astro-ph0211505/ms.tex
1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %%              B68 and L1544 H2O  PAPER                                        %%  
3: %%                                                                              %%
4: %%              EDITED: 03/21/01 EAB                                            %%
5: %%                                                                              %%
6: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
7: %\documentclass{aastex}
8: %\usepackage{emulateapj5}
9: \documentclass[12pt, preprint]{aastex}
10: %\documentclass[preprint2]{aastex}
11: 
12: \input {macro.tex}
13: %\input{macro.tex}
14: 
15: %\def\Hthreep{H$_3^+$}          % H3+
16: %\def\NtwoHp{N$_2$H$^+$}        % N2H+ 
17: %\def\CHthreeOH{CH$_3$OH}        % CH3OH
18: %\def\CeiO{C$^{18}$O}        % N2H+ 
19: %\def\CtfS{C$^{34}$S}        % N2H+ 
20: %\def\Ntwo{N$_2$}               % N2
21: %\def\Htwo{H$_2$}               % N2
22: %\def\av{$A_V$}               % N2
23: \def\swash2o{$1_{10} - 1_{01}$}               % N2
24: \def\swas13co{J~$=~5~-~4$}
25: \def\isoh2o{2$_{12}-1_{01}$}               % N2
26: \def\h2oabun{$x$(o-H$_2$O)}
27: \def\oHtwoO{o-H$_2$O}               % H2O
28: %\newcommand{\nhtwo}{\mbox{n$_{H_{2}}$}}
29: %\newcommand{\cc}{\mbox{cm$^{-3}$}}
30: 
31: 
32: 
33: 
34: \begin{document}
35: \setcounter{figure}{0}
36: 
37: \title{Sensitive Limits on the Water Abundance in Cold Low Mass Molecular Cores}
38: 
39: \author{Edwin A. Bergin}
40: \affil{Harvard-Smithsonian Center for Astrophysics, 60 Garden Street,
41: Cambridge, MA 02138} \email{ebergin@cfa.harvard.edu}
42: 
43: \author{Ronald L. Snell}
44: \affil{
45: Department of Astronomy,
46: University of Massachusetts, Amherst, MA 01003
47: } \email{snell@astro.umass.edu}
48: 
49: 
50: \begin{abstract}
51: We present SWAS observations of water vapor in two cold star-less clouds, B68 and Core D in 
52: $\rho$ Ophiuchus. Sensitive non-detections of the \swash2o\ transition of \oHtwoO\ are
53: reported for each source.  Both molecular cores have been previously examined by
54: detailed observations that have characterized the physical structure.  Using
55: these rather well defined physical properties and a Monte-Carlo radiation transfer
56: model we have removed one of the largest uncertainties from the abundance calculation
57: and set the lowest water abundance limit to date in cold low-mass
58: molecular cores.  These limits are $x$(\oHtwoO ) $< 3 \times 10^{-8}$ (relative to H$_2$) 
59: and $x$(\oHtwoO ) $< 8 \times 10^{-9}$ in B68 and $\rho$ Oph D, respectively.
60: Such low abundances confirm the general lack of ortho-water vapor in cold
61: (T $< 20$ K) cores.  Provided that the ortho/para ratio of water is not near
62: zero, these limits are well below theoretical predictions and appear to support 
63: the suggestion that most of the water in dense low-mass cores is frozen onto
64: the surfaces of cold dust grains.
65: \end{abstract}
66: 
67: 
68: \keywords{
69: astrochemistry -- ISM: molecules -- stars: formation -- submillimeter
70: }
71: 
72: \section{Introduction}
73: 
74: 
75: Water is an important molecule in the interstellar medium (ISM) because it
76: links chemistry in the ISM to comets and planetary systems and provides
77: crucial aid in constraining the chemistry of 
78: astrophysical systems \citep{bergin_imp}.   Recently
79: NASA's Submillimeter Wave Astronomy Satellite (SWAS) and
80: ESA's Infrared Space Observatory (ISO)
81: have provided the first unambiguous glimpse of water in the ISM.
82: Because the SWAS instrument is pre-tuned to the ground state rotational transition
83: of ortho-\HtwoO , it is sensitive primarily to water in the quiescent or coldest regions
84: of molecular clouds.  In contrast, ISO sampled higher energy transitions that consequently
85: probed warm
86: gas found in shocks or in close proximity to young embedded stars. 
87: One of the surprises from the analysis of SWAS results is that the derived water abundance
88: is unexpectedly low, with values \h2oabun\ $\sim 1 - 10 \times 10^{-9}$ (relative to
89: \Htwo ) found in a variety of regions (Snell et al. 2000a,b,c).
90: These abundances conflicted with theoretical expectations and several suggestions
91: have been provided to account for the discrepancy, with the principal solution 
92: for cold gas being the freeze-out of water onto grain surfaces 
93: \citep{bergin_imp, spaans_h2o, viti_h2o, charnley_swasiso}.   Moreover, ISO and SWAS detections of
94: water vapor in absorption
95: produced results in disagreement with the emission line analysis.   Absorption lines are more straightforward to
96: analyze and provide water abundances of \h2oabun\ $\sim 10^{-6}$,
97: in agreement with theory \citep{cernicharo_h2oabs, neufeld_h2oabs, moneti_h2oabs}.
98: These differences can be reconciled
99: provided that the freeze-out of water is greater in denser regions, which are
100: seen in emission, as opposed to the low density gas, with correspondingly longer
101: depletion timescales, seen in absorption \citep{bergin_imp}.
102: 
103: Nonetheless there are some well discussed uncertainties in the analysis of 
104: water emission lines detected by SWAS
105: (Snell et al. 2000a,b,c,).  These arise primarily from the fact that
106: only a small column of water is required to produce optically thick emission.
107: The situation is slightly simplified, because the
108: critical density for the \swash2o\ transition ($\sim 10^{8}$ \cc )
109: is higher than typical densities found in molecular cores.  Under this condition
110: every photon that is collisionally created eventually escapes and the water emission,
111: although optically thick, is effectively thin \citep{wannier_h2o, snell_omc1}.
112: This fact aids in constraining the abundance, but crucial
113: assumptions must be made regarding the source physical structure.  
114: Assuming that the \swash2o\ emission arises predominantly
115: from high density gas, simple single component models were constructed
116: (i.e. single density and temperature along the line of sight), resulting
117: in the above abundance estimates (Snell et al. 2000a,b,c).
118: However, uncertainties in the density are directly reflected in the abundance, and
119: a single density characterizing the entire line of sight is itself an approximation.  
120: 
121: More realistic physical structures were examined for a few 
122: molecular cores using a Monte-Carlo radiative transfer code \citep{ashby_cores}.  
123: This analysis supported the simple single component results.  However,
124: there are complications.  For instance in S140, a separate analysis
125: argues that a clumpy cloud model is more appropriate, although the resulting abundance
126: estimates are similar \citep{spaans_h2o}.  
127: %For $\rho$ Oph A the 
128: %presence of extended molecular emission (e.g. CS), and the expectation of
129: %strong temperature gradients needed to account for the infrared radiation field
130: %complicated the effort to construct a reliable physical model.  
131: Sources that offer the most promise to set good limits or determinations of
132: the water abundance 
133: are therefore ones that are largely 
134: isolated, and have well established physical properties.   In general, giant
135: molecular cloud cores, such as S140, do not fit this criteria 
136: as they have been suggested to have clumpy physical
137: structures which are difficult to model definitively.
138: On the other hand,
139: advances in the resolution and sensitivity of infrared (IR) and submillimeter continuum
140: arrays have begun to provide
141: a wealth of information on the physical structure of low mass cores.   
142: Studies using these instruments effectively apply similar techniques in
143: which azimuthally averaged dust column density
144: profiles are combined with assumptions regarding the geometry and a gas-to-dust ratio to 
145: determine the H$_2$ density profile.  
146: 
147: We present here SWAS observations of two molecular cores, B68 and $\rho$ Oph D, that have been
148: subject to these techniques and thus have rather reliable estimates of the 
149: source physical structure.    Using these physical models and SWAS observations
150: we have set the lowest limit to date on the water abundance in cold low mass 
151: cores.  The limits are below pure gas-phase chemical
152: predictions and support the assertion that most of the water is 
153: likely frozen onto grain surfaces in cold dense cores of molecular clouds.
154:  
155: \section{Observations and Results}
156: 
157: Between Feb. 25, 2001 to Mar. 8, 2001  SWAS observed B68 in the \swash2o\ transition
158: of \oHtwoO\ at 556.936 GHz for a total of 28 hours of on-source integration.  Similarly during 
159: Sept. 7-10, 2001 SWAS observed $\rho$ Oph~D for a total of 24.6 hrs (on-source).    System temperatures
160: were typically 2500 K with minimal scatter around that value.
161: The spacecraft
162: was used in nod mode, involving alternately nodding the spacecraft to an off-source 
163: position free of emission.   For B68 the elliptical  3\farcm3 $\times$ 4\farcm5 beam
164: was centered on $\alpha =$ 17:22:38.2, 
165: $\delta =$ $-$23:49:34 (J2000) and towards $\rho$ Oph~D, $\alpha =$ 16:28:30.4, 
166: $\delta =$ $-$24:18:29 (J2000).
167: All data were reduced with the standard SWAS
168: pipeline described by \citet{melnick_swas}.  
169: Towards both cores, along with \oHtwoO ,
170: SWAS simultaneously observed transitions
171: of [C I] ($^3$P$_1$ $\rightarrow$ $^3$P$_0$), \thCO\ (J = 5 $\rightarrow$ 4),
172: and \Otwo\ ($3_3 \rightarrow 1_2$).  In this work we primarily use the 
173: \oHtwoO\ and \thCO\ data, with a velocity resolution of 1.0 \kms , and a sampling
174: of 0.6 \kms .    All data are presented here on the T$_A^*$ scale and for 
175: subsequent analysis are scaled
176: by the main beam efficiency of 90\% \citep{melnick_swas}.
177: 
178: Figure 1 shows  spectra of the \swash2o\ transition of \oHtwoO\  and \thCO\
179: \swas13co\ taken towards
180: B68 and $\rho$ Oph~D.   
181: It is apparent that there are no water detections toward either source.
182:  In B68 the 3$\sigma$ upper limit is T$_A^* =$
183: 36 mK and in $\rho$ Oph D the limit is 45 mK.   
184: The J $= 5 - 4$ transition of \thCO\ was not detected towards B68 with a similar
185: limit as \oHtwoO , confirming the cold (T $\le$ 10) nature of this source.    \thCO\ \swas13co\
186: is detected towards $\rho$ Oph D with an integrated intensity of 0.30 K km s$^{-1}$,
187: a peak antenna temperature of T$_A^* =$ 150 mK, and $\Delta$v $= 1.86$ km s$^{-1}$.
188: 
189: \section{Monte-Carlo Models of Water Emission}
190: 
191: To estimate the water abundance we draw upon previous determinations of the 
192: density structure derived from observations of dust in emission or 
193: absorption \citep{motte_oph, alves_b68, bacmann_dust}. 
194: %In general to derive the dust column density three techniques
195: %are used:  (1) dust emission maps at 450, 850, and 1300 $\mu$m 
196: %\citep[e.g.,][]{motte_oph, awb_ppiv,  shirley_dust},
197: %(2) infrared star counting techniques at J, H, and K bands (Lada et al. 1999; 
198: %Alves, Lada, \& Lada 2001), and (3) mapping the infrared absorption of cores
199: %against the diffuse mid--infrared background (Bacmann et al. 2000).   
200: %Each of these techniques suffer from various complications, but have greatly
201: %enhanced our understanding of the structure of low mass cores.
202: Towards both clouds, the SWAS beam encompasses the entire area seen in
203: transitions of molecules such as \CeiO , CS, and \NtwoHp .
204: Higher spatial resolution observations of molecular cores in these other tracers
205: are able to determine
206: the extent of emission, which along with the radial 
207: density profile with radius,  permits a more accurate analysis
208: of abundances
209: \citep{jessop_dep, tafalla_dep, bergin_b68, hotzel_b68}.
210: Because of the large beamsize our initial procedure is to assume 
211: a simple model with an abundance of \oHtwoO\ that is constant with cloud depth.
212: However, chemical models of centrally condensed objects 
213: predict that lower abundances should exist in the dense central
214: regions where molecules freeze onto grain surfaces more frequently
215: \citep{rawlings_dep}. 
216: % in the
217: %low density outer layers the gas-grain collision timescales are longer and
218: %higher abundances are predicted 
219: To account for this possibility we have also examined whether a 
220: water abundance profile predicted by theory is in conflict
221: with our observations.  This model is applied to B68 which has
222: a large amount of additional observational constraints.  
223: %We note that SWAS, with its larger beam, is insensitive to
224: %structures smaller than 4$'$ and thus SWAS data alone 
225: %cannot discriminate between simple constant abundance models and models
226: %with complex abundance structure.
227: 
228: The water abundance profile, 
229: along with the adopted density, temperature, and velocity width
230: profiles, is used as input into the spherical one-dimensional
231: Monte-Carlo radiative transfer code discussed by 
232: \citet{ashby_profile}.\footnote{Rates of excitation of water with 
233: ortho-H$_2$ are often an order of magnitude
234: greater than rates for lower energy para-H$_2$ \citep{phillips_h2oxs}.  
235: For our calculations of cold clouds we have assumed 
236: an ortho/para-H$_2$ ratio of 0.1 and extrapolated the rates down
237: to 10 K.  If the ratio is lower our results will be essentially
238: unchanged; a higher ratio would result in lower abundance limits.}
239: The adopted physical profiles for each source are motivated below, but
240: in each case we have assumed a static cloud.
241: The radiative transfer model determines the expected emission spectra, which is
242: compared to the observed 3$\sigma$ upper limit on the antenna temperature.   
243: The water abundance is
244: then iterated until the predicted emission matches the limit.   To account for 
245: oversampling in the SWAS spectrometer,
246: we have convolved the model data with a Gaussian  that has a width of 1.45 MHz.
247: Because of the weak IR continuum radiation field in both cores the effects of infrared
248: pumping are negligible.
249: %We include the effects of dust continuum on the water excitation in these models.
250: %But, as discussed in \citet{ashby_cores}, for these weak infrared sources the effects of infrared pumping are negligible.
251: %
252: \subsection{B68}
253: 
254: The B68 globule is located at a distance of 125 pc \citep{degeus_dist}.  \citet{alves_b68} examined this
255: isolated pre-stellar core using near-infrared extinction techniques.   The resulting 
256: extinction map was used to constrain the radial density profile,
257: which is well fit by a pressure confined, self-gravitating cloud
258: near equilibrium (Bonnor-Ebert sphere).  Recently,  
259: \citet[][hereafter BAHL02]{bergin_b68} and \citet{hotzel_b68} have examined this
260: cloud in the low-J mm-wave transitions of various molecules.  Both found 
261: that in the dense core center,
262: \CeiO\ molecules are systematically depleted onto grain surfaces.
263: 
264: We use the model described by BAHL02 
265: to estimate the water abundance. 
266: In this model the column density is constrained by the 
267: visual extinction data and the line width by \CeiO\ emission
268: (BAHL02).  
269: The assumed density profile is that derived from near-IR extinction measurements, while
270: the temperature structure is that for dust in a Bonnor-Ebert sphere 
271: derived by \citet{zucconi_tdust}.\footnote{However,
272: for the temperature, a uniform reduction of 2 K was required to match the multi-transitional
273: \NtwoHp\ observations of BAHL02.}  
274: The density and temperature structure are
275: provided in Figure~2.  This model assumes a static 
276: cloud, but includes contributions
277: from thermal and turbulent line widths with the latter increasing as a function of
278: radius.   
279: 
280: As a check on the assumptions regarding the cloud physical structure we first
281: examine the non-detection of \thCO\ \swas13co.
282: To model this emission we use the depleted \CeiO\ abundance profile
283: given by BAHL02 scaled by a factor of 7.8 (accounting for 
284: $^{12}$C/$^{13}$C = 75; $^{16}$O/$^{18}$O = 500).  Placing this into the
285: radiative transfer code  produces a peak temperature of  T$_A \sim 40$ mK,
286: which is close to the observed $3\sigma$ limit of 36 mK.    
287: Since the \thCO\ analysis does not significantly conflict with current limits, we then
288: use the iterative procedure outlined previously to estimate 
289: the 3$\sigma$ limit on the water abundance 
290: in B68, which is $x$(\oHtwoO ) $< 3 \times 10^{-8}$.    The resulting
291: emission spectra from the radiative transfer model are shown in Figure~1.\footnote{If 
292: the core radius is allowed to extend beyond 0.06 pc, 
293: as estimated by the near-IR analysis, 
294: then the abundance limit will be reduced.}
295: 
296: BAHL02 linked a radiative transfer code to a chemical model including 
297: the effects of molecular depletion to set limits on the 
298: freeze-out of \CeiO\ molecules.
299: Water is more tightly bound to grain
300: surfaces when compared to CO \citep{sa_binding}. Therefore it is more difficult
301: to remove and can be expected to be significantly depleted in the center, 
302: perhaps with larger depletions than CO.  
303: To see if current chemical theory is in
304: in conflict with our observational results we compare the predictions
305: of a gas-grain chemical model to the observations.   
306: 
307: To this end we use the gas-grain chemical model discussed in BAHL02.
308: The only difference in the chemistry is the 
309: additional inclusion of the grain surface formation of water 
310: via hydrogenation of oxygen as required by previous studies \citep{bergin_imp, viti_h2o,
311: charnley_swasiso}.   This model incorporates the cloud density and temperature profile 
312: shown in Figure~2.
313: We also include the observed cloud extinction 
314: profile which increases to A$_V$ = 17$^m$ at the core center.
315: The chemical model predicts the profile of water abundance with cloud radius which, 
316: along with the same physical profile, are placed as 
317: inputs to the radiative transfer calculations.  
318: Due to the inclusion of gas-grain interactions the chemical model
319: does not reach a steady state and the predicted abundance profile strongly 
320: varies with time.   Fortunately BAHL02 find that the observed C$^{18}$O and
321: N$_2$H$^{+}$ emission in B68 can be simultaneously reproduced by the
322: model at t  $\sim  7.6 \times 10^4$ yr (later times predict increasing depletion
323: and, in consequence, less C$^{18}$O emission than observed).
324: We therefore adopt this time
325: in our analysis to examine whether
326: the \HtwoO\ abundance profile predicted by the same model is
327: in conflict with SWAS observations. 
328:  
329: In Figure~3 we provide the predicted water vapor and ice abundances 
330: as a function of depth.   Also shown are the abundance of atomic oxygen and the
331: best fit constant abundance.  The chemical model predicts
332: that the water vapor abundance sharply declines
333: at the cloud edge due to photodissociation, rises to a peak at A$_V \sim 2^m$,
334: and then uniformly declines towards higher depths. 
335: As seen in Figure~3, nearly all of the water and oxygen
336: (in the form of H$_2$O)
337: is frozen onto dust grains.  
338: Placing the results of this model into the radiation transfer code we predict
339: a peak temperature in the \oHtwoO\ 557 GHz line of T$_A^*$ = 18 mK
340: below the 3$\sigma$ upper limit.  As long as the age is $> 10^{4}$ yr
341: our observations are consistent with this model.
342: Thus current predictions of water abundance profiles by chemical
343: models are not in conflict with observations.
344: %However, it must be stressed that the chemical model assumes that the chemistry
345: %is evolving while the density evolution is static.  Thus all quoted times
346: %represent minimum timescales as we do not account for any prior evolution
347: %at a lower density state.   
348: Given the potential of B68 as a template for testing chemical models 
349: it will be useful in the future
350: to re-examine these results in light of additional molecular observations.
351: 
352: \subsection{$\rho$ Oph D}
353: 
354: The $\rho$ Oph D molecular core is a star-less object
355: located in the Ophiuchus complex at a distance of 160 pc.
356: The core was detected in absorption against the galactic 
357: mid-IR background using ISOCAM \citep{bacmann_dust} and in 1.3 mm continuum emission \citep{motte_oph}.
358: These studies have provided good constraints on the core density profile, although
359: there are differences in the details.   Both agree that the density declines
360: as $\rho (r) \propto r^{-2}$ beyond $r$ $\sim 3300 - 4000$ AU; within this radius the
361: density is found to be constant, ranging from 
362: $n_{cst} = 3 \times 10^{5}$ \cc\ (Bacmann et al. 2000) to
363: $n_{cst} = 9 \times 10^{5}$ \cc\  (Motte et al. 1998).
364: Both studies derive a core radius of $\sim$ 13,000 AU (0.063 pc).  
365: However \citet{motte_oph} find a sharp edge to the South-West but, 
366: in the East-West direction, the dust emission merges with the ambient cloud. 
367: With these differences core mass estimates range from 2 -- 5 M$_{\odot}$.
368: In the following we examine both density distributions.
369: 
370: Unlike B68, where a detailed model exists, there is no 
371: information regarding the radial dependence of the velocity field or temperature in $\rho$ Oph D.
372: For temperature, the sub-millimeter dust continuum  survey of Ophiuchus by \citet{johnstone_oph} 
373: did not include core D, but for numerous other cores in the cloud they find T$_{dust}$ $\sim 10 - 30$ K, which
374: may be considered the expected range.   Given the lack of detailed information, we adopt the expression for the equilibrium dust temperature given by \citet{burton_pdr}, further assuming that T$_{dust} =$ T$_{gas}$.
375: This expression provides an estimate of the dust temperature depending on the 
376: extinction and of the local enhancement of the ultra-violet radiation field, G$_0$.    
377: For the  A$_V(r)$ profile we integrate the observed density 
378: profile with radius in a pencil beam along the line of sight. 
379: Because of the presence of early type stars the local radiation field is enhanced 
380: throughout the $\rho$ Oph cloud and detailed modeling of the 
381: far-infrared emission by  \citet{liseau_oph}  finds G$_0 = 20$ towards  core D. 
382: 
383: The procedure is to first examine how well the physical model(s) reproduce
384: the SWAS $^{13}$CO \swas13co\ emission and then apply the ``best'' model to the water
385: emission.   If we use the \citet{bacmann_dust} density profile, the above
386: temperature structure, and a constant turbulent width of 1.5 km s$^{-1}$ we can
387: match the observed \thCO\ emission provided $x$(\thCO ) $\sim 3 \times
388: 10^{-6}$.  This abundance is higher than the expected range for 
389: undepleted gas.\footnote{Using the \CeiO\ abundance derived in the extended Ophiuchus cloud
390: by \citet{flw82} with plausible isotope ratios ($^{16}$O/$^{18}$O = 500; 
391: $^{12}$C/$^{13}$C = 45 -- 90) the \thCO\ abundance is expected 
392: to be 1 -- 2 $\times 10^{-6}$.}   Since this source is a cold pre-stellar
393: object we expect to see some evidence for \thCO\ depletion.  For instance in
394: B68, and several other similar pre-stellar objects,
395: large depletions are observed \citep{bergin_b68, jessop_dep, tafalla_dep}.
396: 
397: Adopting the denser profile of \citet{motte_oph} we can reproduce observed
398: \thCO\ emission at a lower abundance of $\sim$ 4 $\times 10^{-7}$. 
399: The predicted emission spectra from this model is shown superposed
400: on the observational data in Figure~1.   
401: For reproducing the water emission, the \citet{motte_oph} model is favored due to
402: the anticipation of CO depletion. 
403: Adopting this model for the  \swash2o\  observations we derive 
404: $x$(\oHtwoO ) $< 8 \times 10^{-9}$, relative to \Htwo (3$\sigma$). 
405: The model that matches the 3$\sigma$ limit is also provided in Figure~1.
406: If we use the Bacmann model the abundance limit 
407: is $x$(\oHtwoO ) $\lesssim 1 \times 10^{-7}$.
408:  
409: \section{Conclusions}
410: 
411: We have derived upper limits to the water abundance in two clouds with well
412: described physical properties.   These abundance limits are, 
413: $x$(\oHtwoO ) $< 8 \times 10^{-9}$ for
414: core D in $\rho$ Oph and $x$(\oHtwoO ) $< 3 \times 10^{-8}$ in B68.
415: These limits are below those previously set for a cold star-less object by \citet{snell_h2o}
416: (TMC-1: \h2oabun\ $< 7 \times 10^{-8}$).  In this Letter two cores have 
417: been subject to a more detailed and careful study which essentially 
418: confirms the general lack of ortho-water vapor in cold (T $< 20$ K) molecular cores.     
419: Provided  the ortho/para ratio of water is $>$ 0.03 then these results are in agreement 
420: with the assertion that the water abundance in low mass objects is well below the 
421: predictions of pure gas phase chemistry.  In the case of
422: B68, these observations are compared to theoretical predictions of
423: a gas-grain chemical model, which have been directly placed into the radiative transfer
424: calculations.  
425: We find that these model predictions are not in conflict with observations. 
426: However, due to the large SWAS beam, which encompasses the entire extent of the observed
427: molecular emission seen in other tracers, we cannot discriminate between
428: simple constant abundance models and those with complex abundance 
429: structure constrained by theory.
430: Given the wide-spread 
431: molecular depletion found in B68 and similar cores \citep{bergin_b68, hotzel_b68, 
432: tafalla_dep},  it is likely that most of the water in low mass dense cores 
433: is frozen on the surfaces of cold dust grains.
434: 
435: 
436: \acknowledgements
437: 
438: 
439: We acknowledge Gary Melnick for a thorough reading of the manuscript.
440: E.A.B. is grateful for the help and collaboration with Charlie Lada,
441: Jo\~ao Alves, and Tracy Huard which greatly aided the B68 analysis.
442: For these data we are grateful to entire SWAS team and acknowledge
443: support from NASA's SWAS Grant NAS5-30702.
444: 
445: 
446: 
447: \begin{thebibliography}{}
448: 
449: \bibitem[Alves, Lada, \& Lada(2001)]{alves_b68} Alves, J.~;., Lada, C.~J., \&
450: Lada, E.~A.\ 2001, \nat, 409, 159 (ALL01)
451: 
452: \bibitem[Ashby et al.(2000b)]{ashby_profile} Ashby, M.~L.~N.~et al.\ 2000b, \apjl, 539, L115. \bibitem[Ashby et al.(2000a)]{ashby_cores} Ashby, M.~L.~N.~et al.\ 2000a, \apjl, 539, L119. 
453: 
454: \bibitem[Bacmann et al.(2000)]{bacmann_dust} Bacmann, A., Andr{\' e}, P., Puget, J.-L., Abergel, A., Bontemps, S., \& Ward-Thompson, D.\ 2000, \aap, 361, 555. 
455: 
456: \bibitem[Bergin et al.(2002)]{bergin_b68} Bergin, E.A., Alves, J., Huard, T.L., 
457: \& Lada, C.J. (2002), ApJ, 570, L101 (BAHL02)
458: 
459: \bibitem[Bergin et al.(2000)]{bergin_imp} Bergin, E.~A.~et al.\ 2000, \apjl, 539, L129. 
460: \bibitem[Burton, Hollenbach, \& Tielens(1990)]{burton_pdr} Burton, M.~G., Hollenbach, D.~J., \& Tielens, A.~G.~G.~M.\ 1990, \apj, 365, 620. 
461: 
462: \bibitem[Cernicharo et al.(1997)]{cernicharo_h2oabs} Cernicharo, J.~et al.\ 1997, \aap, 323, L25. \bibitem[Charnley, Rodgers, \& Ehrenfreund(2001)]{charnley_swasiso} Charnley, S.~B., Rodgers, S.~D., \& Ehrenfreund, P.\ 2001, \aap, 378, 1024. 
463: 
464: \bibitem[de Geus, de Zeeuw, \& Lub(1989)]{degeus_dist} de Geus, E.~J., de Zeeuw, P.~T., \& Lub, J.\ 1989, \aap, 216, 44. 
465: 
466: \bibitem[Frerking, Langer, \& Wilson(1982)]{flw82} Frerking, 
467: M.~A., Langer, W.~D., \& Wilson, R.~W.\ 1982, \apj, 262, 590. 
468: 
469: \bibitem[Hotzel et al.(2002)]{hotzel_b68} Hotzel, S., Harju, J., Juvela, M.,
470: Mattila, K., \& Haikala, L.K. 2002, \aap , submitted
471: 
472: \bibitem[Jessop \& Ward-Thompson(2001)]{jessop_dep} Jessop, 
473: N.~E.~\& Ward-Thompson, D.\ 2001, \mnras, 323, 1025 
474: 
475: \bibitem[Johnstone et al.(2000)]{johnstone_oph} Johnstone, D., Wilson, C.~D., Moriarty-Schieven, G., Joncas, G., Smith, G., Gregersen, E., \& Fich, M.\ 2000, \apj, 545, 327. 
476: 
477: \bibitem[Liseau et al.(1999)]{liseau_oph} Liseau, R.~et al.\ 1999, \aap, 344, 342
478: 
479: \bibitem[Melnick et al.(2000)]{melnick_swas} Melnick, G.~J.~et al.\ 2000, \apjl, 539, L77
480: 
481: \bibitem[Moneti, Cernicharo, \& Pardo(2001)]{moneti_h2oabs} Moneti, A., Cernicharo, J.~;, \& Pardo, J.~R.~;.\ 2001, \apjl, 549, L203. 
482: \bibitem[Motte, Andre, \& Neri(1998)]{motte_oph} Motte, F., Andre, P., \& Neri, R.\ 1998, \aap, 336, 150
483: 
484: \bibitem[Neufeld et al.(2000)]{neufeld_h2oabs} Neufeld, D.~A.~et al.\ 2000, \apjl, 539, L111
485: 
486: 
487: \bibitem[Phillips, Maluendes, \& Green(1996)]{phillips_h2oxs} 
488: Phillips, T.~R., Maluendes, S., \& Green, S.\ 1996, \apjs, 107, 467 
489: 
490: 
491: \bibitem[Rawlings, Hartquist, Menten, \& 
492: Williams(1992)]{rawlings_dep} Rawlings, J.~M.~C., Hartquist, T.~W., 
493: Menten, K.~M., \& Williams, D.~A.\ 1992, \mnras, 255, 471 
494: 
495: \bibitem[Sandford \& Allamandola(1988)]{sa_binding} Sandford, 
496: S.~A.~\& Allamandola, L.~J.\ 1988, Icarus, 76, 201 
497: 
498: 
499: \bibitem[Snell et al.(2000a)]{snell_omc1} Snell, R.~L.~et al.\ 2000a, \apjl, 539, L93. 
500: 
501: \bibitem[Snell et al.(2000b)]{snell_m17} Snell, R.~L.~et al.\ 2000b, \apjl, 539, L97. 
502: \bibitem[Snell et al.(2000c)]{snell_h2o} Snell, R.~L.~et al.\ 2000c, \apjl, 539, L101. 
503: 
504: \bibitem[Spaans \& van Dishoeck(2001)]{spaans_h2o} Spaans, M.~\& van Dishoeck, E.~F.\ 2001, \apjl, 548, L217. 
505: 
506: \bibitem[Tafalla et al. (2002)]{tafalla_dep} Tafalla, M., Myers, P.C.,
507: Caselli, P., Walmsley, C.M., \& Comito, C. 2002, ApJ, 569, 815
508: \bibitem[Viti et al.(2001)]{viti_h2o} Viti, S., Roueff, E., Hartquist, T.~W., Pineau des For{\^ e}ts, G., \& Williams, D.~A.\ 2001, \aap, 370, 557. 
509: 
510: \bibitem[Wannier et al.(1991)]{wannier_h2o} Wannier, P.~G.~et al.\ 1991, \apj, 377, 171. 
511: 
512: \bibitem[Zucconi, Walmsley, \& Galli (2001)]{zucconi_tdust} Zucconi, A., Walmsley,
513: C.~M., \& Galli, D.\ 2001, \aap, 376, 650
514: 
515: \end{thebibliography}
516: \begin{center}
517: \begin{figure*}
518: \plotone{f1.eps}
519: \caption{
520: Spectra of the $1_{10} - 1_{01}$ transition of o-H$_2$O and $^{13}$CO J $= 5 - 4$
521: towards the B68 dark cloud (left) and $\rho$ Oph core D (right) shown as a
522: thick solid line.  The thin solid line is the result of the excitation model
523: described in \S3.1 and \S3.2, while the dashed line denotes the systemic velocity.
524: The model is shifted to the proper source velocity.  
525: $^{13}$CO J $= 5 - 4$ is detected in $\rho$ Oph D, but all other observations 
526: are non-detections.   In the figure a linear baseline has been subtracted from
527: the observational data.  The model spectra are shown with continuum included
528: which, due to the small level, has not been subtracted.
529: }
530: \end{figure*}
531: \end{center}
532: 
533: \begin{figure*}
534: \plotone{f2.eps}
535: \caption{Adopted density and temperature profile with radius in B68 (left)
536: and $\rho$ Oph D (right) molecular cores.  In the latter core we present the density
537: profile derived by Motte et al. (1998) and Bacmann et al. (2000). }
538: \end{figure*}
539: 
540: \begin{figure*}
541: \epsscale{0.65}
542: \plotone{f3.eps}
543: \caption{Theoretical predictions of molecular abundance as a function of cloud
544: depth (extinction) from a gas-grain chemical model.  Also shown is the best
545: fit constant abundance model for the B68 cloud.  
546: Details of this model are given in \S3.1.}
547: \end{figure*}
548: 
549: 
550: \end{document}
551: 
552: 
553: 
554: 
555: 
556: 
557: 
558: 
559: