1: % Joshua Eisner
2: % onc_1mm
3: %
4: %
5: %
6:
7: \documentclass[12pt,preprint]{aastex}
8: %\documentclass{emulateapj}
9: %\usepackage{natbib}
10: %\citestyle{aa}
11: \begin{document}
12:
13:
14: \title{Proplyds and Massive Disks in the Orion Nebula Cluster
15: Imaged with CARMA and SMA}
16:
17: \author{J. A. Eisner\altaffilmark{1,2}, R. L. Plambeck}
18: \affil{University of California at Berkeley \\
19: Department of Astronomy \\
20: 601 Campbell Hall \\
21: Berkeley, CA 94720}
22: \email{jae@astro.berkeley.edu,plambeck@astro.berkeley.edu}
23:
24: \author{John M. Carpenter, S. A. Corder}
25: \affil{California Institute of Technology \\
26: Department of Astronomy MC 105-24 \\
27: Pasadena, CA 91125}
28: \email{jmc@astro.caltech.edu,sac@astro.caltech.edu}
29:
30: \and
31:
32: \author{C. Qi, D. Wilner}
33: \affil{Harvard-Smithsonian Center for Astrophysics \\
34: 60 Garden Street, Mail Stop 42 \\
35: Cambridge, MA 02138}
36: \email{cqi@cfa.harvard.edu,dwilner@cfa.harvard.edu}
37:
38: %\author{C. Qi, C. Lada, D. Wilner}
39: %\affil{Harvard-Smithsonian Center for Astrophysics \\
40: %60 Garden Street, Mail Stop 42 \\
41: %Cambridge, MA 02138}
42: %\email{cqi@cfa.harvard.edu,clada@cfa.harvard.edu,dwilner@cfa.harvard.edu}
43:
44:
45: \altaffiltext{1}{Miller Institute for Basic Research in Science, Berkeley,
46: CA 94720}
47: \altaffiltext{2}{Steward Observatory, University of Arizona, Tucson, AZ 85721}
48:
49:
50: %
51: %For submission to {\bf ApJ}}
52:
53:
54: \begin{abstract}
55: We imaged a $2' \times 2'$ region of the Orion Nebula cluster in 1.3 mm
56: wavelength continuum emission with the recently commissioned
57: Combined Array for
58: Research in Millimeter Astronomy (CARMA) and with the Submillimeter Array
59: (SMA)\footnote{The Submillimeter Array is a joint
60: project between the Smithsonian Astrophysical Observatory and the
61: Academia Sinica Institute of Astronomy and Astrophysics, and is
62: funded by the Smithsonian Institution and the Academia Sinica.}.
63: Our mosaics include $\ga 250$
64: known near-IR cluster members, of which 36 are so-called ``proplyds''
65: that have been imaged previously with the Hubble Space Telescope.
66: We detected 40 sources in 1 mm continuum emission (one of which is
67: the BN Object), and several of them
68: are spatially resolved with our observations. 33 detected sources are
69: known near-IR cluster members, of which 11 are proplyds. The 1 mm
70: emission from the majority of detected sources appears to trace
71: warm circumstellar dust. However, for many of the proplyds, which are
72: located close to the Trapezium stars, the millimeter wavelength fluxes
73: are dominated by thermal free-free emission from hot, ionized gas.
74: %Sources detected only at mm wavelengths appear to be deeply embedded
75: %(Class 0 or Class I) protostellar objects.
76: Dust masses inferred for detected sources range from 0.01 to 0.5 M$_{\odot}$.
77: For the $\sim 225$ known near-IR cluster members not detected in our 1 mm
78: observations, images toward the positions of near-IR sources
79: were stacked to constrain the mean 1 mm flux
80: of the ensemble. The average flux is detected at the $\ga$ 4$\sigma$
81: confidence level, and implies an average disk mass of $\sim 0.001$
82: M$_{\odot}$, approximately an order of magnitude smaller than the
83: minimum mass solar nebula. Most stars in the ONC thus do not
84: appear to currently possess sufficient mass in small dust grains to form
85: Jupiter-mass (or larger) planets. Comparison with previous results
86: for younger and older regions indicates that massive disks evolve
87: significantly on $\sim$Myr timescales. We also show that the
88: percentage of stars in Orion surrounded by disks more massive than
89: $\sim 0.01$ M$_{\odot}$ is substantially lower than in Taurus, indicating
90: that environment has an impact on the disk mass distribution.
91: Disks in Orion may be truncated through photoevaporation caused by the
92: intense radiation field of the Trapezium stars, and we see marginal evidence
93: for such a scenario in the spatial distribution of massive disks
94: within the cluster. Our data show no statistically significant
95: correlation between disk and stellar masses, although we see hints of
96: a higher percentage of massive disks around lower mass stars.
97: %This is contrary to previous results for other rich clusters.
98: %Comparison of the average disk mass and number of massive dusty structures
99: %in Orion with similar surveys of the NGC 2024 and IC 348 clusters
100: %is used to constrain the evolutionary timescales of massive circumstellar
101: %disks in clustered environments.
102: \end{abstract}
103:
104:
105: \keywords{Galaxy:Open Clusters and Associations:Individual: Orion,
106: Stars:Planetary Systems:Protoplanetary Disks, Stars: Pre-Main-Sequence}
107:
108:
109: \section{Introduction}
110: The existence of protoplanetary disks around young stars is now firmly
111: established. High resolution images from optical to radio wavelengths
112: have shown disk-like morphologies and Keplerian rotation profiles around
113: a number of young stars \citep[e.g.,][]{OW96,MO96,PADGETT+99,EISNER+04,KS95,
114: DUTREY+96,WILNER+00}. Moreover, observations of near-IR emission from
115: young stars in excess of that expected from their
116: stellar photospheres imply that most stars aged less than a few million years
117: possess inner circumstellar disks \citep[e.g.,][]{STROM+89,HLL01b}.
118:
119: Protoplanetary disks are the birth-sites of planetary systems, and the
120: ubiquity, properties, and lifetimes of disks constrain the
121: timescales and mechanisms of planet formation. The mass distribution of
122: protoplanetary disks is especially important since disk mass is related to the
123: mass of planets that may potentially form. For our own solar system,
124: the masses of planets and other bodies can be used to reconstruct a
125: minimum-mass solar nebula (MMSN) describing the
126: amount of solar-composition material needed to build the solar system.
127: Depending primarily on the core masses (and hence chemical compositions)
128: of Jupiter and Saturn (which are not known precisely), estimates of the
129: MMSN range from $\sim 0.01$--0.1 M$_{\odot}$ \citep{WEID+77}. Such disk masses
130: are also required by planet formation models to build giant planets on
131: timescales shorter than inferred disk lifetimes
132: \citep[e.g.,][]{HAYASHI81,ALIBERT+05}. The MMSN is thus an
133: informative benchmark against which to gauge the potential of disks around
134: other stars to form solar systems like our own.
135:
136: A widely used method \citep[e.g.,][]{BECKWITH+90}
137: for measuring disk masses is to observe emission
138: from optically thin dust, and then use assumed dust grain properties
139: to convert observed fluxes into dust masses. An assumed gas-to-dust ratio
140: is then used to estimate the total (gas+dust) circumstellar mass.
141: At short wavelengths ($\lambda \la 10$ $\mu$m), the dust in protoplanetary
142: disks is optically thick even for masses $<10^{-6}$ M$_{\odot}$.
143: Observations at sub-mm and mm wavelengths are necessary to measure
144: optically thin dust emission, and hence to determine the total mass
145: of dust in the disk.
146:
147: %To measure the mass of a disk, an optically thin tracer is needed.
148: %Ideally one would observe molecular gas directly, since it constitutes the
149: %bulk of the disk mass. However warm H$_2$, the dominant gaseous species, is
150: %difficult to observe due to its lack of dipole transitions. Using potential
151: %gaseous tracers like CO is problematic because the ratio of the tracer
152: %and the more abundant H$_2$ is not precisely known (in large part because
153: %of freeze-out of these tracers onto dust grains).
154:
155: Several investigators have carried out
156: comprehensive single-dish mm and sub-mm continuum surveys toward
157: regions of star formation comprising loose aggregates of stars:
158: Taurus \citep{BECKWITH+90,OB95,MA01,AW05}, $\rho$ Ophiuchi
159: \citep{AM94,NUERNBERGER+98,MAN98,AW07},
160: Lupus \citep{NCZ97}, Chamaeleon I \citep{HENNING+93}, Serpens
161: \citep{TS98}, and MBM 12 \citep{ITOH+03,HOGERHEIJDE+02}.
162: In Taurus, 22\% of stars surveyed appear to possess disks more massive than
163: $\sim 0.01$ M$_{\odot}$, and the median disk mass is
164: $5 \times 10^{-3}$ M$_{\odot}$ \citep{AW05}. The fraction of massive
165: disks\footnote{Here and throughout the text, ``massive disks'' refer
166: to disks with mass comparable to or greater than 0.01 M$_{\odot}$, the
167: lower range of estimates for the minimum-mass solar nebula.} and
168: the median disk mass is comparable in $\rho$ Ophiuchi \citep{AM94,AW07}.
169:
170: Low-density star forming regions are not the typical birth-sites
171: of stars; rather, most stars form in rich clusters like the Orion Nebula
172: \citep{LADA+91,LSM93,CARPENTER00,LL03}. Isotopic abundances in
173: our solar system suggest that it, too, may have formed in
174: a dense, Orion-like environment \citep[e.g.,][]{HD05,WG07}.
175: Expanding millimeter continuum surveys to include rich clusters allows the
176: determination of the frequency and evolution of massive disks in
177: typical star (and planet) formation environments. The high stellar density
178: in rich clusters also allows the assembly of good statistics since many
179: disks can be mapped at once.
180:
181: The main challenge to observing rich clusters at (sub-)mm
182: wavelengths is that very
183: high angular resolution is required to resolve individual sources
184: and to distinguish compact disk emission from the more extended emission
185: of the molecular clouds in which young clusters are typically embedded.
186: Single-aperture mm-wavelength telescopes lack sufficient angular resolution,
187: and to date, only three rich clusters have been observed with mm-wavelength
188: interferometers: the Orion Nebula cluster \citep{MLL95,BALLY+98,WAW05,EC06},
189: IC 348 \citep{CARPENTER02}, and NGC 2024 \citep{EC03}.
190:
191: These interferometric surveys of rich clusters have detected very few disks
192: with $\ga 0.01$--0.1 M$_{\odot}$ of material, in large part because of
193: limited sensitivity and areal coverage. The most recent observations of Orion
194: detected emission from several massive ($\ga 0.01$ M$_{\odot}$) disks
195: \citep{WAW05,EC06}, while upper limits from other surveys range from
196: $\sim 0.025$--0.17 M$_{\odot}$ \citep{MLL95,BALLY+98}.
197: Considering as ensembles the large numbers ($\ga 100$) of young stars included
198: in the cluster surveys allowed estimates of
199: mean disk masses of $\sim 0.002$, 0.005, and 0.005 M$_\odot$ in
200: IC 348, NGC 2024, and the ONC, respectively
201: \citep{CARPENTER00,EC03,EC06}. Thus, it appears that many stars
202: aged $\la 1$ Myr still possess massive circumstellar disks, although
203: more sensitive observations are necessary to detect directly large numbers
204: of massive disks at a range of ages, and thereby constrain the mass
205: distribution and evolutionary timescales.
206:
207: Here we present a new 1.3 mm wavelength interferometric survey of
208: the Orion Nebula cluster (ONC), a young, embedded
209: stellar cluster that includes the bright, massive Trapezium stars.
210: Our observations make use of the Submillimeter Array (SMA) and the recently
211: commissioned Combined Array for Research in Millimeter Astronomy (CARMA).
212: The combination of data from these two instruments yields a map of
213: the mm-wavelength continuum emission in Orion with unprecedented
214: sensitivity, angular resolution, and image fidelity.
215:
216: The Trapezium region contains hundreds of stars within a several
217: arcminute radius, and pre-main-sequence
218: evolutionary models \citep[e.g.,][]{DM94} fitted to spectroscopic
219: and/or photometric data indicate that most stars are less than
220: approximately one million years old \citep[e.g.,][]{PROSSER+94,HILLENBRAND97}.
221: The standard deviation in the distribution of
222: inferred stellar ages is $\la 1$ Myr \citep{HILLENBRAND97}.
223: Our observations thus provide a snapshot of millimeter emission around
224: a large number of roughly coeval young stars.
225:
226: With the large number of stars in the ONC, we can also investigate the
227: correlation of disk properties with stellar and/or environmental
228: properties. Previous investigations of near-IR excess emission
229: showed the inner disk fraction for stars in Orion
230: to be largely independent of stellar age and mass, although there are
231: indications of a paucity of disks around very massive stars
232: \citep{HILLENBRAND+98,LADA+00}. In addition, the inner disk fraction may
233: decrease at larger cluster radii \citep{HILLENBRAND+98}.
234: The millimeter observations presented here enable investigation of
235: how such stellar and environmental properties correlate with disk mass.
236:
237: We adopt a distance to the ONC of 400 pc, based on
238: recent trigonometric parallax measurements of several stars
239: \citep{SANDSTROM+07,MENTEN+07} and orbital fitting for a spectroscopic
240: binary \citep{KRAUS+07}. This is substantially lower than the value
241: of 480 pc computed based on statistical parallax of water maser spots
242: \citep{GENZEL+81}, which was adopted in previous studies of the ONC
243: \citep[e.g.,][]{EC06}. We discuss below the importance of this revised
244: distance on our results.
245:
246:
247: %In the next section, we describe the ONC region and discuss
248: %the stellar and protostellar populations. The observations and results are
249: %presented in \S \ref{sec:obs} and \S \ref{sec:cont_o2},
250: %and we derive constraints on the
251: %circumstellar disk masses in \S \ref{sec:disks}. Finally, we
252: %compare the results for the ONC to those
253: %for NGC 2024, IC 348, and Taurus, and discuss the implications for
254: %disk evolution in rich clusters.
255:
256: %\section{The Orion Nebula Cluster\label {sec:onc}}
257:
258:
259: \section{Observations and Data Reduction \label{sec:obs}}
260:
261: \subsection{CARMA Observations and Calibration \label{sec:carma_obs}}
262: We mosaicked a $2' \times 2'$ region toward the
263: ONC in $\lambda$1.3 mm continuum with CARMA between October and December, 2007.
264: CARMA consists of six 10-m antennas and nine 6-m antennas situated at
265: 2200 meters elevation at Cedar Flat in the Inyo Mountains of California.
266: With a total of 15 antennas, CARMA provides 105 baselines, enabling excellent
267: coverage of the uv plane and hence high image fidelity. Two different array
268: configurations (`C' and `B') were used to obtain antenna separations
269: ranging from 30 to 946 meters.
270:
271: Continuum data were recorded in six $\sim 500$ MHz bands covering the
272: frequency ranges 221.75--223.25 GHz and 227.25--228.75 GHz from the
273: receivers' lower and upper sidebands, respectively. Each band
274: consists of 15 channels. Spectral line emission in Orion is mostly resolved
275: out by these observations; that is, across most of the mapped region
276: it is detected only weakly because
277: it is spatially extended relative to the interferometer fringe
278: spacings. The spectral lines most visible on the shortest baselines,
279: mostly toward the BN/KL region in the northwest corner of the mosaic, are the
280: 11(1,11)-10(0,10) transition of SO$_2$ at 221.965 GHz and the 25-24
281: transition of HC$_3$N at 227.402 GHz. Even these lines are almost
282: completely resolved out for projected antenna spacings $> 70$
283: k$\lambda$ (used to generate the final maps; see \S \ref{sec:mapping}).
284: We therefore assume our bands are effectively line-free.
285:
286: The mosaic consists of 16 pointing centers (Figure \ref{fig:pointings}),
287: separated by
288: $26''$. This separation is comparable to the FWHM beamwidth of the 10-m
289: antennas, but $\sqrt{3}$ smaller than the FWHM beamwidth of the 6-m
290: antennas. A 2-D mosaic is Nyquist sampled if pointings are separated
291: by $\le $FWHM/$\sqrt{3}$. Thus the CARMA mosaic is Nyquist sampled only for
292: the 6-m dishes. However, Nyquist sampling is not crucial since we
293: are interested only in compact sources, rather than extended emission.
294: Simulated CARMA mosaics of a synthetic star field showed that $26''$
295: spacings provided the best balance of sensitivity and areal
296: coverage to maximize the number of sources detected.
297:
298: Each night we interleaved 16-minute observations of the ONC mosaic,
299: with 1 minute integration time on each pointing center, with 3-minute
300: observations of the phase and amplitude calibrator, J0530+135.
301: Observing the mosaic in its entirety every 20 minutes and repeating
302: this multiple times during the night ensures a high quality
303: synthesized beam and equal sensitivity for each pointing center.
304: The total integration time for the maps was 34 minutes per pointing
305: center.
306:
307: Telescope pointing was checked and updated on 20
308: minute intervals using optical counterparts very near to the source.
309: These objects were observed with optical cameras mounted on the dishes,
310: and radio--optical offsets were calibrated periodically.
311: Gain stability, especially in 1 mm observations like those presented here, is
312: found to be enhanced through this method (Corder, Carpenter
313: \& Wright, in prep).
314:
315: We measured a 1.3 mm flux density of $3.0 \pm 0.3$ Jy for J0530+135
316: using Uranus as a primary flux calibrator, based on observations
317: during several nights in October when both sources were observed.
318: J0530+135 was also used to calibrate the passband.
319: All calibrations
320: for these data were performed with the MIRIAD package \citep{STW95}.
321:
322: \subsection{SMA Observations and Calibration \label{sec:sma_obs}}
323: We mosaicked a triangular region approximately $1\rlap{.}'7$ on a side,
324: consisting of three pointings, with the SMA between September, 2005, and
325: February, 2006
326: (Figure \ref{fig:pointings}).
327: The SMA consists of eight 6-m dishes near the summit of Mauna Kea in
328: Hawaii. However, for our observations only seven antennas were available,
329: providing 21 baselines between 10 and 220 m.
330: The $uv$ coverage for
331: our SMA observations is substantially sparser than for our CARMA observations,
332: and thus the image fidelity is worse (which means, for example, that
333: strong emission is scattered more strongly into other regions of the map).
334: However, the high altitude of
335: the SMA enables very low opacity observing conditions. The low opacities,
336: combined with increased observing time per pointing, lead to substantially
337: better sensitivity relative to the CARMA observations.
338:
339: Double sideband receivers were tuned to a local oscillator (LO) frequency of
340: 225.333 GHz. The SMA digital correlator is configured with 24 partially
341: overlapping bands of 104 MHz width in each sideband. Each sideband
342: provides 2 GHz of bandwidth, centered $\pm 5$ GHz away from the LO frequency.
343: The double-sideband
344: (DSB) system temperatures were between 80 and 200 K.
345:
346: As for the CARMA mosaic, we
347: observed the mosaic multiple times throughout the night, obtaining
348: equal integration times for each pointing position.
349: The pointings in the mosaic were separated by $\sim 44''$, the approximate
350: FWHM for the SMA dishes at this observing wavelength. As for the CARMA
351: map, the SMA mosaic is larger-than-Nyquist sampled, which provides
352: enhanced areal coverage compared to a Nyquist-sampled mosaic.
353:
354: We used J0423-013 and 3C120 as gain calibrators for these tracks, with
355: flux densities of 1.4 and 1.0 Jy, respectively, derived using Uranus
356: as a primary flux calibrator. We estimate that the absolute flux scale is
357: uncertain by $\sim 10\%$. We calibrated the passband
358: using the quasar 3C454.3, and Uranus where available.
359: Because the CO(2-1) transitions are present in
360: the observing window, we edited out the parts of the band with
361: strong lines and
362: generated a line-free continuum channel. All calibrations were performed
363: using the SMA adaptation of the IDL-based data reduction package MIR
364: developed at Caltech;
365: calibrated data were then converted into MIRIAD format for further processing.
366:
367: \subsection{Mapping \label{sec:mapping}}
368: We made mosaics of our CARMA and SMA datasets individually, and after combining
369: the two datasets in the $uv$ plane. For the individual and combined
370: datasets, we mosaicked the individual pointings
371: into a single image, weighting the data by system temperature and by $uv$
372: distance (with a ``robust'' parameter of 0.5), then de-convolved and
373: CLEANed (all using MIRIAD). The angular resolution afforded by
374: the longest baseline data in our maps (from the CARMA B-array)
375: is $\sim 0\rlap{.}''3$. Our mosaics have $0\rlap{.}''1$
376: pixels, which ensures adequate sampling of individual resolution elements.
377:
378: Since we are primarily interested in compact disk emission, we
379: eliminated $uv$ spacings shorter than 70 k$\lambda$ (i.e., projected
380: baselines shorter than 93 m) in order to reduce
381: contamination from bright extended emission. The eliminated spacings
382: correspond to size scales larger than $\sim 3''$. The cutoff value was chosen
383: to minimize the RMS background noise in the CLEANed images; we measured
384: the RMS for $uv$ cutoff radii of 50, 60, 70, and 80 k$\lambda$, and found
385: the 70 k$\lambda$ cutoff to be optimal.
386:
387: Mosaics produced from our robust-weighted data with $r_{uv} > 70$ k$\lambda$
388: are shown in Figures \ref{fig:mos}--\ref{fig:map_combo}.
389: In the Figures we have divided by the theoretical sensitivity at each location
390: in the image, in order to visually down-weight the noisier edges of
391: the mosaic (where there are fewer overlapping pointings); we do not
392: divide by the sensitivity in the analysis presented below.
393: We note that even within the uniform (theoretical)
394: sensitivity region of the mosaic,
395: the RMS varies substantially because of emission scattered from bright compact
396: and extended sources in the BN/KL and OMC1-S regions located in the upper and
397: lower right quadrants of the maps.
398:
399: For the SMA mosaic, the unit gain region (within which the theoretical
400: sensitivity does not vary substantially) encompasses a roughly triangular
401: region covering $\sim 2$ square arcminutes. The RMS of pixels
402: within ``clean'' regions of the unit gain contour (i.e., away from the
403: crowded BN/KL and OMC1-S regions) is 0.8 mJy.
404: The unit gain region of the CARMA mosaic
405: encompasses a $2' \times 2'$ area, with an RMS noise level (again, in
406: clean regions of the map) of 2.3 mJy.
407: For the combined map, which will be used for the bulk of our analysis,
408: the unit gain region is slightly larger than for the CARMA-only mosaic, and
409: the RMS noise level is 1.8 mJy. The synthesized beam has dimensions (FWHM) of
410: $0\rlap{.}''69 \times 0\rlap{.}''60$ at a position angle of 72$^{\circ}$.
411:
412: \section{Analysis and Results \label{sec:analysis}}
413:
414: \subsection{Detection Thresholds \label{sec:thresh}}
415: Because the map contains a large number of pixels, we must employ a fairly
416: high detection threshold to avoid random noise spikes if we search the image
417: blindly. The mosaic area is approximately $35,000$ synthesized beams. With
418: this number of independent pixels, one expects $> 1$ noise spike above
419: the 4$\sigma$ level (assuming Gaussian noise).
420: We therefore use a 5$\sigma$ detection limit, at which level $\ll 1$ pixels are
421: expected to show noise spikes. Because the noise varies greatly across the
422: map, we calculate $\sigma$ locally in small sub-regions of the image.
423:
424: Specifically, a ``local'' $\sigma$ is computed
425: in $10'' \times 10''$ ($100 \times 100$ pixels) boxes around each pixel
426: in the mosaic. For our detection thresholds to be meaningful, the
427: noise must be well-characterized. However, poorly sampled extended emission
428: leads to excess noise in the BN/KL and OMC1-S regions. Moreover, the
429: noise increases toward the edges of the unit gain region because there
430: are fewer overlapping mosaic pointings there.
431: Detections in these areas should be treated with some caution.
432:
433: As a test of our detection threshold, we searched the maps
434: for false detections below the $-5\sigma$ level. None were detected,
435: confirming that 5$\sigma$ is a reasonable detection limit.
436: In contrast, 28 sources were seen below the $-4\sigma$ level
437: (most of them toward the edges of the mosaic or in the BN/KL and OMC1-S
438: regions), demonstrating that 4$\sigma$ is not a sufficiently stringent
439: detection threshold (and that the noise across our mosaic is not
440: always Gaussian).
441:
442: Instead of blindly searching for detections, we can also use our prior
443: knowledge of the locations of near-IR cluster members and search only these
444: positions. For these $\sim 250$
445: pre-determined positions, $\sim 0.3$ sources are expected to show
446: emission above the 3$\sigma$ level from Gaussian noise.
447: We can therefore try a 3$\sigma$ detection threshold,
448: where $\sigma$ is the noise determined locally
449: (as above) in $10'' \times 10''$ sub-regions
450: centered on individual cluster member positions. Although the
451: noise in the mosaic is not always Gaussian, this 3$\sigma$ threshold
452: appears reasonable: none of the near-IR source positions
453: were detected below the $-3\sigma$ level.
454:
455: Sources with 1 mm continuum emission at the $>3 \sigma$ level in our maps
456: are deemed to coincide with near-IR cluster members if the mm peaks
457: and near-IR source positions lie within $0\rlap{.}''4$ of each other.
458: The estimated relative positional accuracy of $0\rlap{.}''4$ is the
459: quadrature sum of uncertainties from centroiding the mm images
460: ($\sim 0.5 \theta_{\rm beam}/$signal-to-noise $\approx 0\rlap{.}''1$),
461: uncertainties in the absolute astrometry due to baseline errors
462: ($\sim 0\rlap{.}''2$), and
463: uncertainties in the near-IR source positions ($\sim 0\rlap{.}''3$).
464:
465: We detected 19 sources within the unit gain contour of
466: our mosaic above the 5$\sigma$ level (Table
467: \ref{tab:detections}). 12 of these are coincident with near-IR cluster
468: members listed in \citet{HC00}.
469: An additional 21 objects were detected above the 3$\sigma$
470: level toward positions of near-IR sources. 1 mm continuum images of
471: sources detected in our mosaic are displayed in Figure \ref{fig:detections}.
472:
473: While the BN object is detected, we defer discussion of this
474: high-mass, embedded object \citep[e.g.,][]{GBW98,PLAMBECK+95} to a
475: later paper that examines the BN/KL region in detail\footnote{Although
476: Source I is detected as a
477: strong, individual object in our CARMA B-array data, we do not detect it in
478: our combined SMA+CARMA mosaic because of confusion with the hot core. Source
479: I will also be discussed in the later paper.}.
480: In the remaining discussion, we focus our attention on the sources detected
481: at both infrared and millimeter wavelengths. While most of these
482: objects do not have
483: known stellar masses, they are likely to be low-mass stars based on
484: the stellar mass distribution computed statistically for the ONC as a whole
485: \citep{HC00}.
486:
487: %There are,
488: %however, exceptions to this argument.
489: %Massive stars may be surrounded by hotter dust than low-mass stars, and
490: %for these objects smaller dust masses (with smaller associated extinction of
491: %near-IR light) can produce the observed emission. Thus, we cannot
492: %necessarily infer a flattened distribution of dust around $\theta^1$OriA
493: %or BN based on the observed 3 mm fluxes.
494: %However, most of the stars in our observations have masses
495: %between 0.1 and 1 M$_\odot$ \citep[e.g.,][]{HC00}, and thus the extinction
496: %argument generally applies, suggesting that most objects detected in both
497: %millimeter and near-IR emission trace flattened structures.
498:
499: \subsection{Angular Sizes of Detected Objects \label{sec:sizes}}
500: For each source detected in our 1 mm mosaic,
501: we fitted a 2-D elliptical Gaussian
502: to the emission. The synthesized, clean, beam for the combined mosaic is a
503: 2-D Gaussian with FWHM of $0\rlap{.}''69 \times 0\rlap{.}''60$ at a position
504: angle (north of west) of $72^{\circ}$. At the assumed 400 pc distance to the
505: ONC, the core of the synthesized beam
506: has dimensions of 240 AU by 280 AU (again at a PA of $72^{\circ}$).
507: For simplicity, we approximate this as 240 AU in the East-West direction
508: and 280 AU in the North-South direction.
509:
510: For sources detected at a signal-to-noise ratio of $\sim 5$, the
511: statistical uncertainty
512: in the fitted FWHM is $\sim 10\%$. Baseline errors or phase noise in our
513: mosaics can broaden the apparent source sizes, however. We assume
514: that a source is resolved only if the major or minor axis
515: of the fitted FWHM is 25\% larger than that of the synthesized beam.
516: Objects for
517: which the fitted Gaussian FWHM is smaller are considered to be unresolved.
518:
519: Approximately 25\% (9/39) of detected sources are spatially resolved in
520: our images (Table \ref{tab:results}).
521: An additional nine objects (all proplyds) have been spatially
522: resolved with HST \citep{VA05}. Thus, angular sizes are available for
523: $\sim 50\%$ of our sources. The inferred radii for
524: resolved sources range from $\sim 90$ to $\sim 220$ AU.
525: For unresolved objects we can say only that the emission is
526: confined to radii smaller than\footnote{Larger sources, when
527: convolved with the synthesized beam, would produce measured sizes $>25\%$
528: broader than the beam.} $\sim 100$ AU.
529: For a sample of 134 proplyds with sizes measured with HST,
530: the mean disk radius is 71 AU \citep{VA05}. Since the mean disk diameter is
531: $\sim 1/2$ the size of the linear resolution of our observations, it is not
532: surprising that most of the sources detected in the 1 mm mosaic are unresolved.
533:
534: \subsection{Distinguishing Dust and Free-Free Emission \label{sec:ff}}
535: Since we are interested in using our observations to constrain the mass of
536: circumstellar dust around our sources, we must account for potential
537: contributions to the observed fluxes from sources other than dust emission.
538: Free-free emission arises in hot ionized gas, and in the ONC
539: such conditions may exist either in HII regions around high-mass stars
540: \citep[e.g.,][]{GMR87,PLAMBECK+95} or in the outer regions of disks or
541: envelopes that are
542: irradiated by the hot Trapezium stars \citep[e.g.,][]{OWH93,HO99}.
543: While some sources in the ONC have shown flares of cyclotron emission
544: \citep[e.g.,][]{BOWER+03,FURUYA+03},
545: we expect that such rare events will not contribute
546: significantly to the 1.3 mm fluxes, and we do not consider them here.
547: %Our 1 mm fluxes may thus trace some free-free emission from hot gas
548: %in addition to thermal continuum emission from cool dust.
549:
550: Because the spectral shape of free-free radiation differs
551: from that of thermal dust emission, comparing
552: 1 mm measurements with longer-wavelength data enables us to
553: distinguish these components.
554: We use long-wavelength fluxes from the literature
555: \citep{FELLI+93a,FELLI+93b,MLL95,ZAPATA+04,EC06,FMR07}.
556: In addition, we use 880 $\mu$m fluxes measured by \citet{WAW05} for the few
557: objects where these are available.
558: %For the BN object and for Source I (IRC 2), we also use the long wavelength
559: %fluxes compiled by \citet{PLAMBECK+95}.
560:
561: For a freely expanding, fully ionized wind with constant
562: $\dot{M}$, such as we expect for proplyds,
563: free-free emission will have the following spectrum\footnote{In a previous
564: paper \citep{EC06}, we assumed that free-free emission originated from
565: static HII regions rather than from winds, as in the present work.
566: This choice affects only the long-wavelength behavior of the free-free
567: spectrum, and is relatively unimportant to our analysis.}:
568: \begin{equation}
569: F_{\rm \nu, ff} = \cases{F_{\rm \nu, turn} (\nu / \nu_{\rm turn})^{-0.1} &
570: if $\nu \ge \nu_{\rm turn}$ \cr
571: F_{\rm \nu , turn} \left({\nu} / {\nu_{\rm turn}}\right)^{0.6} &
572: if $\nu \le \nu_{\rm turn}$}.
573: \label{eq:ffwind}
574: \end{equation}
575: Here, $\nu_{\rm turn}$ is the frequency above which the wind is
576: optically thin at all radii.
577: We include a derivation of this result
578: in the appendix, and an alternative derivation can be found in \citet{WB75}.
579:
580: For $\nu < \nu_{\rm turn}$, the inner
581: parts of the wind are optically thick to free-free radiation.
582: If we adopt a simple model for proplyd winds where
583: $\dot{M}=10^{-7}$ M$_{\odot}$ yr$^{-1}$ with spherical wind velocity of 20 km
584: s$^{-1}$ \citep[e.g.,][]{HO99}, and $T_{\rm e} = 10^4$ K, we
585: can estimate the size of the optically thick region. Using Equation
586: \ref{eq:xff}, we obtain $x_{\tau\approx 1} \sim 1$ AU at $\lambda$1 mm and
587: $\sim 30$ AU at $\lambda$10 cm. Even at 10 cm, this is smaller
588: than the likely wind-launching regions
589: for proplyds, and free-free emission from most proplyds is likely to be fully
590: optically thin. For the highest measured mass loss rates of
591: $\sim 10^{-6}$ M$_{\odot}$ yr$^{-1}$ \citep[e.g.,][]{HO99}, the optically thick
592: regions of proplyd winds are $\sim 130$ AU for $\lambda \sim 10$ cm.
593: We therefore do not expect to see a spectral turnover
594: (Equation \ref{eq:ffwind}) for wavelengths $\la 5$ cm.
595:
596: %We could also fit a free-free emission model to the data where
597: %\begin{equation}
598: %F_{\rm \nu, ff} = F_0 \nu^{\alpha},
599: %\label{eq:ffother}
600: %\end{equation}
601: %where $F_0$ and $\alpha$ are free parameters. Ideally, we would combine
602: %also include a turnover frequency in the model, but we typically lack
603: %lack enough data points to constrain larger numbers of free parameters.
604: %For high mass protostellar sources, this may not be an issue: since the
605: %ionization source may be the central star, even high-density circumstellar
606: %material at small stellocentric radii can be ionized, and this high
607: %density gas is more likely to be optically thick even at high frequencies.
608:
609: Emission from cool dust is added to this free-free emission to obtain
610: a model of the observed flux. We assume that
611: \begin{equation}
612: F_{\rm \nu, dust} = F_{\rm 230 GHz, dust} (\nu / 230 {\rm \: GHz})^{(2+\beta)}
613: = F_{\rm 230 GHz, dust} (\nu / 230 {\rm \: GHz})^{3},
614: \label{eq:dust}
615: \end{equation}
616: for $\beta=1$ \citep[e.g.,][]{BECKWITH+90}. Other values of $\beta$ can
617: not be ruled out based on our data for most objects, and $\beta=0$,
618: corresponding to emission from optically thick or large-grained dust,
619: is typically compatible with the data.
620:
621: We estimate the relative contributions of dust and free-free emission
622: by fitting this model, $F_{\nu} = F_{\rm \nu, ff}+F_{\rm \nu, dust}$, to our
623: measured 1.3 mm fluxes and to 880 $\mu$m,
624: 3 mm, 3.6 mm, 1.3 cm, 2 cm, 6 cm, and 20 cm fluxes
625: from the literature. For comparison, we also fit a dust-only model,
626: described by Equation \ref{eq:dust}, with $\beta=1$ and $\beta=0$.
627: For objects detected at centimeter wavelengths, we fit the
628: dust+free-free model to the $\ge 4$ flux measurements
629: for each source, and thus we are able to determine the three free parameters
630: of the model, $\nu_{\rm turn}$, $F_{\rm \nu, turn}$, and $F_{\rm \nu, dust}$.
631: For sources with $\le 3$ flux measurements (i.e., those undetected
632: in centimeter wavelength surveys),
633: we fit only the dust emission model to the data.
634:
635: Given the noise level of previous centimeter observations covering
636: the entire region of our 1 mm mosaic
637: \citep[$\la 0.3$ mJy;][]{FELLI+93a,FELLI+93b} and the measured
638: 1 mm fluxes for detected objects ($\ga 10$ mJy), sources undetected at
639: cm wavelengths are probably dominated by dust emission.
640: For a source with 10 mJy flux at 1 mm, a non-detection at 10 cm
641: implies that $\la 0.2$ mJy, or $\la 2\%$ of the measured 1 mm flux is due
642: to free-free emission. For simplicity, we attribute 100\% of the 1 mm fluxes
643: to dust emission for these objects.
644:
645: Fluxes, from sub-millimeter to radio wavelengths, and models are plotted in
646: Figure \ref{fig:seds}, and the
647: fluxes due to thermal dust emission
648: are listed in Table \ref{tab:results}. Uncertainties for these
649: dust fluxes are given by
650: the 1$\sigma$ uncertainties of the model fits.
651: The majority of detected sources appear to be dominated by dust emission.
652: However, for the subset of the sample seen in optical emission
653: or absorption with HST (the proplyds at the top of Table
654: \ref{tab:detections}), the 1 mm fluxes are dominated by
655: free-free emission. This probably reflects the relative proximity of
656: proplyds to the luminous Trapezium stars.
657:
658: \subsection{Estimating Circumstellar Dust Masses \label{sec:massest}}
659: The mass of circumstellar dust is related to the component of the 1 mm
660: continuum flux due to dust emission. Assuming the dust is
661: optically thin, and following \citet{HILDEBRAND83},
662: \begin{equation}
663: M_{\rm dust} = \frac{S_{\rm \nu,dust} d^2}
664: {\kappa_{\rm \nu,dust} B_{\nu}(T_{\rm dust})}.
665: \label{eq:dustmass}
666: \end{equation}
667: Here, $\nu$ is the observed frequency,
668: $S_{\rm \nu,dust}$ is the observed flux due to cool dust, $d$ is the distance
669: to the source,
670: $\kappa_{\rm \nu, dust} = \kappa_0 (\nu / \nu_0)^{\beta}$ is the
671: dust mass opacity,
672: $T_{\rm dust}$ is the dust temperature, and $B_{\nu}$ is the Planck function.
673: We assume $d \approx 400$ pc,
674: $\kappa_0=0.0002$ cm$^{2}$ g$^{-1}$ at 1.3
675: mm, $\beta=1.0$ \citep{HILDEBRAND83,BECKWITH+90}, and $T_{\rm dust} = 20$ K
676: (based on the average dust temperature inferred for Taurus; Andrews \&
677: Williams 2005; see also the discussion in Carpenter 2002;
678: Williams et al. 2005).
679: The dust mass can be converted into a total circumstellar mass by assuming
680: the canonical gas-to-dust mass ratio:
681: $M_{\rm circumstellar} = M_{\rm dust} \times 100$.
682: Column 3 of Table \ref{tab:results} lists the estimated circumstellar
683: masses for detected objects.
684:
685: Uncertainties in the assumed values of these parameters (notably $\kappa$)
686: lead to uncertainties in the derived masses (in an absolute sense)
687: of at least a factor of $3$ \citep[e.g.,][]{POLLACK+94}, which are not
688: included in the uncertainties listed in the table. Values of
689: $\kappa_{\rm \nu,dust}$ and $T_{\rm dust}$ may also vary across our
690: sample. Since the cluster population in Orion is roughly co-eval
691: \citep[e.g.,][]{HILLENBRAND97}, such effects should be minimal. However,
692: there is some spread in stellar masses, which may lead to some range
693: in these parameters. For example, since some of the objects in Table
694: \ref{tab:detections}
695: may be massive stars, the millimeter flux may contain contributions
696: from dust hotter than the assumed 20 K;
697: %The high luminosity of the BN object
698: %($\sim 2500$ L$_{\odot}$) implies a spectral type of B3-B4 \citep{GBW98},
699: %and the luminosity of $\theta^1$OriA implies a spectral type of O7
700: %\citep{WH77}. In addition, IRc2 is probably a massive star, and the
701: %high luminosity of nearby ``source I'' provides additional dust heating
702: %\citep{GBW98}. For these sources,
703: %the computed dust masses are over-estimated, since smaller masses of
704: %hotter dust can produce the observed 3 mm emission
705: if $T_{\rm dust}=30$ K, then the computed dust masses would be
706: lower by a factor of 1.6.
707: %The circumstellar masses listed in Table \ref{tab:detections}
708: %for BN (HC 705), $\theta^1$OriA (HC 336), and IRc2 (MM 3)
709: %should therefore be treated as upper limits.
710: For the predominantly low-mass cluster population \citep{HC00}, from which
711: the sources listed in Table \ref{tab:detections} are drawn,
712: the assumed values for $\kappa_{\rm \nu, dust}$ and $T_{\rm dust}$
713: presumably do not vary much, and the masses predicted by Equation
714: \ref{eq:dustmass} are probably valid in a relative sense to within a
715: factor of two.
716:
717: \subsection{Constraints on Dust Optical Depth \label{sec:tau}}
718: %In \S \ref{sec:massest}, we used measured 1 mm fluxes to estimate
719: %circumstellar dust masses, under the assumption that the material is
720: %optically thin.
721: %Our modeling of the relative dust and free-free contributions to the
722: %measured fluxes also depended on this assumption. While we assumed $\beta=1$,
723: %if $\tau_{\rm dust} \ga 1$, then $\beta=0$, since the dust simply emits as a
724: %blackbody. It is therefore prudent to examine the assumption that
725: %$\tau_{\rm dust} < 1$ at 1 mm wavelengths.
726:
727: We perform a simple test to determine whether the optical depth ($\tau$)
728: might become
729: comparable to or larger than unity in the systems under study.
730: Using disk sizes measured from
731: either our data or HST data \citep{VA05}, or limits on disk sizes from
732: our observations, we compute the emission expected from optically thick
733: dust with a temperature of 20 K:
734: \begin{equation}
735: F_{\nu, \tau \ga 1} = B_{\nu} (T_{\rm dust})
736: \pi \left(\frac{R_{\rm disk}}{d}\right)^2
737: \cos i \approx 112 {\rm \: mJy} \left(\frac{R_{\rm disk}}{\rm 100 \: AU}
738: \right)^2.
739: \end{equation}
740: Here $R_{\rm disk}$ is the disk radius, $i$ is the inclination and
741: $d$ is the distance. For simplicity, we take $i=0$, which leads to an
742: upper limit on the flux for an optically thick disk of radius $R_{\rm disk}$.
743: Fluxes expected for optically thick dust for our sample are
744: listed in Table \ref{tab:results}.
745:
746: For most objects, the fluxes (or upper limits) expected for optically thick
747: dust are substantially higher than our measured fluxes
748: (Table \ref{tab:results}). These disks
749: must be either highly inclined
750: or optically thin. Since it is unlikely that all of the disks detected
751: in our observations are edge-on (especially since these would be the dimmest
752: portion of a sample of randomly inclined, optically thick disks),
753: we take this as evidence for optically thin material.
754: Although many of the disk sizes, and hence the expected optically thick
755: fluxes, are upper limits, the mean radius for proplyds in Orion of
756: $\sim 70$ AU \citep[e.g.,][]{VA05} would produce an optically thick flux of
757: $\sim 50$ mJy, still higher than the majority of our measured fluxes.
758:
759: We therefore believe that most disks detected in our observation
760: are composed largely of optically thin dust, and the circumstellar dust masses
761: computed in \S \ref{sec:massest} are reasonable for these sources.
762: There are a few among the sources detected only at $\ga $ mm
763: wavelengths (MM8, MM13, and MM21) for which the measured fluxes are
764: comparable to or larger than the expected optically thick fluxes. The
765: dust in these objects is either optically thick or hotter than 20 K,
766: as may occur around higher-mass (proto-)stars.
767:
768: \subsection{Stacking Analysis \label{sec:stacking}}
769: With the large number of young stars contained within our
770: mosaic, we can enhance the effective sensitivity by considering the
771: ensemble of $\sim 225$ sources not detected individually.
772: For each known cluster member within the mosaic that is not detected
773: above the 3$\sigma$ noise level, we make a $10'' \times 10''$ sub-image
774: centered on the stellar position. We weight the sub-image by the
775: local RMS (determined as described in \S \ref{sec:thresh}), sum all
776: of the weighted images, and divide by the sum of the weights to
777: normalize.
778: We exclude any cluster members known to have radio-wavelength
779: emission \citep{FELLI+93a,FELLI+93b}.
780:
781: The weighted image is shown in Figure \ref{fig:avg}. The average flux
782: for the ensemble of non-detected sources is
783: $0.9 \pm 0.2$ mJy, with a significance of $> 4\sigma$.
784: The peak flux is centered on the mean position
785: of the near-IR sources (within the positional uncertainties of
786: $\sim 0\rlap{.}''4$), and resembles the synthesized beam core, indicating
787: that the average source is compact.
788:
789: Since the positional uncertainties are comparable to the half-width
790: half-maximum of the synthesized beam, the average flux seen in Figure
791: \ref{fig:avg} may be slightly reduced because different sources in the
792: ensemble do not lie exactly atop one another. Assuming
793: the positional uncertainties are random and Gaussian-distributed, one
794: would expect a reduction in the measured average flux of $\sim 35\%$.
795: Correcting for the potential flux-smearing, one would obtain an average
796: flux for the ensemble of $1.2 \pm 0.2$ mJy. We verify this by integrating
797: the central region of the average image over a region with four times
798: the area of the synthesized beam; as expected, we find an integrated
799: flux of $\sim 1.2$ mJy.
800:
801: Low-level free-free emission may contaminate the average image, and hence
802: bias the average flux inferred for the ensemble.
803: The 1$\sigma$ sensitivity in cm-wavelength surveys is $\sim 0.3$ mJy
804: \citep[e.g.,][]{FELLI+93b}.
805: We argued in \S \ref{sec:ff} that gaseous winds in the ONC are likely to
806: be optically thin to free-free emission, and hence that
807: $F_{\rm \nu, ff} \propto \nu^{-0.1}$. Thus we would expect free-free emission
808: to be no stronger than $\sim 0.2$ mJy at 1 mm wavelengths for sources
809: undetected in cm-wavelength surveys. This is comparable to the 1$\sigma$
810: sensitivity in our 1.3 mm average image. Since the average image is detected
811: above the 4$\sigma$ level, $>75\%$ of the average flux comes from dust.
812:
813:
814: \section{Discussion \label{sec:disc_o2}}
815:
816: \subsection{Nature of Detected Sources \label{sec:geometry}}
817: The sub-arcsecond resolution of the CARMA observations is enough to
818: marginally resolve some of the detected sources in the ONC, and in
819: principle we could observe flattened, disk-like geometries.
820: For example, Figure \ref{fig:proplyds} shows that 177-341 has a
821: disk-like morphology aligned with the silhouette disk seen by HST.
822: While only a few sources can be well-resolved with our observations,
823: HST observations show that many of the observed proplyds appear disk-like
824: \citep{MO96,BALLY+98a}, some even exhibiting silhouette disks \citep{BOM00}.
825: For the proplyds and well-resolved mm sources, the 1 mm
826: emission evidently arises from disk-like distributions.
827:
828: Mid-IR emission is also observed toward many of
829: the sources detected at 1 mm. 82\% of sources
830: (all except the ``MM'' sources and HC 495) are also
831: seen at 3.6 $\mu$m \citep{LADA+04}, and 48\% are seen at 11.7 $\mu$m
832: \citep{SMITH+05}. While 3.6 $\mu$m emission may trace stellar
833: photospheres and/or infrared excess, the 11.7 $\mu$m emission
834: provides direct evidence for circumstellar material at least out to
835: radii of a few AU. Thus, many detected sources (the majority, if the ``MM''
836: objects are excluded) have evidence for inner circumstellar disks.
837:
838: More generally, where 1 mm emission is detected
839: toward known near-IR cluster members, the fact that the near-IR light
840: is visible despite the high extinctions ($A_{\rm V} \ga 300$) that one
841: would derive based on the amount of material needed to produce the 1 mm
842: emission (for spherically distributed
843: material) implies that the dust lies in flattened, disk-like
844: distributions \citep[see also, e.g.,][]{BECKWITH+90,EC03}.
845:
846: It is interesting to speculate as to the nature of sources detected at
847: $HKL$ bands and at mm wavelengths, but not at 11.7 $\mu$m. 13/32 (40\%) of
848: mm and near-IR detected sources fall into this category.
849: It is possible that some of these are transitional disks.
850: The $HKL$ emission may trace the stellar photosphere of a late-type star
851: while the mm emission traces a remnant outer disk,
852: but large inner clearings may lead to a lack of mid-IR excess.
853: Better coverage of the wavelength range between 10 $\mu$m and
854: 1 mm is needed to test this hypothesis.
855:
856: %Despite the evidence that most detected sources are circumstellar disks,
857: %in the absence of kinematic information we can not be certain that flattened
858: %structures observed around these stars are
859: %geometrically thin disks in Keplerian rotation, as opposed to flattened
860: %disk+envelope structures such as those inferred around some Class I objects
861: %\citep[e.g.,][]{EISNER+05b}.
862:
863: For sources without near-IR detections, the ``MM'' sources in
864: Table \ref{tab:detections}, the arguments presented above do not apply.
865: Although the emission appears to trace circumstellar dust, the fact that
866: no near-IR counterparts are observed suggests high columns of obscuring
867: material. The MM sources all lack
868: mid-IR counterparts as well. These objects appear to be so embedded that they
869: are still highly obscured even at 11.7 $\mu$m.
870: All of the MM sources
871: reside in either the Orion BN/KL or OMC1-S region, both of which
872: are known to contain young, embedded sources, HII regions, and outflows
873: \citep[e.g.,][]{ZWM90,BACHILLER96,ZAPATA+04}.
874:
875: All of the sources in OMC1-S (LMLA162, MM8, MM13, MM21, and MM22)
876: have been detected at 1.3 mm wavelength in previous observations
877: \citep{ZAPATA+05,ZAPATA+07}. Measured fluxes are similar to
878: those listed in Table \ref{tab:detections}, but somewhat lower in most
879: cases, presumably because the poorer $uv$ coverage did not allow large negative
880: sidelobe contributions from extended emission to be fully removed.
881: All of these objects appear to drive molecular outflows
882: traced by CO or SiO emission \citep{ZAPATA+05,ZAPATA+06}.
883:
884: We classify the MM sources as candidate Class 0 or Class I
885: protostars. As discussed above, it appears that the 1.3 mm emission from at
886: least some of these sources may
887: trace dust hotter than 20 K. Such warm dust is expected in the
888: circumstellar environments of massive protostars, suggesting that some of
889: the MM sources trace high-mass protostars.
890:
891: Several sources detected in previous surveys were not detected here.
892: HC178, HC192, HC282, MM3, MM4, MM10, MM15, MM16, MM19, and MM20
893: should have been detected if their 3 mm fluxes \citep{EC06} traced dust
894: emission; however, they would not have been detected if the objects exhibited
895: flat spectra (e.g., from free-free emission). MM7, MM17, and MM18
896: should have been detected even if they showed flat spectra.
897: We detect a 1.3 mm
898: continuum source (MM21) near to HC 178, but find it to be offset by more than
899: the relative positional uncertainties, suggesting that the previous
900: association of HC 178 with a 3 mm source was mistaken.
901: The other 3 mm objects trace either non-dust, potentially time-variable
902: emission, or are spurious, caused by confusion with the BN/KL and OMC1-S
903: regions in which they reside. Because our 1 mm observations have far
904: superior $uv$ coverage than previous observations, they are less prone
905: to such spurious detections. One source detected by \citet{WAW05} at
906: 880 $\mu$m (171-334) is not detected at 1.3 mm; if the emission comes
907: from small-grained dust, then the expected 1.3 mm flux is comparable
908: to our 3$\sigma$ noise level, and hence a non-detection is unsurprising.
909:
910:
911: \subsection{Frequency of Massive Disks \label{sec:disc_mass}}
912: We detected 39 sources in our 1 mm mosaics (excluding the BN object).
913: 32 of these correspond to (presumed) low-mass near-IR cluster members, and
914: 6 (the ``MM'' sources) are detected only at $\ga 1$ mm wavelengths.
915: The remaining detection, LMLA162, while not listed as a near-IR source in
916: \citet{HC00}, is seen at 3.6 $\mu$m \citep{LADA+04};
917: examination of an archival 2MASS
918: image shows a weak 2 $\mu$m source at this position as well.
919: Since the mm-only detections are probably
920: embedded, possibly spherical, protostellar objects (\S \ref{sec:geometry}),
921: we exclude these from our discussion of disk statistics. Of the remaining
922: 33 detections, 100\% of the 1 mm emission can be attributed to hot gas
923: (free-free) for 6 sources. Thus, we are left with 27 sources whose 1 mm
924: emission (probably) traces dust in protoplanetary disks.
925:
926: Since the noise varies across our images, these 27 sources are
927: all detected above slightly different thresholds. To examine the frequency
928: of disks more massive than some value, we make sensitivity cuts at various
929: levels, examining only the statistics of sources detected above chosen
930: noise levels.
931:
932: We consider first the 115 cluster members surveyed to a 1$\sigma$ noise
933: level of 2.7 mJy or less. Sources detected above 3$\sigma$ have
934: a circumstellar (dust+gas) mass of $\ge 0.01$ M$_{\odot}$.
935: Nine sources (8\%) show dust emission of $\ge 8.1$ mJy
936: (i.e., 3$\sigma$ detections at this noise level).
937: If we use a higher noise cutoff of 5 mJy, then we
938: find that 7 out of 193 stars exhibit dust emission
939: above the 3$\sigma$ level of 15 mJy. So, $\sim 4\%$
940: of stars have disks more massive than 0.02
941: M$_{\odot}$. If we extend the sensitivity cutoff further, to 10 mJy,
942: then 3/254, or $\sim 1\%$ of stars are seen with disks more massive than 0.04
943: M$_{\odot}$.
944: All of the 3$\sigma$ mass levels considered here fall within the
945: range of estimates for the MMSN
946: \citep[$\sim 0.01$--0.1 M$_{\odot}$;][]{WEID+77}.
947:
948: The percentage of high-mass disks derived here can be compared to that
949: determined by \citet{EC06}. The observations presented here are
950: substantially more sensitive than previous observations, and we
951: probe the frequency of disks down to lower mass levels; we can therefore
952: only compare statistics for the most massive disks in our sample. \citet{EC06}
953: found that $\le 2\%$ of cluster members in the ONC have disks more massive than
954: 0.1 M$_{\odot}$. They assumed a distance of 480 pc; their mass limit is
955: actually only 0.07 M$_{\odot}$ for the distance of 400 pc assumed here.
956: Here we find that $\la 1\%$ of stars surveyed are surrounded by
957: such massive disks, consistent with the estimate from \citet{EC06}.
958:
959: We emphasize that that results presented above (and in the following
960: sections) depend on the conversion of 1.3 mm flux into mass. As
961: discussed in \S \ref{sec:massest}, there may be some spread in the
962: dust properties of our sample that could lead to variations in the
963: derived circumstellar masses. For the roughly co-eval, predominantly low-mass
964: cluster population in the ONC, we argued that this is a relatively small
965: uncertainty.
966:
967: \subsection{A Typical Disk in the ONC \label{sec:avg}}
968: We computed the average
969: flux for the ensemble of non-detected sources in \S \ref{sec:stacking}.
970: The average flux indicates that a ``typical'' non-detected source in the ONC
971: likely possesses a disk with a mass of $0.0015 \pm 0.0003$ M$_{\odot}$.
972: If we include detected objects (whose dust fluxes are listed in
973: Table \ref{tab:results}) in the ensemble, we find that the
974: average disk mass for near-IR cluster members in the region is
975: $\sim 0.0027 \pm 0.0002$ M$_{\odot}$.
976:
977: This is comparable to the average mass determined for 23
978: proplyds in the ONC at 880 $\mu$m \citep{WAW05}, but substantially
979: lower than the average mass determined for $>300$ stars at 3 mm wavelengths,
980: $0.005 \pm 0.001$ M$_{\odot}$ \citep{EC06}. The discrepancy can be
981: explained in large part by contamination from free-free emission.
982: Because the inferred dust mass is proportional to $\lambda^3 S_{\nu}$
983: (Equation \ref{eq:dustmass}), this contamination has a much greater affect on
984: the masses inferred from the 3 mm data than on our estimates based on
985: 1 mm data. The dust mass attributed to
986: free-free emission is $(2.3)^3 \approx 12$ times larger at 3 mm than at
987: 1.3 mm. If free-free emission is present at the $\la 0.2$ mJy level
988: (\S \ref{sec:ff}) it would add $\la 0.003$ M$_{\odot}$ to the average mass
989: inferred at 3 mm. Furthermore, if we recompute the mean disk mass from
990: \citet{EC06} using a distance of 400 pc, the average mass is decreased
991: by 30\%. With the distance correction and the subtraction of
992: potential free-free contamination, the recomputed average mass
993: from \citet{EC06} is $\ga 0.001$ M$_{\odot}$,
994: in agreement with the estimated average mass inferred from our 1 mm
995: observations.
996:
997: \subsection{Comparison of Disk and Exoplanet Frequencies \label{sec:exop}}
998: Less than $10\%$ of stars in
999: the ONC possess disks comparable to the MMSN
1000: (\S \ref{sec:disc_mass}).
1001: Moreover, the
1002: average mass measured for the ensemble of (individually) non-detected sources
1003: is ten times smaller than even the low end of estimates for the MMSN
1004: (\S \ref{sec:avg}),
1005: indicating that the majority of stars do not possess enough mass to form
1006: Jupiter-mass planets.
1007:
1008: These statistics can be compared with the frequency of Jupiter-mass planets
1009: found around nearby main-sequence stars. 6\% of stars surveyed have a
1010: Jupiter-mass (or larger)
1011: planet within 5 AU, while an extrapolation based on current
1012: results suggests up to 10\% of stars could have a Jupiter-mass planet within
1013: 20 AU \citep{MARCY+05}. The frequency of massive planets is comparable
1014: to the frequency of disks in the ONC with (low-end)
1015: minimum minimum mass solar nebulae.
1016: It appears that the MMSN, applied to disk mass measurements like those
1017: presented here, is a reasonable criteria
1018: for forming massive, Jupiter-like planets in typical star forming regions
1019: like Orion.
1020:
1021: \subsection{Disk Evolution \label{sec:disc_evol}}
1022: The frequency of massive disks in the ONC (aged $\sim 1$ Myr)
1023: can be compared with surveys of
1024: rich clusters of different ages, NGC 2024 (aged $\sim 0.3$ Myr)
1025: and IC 348 (aged $\sim 2$ Myr),
1026: to constrain the evolution of disks in
1027: clustered star forming environments. While this comparison has been made
1028: previously using 3 mm observations \citep[e.g.,][]{EC06}, our 1 mm measurements
1029: are less contaminated by free-free emission and yield
1030: different results (\S \ref{sec:avg}).
1031: Although the surveys of NGC 2024 and IC 348 were at
1032: 3 mm, the lack of O stars in those regions should produce less ionized
1033: gas, and hence less contamination by free-free emission than in the ONC.
1034:
1035: The average disk masses for ``typical'' low-mass stars in the three regions
1036: is plotted as a function of cluster age in Figure \ref{fig:evol}.
1037: For the ONC, we infer a mean disk mass of $0.0027 \pm 0.0002$
1038: (\S \ref{sec:avg}).
1039: In NGC 2024, the mean disk mass is $0.005 \pm 0.001$ M$_{\odot}$
1040: \citep{EC03}, compared to $0.002 \pm 0.001$ M$_{\odot}$ in IC 348
1041: \citep{CARPENTER02}. If the differences
1042: between NGC 2024, the ONC, and IC 348 are due to temporal evolution, these
1043: observations suggest that massive disks/envelopes dissipate on timescales
1044: $\la 1$ Myr, and that the average disk mass
1045: decreases by a factor of $1.9 \pm 0.4$ between $\sim 0.3$ and
1046: 1 Myr.
1047:
1048: %It is important to keep in mind that the total millimeter emission is
1049: %sensitive to dust grain properties in addition to total dust mass. For
1050: %example, a given mass of dust grains larger than the observing wavelength
1051: %emits less radiation than the same mass of small dust particles,
1052: %and the millimeter flux therefore depends on
1053: %dust grain properties in addition to the total mass. Furthermore, the
1054: %presence of hotter stars in the ONC (relative to NGC 2024 and IC 348)
1055: %may potentially lead to systematically higher disk temperatures, which would
1056: %lead to smaller inferred masses (Equation \ref{eq:dustmass}):
1057: %if $T_{\rm dust}=50$
1058: %K, disk masses would be reduced by a factor of 3.3.
1059: %Thus, observed evolution in the millimeter flux may indicate that one
1060: %or more of the assumed quantities in Equation \ref{eq:dustmass}
1061: %(e.g., temperature or opacity) is different in the three regions.
1062:
1063: %Regardless of the underlying factors, our measurements
1064: %provide a 2$\sigma$ suggestion of evolution between $\sim 0.3$ and 2 Myr.
1065: %Future,
1066: %more sensitive measurements of the disk mass distributions in larger
1067: %numbers of clusters will decrease
1068: %the uncertainties in Figure \ref{fig:evol}, enabling more concrete
1069: %constraints on the evolutionary timescales of massive disks.
1070: %Moreover, follow-up observations at multiple wavelengths will begin
1071: %to break degeneracies between dust grain properties and temperatures,
1072: %and total disk masses.
1073:
1074:
1075: \subsection{Dependence of Disk Properties on Environment \label{sec:disc_env}}
1076: It has been suggested that circumstellar disks in clustered environments
1077: may be truncated due to close encounters with massive stars resulting
1078: in either tidal stripping or photo-evaporation of outer disk material
1079: \citep[e.g.,][]{SC01}. Indeed, photoevaporative mass loss has been
1080: observed from many proplyds, suggesting mass loss rates as high as
1081: $10^{-7}$--$10^{-6}$ M$_{\odot}$ yr$^{-1}$
1082: \citep[e.g.,][]{HO99}, which would severely
1083: deplete the masses of disks over the $\sim 1$ Myr lifetime of the cluster.
1084: More detailed models have shown that the mass loss rate should be substantially
1085: lower for disks with smaller outer radii, since disk material at smaller
1086: radii is more tightly gravitationally bound \citep{CLARKE07}. A
1087: prediction of these models is that larger disks will also be the most
1088: massive, since they have to withstand higher photoevaporative mass
1089: loss rates.
1090:
1091: The proplyds detected in our observations are in the top
1092: $\sim 1/3$ of the size distribution inferred by \citet{VA05}. However,
1093: the emission from most of these is dominated by free-free emission,
1094: and even for objects where some component of the flux is due to dust,
1095: inferred masses are $\la 0.01$ M$_{\odot}$.
1096: Furthermore, there are many other proplyds whose diameters are in the
1097: top 30\% that are not detected in our observations. We also see no
1098: obvious trend of increasing flux with increasing angular size in our data
1099: (Table \ref{tab:results}). Thus, we
1100: find little evidence that the most extended disks are the most massive.
1101:
1102: Environmental effects on massive disks can also be investigated through
1103: the dependence of disk properties on cluster radius.
1104: We consider the positions of the disks detected in our observations
1105: (i.e., detected sources corresponding to known near-IR cluster member
1106: positions) relative to the cluster center, which we define to lie
1107: roughly in the middle of the four bright Trapezium
1108: stars at $(\alpha,\delta)_{\rm J2000} = (5^{\rm h}35^{\rm m}16.34^{\rm s},
1109: -5^{\circ}23'15\rlap{.}''6)$. Figure \ref{fig:radii}
1110: shows that more massive disks tend to be found further away from the
1111: Trapezium stars. If we consider only those 194 cluster members where we could
1112: have detected disks more massive than 0.02 M$_{\odot}$ above the 3$\sigma$
1113: noise level, we find disks around 1/84 stars
1114: ($\sim 1\%$ ) within $30''$ and 6/110 stars ($\sim 5\%$) at radii larger than
1115: $30''$. Fisher's exact test indicates 86\% probability (1.5$\sigma$)
1116: that the small and large cluster radii sources have different
1117: frequencies of massive disks.
1118:
1119: Finally, comparison of the ONC with the lower stellar density Taurus region
1120: provides another test of whether the massive O stars
1121: and high stellar density in the Trapezium
1122: region lead to different disk properties than in more ``benign'' environments.
1123: As discussed in \S \ref{sec:disc_mass}, we
1124: detected disks more massive than $0.01$ M$_{\odot}$
1125: around 9/115 ($\sim 8\%$) low-mass ONC cluster members.
1126: For comparison, 34/153 ($\sim 22\%$) of Taurus stars possess such
1127: massive disks \citep{AW05}. Fisher's Exact Test
1128: yields $> 99\%$ probability ($3\sigma$) that the frequencies
1129: of 0.01 M$_{\odot}$ disks in Taurus and the ONC are different.
1130: For a slightly higher mass cutoff of
1131: 0.04 M$_{\odot}$, such massive disks are found around $< 1\%$ of
1132: stars in the ONC compared to $\sim 5\%$ for Taurus. These percentages
1133: indicate $>99\%$ probability that the underlying distribution of 0.04
1134: M$_{\odot}$ disks in Taurus and the ONC differ.
1135: The fraction of approximately MMSN-massed disks in Orion is substantially
1136: smaller than in Taurus, arguing that the rich cluster environment may play a
1137: role in limiting the number of massive disks.
1138:
1139: This conclusion differs
1140: from that of \citet{EC06}, where the statistics of disks more massive than
1141: 0.1 M$_{\odot}$ were found to be statistically indistinguishable in Taurus
1142: and the ONC. Using the revised distance of 400 pc changes the conclusion
1143: from \citet{EC06}, because if statistics of (distance-corrected)
1144: 0.07 M$_{\odot}$ disks are
1145: compared, they are found to be substantially more common in Taurus
1146: (Fisher's Exact Test indicates only $\sim 1.3 \%$ probability that the two
1147: distributions are the same). Furthermore, the 1 mm observations
1148: presented here are sensitive to much more of the disk mass distribution,
1149: allowing a more robust comparison between Taurus and the ONC.
1150:
1151: \subsection{Correlation of Circumstellar and Stellar Masses \label{sec:mstars}}
1152: Several spectroscopic
1153: surveys have provided accurate masses for a subset of the stellar population
1154: encompassed by our mosaics
1155: \citep{HILLENBRAND97,LUHMAN+00,SLESNICK+04}. By examining these
1156: surveys, after registration of the positions of detected sources to
1157: the 2MASS grid, we find $\sim 130$ objects with spectroscopically-determined
1158: masses within the unit gain contour of our mosaic. While the stellar masses
1159: of the remaining cluster members contained in our mosaic have
1160: been estimated statistically by de-reddening stars so that they fall on
1161: the expected isochrone for the ONC \citep{HC00} the masses of individual
1162: stars determined in this way have large uncertainties and we do not use them
1163: here.
1164:
1165: To examine how disk mass depends on stellar mass, we divide
1166: the $\sim 130$ sources with spectroscopically-determined masses into
1167: three mass bins containing roughly equal numbers of stars.
1168: The first bin contains stars less massive than
1169: 0.3 M$_{\odot}$, the second bin includes stars with masses between 0.3 and
1170: 1.0 M$_{\odot}$, and the third bin contains stars with masses between
1171: 1.0 and 10.0 M$_{\odot}$. We then make a further cut by excluding all
1172: objects for which the noise in our 1 mm mosaic at the source position
1173: is greater than some cutoff value. As in \S \ref{sec:disc_mass}, we
1174: consider a noise cutoff of 2.7 mJy, which provides a corresponding
1175: 3$\sigma$ circumstellar mass threshold of 0.01 M$_{\odot}$.
1176: Ideally, we would also bin this sample by cluster radius to control
1177: for potential mass segregation in the inner regions of the ONC
1178: \citep[e.g.,][]{HILLENBRAND97}. Unfortunately we lack a sufficiently large
1179: sample to do this here.
1180:
1181: If we use the raw image fluxes (Table \ref{tab:detections}), we find
1182: more sources detected in the highest stellar mass bin. For 65 stars with
1183: spectroscopically-determined masses, surveyed with a noise level of 2.7 mJy
1184: or lower, we detect 1/21 (5\%) stars with
1185: $M_{\ast}<0.3$ M$_{\odot}$, 2/24 (8\%) stars with
1186: 0.3 M$_{\odot} < M_{\ast}<1$ M$_{\odot}$, and 6/20 (30\%) stars with
1187: $M_{\ast}>1$ M$_{\odot}$. However, the higher percentage of detected
1188: sources in the highest stellar-mass bin is due entirely to contamination
1189: by free-free emission: if we use dust fluxes from Table \ref{tab:results},
1190: then 0/20 stars in the highest stellar mass bin are detected.
1191: We infer, therefore, that higher mass stars are more likely to
1192: exhibit free-free emission.
1193:
1194: In contrast, it seems that more massive stars may be less likely to
1195: possess massive circumstellar disks. Using the dust fluxes from Table
1196: \ref{tab:results}, we find that out of the 65 stars discussed above,
1197: we detect {\it dust} emission above the 3$\sigma$ level toward
1198: 1/21 (5\%) stars with $M_{\ast}<0.3$ M$_{\odot}$, 2/24 (8\%) stars with
1199: 0.3 M$_{\odot} < M_{\ast}<1$ M$_{\odot}$, and 0/20 (0\%) stars with
1200: $M_{\ast}>1$ M$_{\odot}$. This suggests a lower frequency of disks
1201: around stars more massive
1202: than 1 M$_{\odot}$, but these small number statistics do not allow
1203: a definite conclusion. A plot of inferred circumstellar disk masses versus
1204: stellar masses (where available) supports the hypothesis
1205: that the most massive disks are found
1206: around the lowest mass stars. This trend is not, however, statistically
1207: significant for the small number of objects in Figure \ref{fig:mstars}.
1208: A similar picture was seen in Taurus \citep{AW05}: no
1209: correlation between stellar mass and disk mass could be established,
1210: although the most massive disks were found around stars less massive than
1211: 1 M$_{\odot}$. We note that \citet{NGM00} claimed to see
1212: a correlation between disk and stellar masses around early-type stars;
1213: however, the dispersion is large and the significance of the trend is marginal.
1214:
1215:
1216: %Non-detected sources generally not coincident with X-ray sources in
1217: %the COUP sample \citep{GETMAN+05}. Several proplyds also have embedded
1218: %silhouette disks \citep{BOM00}: 147-323, 158-327, 171-340, and 177-341.
1219: %Most of these also have detected 1mm dust emission. Disk sizes have been
1220: %measured for almost all of the proplyds detected in our observations
1221: %using HST images \citep{VA05}.
1222:
1223: \section{Conclusions \label{sec:conc}}
1224: We imaged a $2' \times 2'$ region of the Orion Nebula cluster in 1.3 mm
1225: wavelength continuum emission with CARMA and the SMA. Out of $>250$ known
1226: near-IR cluster members, we detected 1.3 mm emission above the $3\sigma$
1227: noise level toward 33. In addition, we detected 1 mm emission above the
1228: 5$\sigma$ noise level from six sources not associated with
1229: shorter-wavelength counterparts.
1230: Several of these detected objects are spatially resolved with our observations,
1231: indicating sizes of $\sim 250$--450 AU.
1232:
1233: Modeling of long-wavelength fluxes for our targets allowed separation of
1234: dust and free-free emission components in the measured fluxes.
1235: We showed that for the majority of detected sources, the 1 mm emission appears
1236: to trace warm, optically thin,
1237: circumstellar dust. However, for many of the proplyds, which are
1238: located close to the Trapezium stars, the millimeter wavelength emission
1239: is dominated by thermal free-free emission from hot, ionized gas.
1240:
1241: Dust masses inferred for detected sources range from 0.01 to 0.5 M$_{\odot}$.
1242: For the $\sim 225$ known near-IR cluster members not detected in our 1 mm
1243: observations, images toward the positions of near-IR sources
1244: were stacked to constrain the mean flux, and circumstellar mass,
1245: of the ensemble. The average flux is detected at the $> 4\sigma$
1246: confidence level, and implies an average disk mass of $\sim 0.001$
1247: M$_{\odot}$, approximately an order of magnitude smaller than the
1248: minimum mass solar nebula. Even when detected sources are included,
1249: the average mass is $<0.003$ M$_{\odot}$.
1250: While the derived masses are uncertain by
1251: a factor of 3 or so (mostly due to uncertainties in the dust opacity),
1252: the range of possible average disk masses is still smaller than even the
1253: low end of estimates for the MMSN. A ``typical'' star in the ONC does
1254: not appear to possess sufficient mass in small dust grains to form
1255: Jupiter-mass (or larger) planets. Evidently, giant planet formation
1256: is either advanced (having thus depleted the small dust grains in the disk)
1257: or impossible around most stars in the ONC.
1258:
1259: We compared the average disk mass inferred for the ONC with similarly
1260: determined average masses in older and younger clusters. We find evidence for
1261: evolution of the dust (most likely depletion or agglomeration) on
1262: $\sim 1$ Myr timescales. Between $\sim 0.3$ Myr and $\sim 1$ Myr, the
1263: average disk mass decreases by a factor of $1.9 \pm 0.4$.
1264:
1265: The percentage of stars in Orion surrounded by disks more massive than the
1266: minimum mass solar nebula is $<10\%$. This is significantly lower than in
1267: Taurus, indicating that environment has an impact on the disk mass
1268: distribution. Our data suggest (with marginal statistical
1269: significance) that the most massive disks may be located
1270: further from the Trapezium stars, supporting the hypothesis that
1271: photoevaporation may be truncating disks near to the cluster center.
1272:
1273: Finally, our observations show no clear correlation between stellar mass and
1274: disk mass, but suggest that massive disks may be more likely
1275: to be found around lower mass stars. The percentage of detected disks
1276: is lower for stars more massive than 1 M$_{\odot}$, and the most massive
1277: disks detected are associated with the relatively low stellar mass stars
1278: in the sample.
1279: However, larger numbers of stellar and disk mass measurements in the ONC
1280: are needed to build up better statistics and further
1281: constrain the relationship between stellar and disk properties.
1282:
1283:
1284:
1285: \acknowledgements
1286: JAE acknowledges support from a
1287: Miller Research Fellowship. The authors thank M. Wright for his assistance
1288: in simulating CARMA mosaics in order to choose the optimal mosaic
1289: spacings for the observations. This publication makes use of data products
1290: from the Two Micron All Sky Survey, which is a joint project of the
1291: University of Massachusetts and the Infrared Processing and Analysis Center
1292: (IPAC), funded by the National Aeronautics and Space Administration and the
1293: National Science Foundation. 2MASS science data and information services were
1294: provided by the Infrared Science Archive at IPAC.
1295: Support for CARMA construction was derived from the states of Illinois,
1296: California, and Maryland, the Gordon and Betty Moore Foundation, the Kenneth
1297: T. and Eileen L. Norris Foundation, the Associates of the California Institute
1298: of Technology, and the National Science Foundation. Ongoing CARMA development
1299: and operations are supported by the National Science Foundation under a
1300: cooperative agreement and by the CARMA partner universities.
1301:
1302: %This publication makes use of data products from the Two Micron
1303: %All Sky Survey, which is a joint project of the University of Massachusetts
1304: %and the Infrared Processing and Analysis Center, funded by the National
1305: %Aeronautics and Space Administration and the National Science Foundation.
1306: %2MASS science data and information services were provided by the Infrared
1307: %Science Archive (IRSA) at IPAC.
1308:
1309: \appendix
1310: \section{Free-free emission spectrum for an ionized wind \label{sec:ffapp}}
1311:
1312: We compute the spectrum expected for free-free emission from ionized gas
1313: whose density depends on stellocentric radius as
1314: \begin{equation}
1315: n_{\rm gas} = n_0 (R/R_{\rm 0})^{-\alpha}.
1316: \label{eq:ffdens}
1317: \end{equation}
1318: For an ionized wind with constant $\dot{M}$, such as we expect for proplyds,
1319: $\alpha=2$. However, massive stars may ionize their circumstellar environments
1320: directly, in which case the density distribution may differ from this
1321: $R^{-2}$ power-law.
1322:
1323: For a power-law density profile (Equation \ref{eq:ffdens}), the optical depth
1324: can be approximated as \citep[e.g.,][]{ALTENHOFF+60},
1325: \begin{equation}
1326: \tau_{\rm \nu,ff} \approx \int_{0}^{\infty}
1327: \frac{0.16 n_{\rm 0}^2}{\nu^{2.1}T_{\rm e}^{1.35}} \frac{dz}
1328: {(x^2+z^2)^{\alpha}}.
1329: \end{equation}
1330: Here, $z$ is the line of sight through the ionized gas in pc, $x$ is the
1331: impact parameter in pc, $n_0$ is the normalization of the gas density
1332: in cm$^{-3}$ (assumed to represent the electron and ion densities),
1333: $T_{\rm e}$ is the electron temperature,
1334: and $\nu$ is the frequency in GHz. This expression can be integrated
1335: directly as long as $\alpha \ge 1.5$, with the result
1336: \begin{equation}
1337: \tau_{\rm \nu,ff} \approx \frac{0.08 n_{0}^2 R_0^4}{\nu^{2.1}T_{\rm e}^{1.35}}
1338: x^{-(2\alpha -1)}
1339: \frac{\pi (-1)^{\alpha-1} \: \Gamma(1/2)}{\sin(\pi/2) (\alpha-1)!
1340: \: \Gamma(1/(2-\alpha+1))}.
1341: \end{equation}
1342: The last fraction has order unity, and we can thus approximate $\tau$ as
1343: \begin{equation}
1344: \tau_{\rm \nu,ff} \approx \frac{0.08 n_{0}^2 R_0^4}{\nu^{2.1}T_{\rm e}^{1.35}}
1345: x^{-(2\alpha -1)}.
1346: \label{eq:fftau}
1347: \end{equation}
1348:
1349: We can now invert Equation \ref{eq:fftau} to determine the maximum impact
1350: parameter for which $\tau \ga 1$:
1351: \begin{equation}
1352: x_{\tau\approx 1} \approx \left(\frac{0.08 n_0^2 R_0^4}{\nu^{2.1}
1353: T_{\rm e}^{1.35}}\right)^{\frac{1}{2\alpha-1}}.
1354: \label{eq:xff}
1355: \end{equation}
1356: For impact parameters larger than $x_{\tau\approx 1}$, the gas is
1357: optically thin, and at smaller impact parameters the gas is optically thick.
1358: The total spectrum of free-free emission from the source can be approximated
1359: by the blackbody flux times the solid angle of the optically thick region,
1360: as long as the optically thick region is finite. When the entire wind
1361: becomes optically thin, the spectrum flattens.
1362:
1363: We can thus parameterize the free-free emission from a wind as
1364: \begin{equation}
1365: F_{\rm \nu, ff} = \cases{F_{\rm \nu, turn} (\nu / \nu_{\rm turn})^{-0.1} &
1366: if $\nu \ge \nu_{\rm turn}$ \cr
1367: F_{\rm \nu , turn} \left({\nu} / {\nu_{\rm turn}}\right)^{\frac{4\alpha-6.2}
1368: {2\alpha-1}} &
1369: if $\nu \le \nu_{\rm turn}$}
1370: \label{eq:ff}
1371: \end{equation}
1372: \citep[see][for a somewhat different derivation of this result]{WB75}.
1373: For a spherical wind with constant $\dot{M}$, $\alpha=2$ and the
1374: optically thick
1375: material emits with $F_{\nu} \propto \nu^{0.6}$. Steeper exponents can result
1376: if the ionized gas density drops off more steeply than $R^{-2}$, for example
1377: as might occur in a centrally illuminated wind from a massive star like BN
1378: \citep[e.g.,][]{PLAMBECK+95}.
1379: Alternatively, non-spherical wind geometries can lead to shallower
1380: radial density profiles, and hence shallower spectral slopes of the free-free
1381: emission \citep{WB75}. For simplicity, since we generally have a limited
1382: number of data points with which to constrain the free-free emission spectrum,
1383: we will assume $\alpha=2$ in the present analysis.
1384:
1385:
1386: \begin{thebibliography}{83}
1387: \bibitem[{{Ali}(1996)}]{ALI96}
1388: {Ali}, B. 1996, PhD thesis, The Ohio State University, Ohio, USA
1389:
1390: \bibitem[{{Alibert} {et~al.}(2005){Alibert}, {Mordasini}, {Benz}, \&
1391: {Winisdoerffer}}]{ALIBERT+05}
1392: {Alibert}, Y., {Mordasini}, C., {Benz}, W., \& {Winisdoerffer}, C. 2005, \aap,
1393: 434, 343
1394:
1395: \bibitem[{{Altenhoff} {et~al.}(1960){Altenhoff}, {Mezger}, {Wendker}, \&
1396: {Westerhout}}]{ALTENHOFF+60}
1397: {Altenhoff}, W.~J., {Mezger}, P.~G., {Wendker}, H., \& {Westerhout}, G. 1960,
1398: {\it Ver\"{o}f. Sternw\"{a}rte}, Bonn, 59, 48
1399:
1400: \bibitem[{{Andr\'{e}} \& {Montmerle}(1994)}]{AM94}
1401: {Andr\'{e}}, P. \& {Montmerle}, T. 1994, \apj, 420, 837
1402:
1403: \bibitem[{{Andrews} \& {Williams}(2005)}]{AW05}
1404: {Andrews}, S.~M. \& {Williams}, J.~P. 2005, \apj, 631, 1134
1405:
1406: \bibitem[{{Andrews} \& {Williams}(2007)}]{AW07}
1407: ---. 2007, \apj, 671, 1800
1408:
1409: \bibitem[{{Bachiller}(1996)}]{BACHILLER96}
1410: {Bachiller}, R. 1996, \araa, 34, 111
1411:
1412: \bibitem[{{Bally} {et~al.}(2000){Bally}, {O'Dell}, \& {McCaughrean}}]{BOM00}
1413: {Bally}, J., {O'Dell}, C.~R., \& {McCaughrean}, M.~J. 2000, \aj, 119, 2919
1414:
1415: \bibitem[{{Bally} {et~al.}(1998{\natexlab{a}}){Bally}, {Sutherland}, {Devine},
1416: \& {Johnstone}}]{BALLY+98a}
1417: {Bally}, J., {Sutherland}, R.~S., {Devine}, D., \& {Johnstone}, D.
1418: 1998{\natexlab{a}}, \aj, 116, 293
1419:
1420: \bibitem[{{Bally} {et~al.}(1998{\natexlab{b}}){Bally}, {Testi}, {Sargent}, \&
1421: {Carlstrom}}]{BALLY+98}
1422: {Bally}, J., {Testi}, L., {Sargent}, A., \& {Carlstrom}, J. 1998{\natexlab{b}},
1423: \aj, 116, 854
1424:
1425: \bibitem[{{Beckwith} {et~al.}(1990){Beckwith}, {Sargent}, {Chini}, \&
1426: {Guesten}}]{BECKWITH+90}
1427: {Beckwith}, S.~V.~W., {Sargent}, A.~I., {Chini}, R.~S., \& {Guesten}, R. 1990,
1428: \aj, 99, 924
1429:
1430: \bibitem[{{Bower} {et~al.}(2003){Bower}, {Plambeck}, {Bolatto}, {McCrady},
1431: {Graham}, {de Pater}, {Liu}, \& {Baganoff}}]{BOWER+03}
1432: {Bower}, G.~C., {Plambeck}, R.~L., {Bolatto}, A., {McCrady}, N., {Graham},
1433: J.~R., {de Pater}, I., {Liu}, M.~C., \& {Baganoff}, F.~K. 2003, \apj, 598,
1434: 1140
1435:
1436: \bibitem[{{Carpenter}(2000)}]{CARPENTER00}
1437: {Carpenter}, J.~M. 2000, \aj, 120, 3139
1438:
1439: \bibitem[{{Carpenter}(2002)}]{CARPENTER02}
1440: ---. 2002, \aj, 124, 1593
1441:
1442: \bibitem[{{Clarke}(2007)}]{CLARKE07}
1443: {Clarke}, C.~J. 2007, \mnras, 376, 1350
1444:
1445: \bibitem[{{D'Antona} \& {Mazzitelli}(1994)}]{DM94}
1446: {D'Antona}, F. \& {Mazzitelli}, I. 1994, \apjs, 90, 467
1447:
1448: \bibitem[{{Dutrey} {et~al.}(1996){Dutrey}, {Guilloteau}, {Duvert}, {Prato},
1449: {Simon}, {Schuster}, \& {Menard}}]{DUTREY+96}
1450: {Dutrey}, A., {Guilloteau}, S., {Duvert}, G., {Prato}, L., {Simon}, M.,
1451: {Schuster}, K., \& {Menard}, F. 1996, \aap, 309, 493
1452:
1453: \bibitem[{{Eisner} \& {Carpenter}(2003)}]{EC03}
1454: {Eisner}, J.~A. \& {Carpenter}, J.~M. 2003, \apj, 598, 1341
1455:
1456: \bibitem[{{Eisner} \& {Carpenter}(2006)}]{EC06}
1457: ---. 2006, \apj, 641, 1162
1458:
1459: \bibitem[{{Eisner} {et~al.}(2004){Eisner}, {Lane}, {Hillenbrand}, {Akeson}, \&
1460: {Sargent}}]{EISNER+04}
1461: {Eisner}, J.~A., {Lane}, B.~F., {Hillenbrand}, L., {Akeson}, R., \& {Sargent},
1462: A. 2004, \apj, 613, 1049
1463:
1464: \bibitem[{{Felli} {et~al.}(1993{\natexlab{a}}){Felli}, {Churchwell}, {Wilson},
1465: \& {Taylor}}]{FELLI+93a}
1466: {Felli}, M., {Churchwell}, E., {Wilson}, T.~L., \& {Taylor}, G.~B.
1467: 1993{\natexlab{a}}, \aaps, 98, 137
1468:
1469: \bibitem[{{Felli} {et~al.}(1993{\natexlab{b}}){Felli}, {Taylor}, {Catarzi},
1470: {Churchwell}, \& {Kurtz}}]{FELLI+93b}
1471: {Felli}, M., {Taylor}, G.~B., {Catarzi}, M., {Churchwell}, E., \& {Kurtz}, S.
1472: 1993{\natexlab{b}}, \aaps, 101, 127
1473:
1474: \bibitem[{{Forbrich} {et~al.}(2008){Forbrich}, {Menten}, \& {Reid}}]{FMR07}
1475: {Forbrich}, J., {Menten}, K.~M., \& {Reid}, M.~J. 2008, \aap, 477, 267
1476:
1477: \bibitem[{{Furuya} {et~al.}(2003){Furuya}, {Shinnaga}, {Nakanishi}, {Momose},
1478: \& {Saito}}]{FURUYA+03}
1479: {Furuya}, R.~S., {Shinnaga}, H., {Nakanishi}, K., {Momose}, M., \& {Saito}, M.
1480: 2003, \pasj, 55, L83
1481:
1482: \bibitem[{{Garay} {et~al.}(1987){Garay}, {Moran}, \& {Reid}}]{GMR87}
1483: {Garay}, G., {Moran}, J.~M., \& {Reid}, M.~J. 1987, \apj, 314, 535
1484:
1485: \bibitem[{{Genzel} {et~al.}(1981){Genzel}, {Reid}, {Moran}, \&
1486: {Downes}}]{GENZEL+81}
1487: {Genzel}, R., {Reid}, M.~J., {Moran}, J.~M., \& {Downes}, D. 1981, \apj, 244,
1488: 884
1489:
1490: \bibitem[{{Gezari} {et~al.}(1998){Gezari}, {Backman}, \& {Werner}}]{GBW98}
1491: {Gezari}, D.~Y., {Backman}, D.~E., \& {Werner}, M.~W. 1998, \apj, 509, 283
1492:
1493: \bibitem[{{Haisch} {et~al.}(2001){Haisch}, {Lada}, \& {Lada}}]{HLL01b}
1494: {Haisch}, K.~E., {Lada}, E.~A., \& {Lada}, C.~J. 2001, \aj, 121, 2065
1495:
1496: \bibitem[{{Hayashi}(1981)}]{HAYASHI81}
1497: {Hayashi}, C. 1981, Progress of Theoretical Physics, 70, 35
1498:
1499: \bibitem[{{Henney} \& {O'Dell}(1999)}]{HO99}
1500: {Henney}, W.~J. \& {O'Dell}, C.~R. 1999, \aj, 118, 2350
1501:
1502: \bibitem[{{Henning} {et~al.}(1993){Henning}, {Pfau}, {Zinnecker}, \&
1503: {Prusti}}]{HENNING+93}
1504: {Henning}, T., {Pfau}, W., {Zinnecker}, H., \& {Prusti}, T. 1993, \aap, 276,
1505: 129
1506:
1507: \bibitem[{{Hester} \& {Desch}(2005)}]{HD05}
1508: {Hester}, J.~J. \& {Desch}, S.~J. 2005, in Astronomical Society of the Pacific
1509: Conference Series, Vol. 341, Chondrites and the Protoplanetary Disk, ed.
1510: A.~N. {Krot}, E.~R.~D. {Scott}, \& B.~{Reipurth}, 107
1511:
1512: \bibitem[{{Hildebrand}(1983)}]{HILDEBRAND83}
1513: {Hildebrand}, R.~H. 1983, \qjras, 24, 267
1514:
1515: \bibitem[{{Hillenbrand}(1997)}]{HILLENBRAND97}
1516: {Hillenbrand}, L.~A. 1997, \aj, 113, 1733
1517:
1518: \bibitem[{{Hillenbrand} \& {Carpenter}(2000)}]{HC00}
1519: {Hillenbrand}, L.~A. \& {Carpenter}, J.~M. 2000, \apj, 540, 236
1520:
1521: \bibitem[{{Hillenbrand} {et~al.}(1998){Hillenbrand}, {Strom}, {Calvet},
1522: {Merrill}, {Gatley}, {Makidon}, {Meyer}, \& {Skrutskie}}]{HILLENBRAND+98}
1523: {Hillenbrand}, L.~A., {Strom}, S.~E., {Calvet}, N., {Merrill}, K.~M., {Gatley},
1524: I., {Makidon}, R.~B., {Meyer}, M.~R., \& {Skrutskie}, M.~F. 1998, \aj, 116,
1525: 1816
1526:
1527: \bibitem[{{Hogerheijde} {et~al.}(2002){Hogerheijde}, {Jayawardhana},
1528: {Johnstone}, {Blake}, \& {Kessler}}]{HOGERHEIJDE+02}
1529: {Hogerheijde}, M.~R., {Jayawardhana}, R., {Johnstone}, D., {Blake}, G.~A., \&
1530: {Kessler}, J.~E. 2002, \aj, 124, 3387
1531:
1532: \bibitem[{{Itoh} {et~al.}(2003){Itoh}, {Sugitani}, {Fukuda}, {Nakanishi},
1533: {Ogura}, {Tamura}, {Marui}, {Fujita}, {Oasa}, \& {Fukagawa}}]{ITOH+03}
1534: {Itoh}, Y., {Sugitani}, K., {Fukuda}, N., {Nakanishi}, K., {Ogura}, K.,
1535: {Tamura}, M., {Marui}, K., {Fujita}, K., {Oasa}, Y., \& {Fukagawa}, M. 2003,
1536: \apjl, 586, L141
1537:
1538: \bibitem[{{Koerner} \& {Sargent}(1995)}]{KS95}
1539: {Koerner}, D.~W. \& {Sargent}, A.~I. 1995, \aj, 109, 2138
1540:
1541: \bibitem[{{Kraus} {et~al.}(2007){Kraus}, {Balega}, {Berger}, {Hofmann},
1542: {Millan-Gabet}, {Monnier}, {Ohnaka}, {Pedretti}, {Preibisch}, {Schertl},
1543: {Schloerb}, {Traub}, \& {Weigelt}}]{KRAUS+07}
1544: {Kraus}, S., {Balega}, Y.~Y., {Berger}, J.-P., {Hofmann}, K.-H.,
1545: {Millan-Gabet}, R., {Monnier}, J.~D., {Ohnaka}, K., {Pedretti}, E.,
1546: {Preibisch}, T., {Schertl}, D., {Schloerb}, F.~P., {Traub}, W.~A., \&
1547: {Weigelt}, G. 2007, \aap, 466, 649
1548:
1549: \bibitem[{{Lada} {et~al.}(1991){Lada}, {Depoy}, {Merrill}, \&
1550: {Gatley}}]{LADA+91}
1551: {Lada}, C.~J., {Depoy}, D.~L., {Merrill}, K.~M., \& {Gatley}, I. 1991, \apj,
1552: 374, 533
1553:
1554: \bibitem[{{Lada} \& {Lada}(2003)}]{LL03}
1555: {Lada}, C.~J. \& {Lada}, E.~A. 2003, \araa, 41, 57
1556:
1557: \bibitem[{{Lada} {et~al.}(2000){Lada}, {Muench}, {Haisch}, {Lada}, {Alves},
1558: {Tollestrup}, \& {Willner}}]{LADA+00}
1559: {Lada}, C.~J., {Muench}, A.~A., {Haisch}, K.~E., {Lada}, E.~A., {Alves}, J.~F.,
1560: {Tollestrup}, E.~V., \& {Willner}, S.~P. 2000, \aj, 120, 3162
1561:
1562: \bibitem[{{Lada} {et~al.}(2004){Lada}, {Muench}, {Lada}, \& {Alves}}]{LADA+04}
1563: {Lada}, C.~J., {Muench}, A.~A., {Lada}, E.~A., \& {Alves}, J.~F. 2004, \aj,
1564: 128, 1254
1565:
1566: \bibitem[{{Lada} {et~al.}(1993){Lada}, {Strom}, \& {Myers}}]{LSM93}
1567: {Lada}, E.~A., {Strom}, K.~M., \& {Myers}, P.~C. 1993, in Protostars and
1568: Planets III, 245--277
1569:
1570: \bibitem[{{Luhman}(1999)}]{LUHMAN99}
1571: {Luhman}, K.~L. 1999, \apj, 525, 466
1572:
1573: \bibitem[{{Luhman} {et~al.}(1998){Luhman}, {Rieke}, {Lada}, \&
1574: {Lada}}]{LUHMAN+98}
1575: {Luhman}, K.~L., {Rieke}, G.~H., {Lada}, C.~J., \& {Lada}, E.~A. 1998, \apj,
1576: 508, 347
1577:
1578: \bibitem[{{Luhman} {et~al.}(2000){Luhman}, {Rieke}, {Young}, {Cotera}, {Chen},
1579: {Rieke}, {Schneider}, \& {Thompson}}]{LUHMAN+00}
1580: {Luhman}, K.~L., {Rieke}, G.~H., {Young}, E.~T., {Cotera}, A.~S., {Chen}, H.,
1581: {Rieke}, M.~J., {Schneider}, G., \& {Thompson}, R.~I. 2000, \apj, 540, 1016
1582:
1583: \bibitem[{{Marcy} {et~al.}(2005){Marcy}, {Butler}, {Fischer}, {Vogt}, {Wright},
1584: {Tinney}, \& {Jones}}]{MARCY+05}
1585: {Marcy}, G., {Butler}, R.~P., {Fischer}, D., {Vogt}, S., {Wright}, J.~T.,
1586: {Tinney}, C.~G., \& {Jones}, H.~R.~A. 2005, Progress of Theoretical Physics
1587: Supplement, 158, 24
1588:
1589: \bibitem[{{McCaughrean} \& {O'Dell}(1996)}]{MO96}
1590: {McCaughrean}, M.~J. \& {O'Dell}, C.~R. 1996, \aj, 111, 1977
1591:
1592: \bibitem[{{Menten} {et~al.}(2007){Menten}, {Reid}, {Forbrich}, \&
1593: {Brunthaler}}]{MENTEN+07}
1594: {Menten}, K.~M., {Reid}, M.~J., {Forbrich}, J., \& {Brunthaler}, A. 2007, \aap,
1595: 474, 515
1596:
1597: \bibitem[{{Meyer}(1996)}]{MEYER96}
1598: {Meyer}, M.~R. 1996, Ph.D.~Thesis
1599:
1600: \bibitem[{{Motte} \& {Andr{\' e}}(2001)}]{MA01}
1601: {Motte}, F. \& {Andr{\' e}}, P. 2001, \aap, 365, 440
1602:
1603: \bibitem[{{Motte} {et~al.}(1998){Motte}, {Andr\'{e}}, \& {Neri}}]{MAN98}
1604: {Motte}, F., {Andr\'{e}}, P., \& {Neri}, R. 1998, \aap, 336, 150
1605:
1606: \bibitem[{{Mundy} {et~al.}(1995){Mundy}, {Looney}, \& {Lada}}]{MLL95}
1607: {Mundy}, L.~G., {Looney}, L.~W., \& {Lada}, E.~A. 1995, \apjl, 452, L137
1608:
1609: \bibitem[{{Natta} {et~al.}(2000){Natta}, {Grinin}, \& {Mannings}}]{NGM00}
1610: {Natta}, A., {Grinin}, V., \& {Mannings}, V. 2000, Protostars and Planets IV,
1611: 559
1612:
1613: \bibitem[{{Nuernberger} {et~al.}(1998){Nuernberger}, {Brandner}, {Yorke}, \&
1614: {Zinnecker}}]{NUERNBERGER+98}
1615: {Nuernberger}, D., {Brandner}, W., {Yorke}, H.~W., \& {Zinnecker}, H. 1998,
1616: \aap, 330, 549
1617:
1618: \bibitem[{{Nuernberger} {et~al.}(1997){Nuernberger}, {Chini}, \&
1619: {Zinnecker}}]{NCZ97}
1620: {Nuernberger}, D., {Chini}, R., \& {Zinnecker}, H. 1997, \aap, 324, 1036
1621:
1622: \bibitem[{{O'Dell} {et~al.}(1993){O'Dell}, {Wen}, \& {Hu}}]{OWH93}
1623: {O'Dell}, C.~R., {Wen}, Z., \& {Hu}, X. 1993, \apj, 410, 696
1624:
1625: \bibitem[{{O'Dell} \& {Wong}(1996)}]{OW96}
1626: {O'Dell}, C.~R. \& {Wong}, K. 1996, \aj, 111, 846
1627:
1628: \bibitem[{{Osterloh} \& {Beckwith}(1995)}]{OB95}
1629: {Osterloh}, M. \& {Beckwith}, S.~V.~W. 1995, \apj, 439, 288
1630:
1631: \bibitem[{{Padgett} {et~al.}(1999){Padgett}, {Brandner}, {Stapelfeldt},
1632: {Strom}, {Terebey}, \& {Koerner}}]{PADGETT+99}
1633: {Padgett}, D.~L., {Brandner}, W., {Stapelfeldt}, K.~R., {Strom}, S.~E.,
1634: {Terebey}, S., \& {Koerner}, D. 1999, \aj, 117, 1490
1635:
1636: \bibitem[{{Plambeck} {et~al.}(1995){Plambeck}, {Wright}, {Mundy}, \&
1637: {Looney}}]{PLAMBECK+95}
1638: {Plambeck}, R.~L., {Wright}, M.~C.~H., {Mundy}, L.~G., \& {Looney}, L.~W. 1995,
1639: \apjl, 455, L189
1640:
1641: \bibitem[{{Pollack} {et~al.}(1994){Pollack}, {Hollenbach}, {Beckwith},
1642: {Simonelli}, {Roush}, \& {Fong}}]{POLLACK+94}
1643: {Pollack}, J.~B., {Hollenbach}, D., {Beckwith}, S., {Simonelli}, D.~P.,
1644: {Roush}, T., \& {Fong}, W. 1994, \apj, 421, 615
1645:
1646: \bibitem[{{Prosser} {et~al.}(1994){Prosser}, {Stauffer}, {Hartmann},
1647: {Soderblom}, {Jones}, {Werner}, \& {McCaughrean}}]{PROSSER+94}
1648: {Prosser}, C.~F., {Stauffer}, J.~R., {Hartmann}, L., {Soderblom}, D.~R.,
1649: {Jones}, B.~F., {Werner}, M.~W., \& {McCaughrean}, M.~J. 1994, \apj, 421, 517
1650:
1651: \bibitem[{{Sandstrom} {et~al.}(2007){Sandstrom}, {Peek}, {Bower}, {Bolatto}, \&
1652: {Plambeck}}]{SANDSTROM+07}
1653: {Sandstrom}, K.~M., {Peek}, J.~E.~G., {Bower}, G.~C., {Bolatto}, A.~D., \&
1654: {Plambeck}, R.~L. 2007, \apj, 667, 1161
1655:
1656: \bibitem[{{Sault} {et~al.}(1995){Sault}, {Teuben}, \& {Wright}}]{STW95}
1657: {Sault}, R.~J., {Teuben}, P.~J., \& {Wright}, M.~C.~H. 1995, in ASP Conf. Ser.
1658: 77: Astronomical Data Analysis Software and Systems IV, 433
1659:
1660: \bibitem[{{Scally} \& {Clarke}(2001)}]{SC01}
1661: {Scally}, A. \& {Clarke}, C. 2001, \mnras, 325, 449
1662:
1663: \bibitem[{{Slesnick} {et~al.}(2004){Slesnick}, {Hillenbrand}, \&
1664: {Carpenter}}]{SLESNICK+04}
1665: {Slesnick}, C.~L., {Hillenbrand}, L.~A., \& {Carpenter}, J.~M. 2004, \apj, 610,
1666: 1045
1667:
1668: \bibitem[{{Smith} {et~al.}(2005){Smith}, {Bally}, {Shuping}, {Morris}, \&
1669: {Kassis}}]{SMITH+05}
1670: {Smith}, N., {Bally}, J., {Shuping}, R.~Y., {Morris}, M., \& {Kassis}, M. 2005,
1671: \aj, 130, 1763
1672:
1673: \bibitem[{{Strom} {et~al.}(1989){Strom}, {Strom}, {Edwards}, {Cabrit}, \&
1674: {Skrutskie}}]{STROM+89}
1675: {Strom}, K.~M., {Strom}, S.~E., {Edwards}, S., {Cabrit}, S., \& {Skrutskie},
1676: M.~F. 1989, \aj, 97, 1451
1677:
1678: \bibitem[{{Testi} \& {Sargent}(1998)}]{TS98}
1679: {Testi}, L. \& {Sargent}, A.~I. 1998, \apjl, 508, L91
1680:
1681: \bibitem[{{Vicente} \& {Alves}(2005)}]{VA05}
1682: {Vicente}, S.~M. \& {Alves}, J. 2005, \aap, 441, 195
1683:
1684: \bibitem[{{Weidenschilling}(1977)}]{WEID+77}
1685: {Weidenschilling}, S.~J. 1977, \apss, 51, 153
1686:
1687: \bibitem[{{Williams} {et~al.}(2005){Williams}, {Andrews}, \& {Wilner}}]{WAW05}
1688: {Williams}, J.~P., {Andrews}, S.~M., \& {Wilner}, D.~J. 2005, \apj, 634, 495
1689:
1690: \bibitem[{{Williams} \& {Gaidos}(2007)}]{WG07}
1691: {Williams}, J.~P. \& {Gaidos}, E. 2007, \apjl, 663, L33
1692:
1693: \bibitem[{{Wilner} {et~al.}(2000){Wilner}, {Ho}, {Kastner}, \&
1694: {Rodr{\'{\i}}guez}}]{WILNER+00}
1695: {Wilner}, D.~J., {Ho}, P.~T.~P., {Kastner}, J.~H., \& {Rodr{\'{\i}}guez}, L.~F.
1696: 2000, \apjl, 534, L101
1697:
1698: \bibitem[{{Wright} \& {Barlow}(1975)}]{WB75}
1699: {Wright}, A.~E. \& {Barlow}, M.~J. 1975, \mnras, 170, 41
1700:
1701: \bibitem[{{Zapata} {et~al.}(2006){Zapata}, {Ho}, {Rodr{\'{\i}}guez}, {O'Dell},
1702: {Zhang}, \& {Muench}}]{ZAPATA+06}
1703: {Zapata}, L.~A., {Ho}, P.~T.~P., {Rodr{\'{\i}}guez}, L.~F., {O'Dell}, C.~R.,
1704: {Zhang}, Q., \& {Muench}, A. 2006, \apj, 653, 398
1705:
1706: \bibitem[{{Zapata} {et~al.}(2007){Zapata}, {Ho}, {Rodr{\'{\i}}guez}, {Schilke},
1707: {Kurtz}, \& {et al.}}]{ZAPATA+07}
1708: {Zapata}, L.~A., {Ho}, P.~T.~P., {Rodr{\'{\i}}guez}, L.~F., {Schilke}, P.,
1709: {Kurtz}, S., \& {et al.} 2007, \aap, 471, L59
1710:
1711: \bibitem[{{Zapata} {et~al.}(2005){Zapata}, {Rodr{\'{\i}}guez}, {Ho}, {Zhang},
1712: {Qi}, \& {Kurtz}}]{ZAPATA+05}
1713: {Zapata}, L.~A., {Rodr{\'{\i}}guez}, L.~F., {Ho}, P.~T.~P., {Zhang}, Q., {Qi},
1714: C., \& {Kurtz}, S.~E. 2005, \apjl, 630, L85
1715:
1716: \bibitem[{{Zapata} {et~al.}(2004){Zapata}, {Rodr{\'{\i}}guez}, {Kurtz},
1717: {O'Dell}, \& {Ho}}]{ZAPATA+04}
1718: {Zapata}, L.~A., {Rodr{\'{\i}}guez}, L.~F., {Kurtz}, S.~E., {O'Dell}, C.~R., \&
1719: {Ho}, P.~T.~P. 2004, \apjl, 610, L121
1720:
1721: \bibitem[{{Ziurys} {et~al.}(1990){Ziurys}, {Wilson}, \& {Mauersberger}}]{ZWM90}
1722: {Ziurys}, L.~M., {Wilson}, T.~L., \& {Mauersberger}, R. 1990, \apjl, 356, L25
1723:
1724: \end{thebibliography}
1725:
1726:
1727: \clearpage
1728: \begin{deluxetable}{lccccccccccc}
1729: \tabletypesize{\scriptsize}
1730: \rotate
1731: %\tabletypesize{\small}
1732: \tablewidth{0pt}
1733: \tablecaption{Long-wavelength fluxes for sources detected in $\lambda$1.3 mm
1734: continuum \label{tab:detections}}
1735: \tablehead{\colhead{ID} & \colhead{$\alpha$} &
1736: \colhead{$\delta$} & \colhead{$S_{\rm 880 \mu m}$ } &
1737: \colhead{$S_{\rm 1.3mm}$ } & \colhead{$S_{\rm 3mm}$ } &
1738: \colhead{$S_{\rm 3.6mm}$ } & \colhead{$S_{\rm 1.3 cm}$ } &
1739: \colhead{$S_{\rm 2cm}$ } & \colhead{$S_{\rm 6cm}$ } &
1740: \colhead{$S_{\rm 20cm}$ } \\
1741: & (J2000) & (J2000) & (mJy) & (mJy) & (mJy) & (mJy) & (mJy) & (mJy) & (mJy)
1742: & (mJy)}
1743: \startdata
1744: 147-323 & 5 35 14.73 & -5 23 22.91 & $ $ & $ 7.5 \pm 2.2$ & $< 5.3$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1745: 155-338 & 5 35 15.51 & -5 23 37.52 & $ $ & $ 9.2 \pm 2.4$ & $< 10.6$ & $ $ & $ $ & $ 9.3 \pm 4.6$ & $ 11.2 \pm 3.8$ & $ 3.5 \pm 0.8$ & $ $ \\
1746: 158-323 & 5 35 15.82 & -5 23 22.50 & $ $ & $ 9.5 \pm 2.3$ & $< 9.7$ & $ 11.4 \pm 2.0$ & $ 10.6 \pm 2.9$ & $ 11.2 \pm 1.5$ & $ 10.6 \pm 4.8$ & $ 7.3 \pm 0.7$ & $ $ \\
1747: 158-327 & 5 35 15.79 & -5 23 26.61 & $ $ & $ 12.8 \pm 2.4$ & $ 18.6 \pm 3.2$ & $ $ & $ $ & $ 9.2 \pm 1.4$ & $ 13.0 \pm 5.7$ & $ 7.5 \pm 1.3$ & $ $ \\
1748: 159-350 & 5 35 15.93 & -5 23 49.96 & $ $ & $ 42.7 \pm 3.7$ & $ 11.9 \pm 3.3$ & $ 13.1 \pm 2.0$ & $ $ & $ 10.5 \pm 4.2$ & $ 16.2 \pm 5.7$ & $ 7.6 \pm 0.8$ & $ $ \\
1749: 163-317 & 5 35 16.27 & -5 23 16.72 & $ 33.3 \pm 3.8$ & $ 7.6 \pm 2.0$ & $ 11.3 \pm 2.7$ & $ $ & $ 10.1 \pm 2.5$ & $ 11.1 \pm 1.2$ & $ 9.5 \pm 2.8$ & $ 10.8 \pm 2.2$ & $ $ \\
1750: 167-317 & 5 35 16.73 & -5 23 16.63 & $ 21.9 \pm 4.1$ & $ 15.0 \pm 2.2$ & $ 19.1 \pm 3.3$ & $ 25.8 \pm 2.0$ & $ 23.3 \pm 4.2$ & $ 25.5 \pm 5.0$ & $ 19.8 \pm 6.0$ & $ 6.8 \pm 0.8$ & $ $ \\
1751: 168-326NS & 5 35 16.82 & -5 23 26.21 & $ $ & $ 13.6 \pm 2.3$ & $< 7.6$ & $ 10.1 \pm 2.0$ & $ 14.8 \pm 5.0$ & $ 8.4 \pm 0.3$ & $ $ & $ $ & $ $ \\
1752: 170-337 & 5 35 16.96 & -5 23 37.04 & $ 38.1 \pm 5.2$ & $ 12.9 \pm 2.2$ & $< 5.9$ & $ $ & $ $ & $ 6.6 \pm 2.4$ & $ 13.6 \pm 3.3$ & $ 6.8 \pm 0.7$ & $ $ \\
1753: 171-340 & 5 35 17.05 & -5 23 39.59 & $ 18.3 \pm 4.6$ & $ 13.0 \pm 2.3$ & $< 6.1$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1754: 177-341 & 5 35 17.67 & -5 23 40.96 & $ $ & $ 15.8 \pm 2.1$ & $ 16.7 \pm 2.8$ & $ $ & $ $ & $ 10.8 \pm 3.7$ & $ 14.4 \pm 4.0$ & $ 7.7 \pm 0.7$ & $ $ \\
1755: HC180 & 5 35 17.39 & -5 24 0.30 & $ $ & $ 10.7 \pm 3.0$ & $< 4.6$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1756: HC189 & 5 35 14.53 & -5 23 56.00 & $ $ & $ 99.6 \pm 8.4$ & $< 9.1$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1757: HC246 & 5 35 15.68 & -5 23 39.10 & $ $ & $ 17.8 \pm 2.4$ & $< 7.5$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1758: HC254 & 5 35 13.86 & -5 23 35.00 & $ $ & $ 17.7 \pm 3.8$ & $< 14.6$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1759: HC295 & 5 35 17.57 & -5 23 24.90 & $ $ & $ 11.1 \pm 2.1$ & $< 5.9$ & $ $ & $ $ & $ 4.4 \pm 0.6$ & $ 5.9 \pm 2.6$ & $ $ & $ $ \\
1760: HC336 & 5 35 15.81 & -5 23 14.30 & $ $ & $ 6.7 \pm 2.0$ & $ 19.7 \pm 3.1$ & $ $ & $ 13.8 \pm 1.0$ & $ 10.0 \pm 5.0$ & $ 10.0 \pm 5.0$ & $ 3.7 \pm 0.7$ & $ $ \\
1761: HC350 & 5 35 16.06 & -5 23 7.30 & $ $ & $ 8.8 \pm 2.1$ & $< 6.1$ & $ $ & $ 3.5 \pm 2.2$ & $ 4.1 \pm 0.8$ & $ 6.6 \pm 2.0$ & $ $ & $ $ \\
1762: HC351 & 5 35 19.07 & -5 23 7.50 & $ $ & $ 8.7 \pm 2.8$ & $< 4.6$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1763: HC361 & 5 35 14.29 & -5 23 4.30 & $ $ & $ 19.2 \pm 3.2$ & $< 7.6$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1764: HC383 & 5 35 17.84 & -5 22 58.20 & $ $ & $ 7.0 \pm 2.3$ & $< 5.2$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1765: HC401 & 5 35 16.08 & -5 22 54.10 & $ $ & $ 7.3 \pm 2.1$ & $< 5.6$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1766: HC412 & 5 35 16.34 & -5 22 49.10 & $ $ & $ 7.7 \pm 2.3$ & $< 5.6$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1767: HC414 & 5 35 16.98 & -5 22 48.50 & $ $ & $ 9.1 \pm 2.2$ & $< 5.2$ & $ $ & $ $ & $ 3.8 \pm 0.9$ & $ 7.0 \pm 5.6$ & $ 10.5 \pm 0.5$ & $ $ \\
1768: HC418 & 5 35 18.08 & -5 22 47.10 & $ $ & $ 7.6 \pm 2.5$ & $< 5.5$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1769: HC436 & 5 35 18.38 & -5 22 37.50 & $ $ & $ 8.7 \pm 2.8$ & $< 5.5$ & $ $ & $ 16.6 \pm 1.3$ & $ 9.0 \pm 0.3$ & $ $ & $ 11.2 \pm 0.4$ & $ $ \\
1770: HC438 & 5 35 14.09 & -5 22 36.60 & $ $ & $ 67.8 \pm 14.2$ & $< 12.1$ & $ $ & $ $ & $ 2.0 \pm 0.2$ & $ $ & $ $ & $ $ \\
1771: HC440 & 5 35 17.36 & -5 22 35.80 & $ $ & $ 13.0 \pm 2.8$ & $< 7.0$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1772: HC495 & 5 35 13.52 & -5 22 19.60 & $ $ & $ 22.1 \pm 6.6$ & $< 6.4$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1773: HC498 & 5 35 18.96 & -5 22 18.80 & $ $ & $ 15.3 \pm 4.6$ & $< 4.3$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1774: HC514 & 5 35 16.43 & -5 22 12.20 & $ $ & $ 23.7 \pm 7.4$ & $< 4.6$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1775: HC771 & 5 35 14.86 & -5 22 44.10 & $ $ & $ 16.1 \pm 4.9$ & $< 6.7$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1776: LMLA162 & 5 35 14.40 & -5 23 50.84 & $ $ & $103.3 \pm 7.2$ & $< 10.0$ & $ $ & $ 1.0 \pm 0.1$ & $ $ & $ $ & $ $ & $ $ \\
1777: MM8 & 5 35 13.73 & -5 23 46.84 & $ $ & $407.2 \pm 27.5$ & $ 28.8 \pm 5.9$ & $ $ & $ 0.8 \pm 0.1$ & $ $ & $ $ & $ $ & $ $ \\
1778: MM13 & 5 35 13.75 & -5 24 7.74 & $ $ & $317.5 \pm 25.1$ & $ 36.9 \pm 8.3$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1779: MM21 & 5 35 13.57 & -5 23 59.04 & $ $ & $153.6 \pm 13.2$ & $ 15.6 \pm 4.7$ & $ $ & $ 0.9 \pm 0.1$ & $ $ & $ $ & $ $ & $ $ \\
1780: MM22 & 5 35 13.66 & -5 23 54.94 & $ $ & $123.0 \pm 12.7$ & $ 17.1 \pm 4.2$ & $ $ & $ 0.4 \pm 0.1$ & $ $ & $ $ & $ $ & $ $ \\
1781: MM23 & 5 35 14.00 & -5 22 45.04 & $ $ & $ 61.2 \pm 7.4$ & $ 9.6 \pm 3.1$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1782: MM24 & 5 35 14.62 & -5 22 28.94 & $ $ & $167.4 \pm 27.0$ & $< 32.7$ & $ $ & $ $ & $ $ & $ $ & $ $ & $ $ \\
1783: \enddata
1784: \tablecomments{
1785: All quoted uncertainties are
1786: 1$\sigma$; uncertainties (of 10\%) in the overall flux scale
1787: are not included in this table. The MM sources are detected only at
1788: wavelengths $\ga 1$ mm. HC objects are known near-IR cluster members
1789: from \citet{HC00} that are also detected at mm wavelengths, and the sources
1790: with numerical labels are a subset of this sample that are also detected
1791: optically as proplyds \citep[e.g.,][]{OWH93}.
1792: LMLA162 is absent from \citet{HC00},
1793: but seen at 3.6 $\mu$m by \citet{LADA+04}; we also see a faint 2 $\mu$m point
1794: source at this location in a 2MASS image. MM8 and MM13
1795: were detected by \citet{EC06}, and MM21 was detected by \citet{EC06}
1796: but associated with HC 178 (we find that MM21 and HC 178 are separated
1797: by an angle larger than our relative positional uncertainties).
1798: LMLA162, MM8, MM13, MM21, and MM22 were also detected in previous
1799: 1.3 mm observations by \citet{ZAPATA+05}. MM23 and MM24 are newly detected
1800: here.}
1801: \tablerefs{References for the fluxes are as follows: 880 $\mu$m
1802: \citep{WAW05}; 1.3 mm (this work); 3 mm \citep{EC06}; 3.6 mm \citep{MLL95};
1803: 1.3 cm \citep{ZAPATA+04}; 2 and 6 cm \citep{FELLI+93b}; 20 cm
1804: \citep{FELLI+93a}. Quoted upper limits for the 3 mm fluxes are 3$\sigma$.}
1805: \end{deluxetable}
1806:
1807:
1808: \begin{deluxetable}{lcccc}
1809: \tabletypesize{\scriptsize}
1810: %\tabletypesize{\small}
1811: \tablewidth{0pt}
1812: \tablecaption{Derived quantities for detected sources
1813: \label{tab:results}}
1814: \tablehead{\colhead{ID} & {$S_{\rm \nu, dust}$ (mJy)} &
1815: \colhead{$M_{\rm circumstellar}$ (M$_{\odot}$)} &
1816: \colhead{$R_{\rm disk}$ (AU)} & \colhead{$S_{\nu, \tau \ga 1}$ (mJy)}}
1817: \startdata
1818: 147-323 & $ 7.5 \pm 2.2$ & 0.009 $\pm$ 0.003 & 88 & 83 \\
1819: 155-338 & $ 4.0 \pm 4.0$ & 0.005 $\pm$ 0.005 & 102 & 112 \\
1820: 158-323 & $ 1.0 \pm 1.0$ & 0.001 $\pm$ 0.001 & 105 & 119 \\
1821: 158-327 & $ 5.0 \pm 4.0$ & 0.006 $\pm$ 0.005 & 122 & 161 \\
1822: 159-350 & $ 33.0 \pm 7.0$ & 0.042 $\pm$ 0.009 & 152 & 250 \\
1823: 163-317 & $ 6.0 \pm 2.0$ & 0.008 $\pm$ 0.003 & 93 & 93 \\
1824: 167-317 & $ 2.0 \pm 2.0$ & 0.003 $\pm$ 0.003 & 122 & 161 \\
1825: 168-326NS & $ 7.0 \pm 4.0$ & 0.009 $\pm$ 0.005 & $<$100 & $<108$ \\
1826: 170-337 & $ 10.0 \pm 2.5$ & 0.013 $\pm$ 0.003 & 126 & 171 \\
1827: 171-340 & $ 13.0 \pm 2.3$ & 0.016 $\pm$ 0.003 & 80 & 69 \\
1828: 177-341 & $ 4.0 \pm 4.0$ & 0.005 $\pm$ 0.005 & 177 & 339 \\
1829: HC180 & $ 10.7 \pm 3.0$ & 0.013 $\pm$ 0.004 & $<$100 & $<108$ \\
1830: HC189 & $ 99.5 \pm 8.4$ & 0.125 $\pm$ 0.011 & $<$100 & $<108$ \\
1831: HC246 & $ 17.8 \pm 2.4$ & 0.022 $\pm$ 0.003 & $<$100 & $<108$ \\
1832: HC254 & $ 17.7 \pm 3.8$ & 0.022 $\pm$ 0.005 & $<$100 & $<108$ \\
1833: HC295 & $ 8.0 \pm 3.5$ & 0.010 $\pm$ 0.004 & 187 & 381 \\
1834: HC336 & $ 0.0 \pm 0.0$ & 0.000 $\pm$ 0.000 & $<$100 & $<108$ \\
1835: HC350 & $ 5.0 \pm 3.5$ & 0.006 $\pm$ 0.004 & $<$100 & $<108$ \\
1836: HC351 & $ 8.7 \pm 2.8$ & 0.011 $\pm$ 0.003 & $<$100 & $<108$ \\
1837: HC361 & $ 19.2 \pm 3.2$ & 0.024 $\pm$ 0.004 & $<$100 & $<108$ \\
1838: HC383 & $ 7.0 \pm 2.3$ & 0.009 $\pm$ 0.003 & 202 & 441 \\
1839: HC401 & $ 7.3 \pm 2.1$ & 0.009 $\pm$ 0.003 & $<$100 & $<108$ \\
1840: HC412 & $ 7.7 \pm 2.3$ & 0.010 $\pm$ 0.003 & 161 & 283 \\
1841: HC414 & $ 3.0 \pm 3.0$ & 0.004 $\pm$ 0.004 & $<$100 & $<108$ \\
1842: HC418 & $ 7.6 \pm 2.5$ & 0.010 $\pm$ 0.003 & $<$100 & $<108$ \\
1843: HC436 & $ 1.0 \pm 1.0$ & 0.001 $\pm$ 0.001 & $<$100 & $<108$ \\
1844: HC438 & $ 67.8 \pm 14.2$ & 0.085 $\pm$ 0.018 & 200 & 433 \\
1845: HC440 & $ 13.0 \pm 2.8$ & 0.016 $\pm$ 0.004 & $<$100 & $<108$ \\
1846: HC495 & $ 22.1 \pm 6.6$ & 0.028 $\pm$ 0.008 & $<$100 & $<108$ \\
1847: HC498 & $ 15.3 \pm 4.6$ & 0.019 $\pm$ 0.006 & $<$100 & $<108$ \\
1848: HC514 & $ 23.7 \pm 7.4$ & 0.030 $\pm$ 0.009 & 166 & 299 \\
1849: HC771 & $ 16.1 \pm 4.9$ & 0.020 $\pm$ 0.006 & 176 & 337 \\
1850: LMLA162 & $103.3 \pm 7.2$ & 0.130 $\pm$ 0.009 & $<$100 & $<108$ \\
1851: MM8 & $407.2 \pm 27.5$ & 0.512 $\pm$ 0.035 & $<$100 & $<108$ \\
1852: MM13 & $317.5 \pm 25.1$ & 0.399 $\pm$ 0.032 & $<$100 & $<108$ \\
1853: MM21 & $153.6 \pm 13.2$ & 0.193 $\pm$ 0.017 & $<$100 & $<108$ \\
1854: MM22 & $123.0 \pm 12.7$ & 0.155 $\pm$ 0.016 & $<$100 & $<108$ \\
1855: MM23 & $ 61.2 \pm 7.4$ & 0.077 $\pm$ 0.009 & $<$100 & $<108$ \\
1856: MM24 & $167.3 \pm 27.0$ & 0.211 $\pm$ 0.034 & 217 & 514 \\
1857: \enddata
1858: \tablecomments{$S_{\rm \nu,dust}$ is the component of the observed
1859: 1.3 mm emission due to cool dust, determined from a fit to long-wavelength
1860: fluxes of a model including thermal free-free emission as well as dust
1861: emission (see Figure \ref{fig:seds}). Quoted uncertainties are
1862: 1$\sigma$, and do not include systematic uncertainties associated with the
1863: overall flux calibration or conversion from flux to mass.}
1864: \end{deluxetable}
1865:
1866:
1867:
1868: %\epsscale{0.7}
1869: %\begin{figure}
1870: %\plotone{figs/orion_vlt_jhk.epsi}
1871: %\caption[$JHK$ mosaic of the Orion Nebula cluster]{A $7' \times 7'$
1872: %$JHK$ mosaic of the Orion Nebula cluster from VLT/ISAAC (courtesy of
1873: %Mark McCaughrean and the European Southern Observatory).
1874: %The bright, hot Trapezium stars are seen toward the center of the image,
1875: %and hundreds of low-mass stars are seen in the central few arcminutes.
1876: %\label{fig:onc_jhk}}
1877: %\end{figure}
1878:
1879: \epsscale{1.0}
1880: \begin{figure}
1881: \plotone{f1.eps}
1882: \caption[Pointing positions for the ONC mosaics]
1883: {Pointing positions for the CARMA mosaic (triangle symbols) and
1884: the SMA mosaic (square symbols) plotted over the positions of $K$-band sources
1885: in the ONC (open circles). These $K$-band source positions are drawn
1886: from \citet{HC00}, and have been registered to the 2MASS astrometric grid.
1887: The unit gain contour of the CARMA mosaic (dotted curve), the SMA mosaic
1888: (dashed curve), and the mosaic produced from the combined SMA+CARMA dataset
1889: (solid curve) are also indicated.
1890: \label{fig:pointings}}
1891: \end{figure}
1892:
1893: \epsscale{1.0}
1894: \begin{figure}
1895: \plotone{f2.eps}
1896: \caption{The Orion Nebula cluster, imaged in $\lambda$1.3 mm
1897: continuum with CARMA. Only data
1898: observed on long baselines ($r_{uv}>70$ k$\lambda$) were
1899: used to create this image. The unit gain region of the mosaic
1900: is indicated by a solid curve. The synthesized beam
1901: has FWHM of $0\rlap{.}''61 \times 0\rlap{.}''52$ at a PA of $70^{\circ}$.
1902: The RMS noise varies within the unit gain contour from 2.3 to 23 mJy
1903: (1$\sigma$), with a mean RMS of 4.6 mJy. Partially resolved extended
1904: emission in the BN/KL and OMC1-S regions increases the noise level in
1905: the upper and lower right corners of the image. The map has been divided
1906: by the theoretical RMS, to visually down-weight the noisier edges.
1907: \label{fig:mos}}
1908: \end{figure}
1909:
1910: \epsscale{1.0}
1911: \begin{figure}
1912: \plotone{f3.eps}
1913: \caption{The Orion Nebula cluster, imaged in $\lambda$1.3 mm
1914: continuum with the SMA. Only data
1915: observed on long baselines ($r_{uv}>70$ k$\lambda$) were
1916: used to create the image. The unit gain region of the mosaic
1917: is indicated by a solid curve. The synthesized beam has a roughly circularly
1918: symmetric core with a FWHM of $0\rlap{.}''98$. The RMS within the unit gain
1919: contour varies from 0.8 to 28 mJy, with a mean value of 2.7 mJy. The BN/KL
1920: and OMC1-S regions produce the strong emission visible in the upper and lower
1921: right corners of the map.
1922: The map has been divided
1923: by the theoretical RMS, to visually down-weight the noisier edges.
1924: \label{fig:sma}}
1925: \end{figure}
1926:
1927:
1928: \epsscale{1.0}
1929: \begin{figure}
1930: \plotone{f4.eps}
1931: \caption[$\lambda$1 mm continuum mosaic of the ONC]
1932: {The Orion Nebula cluster, imaged in $\lambda$1.3 mm
1933: continuum with CARMA and the SMA (greyscale). CARMA and SMA
1934: data were combined in the $uv$ plane, and only data
1935: observed on long baselines ($r_{uv}>70$ k$\lambda$) were
1936: used to create this image. The angular resolution is
1937: $\sim 0\rlap{.}''7 \times 0\rlap{.}''6$. The unit gain region of the mosaic
1938: encompasses a $2' \times 2'$ area, as indicated by the solid contour,
1939: and the RMS residuals (1$\sigma$) within this region vary from
1940: 1.8 mJy in regions devoid of bright emission to $\ga 10$ mJy in the
1941: crowded regions toward the right of the map.
1942: Millimeter detections above the 3$\sigma$ level coincident with
1943: near-IR cluster members and proplyds, and sources without counterparts
1944: detected above the 5$\sigma$ level (\S \ref{sec:thresh}),
1945: are indicated by gray circles.
1946: The map has been divided
1947: by the theoretical RMS, to visually down-weight the noisier edges.
1948: \label{fig:map_combo}}
1949: \end{figure}
1950:
1951: %\epsscale{0.7}
1952: %\begin{figure}
1953: %\plotone{figs/noise_image.epsi}
1954: %\caption[Noise distribution in the ONC OVRO mosaic]
1955: %{Distribution of 3 mm continuum fluxes for all pixels within the unit gain
1956: %contour of a residual image with bright point sources removed using CLEAN
1957: %(solid line), and the frequency distribution expected for Gaussian
1958: %noise with a mean of zero and a standard deviation of 1.75 mJy (dotted line).
1959: %The agreement between the
1960: %measured and Gaussian distributions suggests that the noise in the OVRO mosaic
1961: %is represented well by Gaussian statistics, despite variations in the
1962: %RMS across the image.
1963: %\label{fig:noise}}
1964: %\end{figure}
1965:
1966: \epsscale{1.0}
1967: \begin{figure}
1968: \plotone{f5a.eps}
1969: \caption{Contour images of sources detected in 1 mm continuum emission.
1970: Contour increments are 1$\sigma$, beginning at $\pm 2 \sigma$, where
1971: $\sigma$ is determined locally for each object (see \S \ref{sec:thresh}).
1972: Solid contours represent positive emission, and dotted contours trace negative
1973: features. The BN object is excluded from these plots, as it is
1974: discussed in more detail in a later paper.
1975: For sources detected at infrared wavelengths, individual images are
1976: centered on the previously measured near-IR coordinates. For the MM
1977: sources, detected only at $\ga 1$ mm wavelengths, images are
1978: centered on the peak fluxes.
1979: %The BN object
1980: %(HC 705) and the massive star $\theta^1$OriA (HC 336) are also included
1981: %in this plot, as are the two sources for which the 3 mm flux is
1982: %dominated by free-free emission (HC 322 and HC 323).
1983: \label{fig:detections}}
1984: \end{figure}
1985:
1986: \clearpage
1987: \epsscale{1.0}
1988: {\plotone{f5b.eps}}\\[5mm]
1989: \centerline{Fig. 5. --- continued.}
1990: \clearpage
1991:
1992: \epsscale{1.0}
1993: \begin{figure}
1994: \plotone{f6a.eps}
1995: \caption{Millimeter and radio fluxes for our sample (points), along with
1996: best-fit models including free-free and thermal dust emission.
1997: Models including free-free and dust emission are indicated by solid lines, and
1998: dotted and dashed lines show dust-only models with $\beta=0$ and $\beta=1$,
1999: respectively. The free-free flux density
2000: is proportional to $\nu^{-0.1}$ for gas that is optically thin for all radii,
2001: and to $\nu^{0.6}$ for a partially optically thick gaseous wind; the
2002: emission is
2003: thus parameterized by the flux at a single wavelength and a turnover
2004: frequency. The emission from cool dust is proportional to $\nu^{2+\beta}$.
2005: \label{fig:seds}}
2006: \end{figure}
2007:
2008: \clearpage
2009: \epsscale{1.0}
2010: {\plotone{f6b.eps}}\\[5mm]
2011: \centerline{Fig. 6. --- continued.}
2012: \clearpage
2013:
2014: \epsscale{1.0}
2015: \begin{figure}
2016: \plotone{f7.eps}
2017: \caption{
2018: Average image, obtained by stacking the 1 mm continuum emission observed
2019: toward each of 226 low-mass near-IR sources not detected
2020: individually above the 3$\sigma$ level. Contour levels begin at
2021: $\pm 2 \sigma$ and the contour interval is $1 \sigma$ (negative contours
2022: are shown as dotted lines). Emission is detected
2023: for the ensemble at a significance of $\ga 4\sigma$, and exhibits
2024: a compact (and beam-like) morphology approximately centered on the origin.
2025: The degree to which the emission is smeared out is consistent with
2026: the $\sim 0\rlap{.}''4$ positional errors in the near-IR source positions.
2027: %In addition, the average
2028: %image shows negative emission that resembles sidelobe
2029: %features in the OVRO beam, suggesting that the average emission is
2030: %point-like.
2031: \label{fig:avg}}
2032: \end{figure}
2033:
2034: \epsscale{1.0}
2035: \begin{figure}
2036: \plotone{f8.eps}
2037: \caption{Contour image of the 1.3 mm wavelength
2038: continuum emission toward the proplyd
2039: 177-341, overlaid on an HST map \citep{BOM00}
2040: of the same position. Positive contours
2041: are white, solid curves and negative contours are dotted black curves.
2042: Contours are $\pm 2,4$ and 6$\sigma$. The FWHM of the synthesized beam is
2043: shown as a filled back symbol.
2044: \label{fig:proplyds}}
2045: \end{figure}
2046:
2047: \epsscale{1.0}
2048: \begin{figure}
2049: \plotone{f9.eps}
2050: \caption[Disk mass, as a function of cluster age]
2051: {Average disk mass as a function of age for the NGC 2024, ONC, and
2052: IC 348 clusters. The disk masses are taken from this work, \citet{EC03},
2053: and \citet{CARPENTER02}, and estimated cluster ages and uncertainties are from
2054: \citet{MEYER96}, \citet{ALI96}, \citet{HILLENBRAND97}, \citet{LUHMAN+98},
2055: and \citet{LUHMAN99}. Average disk masses for NGC 2024 and IC 348 were
2056: measured at 3 mm, where potential contributions from free-free
2057: emission would be stronger than for the average mass measured here for
2058: the ONC at 1 mm. We argue in \S \ref{sec:disc_evol}, however, that free-free
2059: contamination is unlikely in NGC 2024 and IC 348.
2060: \label{fig:evol}}
2061: \end{figure}
2062:
2063: %\epsscale{1.0}
2064: %\begin{figure}
2065: %\plotone{figs/flux_size.eps}
2066: %\caption{The angular size (or upper limits) of detected sources, plotted
2067: %against inferred circumstellar masses. Disk radii are either taken from
2068: %\citet{VA05}, or are equal to half the sizes measured from our 1 mm
2069: %images. We assume error bars of 10\% for the plotted sizes.
2070: %\label{fig:sizes}}
2071: %\end{figure}
2072:
2073: \epsscale{1.0}
2074: \begin{figure}
2075: \plotone{f10.eps}
2076: \caption{Circumstellar mass as a function of radial distance from the
2077: center of the Trapezium stars for detected objects.
2078: \label{fig:radii}}
2079: \end{figure}
2080:
2081: \epsscale{1.0}
2082: \begin{figure}
2083: \plotone{f11.eps}
2084: \caption{Circumstellar mass as a function of stellar mass for the
2085: subset of detected objects where spectroscopically determined stellar masses
2086: are available \citep{HILLENBRAND97,LUHMAN+00,SLESNICK+04}.
2087: \label{fig:mstars}}
2088: \end{figure}
2089:
2090:
2091: \end{document}
2092:
2093:
2094: \section{Results \label{sec:detections}}
2095:
2096:
2097:
2098:
2099:
2100:
2101:
2102: \subsection{Ensemble of Non-Detected Sources \label{sec:nondet}}
2103: Only 3\% of the known near-IR cluster members in the Trapezium region
2104: are detected in 3 mm continuum emission, and thus the vast majority of objects
2105: are undetected in our observations. In order to examine the flux
2106: distribution for ``typical'' low-mass stars in the ONC, we consider the
2107: distribution of 3 mm fluxes observed for the ensemble of the
2108: remaining 326 undetected near-IR cluster members; the large number
2109: of sources allows an enhanced sensitivity for the ensemble compared to
2110: that for an individual object.
2111:
2112: Figure \ref{fig:hist_o2} shows the distribution of mm-wavelength
2113: fluxes observed toward 326 $K$-band sources in the ONC, none of which are
2114: detected individually above the 3$\sigma$ level.
2115: The 30 bright point sources visible in Figure \ref{fig:map_o2}
2116: were removed using CLEAN before computing this histogram.
2117: We also plot the flux distribution measured for all other pixels within the
2118: unit gain contour (after subtraction of bright point sources).
2119: The mean flux observed for the ensemble of 326 near-IR cluster members is
2120: 0.267 mJy, and the standard deviation in the mean is $8.75 \times 10^{-5}$ Jy.
2121: The flux distribution is thus biased to positive values with respect to the
2122: noise distribution, and the significance of this bias is 3$\sigma$.
2123: Thus, it appears
2124: that while none of these $K$-band objects are detected in 3 mm continuum
2125: emission above the 3$\sigma$ level, there may be weak emission below the
2126: 3$\sigma$ level from circumstellar disks.
2127:
2128: This positive bias is also illustrated in the right panel of
2129: Figure \ref{fig:hist_o2}, which shows an
2130: average image of the 3 mm flux observed toward $K$-band sources,
2131: obtained by averaging $10'' \times 10''$ images centered around each object.
2132: We CLEANed this average image in order to remove negative features
2133: associated with sidelobe emission. However, some negative features
2134: remain as a result of residual calibration errors which prevent
2135: perfect CLEANing.
2136:
2137: The ``average'' disk is detected at a significance of $\sim 3\sigma$, and
2138: is centered on the mean position of $K$-band sources (within positional
2139: uncertainties). In addition, the FWHM of the emission in Figure
2140: \ref{fig:hist_o2}, $2\rlap{.}''5 \times 1\rlap{.}''3$, is comparable to
2141: the FWHM of the beam, suggesting
2142: that the average flux is dominated by small-scale emission such as that
2143: expected for circumstellar disks.
2144: %However, the average image is slightly
2145: %extended with respect to the beam (which has a FWHM of
2146: %$1\rlap{.}''9 \times 1\rlap{.}''5$), suggesting that larger structures
2147: %(of diameter $\sim 1200$ AU) may also contribute to the average image.
2148: Thus we conclude that the positive bias observed in Figure
2149: \ref{fig:hist_o2} probably represents underlying
2150: weak mm-wavelength emission from point sources.
2151:
2152: Because the sensitivity of previous radio-wavelength surveys
2153: \citep{FELLI+93a,FELLI+93b} is comparable to this average flux, we can
2154: not rule out some contribution of free-free emission to the mean flux.
2155: While the average flux arising from cool dust emission may therefore be somewhat
2156: lower than 0.27 mJy, we ignore this effect here, and await
2157: more sensitive long-wavelength observations to provide better
2158: constraints on low levels of free-free emission.
2159: For the 326 known $K$-band sources, which probably have stellar masses between
2160: 0.1 and 1 M$_{\odot}$ \citep{HC00}, our assumptions in Equation \ref{eq:dustmass}
2161: should be valid, and the conversion from 3 mm flux into mass should be reliable
2162: (modulo the uncertainties concerning free-free emission discussed above).
2163: The mean
2164: circumstellar mass among the 326 low-mass $K$-band sources within the unit
2165: gain contour of our OVRO mosaic is $0.0055 \pm 0.0018$ M$_{\odot}$.
2166:
2167: