as

2347:   upper-limits in the catalog (usually due to confusion). Values from

2348:   Table~\ref{tab:mapphot} are used instead.}

2349: \tablenotetext{c}{IRAS 19404+2342 is close to both V19 and V23; the

2350:   ambiguity is resolved with the improved resolution of the MIPS

2351:   70\,\micron\ map.}

2352: \tablenotetext{d}{Although IRAS 19413+2349 and 19416+2312 are close to

2353:   V38 and V53 respectively, their SEDs are not compatible with the

2354:   BLAST photometry and the identifications are not used in subsequent

2355:   analyses.}

2356: \end{deluxetable}

2357: 

2358: % IRAS/MIPS map photometry table

2359: \begin{deluxetable}{lrcrcrcrcrc}

2360: \tablewidth{0pt}

2361: \small

2362: \tablecaption{{\it IRAS} and {\it Spitzer} MIPS Photometry 

2363: \label{tab:mapphot}}

2364: \tablehead{

2365: \colhead{BLAST} &

2366: \colhead{$F_{12}$ } & 

2367: \colhead{Flag} &

2368: \colhead{$F_{25}$} &

2369: \colhead{Flag} &

2370: \colhead{$F_{60}$} &

2371: \colhead{Flag} &

2372: \colhead{$F_{100}$} &

2373: \colhead{Flag} &

2374: \colhead{$F_{70}$} &

2375: \colhead{Flag} \\

2376: \colhead{ID} &

2377: \colhead{(Jy)} &

2378: \colhead{} &

2379: \colhead{(Jy)} &

2380: \colhead{} &

2381: \colhead{(Jy)} &

2382: \colhead{} &

2383: \colhead{(Jy)} &

2384: \colhead{} &

2385: \colhead{(Jy)} &

2386: \colhead{}

2387: }

2388: \startdata

2389: %\input{irastable.dat}

2390: V01 & 1.2\tablenotemark{a} & g & 7.2\tablenotemark{a} & g & 46.5 & g & 120.5 & u & 33.0  & s \\

2391: V02 & 12.1\tablenotemark{a} & g & 60.9\tablenotemark{a} & g & 437.7\tablenotemark{a} & g & 733.6\tablenotemark{a} & g & 211.1  & s \\

2392: V03 & 2.4\tablenotemark{a} & g & 12.4\tablenotemark{a} & g & 114.5\tablenotemark{a} & g & 243.2\tablenotemark{a} & g & 78.2  & s \\

2393: V04 & 0.2 & n & 0.2 & n & 0.2 & n & 31.3 & n & 0.4  & n \\

2394: V05 & 5.9\tablenotemark{a} & g & 28.6\tablenotemark{a} & g & 410.5\tablenotemark{a} & g & 772.9\tablenotemark{a} & g & 218.6  & s \\

2395: V06 & $-$0.1 & n & $-$0.1 & n & 4.2 & g & 24.7 & u & 2.5  & g \\

2396: V07 & 1.7\tablenotemark{a} & g & 4.7\tablenotemark{a} & g & 66.7\tablenotemark{a} & g & 135.7\tablenotemark{a} & g & 50.4  & s \\

2397: V08 & 2.3\tablenotemark{a} & g & 12.1\tablenotemark{a} & g & 259.9\tablenotemark{a} & g & 569.6\tablenotemark{a} & g & 206.7  & s \\

2398: V09 & 1.5\tablenotemark{a} & g & 2.9 & g & 44.1\tablenotemark{a} & g & 256.8 & u & 44.1  & s \\

2399: V10 & 0.0 & n & 0.0 & g & 2.4 & u & 48.3 & u & 0.9  & n \\

2400: V11 & $-$0.1 & n & $-$0.2 & n & $-$0.2 & n & 18.2 & n & 0.8  & n \\

2401: V12 & 1.7 & g & 4.0 & g & 14.8 & g & 70.9 & u & 9.4  & g \\

2402: V13 & $-$0.4 & n & $-$0.4 & n & $-$3.6 & n & 13.0 & n & $-$1.5  & n \\

2403: V14 & 0.3 & n & 0.5 & u & 4.0\tablenotemark{a} & g & 28.0\tablenotemark{a} & u & 2.7  & g \\

2404: V15 & $-$0.2 & n & 0.2 & n & 9.4 & u & 83.3 & u & 3.1  & n \\

2405: V16 & 1.2 & u & 2.0 & n & 11.7 & u & 91.9 & u & 3.5  & g \\

2406: V17 & 0.2 & n & 0.9 & u & 5.9 & u & 14.5 & n & 3.4  & n \\

2407: V18 & 17.2\tablenotemark{a} & g & 108.5\tablenotemark{a} & g & 570.0\tablenotemark{a} & g & 699.1\tablenotemark{a} & g & 275.4  & s \\

2408: V19 & 17.6 & u & 52.8 & u & 460.1 & u & 979.5 & u & 148.6  & s \\

2409: V20 & 0.0 & g & 0.3\tablenotemark{a} & g & 6.6 & u & 19.4 & n & 6.3  & g \\

2410: V21 & $-$0.1 & n & $-$0.3 & n & $-$1.9 & n & $-$0.1 & n & 4.5  & g \\

2411: V22 & 0.7 & n & 0.5 & n & 7.4 & u & 97.7 & u & 6.0  & n \\

2412: V23 & 15.3\tablenotemark{a} & u & 53.3\tablenotemark{a} & u & 506.5\tablenotemark{a} & g & 916.8\tablenotemark{a} & g & 192.2  & s \\

2413: V24 & 1.4\tablenotemark{a} & u & 3.5\tablenotemark{a} & g & 32.0\tablenotemark{a} & g & 110.6\tablenotemark{a} & u & 25.5  & s \\

2414: V25 & 0.2 & n & 0.9 & u & 3.8 & g & 47.0 & g & 0.7  & g \\

2415: V26 & 2.0 & u & 4.0\tablenotemark{a} & g & 21.9\tablenotemark{a} & g & 91.4 & u & 7.7  & g \\

2416: V27 & 0.4 & u & 0.7 & u & 7.7 & u & 51.7 & u & 0.7  & n \\

2417: V28 & 2.6 & u & 4.2 & u & $-$3.3 & n & 36.9 & n & 3.2  & n \\

2418: V29 & $-$0.7 & n & 0.1 & n & $-$0.5 & n & $-$6.2 & n & $-$0.9  & g \\

2419: V30 & 18.5\tablenotemark{a} & g & 118.0\tablenotemark{a} & g & 892.2\tablenotemark{a} & g & 1463.9\tablenotemark{a} & g & 378.2  & s \\

2420: V31 & 0.7 & n & 0.9 & u & 2.0 & n & $-$7.8 & n & $-$0.3  & n \\

2421: V32 & 6.1\tablenotemark{a} & g & 37.3\tablenotemark{a} & g & 151.2\tablenotemark{a} & g & 478.2\tablenotemark{a} & g & 62.6  & s \\

2422: V33 & 1.2 & n & 1.9 & u & 21.5 & u & 37.0 & u & $-$1.3  & n \\

2423: V34 & 3.2 & u & 5.4 & u & 83.9 & u & 334.1 & u & 2.2  & n \\

2424: V35 & 6.2 & u & 22.9 & u & 87.9 & u & 461.9 & u & 6.6  & n \\

2425: V36 & 5.6 & u & 8.6 & g & 88.4 & u & 472.2 & u & 42.6  & s \\

2426: V37 & 0.3 & n & 0.8 & n & 4.4\tablenotemark{a} & g & 24.6 & u & 3.3  & g \\

2427: V38 & 0.9 & u & 5.3 & u & 34.0 & u & 120.9 & u & 22.9  & s \\

2428: V39 & 3.8 & u & 3.7 & g & 40.4 & u & 183.2 & u & 19.4  & g \\

2429: V40 & 7.2\tablenotemark{a} & g & 44.4\tablenotemark{a} & g & 214.3\tablenotemark{a} & g & 400.5 & g & 152.3  & s \\

2430: V41 & 0.6 & n & 5.2 & u & 33.5 & u & 131.7 & u & 17.8  & u \\

2431: V42 & $-$0.8 & n & $-$0.6 & n & $-$0.6 & n & 223.5 & u & $-$3.3  & n \\

2432: V43 & 1.8 & n & 1.6 & n & 21.1 & u & 147.6 & u & 7.8  & n \\

2433: V44 & 0.0 & n & 0.1 & n & 3.6 & u & 35.2 & n & 2.0  & u \\

2434: V45 & 0.0 & n & 3.9 & u & 1.7 & n & 17.7 & n & 6.8  & n \\

2435: V46 & 0.0 & n & $-$0.2 & n & 9.5 & u & 137.3 & u & 12.3  & u \\

2436: V47 & 0.9 & g & 10.5 & g & 60.5 & g & 174.6 & g & 21.0  & s \\

2437: V48 & 0.8 & n & 1.2 & n & 9.0 & u & 61.0 & u & 8.2  & g \\

2438: V49 & $-$0.9 & n & 1.0 & n & 16.7 & u & 149.0 & u & 17.6  & g \\

2439: V50 & 0.2 & g & 1.0 & g & 6.7 & u & 51.5 & u & 4.2  & g \\

2440: V51 & 0.3 & n & 1.9\tablenotemark{a} & n & 15.9 & u & 103.5 & u & 3.2  & g \\

2441: V52 & 4.4 & u & 27.6 & u & 278.2\tablenotemark{a} & g & 675.1\tablenotemark{a} & g & 211.7  & s \\

2442: V53 & $-$0.5\tablenotemark{a} & n & $-$0.3 & n & 5.5 & n & 227.9\tablenotemark{a} & u & 0.7  & n \\

2443: V54 & 0.5 & u & 0.9 & u & 11.2 & u & 82.4 & u & 12.9  & g \\

2444: V55 & 1.9 & u & 5.6 & u & 11.3 & g & 369.6 & u & 8.1  & n \\

2445: V56 & $-$0.1 & n & $-$0.3 & n & 1.1 & n & 31.4 & u & $-$0.4  & n \\

2446: V57 & 1.7\tablenotemark{a} & g & 5.3\tablenotemark{a} & g & 4.5 & n & 423.2 & u & 5.7  & g \\

2447: V58 & $-$0.1 & n & $-$0.2 & n & 7.2 & u & 90.1 & u & 2.5  & g \\

2448: V59 & 1.4\tablenotemark{a} & g & 2.3\tablenotemark{a} & g & 26.4\tablenotemark{a} & g & 57.4\tablenotemark{a} & g & 24.7  & s \\

2449: V60 & 0.8 & n & 1.3 & u & 10.6 & u & 75.4 & u & 2.2  & n \\

2450: 

2451: \enddata

2452: \tablecomments{{\it IRAS} and {\it Spitzer} MIPS map photometry are

2453:   described in \S\ref{sec:iras} and \S\ref{sec:mips70}

2454:   respectively.  Each measurement is flagged upon visual inspection:

2455:   ``g'' if it is good; ``u'' for upper limit; ``n'' for no visible

2456:   source; and ``s'' for saturated (unreliable) values in the MIPS

2457:   70\,\micron\ map.}

2458: \tablenotetext{a}{Clear detections from the {\it IRAS} PSC

2459:   (Table~\ref{tab:iras}) are used in favor of these values when

2460:   available.}

2461: \end{deluxetable}

2462: 

2463: 

2464: % MSX table

2465: \begin{deluxetable}{lrrrrrr}

2466: \tablewidth{0pt}

2467: \small

2468: \tablecaption{MSX counterparts \label{tab:msx}}

2469: \tablehead{

2470: \colhead{BLAST} &

2471: \colhead{$\Delta \alpha$\tablenotemark{a}} &

2472: \colhead{$\Delta \delta$\tablenotemark{a}} &

2473: \colhead{$F_8.3$} &

2474: \colhead{$F_{12.1}$} &

2475: \colhead{$F_{14.7}$} &

2476: \colhead{$F_{21.3}$} \\

2477: \colhead{ID} &

2478: \colhead{(\arcsec)} &

2479: \colhead{(\arcsec)} &

2480: \colhead{(Jy)} &

2481: \colhead{(Jy)} &

2482: \colhead{(Jy)} &

2483: \colhead{(Jy)}

2484: }

2485: \startdata

2486: %\input{msxtable.dat}

2487: V01 & $-$6.9 & 1.4 &   0.9 &   0.8 &   0.8 &   4.7\\

2488: V02 & $-$11.0 & $-$1.3 &   4.3 &   5.3 &   3.7 &  23.9\\

2489: V03 & 18.0 & 8.1 &   0.9 &   1.1 &   0.6 &   1.6\\

2490:  & $-$20.7 & $-$9.9 &   0.8 &   1.0 &   1.4 &   3.4\\

2491: V05 & $-$6.9 & 26.0 &   1.7 &   1.8 &   1.2 &   2.7\\

2492:  & $-$2.7 & 2.9 &   0.9 &   0.8 &   1.2 &   6.2\\

2493: V06 & 19.3 & $-$63.1 &   0.4 & \nodata & \nodata & \nodata\\

2494: V07 & $-$0.0 & 11.2 &   1.0 &   1.4 &   0.8 &   2.0\\

2495:  & $-$26.0 & $-$53.6 &   0.1 &   1.2 & \nodata & \nodata\\

2496: V08 & $-$0.0 & $-$23.9 &   0.6 &   1.4 &   0.9 & \nodata\\

2497:  & $-$8.2 & 6.7 &   0.5 &   1.3 &   0.8 &   3.6\\

2498: V12 & $-$48.2 & $-$41.7 &   1.7 &   1.2 &   0.7 & \nodata\\

2499:  & 8.3 & 9.8 &   0.5 &   2.0 &   3.1 &   4.7\\

2500: V13 & $-$4.1 & 21.1 &   0.1 & \nodata & \nodata & \nodata\\

2501: V14 & $-$11.0 & 16.7 &   0.2 & \nodata &   0.8 & \nodata\\

2502: V18 & $-$8.3 & 16.9 &   8.4 &   9.0 &   6.2 &  57.5\\

2503: V19 & 11.0 & 28.5 &   0.9 &   1.8 &   2.2 &   3.9\\

2504:  & 37.0 & 38.2 &   0.7 &   2.6 &   3.2 &   7.9\\

2505:  & $-$9.6 & 49.0 &   0.1 &   0.6 & \nodata &   1.3\\

2506: V20 & $-$0.0 & $-$0.2 &   0.2 & \nodata & \nodata & \nodata\\

2507: V21 & $-$15.1 & 13.9 &   0.1 & \nodata & \nodata & \nodata\\

2508: V22 & $-$20.7 & $-$3.6 &   0.2 & \nodata & \nodata & \nodata\\

2509: V23 & $-$15.1 & $-$11.7 &   0.7 &   2.6 &   3.2 &   7.9\\

2510:  & 17.8 & $-$5.2 &   1.5 &   2.5 &   1.3 &   3.1\\

2511:  & $-$19.2 & 17.1 &   0.8 &   2.2 &   2.7 &   5.2\\

2512: V24 & $-$5.5 & 9.9 &   0.4 & \nodata & \nodata &   2.5\\

2513: V26 & 15.1 & 6.7 &   1.4 &   2.1 &   2.2 &   3.6\\

2514: V28 & $-$23.4 & $-$28.5 &   0.9 &   1.4 &   1.7 &   1.9\\

2515: V30 & $-$0.0 & $-$0.2 &   5.3 &   8.5 &  14.9 &  47.4\\

2516: V31 & $-$15.1 & 17.8 &   0.2 & \nodata & \nodata & \nodata\\

2517: V32 & 8.3 & 8.5 &   1.8 &   5.2 &   8.7 &  25.6\\

2518: V35 & 66.1 & 25.8 &   0.2 & \nodata & \nodata &   2.7\\

2519: V36 & 31.6 & $-$15.4 &   1.7 &   1.6 &   1.0 &   2.1\\

2520: V37 & 20.5 & 60.9 &   0.9 &   0.6 &   0.5 & \nodata\\

2521: V38 & 2.7 & 16.0 &   0.7 &   0.9 &   1.2 &   2.7\\

2522: V39 & 2.8 & 10.5 &   0.9 &   1.9 &   2.1 &   2.3\\

2523: V40 & $-$9.6 & $-$21.8 &   2.2 &   2.5 &   1.1 &   6.1\\

2524:  & 6.9 & 11.7 &   2.0 &   2.2 &   2.7 &  18.3\\

2525: V43 & $-$6.9 & $-$20.0 &   0.2 & \nodata & \nodata & \nodata\\

2526: V46 & 8.2 & 4.9 &   0.1 & \nodata & \nodata & \nodata\\

2527:  & 4.1 & $-$39.0 &   0.3 &   1.8 & \nodata & \nodata\\

2528: V47 & 4.1 & 8.8 &   0.1 &   1.1 &   1.7 &   4.7\\

2529: V49 & $-$19.3 & $-$24.1 &   0.4 &   0.8 & \nodata & \nodata\\

2530: V50 & 15.1 & 48.0 &   0.2 & \nodata & \nodata & \nodata\\

2531: V51 & $-$2.8 & $-$7.9 &   0.2 & \nodata & \nodata & \nodata\\

2532: V52 & $-$4.1 & 0.0 &   0.2 & \nodata & \nodata &   3.3\\

2533: V53 & $-$20.7 & $-$29.8 &   0.1 & \nodata & \nodata & \nodata\\

2534: V55 & $-$17.9 & 11.8 &   0.2 & \nodata & \nodata & \nodata\\

2535: V57 & $-$1.4 & $-$11.7 &   1.3 &   1.3 &   1.2 &   1.2\\

2536: V59 & 1.4 & 1.5 &   0.8 &   1.1 &   0.6 & \nodata\\

2537: V60 & $-$52.2 & 5.7 &   0.1 &   1.0 & \nodata & \nodata\\

2538: 

2539: \enddata

2540: \tablecomments{All candidate {\it MSX} counterparts are listed for

2541: each BLAST source (\S\ref{sec:msx}).}

2542: \tablenotetext{a}{BLAST source tangent plane offsets (E and N)

2543:   compared to {\it MSX} source position.}

2544: \end{deluxetable}

2545: 

2546: % SED data table

2547: \begin{deluxetable}{lcrrcc}

2548: \tablewidth{0pt}

2549: \small

2550: \tablecaption{Results from SED Fits \label{tab:sed}}

2551: \tablehead{

2552: \colhead{BLAST} &

2553: \colhead{Temperature} &

2554: \colhead{$M_\mathrm{c}$} &

2555: \colhead{$L_{\mathrm{FIR}}$} &

2556: \colhead{Luminosity Fraction} &

2557: \colhead{$L_{\mathrm{FIR}}/M_\mathrm{c}$} \\

2558: \colhead{ID} &

2559: \colhead{(K)} &

2560: \colhead{(M$_\odot$)} &

2561: \colhead{(L$_\odot$)} &

2562: \colhead{(\%)} &

2563: \colhead{(L$_\odot$\,M$_\odot^{-1}$)}

2564: }

2565: \startdata

2566: %\input{sedtable.dat}

2567: V01 & $ 27.2 \pm  1.5 $ & $ 44 \pm 7 $ & $ 750 \pm 130 $ & $ 76 $ &  $   17.5 \pm   5.0 $ \\

2568: V02 & $ 31.9 \pm  2.3 $ & $ 114 \pm 22 $ & $ 4500 \pm 1000 $ & $ 78 $ &  $   40.5 \pm  15.8 $ \\

2569: V03 & $ 22.2 \pm  0.9 $ & $ 273 \pm 40 $ & $ 1470 \pm 150 $ & $ 67 $ &  $    5.6 \pm   1.2 $ \\

2570: V04 & $ 20.0 $ & $ 38 \pm 5 $ & $ 121 \pm 15 $ & \nodata &  $    3.2 $ \\

2571: V05 & $ 31.3 \pm  1.9 $ & $ 130 \pm 27 $ & $ 4910 \pm 880 $ & $ 89 $ &  $   36.1 \pm  12.2 $ \\

2572: V06 & $ 20.2 \pm  2.4 $ & $ 55 \pm 35 $ & $ 96 \pm 57 $ & $ 62 $ &  $    1.9 \pm   1.9 $ \\

2573: V07 & $ 32.6 \pm  1.9 $ & $ 579 \pm 122 $ & $ 28200 \pm 4900 $ & $ 83 $ &  $   49.0 \pm  16.5 $ \\

2574: V08 & $ 29.7 \pm  1.7 $ & $ 97 \pm 19 $ & $ 2610 \pm 430 $ & $ 84 $ &  $   26.5 \pm   8.8 $ \\

2575: V09 & $ 22.8 \pm  1.7 $ & $ 233 \pm 45 $ & $ 1560 \pm 350 $ & $ 80 $ &  $    6.4 \pm   2.4 $ \\

2576: V10 & $ 15.8 \pm  2.6 $ & $ 89 \pm 50 $ & $ 53 \pm 22 $ & $ 93 $ &  $    0.9 \pm   0.9 $ \\

2577: V11 & $ 12.3 \pm  1.6 $ & $ 213 \pm 125 $ & $ 42 \pm 15 $ & \nodata &  $    0.2 \pm   0.1 $ \\

2578: V12 & $ 21.9 \pm  1.0 $ & $ 42 \pm 8 $ & $ 220 \pm 30 $ & $ 51 $ &  $    5.3 \pm   1.3 $ \\

2579: V13 & $ 16.5 \pm  1.5 $ & $ 85 \pm 28 $ & $ 87 \pm 20 $ & $ 70 $ &  $    1.0 \pm   0.5 $ \\

2580: V14 & $ 22.6 \pm  1.0 $ & $ 17 \pm 5 $ & $ 107 \pm 11 $ & $ 70 $ &  $    6.0 \pm   1.6 $ \\

2581: V15 & $ 17.8 \pm  1.5 $ & $ 93 \pm 23 $ & $ 145 \pm 43 $ & \nodata &  $    1.5 \pm   0.7 $ \\

2582: V16 & $ 20.2 \pm  2.0 $ & $ 39 \pm 16 $ & $ 117 \pm 34 $ & \nodata &  $    2.9 \pm   1.6 $ \\

2583: V17 & $ 19.0 \pm  1.1 $ & $ 651 \pm 171 $ & $ 1400 \pm 280 $ & \nodata &  $    2.3 \pm   0.8 $ \\

2584: V18 & $ 30.3 \pm  1.5 $ & $ 148 \pm 24 $ & $ 4590 \pm 740 $ & $ 58 $ &  $   30.9 \pm   9.3 $ \\

2585: V19 & $ 20.0 $ & $ 281 \pm 67 $ & $ 890 \pm 210 $ & \nodata &  $    3.2 $ \\

2586: V20 & $ 16.6 \pm  3.1 $ & $ 1274 \pm 739 $ & $ 1250 \pm 590 $ & $ 75 $ &  $    0.9 \pm   0.9 $ \\

2587: V21 & $ 15.8 \pm  3.2 $ & $ 1109 \pm 762 $ & $ 920 \pm 540 $ & $ 66 $ &  $    0.9 \pm   0.9 $ \\

2588: V22 & $ 17.7 \pm  1.3 $ & $ 145 \pm 34 $ & $ 227 \pm 44 $ & $ 72 $ &  $    1.6 \pm   0.6 $ \\

2589: V23 & $ 39.4 \pm  1.7 $ & $ 526 \pm 65 $ & $ 71000 \pm 11000 $ & $ 89 $ &  $  130.4 \pm  31.6 $ \\

2590: V24 & $ 23.2 \pm  0.8 $ & $ 970 \pm 121 $ & $ 6770 \pm 640 $ & $ 68 $ &  $    6.7 \pm   1.0 $ \\

2591: V25 & $ 17.6 \pm  2.6 $ & $ 93 \pm 44 $ & $ 112 \pm 61 $ & $100 $ &  $    1.3 \pm   1.1 $ \\

2592: V26 & $ 18.9 \pm  1.1 $ & $ 150 \pm 27 $ & $ 346 \pm 62 $ & $ 58 $ &  $    2.4 \pm   0.7 $ \\

2593: V27 & $ 14.9 \pm  2.6 $ & $ 105 \pm 64 $ & $ 54 \pm 26 $ & \nodata &  $    0.6 \pm   0.6 $ \\

2594: V28 & $ 15.9 \pm  1.8 $ & $ 90 \pm 46 $ & $ 60 \pm 25 $ & $ 39 $ &  $    0.8 \pm   0.7 $ \\

2595: V29 & $ 20.0 $ & $ 23 \pm 4 $ & $ 73 \pm 14 $ & \nodata &  $    3.2 $ \\

2596: V30 & $ 25.9 \pm  1.0 $ & $ 680 \pm 115 $ & $ 9020 \pm 700 $ & $ 63 $ &  $   13.0 \pm   2.9 $ \\

2597: V31 & $ 16.8 \pm  1.7 $ & $ 43 \pm 14 $ & $ 45 \pm 22 $ & $ 50 $ &  $    1.0 \pm   0.6 $ \\

2598: V32 & $ 26.2 \pm  1.3 $ & $ 208 \pm 30 $ & $ 2820 \pm 410 $ & $ 69 $ &  $   13.7 \pm   3.7 $ \\

2599: V33 & $ 14.2 \pm  2.3 $ & $ 107 \pm 64 $ & $ 49 \pm 21 $ & \nodata &  $    0.4 \pm   0.4 $ \\

2600: V34 & $ 20.0 $ & $ 52 \pm 6 $ & $ 166 \pm 18 $ & \nodata &  $    3.2 $ \\

2601: V35 & $ 20.7 \pm  1.2 $ & $ 22 \pm 7 $ & $ 84 \pm 19 $ & $ 54 $ &  $    4.0 \pm   1.1 $ \\

2602: V36 & $ 20.0 $ & $ 118 \pm 12 $ & $ 374 \pm 39 $ & $ 54 $ &  $    3.2 $ \\

2603: V37 & $ 22.6 \pm  0.9 $ & $ 24 \pm 6 $ & $ 139 \pm 18 $ & $ 73 $ &  $    5.8 \pm   1.5 $ \\

2604: V38 & $ 20.0 $ & $ 132 \pm 13 $ & $ 417 \pm 42 $ & $ 72 $ &  $    3.2 $ \\

2605: V39 & $ 21.7 \pm  0.6 $ & $ 78 \pm 8 $ & $ 390 \pm 32 $ & $ 70 $ &  $    5.0 \pm   0.7 $ \\

2606: V40 & $ 28.0 \pm  1.6 $ & $ 142 \pm 26 $ & $ 2670 \pm 610 $ & $ 66 $ &  $   18.3 \pm   6.9 $ \\

2607: V41 & $ 20.8 \pm  0.8 $ & $ 110 \pm 14 $ & $ 423 \pm 43 $ & \nodata &  $    3.8 \pm   0.8 $ \\

2608: V42 & $ 20.0 $ & $ 49 \pm 6 $ & $ 155 \pm 18 $ & \nodata &  $    3.2 $ \\

2609: V43 & $ 20.0 $ & $ 45 \pm 5 $ & $ 142 \pm 15 $ & $ 86 $ &  $    3.2 $ \\

2610: V44 & $ 18.3 \pm  1.3 $ & $ 53 \pm 15 $ & $ 103 \pm 21 $ & \nodata &  $    2.0 \pm   0.7 $ \\

2611: V45 & $ 20.0 $ & $ 14 \pm 4 $ & $ 46 \pm 13 $ & \nodata &  $    3.2 $ \\

2612: V46 & $ 21.4 \pm  1.7 $ & $ 37 \pm 12 $ & $ 163 \pm 35 $ & $ 66 $ &  $    4.5 \pm   1.4 $ \\

2613: V47 & $ 20.2 \pm  2.9 $ & $ 219 \pm 70 $ & $ 830 \pm 310 $ & $ 61 $ &  $    3.4 \pm   2.3 $ \\

2614: V48 & $ 22.5 \pm  1.7 $ & $ 25 \pm 7 $ & $ 154 \pm 34 $ & \nodata &  $    5.8 \pm   2.3 $ \\

2615: V49 & $ 20.0 $ & $ 64 \pm 6 $ & $ 203 \pm 18 $ & $ 58 $ &  $    3.2 $ \\

2616: V50 & $ 21.1 \pm  2.1 $ & $ 25 \pm 10 $ & $ 101 \pm 31 $ & $ 72 $ &  $    4.0 \pm   2.4 $ \\

2617: V51 & $ 16.6 \pm  3.4 $ & $ 1088 \pm 727 $ & $ 1050 \pm 570 $ & $ 70 $ &  $    1.2 \pm   1.2 $ \\

2618: V52 & $ 24.6 \pm  1.4 $ & $ 390 \pm 75 $ & $ 3850 \pm 610 $ & $ 84 $ &  $    9.7 \pm   3.1 $ \\

2619: V53 & $ 18.5 \pm  1.3 $ & $ 80 \pm 17 $ & $ 156 \pm 34 $ & $ 80 $ &  $    1.8 \pm   0.7 $ \\

2620: V54 & $ 24.9 \pm  1.3 $ & $ 210 \pm 48 $ & $ 2320 \pm 410 $ & \nodata &  $   10.5 \pm   3.2 $ \\

2621: V55 & $ 20.0 $ & $ 48 \pm 6 $ & $ 153 \pm 17 $ & $ 59 $ &  $    3.2 $ \\

2622: V56 & $ 16.1 \pm  1.3 $ & $ 180 \pm 42 $ & $ 176 \pm 49 $ & \nodata &  $    0.9 \pm   0.5 $ \\

2623: V57 & $ 20.1 \pm  1.1 $ & $ 58 \pm 9 $ & $ 181 \pm 31 $ & $ 60 $ &  $    3.1 \pm   1.0 $ \\

2624: V58 & $ 16.7 \pm  3.9 $ & $ 757 \pm 510 $ & $ 540 \pm 330 $ & \nodata &  $    0.8 \pm   0.7 $ \\

2625: V59 & $ 25.4 \pm  1.5 $ & $ 297 \pm 67 $ & $ 3400 \pm 390 $ & $ 63 $ &  $   12.2 \pm   4.4 $ \\

2626: V60 & $ 15.8 \pm  2.3 $ & $ 65 \pm 28 $ & $ 42 \pm 21 $ & $ 34 $ &  $    0.6 \pm   0.5 $ \\

2627: \enddata

2628: \tablecomments{Results from modified blackbody SED

2629:   (Equation~\ref{eq:sed}) fits to submillimeter and FIR photometry

2630:   (\S\ref{sec:coldsed}). These fits assume a dust emissivity index

2631:   $\beta=1.5$.  Cloud masses $M_\mathrm{c}$ are derived from the

2632:   amplitude $A$ of the SED using Equation~\ref{eq:mass} and distances

2633:   from \S\ref{sec:dist}.  FIR luminosities, $L_{\mathrm{FIR}}$,

2634:   are the integrated luminosities from the modified blackbody. Objects

2635:   with MIR photometry have estimates for the total bolometric

2636:   luminosity, which may be obtained by dividing the FIR luminosity by

2637:   the Luminosity Fraction. Quoted uncertainties are statistical errors

2638:   derived from Monte Carlo simulations (\S\ref{sec:sedfits}).

2639:   Other systematic errors are not included. For example, adopting

2640:   $\beta=2.0$ decreases temperatures by $\sim5$\,K, and increases

2641:   cloud masses by a factor $\sim2$ (\S\ref{sec:sedfits}). There

2642:   are also significant distance uncertainties

2643:   (\S\ref{sec:dist}), and a factor of $\sim2$ uncertainty

2644:   introduced through choices of $\kappa_0$ and $R$ in

2645:   Equation~\ref{eq:mass}. }

2646: \end{deluxetable}

2647: 

2648: \end{document}

2649: