0803.1504/ms.tex
1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2005 December 5
5: %%
6: %% This is a sample manuscript marked up using the
7: %% AASTeX v5.x LaTeX 2e macros.
8: 
9: %% The first piece of markup in an AASTeX v5.x document
10: %% is the \documentclass command. LaTeX will ignore
11: %% any data that comes before this command.
12: 
13: %% The command below calls the preprint style
14: %% which will produce a one-column, single-spaced document.
15: %% Examples of commands for other substyles follow. Use
16: %% whichever is most appropriate for your purposes.
17: %%
18: \documentclass[12pt,preprint]{aastex}
19: 
20: %% manuscript produces a one-column, double-spaced document:
21: 
22: %\documentclass[article]{aastex}
23: %\documentclass{emulateapj}
24: 
25: %% preprint2 produces a double-column, single-spaced document:
26: 
27: %\documentclass[preprint2]{aastex}
28: 
29: %% Sometimes a paper's abstract is too long to fit on the
30: %% titlepau page in preprint2 mode. When that is the case,
31: %% use the longabstract style option.
32: 
33: %% \documentclass[preprint2,longabstract]{aastex}
34: 
35: %% If you want to create your own macros, you can do so
36: %% using \newcommand. Your macros should appear before
37: %% the \begin{document} command.
38: %%
39: %% If you are submitting to a journal that translates manuscripts
40: %% into SGML, you need to follow certain guidelines when preparing
41: %% your macros. See the AASTeX v5.x Author Guide
42: %% for information.
43: 
44: %\newcommand{\vdag}{(v)^\dagger}
45: %\newcommand{\myemail}{skywalker@galaxy.far.far.away}
46: 
47: %% You can insert a short comment on the title page using the command below.
48: 
49: \slugcomment{c2d Lupus Synthesis \today}
50: 
51: %% If you wish, you may supply running head information, although
52: %% this information may be modified by the editorial offices.
53: %% The left head contains a list of authors,
54: %% usually a maximum of three (otherwise use et al.).  The right
55: %% head is a modified title of up to roughly 44 characters.
56: %% Running heads will not print in the manuscript style.
57: 
58: \shorttitle{Lupus synthesis paper}
59: \shortauthors{Mer\'{\i}n et al.}
60: 
61: %\usepackage{deluxetable}
62: %\usepackage{rotate}
63: 
64: %% This is the end of the preamble.  Indicate the beginning of the
65: %% paper itself with \begin{document}.
66: 
67: \begin{document}
68: 
69: %% LaTeX will automatically break titles if they run longer than
70: %% one line. However, you may use \\ to force a line break ifnote
71: %% you desire.
72: 
73: \title{The Spitzer c2d Survey of Large, Nearby, Interstellar Clouds.
74: XI. \\Lupus Observed With IRAC and MIPS}
75: 
76: %% Use \author, \affil, and the \and command to format
77: %% author and affiliation information.
78: %% Note that \email has replaced the old \authoremail command
79: %% from AASTeX v4.0. You can use \email to mark an email address
80: %% anywhere in the paper, not just in the front matter.
81: %% As in the title, use \\ to force line breaks.
82: 
83: \author{Bruno Mer\'{\i}n\altaffilmark{1,2}, 
84: Jes J{\o}rgensen\altaffilmark{3},
85: Loredana Spezzi\altaffilmark{4}, 
86: Juan M.~Alcal\'a\altaffilmark{4}, 
87: Neal J. Evans II\altaffilmark{5}, 
88: Paul M. Harvey\altaffilmark{5}, 
89: Nicholas Chapman\altaffilmark{6},
90: Tracy Huard\altaffilmark{7},
91: Ewine F. van Dishoeck\altaffilmark{2},
92: Fernando Comer\'on\altaffilmark{8}
93: }
94: 
95: \altaffiltext{1}{Research and Scientific Support Department, European Space Agency (ESTEC), PO Box 299, 2200 AG Noordwijk, The Netherlands }
96: \altaffiltext{2}{Leiden Observatory, Leiden University, PO Box 9513, 2300 RA Leiden, The Netherlands }
97: \altaffiltext{3}{Argelander-Institut f\"ur Astronomie, University of Bonn,
98: Auf dem H\"ugel 71, 53121 Bon, Germany}
99: \altaffiltext{4}{INAF - Osservatorio Astronomico di Capodimonte, via Moiariello 16, I-80131, Naples, Italy}
100: \altaffiltext{5}{Department of Astronomy, University of Texas at Austin, 
101: 	    1 University Station C1400 Austin, TX 78712-0259, USA}
102: \altaffiltext{6}{Astronomy Department, University of Maryland, College Park, 
103: 	    MD 20742, USA}
104: \altaffiltext{7}{Smithsonian Astrophysical Observatory, 60 Garden Street, MS42, Cambridge, MA 02138, USA}
105: \altaffiltext{7}{European Southern Observatory, Karl-Schwarzschild-Strasse 2, 
106: 85748 Garching bei M\"unchen, Germany}
107: 
108: \begin{abstract}
109:  We present c2d Spitzer/IRAC observations of the Lupus I, III and IV dark clouds 
110:   and discuss them in combination with optical and near-infrared and c2d MIPS data.  
111:   With the Spitzer data, the new sample contains 159 stars, 4 times larger than the previous
112:   one. It is dominated by low- and very-low mass stars and it is complete down to
113:   M $\approx$ 0.1M$_\odot$. We find 30-40\% binaries with separations between
114:   100 to 2000 AU with no apparent effect in the disk
115:   properties of the members. A large majority of the objects are Class II or Class III
116:   objects, with only 20 (12\%) of Class I or Flat spectrum
117:   sources. The disk sample is complete down to ``debris''-like systems in stars as
118:     small as M $\approx$ 0.2 M$_\odot$ and includes sub-stellar objects with
119:     larger IR excesses. The disk fraction in Lupus is 70 -- 80\%, consistent
120:     with an age of 1 -- 2 Myr. However, the young population contains 20\% optically
121:     thick accretion disks and 40\% relatively less flared disks. A growing variety of inner 
122:     disk structures is found for larger inner disk clearings for equal disk masses. 
123:     Lupus III is the most centrally populated and rich, followed by Lupus I with
124:     a filamentary structure and by Lupus IV, where a very high density core with little 
125:     star-formation activity has been found. We estimate star formation rates in Lupus 
126:     of 2 -- 10 M$_\odot$ Myr$^{-1}$ and star formation efficiencies of a few percent, 
127:     apparently correlated with the associated cloud masses.
128:     \end{abstract}
129: 
130: \keywords{stars: formation -- stars: pre-main sequence -- stars: low-mass
131: star-forming-regions:individual (Lupus I, Lupus III, Lupus IV)}
132: 
133: \section{Introduction}
134: 
135: The Lupus dark cloud complex is one of the main low-mass star forming
136: regions within 200 pc of the sun. Located in the Scorpius-Centaurus OB
137: association, it consists of a loosely connected group of dark clouds
138: and low-mass pre-main sequence stars located between galactic
139: coordinates 334$^{\rm o} < l < 352^{\rm o}$ and latitudes $+5^{\rm o}
140: < b < +25^{\rm o}$ (RA $\sim$ 16$^h$20$^m$ -- 15$^h$30$'$ and DEC
141: $\sim$ -43$^{\rm o}$00$'$ -- -33$^{\rm o}$00$'$ )
142: \citep{Krautter1991}.  The low ecliptic latitudes ($\sim$ -33 -- -41)
143: of the clouds guarantee a low abundance of asteroids. Given its large
144: size ($\sim$ 20$^{\rm o}$) and proximity, it has been subject of many
145: studies. Large-scale $^{\rm 12}{\rm CO}$ ($J = 1 \to 0$) millimeter
146: maps of the whole complex give total molecular gas masses of several
147: times 10$^4 M_\odot$ \citep{Murphy1986,Tachihara2001} and a relatively
148: small spread of cloud velocities among the different sub-clouds
149: ($\leq$ 3 km s$^{-1}$ \citealt{Vilas-Boas2000}). It hosts four active
150: star-forming regions, including the rich T Tauri association in Lupus
151: 3, plus five other looser dark clouds with signs of moderate
152: star-formation activity \citep{Hara1999}. Objects in all
153: evolutionary phases, from embedded Class I objects to evolved Class
154: III stars, are found in the Lupus clouds.  A comprehensive review about the
155: Lupus clouds by \cite{Comeron2007a} provides more details on previous
156: observations and is used throughout this work to assess the new results
157: from the Spitzer data. The Lupus dark cloud is one of the five large
158: nearby star-forming regions observed as part of the Spitzer Legacy
159: Project ``From Molecular Cores to Planet-forming Disks'' (c2d)
160: \cite{Evans2003}. From the several Lupus subclouds
161: \citep{The1962,Murphy1986}, only Lupus I, III and IV were observed by
162: c2d and are discussed in this paper.
163: 
164: Optical and near-infrared (near-IR) ground-based observations of Lupus
165: identified three regions, denominated as Lupus I, II and III, with
166: large numbers of classical T Tauri stars
167: \citep{Henize1954,The1962}. \cite{Schwartz1977} made a catalog of
168: H$\alpha$ emitting young stars which contained the vast majority of
169: the objects in the clouds known until today. His observations also showed
170: that Lupus 3 (roughly consistent with Cambr\'esy's Lupus III) is one
171: of the most active star-forming regions in the southern sky, followed
172: by Lupus I, II and IV and that a large number of the stars in these
173: clouds are low-mass stars. In addition to these classical T Tauri
174: stars, \cite{Krautter1997} reported a much larger number of weak-line
175: T Tauri stars with X-ray emission, which are believed to be older
176: stars not physically bound to the dark clouds but rather belonging to
177: a more nearby structure (\citealt{Wichmann1999}, see also
178: \citealt{Cieza2007} for a confirmation of this conclusion from c2d
179: observations). Finally, there is evidence for different distances to
180: the different subclouds: {\sl Hipparcos} parallaxes and extinction
181: source counts yield reasonable distance estimates of 150 $\pm$ 20 pc
182: for Lupus I and IV and 200 $\pm$ 20 pc for Lupus III (Comer\'on 2008
183: in prep. and references therein), which we assume for this work.
184: 
185: %Out of the three clouds observed by Spitzer, Lupus III is the most
186: %massive star-forming region in Lupus and the only one containing two
187: %Herbig Ae stars (HR 5999 and HR 6000) together with a great number of
188: %young low-mass and very-low mass stars \citep{Comeron2003}. Lupus I
189: %and IV represent regions of small and moderate masses and
190: %star-formation activity, respectively. Therefore, the present
191: %observations are optimal for a comparative study of star formation and
192: %disk evolution across three clouds with well defined young stellar
193: %populations and different overall evolutionary phases. Also, the high
194: %sensitivity of the Spitzer detectors makes it specially suited for
195: %identifying faint objects and therefore for extending substantially
196: %the census of brown dwarfs and very-low mass stars in the clouds (see
197: %e.g. \citealt{Allen2007,Merin2007}).
198: 
199: Following previous c2d standards, observational results from the
200: Spitzer Space Telescope (\citealt{Werner2004}) Infrared Array
201: Camera (IRAC, \citealt{Fazio2004}) and from the Multiband Imaging
202: Photometer for Spitzer (MIPS, \citealt{Rieke2004}) are reported
203: separately for each cloud, followed by a ``synthesis'' paper which
204: combines all the c2d Spitzer and complementary observations made of
205: the region (see e.g. \citealt{Harvey2006,Harvey2007a,Harvey2007b} for
206: Serpens or \citealt{Young2005,Porras2007}, and \citealt{Alcala2008}
207: for Chamaeleon II). For Lupus, the observational results obtained with
208: MIPS have been reported by \cite{Chapman2007} (Paper I) and this paper
209: describes the IRAC observations and analyzes the complete data set for
210: the region. In that sense, this paper merges the contents of the IRAC
211: and ``synthesis'' papers of the c2d observations in Lupus: it
212:   presents the 3.6 to 8.0 $\mu$m IRAC observations of the clouds for
213:   the first time and combines it with all previously available
214:   information from the optical to the millimeter.
215: 
216:   All the observations analysed in this study are presented in
217:   \S~\ref{observations}, with a discussion of the complementary 
218:   observations and stellar multiplicity in the
219:   optical in \S~\ref{binaries} and a detailed description of the
220:   Spitzer IRAC observations in \S~\ref{IRAC_BDP} and \ref{diffcounts}. 
221:   These data are used
222:   to construct a high-reliability catalog of young stars in the clouds
223:   in \S~\ref{YSOc_list}. This catalog is composed of Young Stellar
224:   Objects (YSO) identified with the Spitzer data (selected in
225:   \S~\ref{yso_sel}) and previously known Pre-Main Sequence stars (PMS,
226:   discussed in \S~\ref{pms_sel}). The rest of the paper is organized
227:   in two main sections, which deal with the individual sources
228:   (\S~\ref{disks}) and with the global cloud properties
229:   (\S~\ref{clouds}), respectively. In the first one, the disk
230:   populations are studied with color-color diagrams
231:   (\S~\ref{colorcolor}), multi-wavelength Spectral Energy
232:   Distributions (SEDs, \S~\ref{SEDs}), stellar and disk luminosity
233:   functions (\S~\ref{star_luminosities} and
234:   \ref{fractional_luminosities}), and finally with a new `2D'
235:   classification system of the SEDs (\S~\ref{2D_param}). The Spitzer
236:   observations of outflows and Herbig-Haro objects in the region are
237:   also described in \S~\ref{outflow_sources}. The second section
238:   describes the structures of the clouds with the help of
239:   Spitzer-derived extinction maps (\S~\ref{extinction}), the spatial
240:   distribution of the YSOs in the clouds and a nearest-neighbor
241:   analysis of their clustering properties
242:   (\S~\ref{spatial_distribution}). Finally, the Star Formation Rates
243:   and Star Formation Efficiencies are computed and discussed in
244:   \S~\ref{SFR} and \ref{SFE}, respectively. A complete summary
245:   of this work is given in \S~\ref{summary}.
246: 
247: \section{Observations and data analysis}
248: \label{observations}
249: 
250: The observations discussed in this paper come from the Spitzer Space
251: Telescope's Infrared Array Camera (IRAC hereafter) and Multiband
252: Imaging Photometer for Spitzer (MIPS) observations of Lupus I, III and
253: IV made by the c2d Spitzer Legacy Program \citep{Evans2003}, together
254: with an optical coordinated survey of the three clouds (F. Comer\'on
255: et al., in preparation) and data from the literature. Detailed
256: information about the MIPS observations can be found in
257: \cite{Chapman2007}. 
258: 
259: \subsection{Complementary data and multiple visual systems}
260: \label{binaries}
261: 
262: An optical survey of the three clouds was performed with the
263: Wide-Field Imager (WFI) attached to the ESO 2.2m telescope, at La
264: Silla Observatory in Chile. The areas observed in the $R_C$ (0.652
265: $\mu$m), $I_C$ (0.784 $\mu$m), and $z_{WFI}$ (0.957 $\mu$m) optical
266: bands in Lupus I, III and IV were defined to overlap completely with
267: the areas observed with Spitzer shown in Figures \ref{lupI} to
268: \ref{lupIV}.  The observational strategy, data reduction and source
269: extraction are described in detail elsewhere (F. Comer\'on et al., in
270: preparation). These observations were complemented
271: with photometry from the {\sl NOMAD} optical and near-IR catalog
272: \citep{Zacharias2005}, which contains $B$, $V$ and $R_C$ magnitudes
273: and from the $J$, $H$ and $K_S$ 2MASS near-IR all-sky catalog
274: \citep{Cutri2003}. This paper also includes the c2d MIPS observations of the Lupus
275: clouds. Figures \ref{spa_distr_I} to \ref{spa_distr_IV} also show the
276: areas mapped with MIPS at 24, 70 and 160 $\mu$m as part of the c2d
277: project. They were defined to provide the maximum overlap and minimum
278: observing time and cover completely the areas mapped with IRAC. 
279: The MIPS data acquisition, reduction and
280: source extraction is presented and discussed in detail in
281: \citet{Chapman2007}.
282: 
283: 
284: The optical images were inspected to search for visual binaries in
285: the sample. Table \ref{binaries_tab} reports all
286: apparent companions detected in the images. The WFI images in $Rc$,
287: $Ic$ and $z_{WFI}$ bands were inspected for all sources. Those
288: observations were performed in service mode over several nights in two
289: different observing seasons and have a range of seeing values, but an
290: average seeing of 1.5$^{''}$ limits the smallest separation detectable
291: to $\sim$ 0.7$^{''}$ in our case. The range of separations for which
292: we report the presence of possible companions is 0.7 to 10$^{''}$ (140
293: to 2000 AU in Lupus III and 105 to 1500 AU at the distance of Lupus I
294: and IV).
295: 
296: There are several pairs of stars with disks which are binaries. Inspection 
297: of the optical
298: images indicates that the probability of projection effects might be
299: small given the relatively low number of optical sources at distances
300: smaller than 10$^{''}$ in most of the examined stars. We recover all
301: binaries listed by \cite{Ghez1997} which fell in our field except HR
302: 5999, which was too bright to allow a shape analysis. We also recover
303: the two objects in common with a yet unpublished AO/ADONIS survey of
304: multiplicity in Lupus which probe at angular distances as close as
305: 0.2$^{''}$ (A. Knockx et al., in prep.). The total binary fractions of
306: 41$\pm$12\% (7/17), 29$\pm$9\% (39/124) and 44$\pm$13\% (8/18) in the
307: three clouds, respectively, compare well with the numbers given for
308: Taurus and Ophiuchus in a similar range of angular distances (e.g.
309: \citealt{Padgett1997}).
310: 
311: In any case, given the difficulty in apportioning the excess among the
312: companions with separations smaller than 2.0$^{''}$, in which the IRAC
313: fluxes could have been merged, and those with separations smaller than
314: 4.0$^{''}$, which could have merged fluxes in MIPS, we set those
315: fluxes as upper limits in the SEDs for the disk evolution studies.
316: More discussion on the effects of binarity in the disks is presented
317: in \S~\ref{fractional_luminosities}.
318: 
319: \subsection{Spitzer IRAC data}
320: \label{IRAC_BDP}
321: 
322: Lupus I, Lupus III and Lupus IV were observed with all IRAC bands
323: (3.6, 4.5, 5.8 and 8.0 $\mu$m) on the 3rd and 4th of September of 2004
324: as part of the c2d Spitzer Legacy Program (P.I.: N. Evans, program ID:
325: 174) and as part of GTO observations (P.I.: G.  Fazio, program ID:
326: 6). The total observed areas were defined to cover all the
327: regions with extinctions $A_V \ge 2$ for Lupus III and IV and $A_V \ge
328: 3$ for Lupus I \citep[see Fig.~2 of][]{Evans2003} as measured in the
329: extinction maps by \cite{Cambresy1999}. The combined mosaics of the three regions overlap in
330: all IRAC bands with areas of 1.39, 1.34 and 0.37 deg$^2$ in Lupus I,
331: Lupus III and Lupus IV, respectively. Furthermore 4 off-cloud regions,
332: each of $\sim$ 0.08 deg$^2$ were observed in low-extinction regions
333: with a range of galactic latitudes for statistical comparison with the
334: on-cloud fields. Each of them contains a 0.05 deg$^2$ overlap between
335: all IRAC bands. All c2d maps were observed in two epochs at least 6
336: hours apart to guarantee the production of highly reliable catalogs of
337: sources clean of asteroids and transient artifacts. The GTO
338: observation of Lupus III was only observed once. For more information
339: about the c2d mapping strategies, methods and results, consult
340: the delivery documentation \citep{Evans2007}.
341: 
342: Table \ref{obslog} shows the details of the individual pointings for
343: all three clouds and the off cloud regions. Figures \ref{spa_distr_I} to
344: \ref{spa_distr_IV} show the different coverages of the c2d IRAC, MIPS 
345: and optical mosaic areas overlaid on the optical extinction map of Lupus
346: from \cite{Cambresy1999}.
347: 
348: \begin{table}
349: \caption{c2d Spitzer IRAC Observations summary in Lupus}             % title of Table
350: %\centering                          % used for centering table
351: \begin{tabular}{l c c c c}        % centered columns (3 columns)
352: \hline\hline                 % inserts double horizontal lines
353: Field & Position ($\alpha$,$\delta$)$_{J2000}$ & Date (2004) & AOR epoch 1 & AOR epoch 2 \\\hline
354: \multicolumn{5}{c}{\sl Lupus I c2d Observations} \\
355: LupI\_1 & (15:45:09.0, -34:23:11.0) & Sept 3 & 0005717248 & 0005719808 \\
356: LupI\_2 & (15:43:26.0, -34:06:27.0) & Sept 3 & 0005717504 & 0005720064 \\
357: LupI\_2a& (15:42:23.0, -34:00:55.0) & Sept 3 & 0005717760 & 0005720320 \\
358: LupI\_3 & (15:41:16.0, -33:43:33.0) & Sept 3 & 0005718272 & 0005720832 \\
359: LupI\_4 & (15:40:08.0, -33:35:16.0) & Sept 3 & 0005718528 & 0005721088 \\
360: LupI\_5 & (15:38:34.0, -33:20:18.0) & Sept 3 & 0005718784 & 0005721344 \\
361: LupI\_6 & (15:39:09.0, -34:23:11.0) & Sept 3 & 0005717248 & 0005721600 \\
362: LupI\_7 & (15:40:34.0, -34:36:34.0) & Sept 3,4 & 0005719296 & 0005721856 \\
363: LupI\_8 & (15:38:08.0, -34:39:27.0) & Sept 3 & 0005719552 & 0005722112 \\
364: LupI\_9 & (15:42:18.0, -34:33:50.0) & Sept 3 & 0005737472 & 0005737728 \\
365: LupI\_10 & (15:40:55.0, -34:06:31.0) & Sept 3,4 & 0005739008 & 0005739264 \\
366: \multicolumn{5}{c}{\sl Lupus III c2d Observations} \\
367: LupIII\_1 & (16:12:28.0, -38:05:43.0) & Sept 4 & 0005724416 & 0005725696 \\
368: LupIII\_2 & (16:10:35.0, -37:50:04.0) & Sept 4 & 0005724672 & 0005725952 \\
369: LupIII\_3 & (16:11:59.0, -38:58:30.0) & Sept 4 & 0005724928 & 0005726208 \\
370: LupIII\_4 & (16:10:32.0, -39:07:18.0) & Sept 4 & 0005725184 & 0005726464 \\
371: LupIII\_6 & (16:07:24.0, -39:07:03.0) & Sept 4 & 0005725440 & 0005726976 \\
372: LupIII\_7 & (16:12:45.0, -38:31:31.0) & Sept 4 & 0005737984 & 0005738240 \\
373: LupIII\_8 & (16:10:38.0, -38:35:24.0) & Sept 4 & 0005738496 & 0005738752 \\
374: \multicolumn{5}{c}{\sl Lupus III GTO \& c2d Observations} \\
375: LupIII\_5 & (16:08:55.8, -39:08:33.2) & Sept 4 & 0003652608 & 0005726720 \\
376: \multicolumn{5}{c}{\sl Lupus IV c2d Observations} \\
377: LupIV\_1 & (16:02:46.0, -41:59:49.0) & Sept 4 & 0005728768 & 0005729536 \\
378: LupIV\_2 & (16:01:27.0, -41:41:00.0) & Sept 4 & 0005729024 & 0005729792 \\
379: LupIV\_3 & (16:00:43.0, -42:02:37.0) & Sept 4 & 0005729280 & 0005730048 \\
380: \multicolumn{5}{c}{\sl Off-Cloud c2d Observations} \\
381: LupOC\_2 & (16:13:00.0, -34:00:00.0) & Sept 4 & 0005731072 & 0005735680 \\
382: LupOC\_6 & (16:40:00.0, -40:30:00.0) & Sept 4 & 0005732096 & 0005736704 \\
383: LupOC\_7 & (15:50:00.0, -41:00:00.0) & Sept 3 & 0005732352 & 0005736960 \\
384: LupOC\_8 & (16:07:00.0, -43:00:00.0) & Sept 4 & 0005732608 & 0005737216 \\
385: \hline                                   %inserts single line
386: \end{tabular}
387: \label{obslog}
388: \end{table}
389: 
390: %\subsubsection{Data reduction}
391: 
392: The IRAC images were processed by the Spitzer Science Center (SSC)
393: using the standard pipeline version S13 to produce the Basic
394: Calibrated Data (BCDs). These images were then processed and combined
395: into mosaics and source catalogs by the c2d pipeline version
396: 2007/January. \cite{Harvey2006} and \cite{Jorgensen2006} describe in
397: detail the processing of the IRAC images for Serpens and Perseus,
398: respectively. The observations presented here were processed in
399: exactly the same way so we refer the reader to those articles and to the
400: Delivery Documentation for details about the data
401: reduction. Here we will just describe briefly the basic steps carried
402: out to produce the presented images and catalogs.
403: 
404: %To arrive at the final source catalogs our data reduction takes the
405: %BCD images and performs four further steps. The first one is
406: %correcting for additional background and image effects, which includes
407: %extra bad-pixel masking, correcting for muxbleed, column pull-downs or
408: %pull-ups as well as correcting for the ``first frame'' effect in the
409: %third IRAC band \citep{Harvey2006}. The second step combines all
410: %individual frames from the two epochs into large scale mosaics for
411: %each region. This is done with the SSC's ``Mopex'' software suite
412: %\citep{Makovoz2005} and eliminates transient effects like asteroids
413: %from the mosaics. In the third step, the point and extended sources
414: %are extracted from the mosaic with the c2d developed tool {\sl
415: %c2dphot} \citep{Harvey2006} for both epochs. This software extracts
416: %photometry and calculates the uncertainties using digitized point
417: %spread functions. It also classifies the objects into point-source or
418: %extended sources. A further check for consistency is done by comparing
419: %the extracted fluxes from the epochs and the final flux of each source
420: %combines both measurements. The accuracy of the flux calibration is
421: %estimated to be $\sim$ 15 \%. The fourth step is to cross-identify the
422: %IRAC sources between bands and merging the final catalog with the MIPS
423: %24, 70 and 160 $\mu$m catalogs \citep{Chapman2007} and with the near-IR 2
424: %$\mu$m All Sky Survey (2MASS) catalog \citep{Cutri2003}. For the band
425: %merging, we use a 2.0$''$ match radius among IRAC bands and 2MASS and
426: %4.0$''$ and 8.0$''$ for matching with MIPS bands 1 and 2 respectively
427: %to account for the larger point spread functions of the MIPS bands.
428: 
429: %\subsubsection{Results: IRAC and MIPS mosaics and catalogs}
430: 
431: In total 63384, 138270 and 42737 sources were detected with at least
432: IRAC in Lupus I, III and IV, respectively. Table \ref{detection_stat}
433: summarizes the number of sources detected in the three surveyed areas
434: with S/N of at least 5. This corresponds to selecting all sources with
435: detections of quality `A' or `B' in any of the IRAC bands from the
436: delivered c2d catalogs. Most of the sources in this
437: catalog can be well fitted with reddened stellar atmospheres from
438: field stars. IRAC bands 1 and 2 are the most sensitive in all three
439: catalogs and produce an average of 7 times more high quality
440: detections than those in IRAC bands 3 and 4. About 300, 400, and 350
441: sources in Lupus I, III and IV respectively of the four band
442: detections are found to be extended in one or more bands (classified
443: with the ``image type'' 2 in the c2d catalogs). Our final catalogs
444: contain 2776, 7539 and 2403 four-band IRAC sources in the three Lupus
445: clouds, respectively (see Table \ref{detection_stat}).
446: 
447: \begin{table}[h]
448:   \caption{Total numbers of IRAC sources detected with S/N $\ge$ 5$\sigma$ in Lupus }\label{detection_stat}
449: \begin{tabular}{llll}\hline\hline
450:                                     & Lupus I   & Lupus III & Lupus IV \\\hline
451: Detection in at least one IRAC band & 63384   & 138270 & 42737  \\
452: Detection in all four IRAC bands       & 2834      & 7638      & 2418  \\
453: Detection in three IRAC bands           & 3012      & 6559     & 2001  \\
454: Detection in two IRAC bands           & 29301     & 67039    & 21125 \\
455: Detection in one IRAC band            & 28237     &  57034   & 17193 \\
456: \\
457: Detection in 2MASS only$^a$         & 0         & 7        & 1  \\
458: Detection in IRAC only              & 56219     &  122166  & 37904 \\
459: Detection in 4 IRAC bands and not 2MASS$^a$ &136  & 268 & 80 \\\hline
460: {\sl Excluding extended sources} &&&\\
461: Four band detections   &   2776 & 7536 & 2403 \\
462: Four band detections with 2MASS association$^a$ & 2637 & 7244  & 2262 \\
463: Detected in IRAC1+2 and 2MASS$^a$ & 6913 & 15702 & 4581 \\\hline
464: \end{tabular}
465: 
466: $^a$ A source is counted as detected in 2MASS if it has a S/N of at least
467: 10 in both $H$ and $K_s$.
468: \end{table}
469: 
470: 
471: Figures \ref{lupI} to \ref{lupIV} show RGB color-composite images of
472: Lupus I, III and IV respectively, with IRAC2 band at 4.5 $\mu$m in
473: blue, IRAC4 band at 8 $\mu$m in green and MIPS1 band at 24 $\mu$m in
474: red. The images show the overlapping areas between the mosaics done
475: with both instruments and include the GTO-imaged area in the core of
476: Lupus III at RA $\sim$ 16$^h$09$^m$ and DEC $\sim$ -39$^{\rm o}$10$'$
477: \citep{Allen2007}. The images show a large and dense concentration of
478: cold dust in the North-West of Lupus I and South of Lupus III. The
479: cloud emissions in Lupus are consistent with the Cambr\'esy's
480: extinction maps, but now shown in much more detail. A general 8 $\mu$m
481: emission gradient coming from the nearby Galactic plane was
482: compensated in the color scales to bring up the cloud structure. The
483: bright green emission to the North-East of the core in Lupus III is
484: likely produced by interstellar PAH molecules, which emit strongly at
485: 8 $\mu$m, possibly illuminated by the two bright Herbig Ae/Be stars HR
486: 5999 and HR 6000 in the core.
487: 
488:    \begin{figure}
489:    \centering
490:    \includegraphics[angle=0,width=17cm]{fig1bm.ps}
491:    \caption{Color-composite image of the mapped area in Lupus I by c2d.  The
492:      color mapping is blue for IRAC2 at 4.5 $\mu$m, green for IRAC4 at
493:      8.0 $\mu$m and red for MIPS1 at 24 $\mu$m. The bright red
494:        emission in the North-West of the cloud is produced by cold
495:        dust.}  \label{lupI}
496:    \end{figure}
497: 
498:    \begin{figure} \centering
499:    \includegraphics[angle=-90,width=17cm]{fig2bm.ps}
500:    \caption{Color-composite image of the mapped area in Lupus III by c2d.  The
501:      color mapping is the same as in Figure \ref{lupI}. The figure
502:      shows the very active star-forming core, which contains the two
503:      Herbig Ae/Be stars HR 5999 and HR 6000 at RA $\sim$ 16$^h$09$^m$
504:      and DEC $\sim$ -39$^{\rm o}$10$'$ and a stream of PAH emission to
505:      the North of the core (RA $\sim$ 16$^h$10$^m$ -- 16$^h$08$^m$ and
506:      DEC $\sim$ -39$^{\rm o}$).}  \label{lupIII} \end{figure}
507: 
508:    \begin{figure} \centering
509:    \includegraphics[angle=-90,width=17cm]{fig3bm.ps}
510:    \caption{Color-composite image of the mapped area in Lupus IV by c2d.  The
511:      color mapping is the same as in Figure \ref{lupI}. The figure
512:      shows the remnant cloud structure emitting at long wavelengths
513:      and many young stellar objects covering a large range of
514:      evolutionary types.  The very red source in the South of the
515:      mosaic is the unrelated Planetary Nebula PN
516:      G336.3+08.0.}  \label{lupIV} \end{figure}
517: 
518: \subsection{Differential source counts}
519: \label{diffcounts}
520: 
521: Given the proximity to the Galactic plane and center (see
522: Table~\ref{cloudinfo}), the Lupus clouds are expected to show a
523: relatively high number of background stars compared to the other
524: clouds surveyed by c2d. Figures \ref{diffcount} to \ref{diffcount2}
525: show the differential source counts for the Lupus regions. Since Lupus
526: I, at $+17^\circ$, is further from the Galactic plane than Lupus III
527: and IV (at 8--9$^\circ$) and likewise the two off-cloud fields with
528: both IRAC and MIPS observations are at $+4^\circ$ and $+12^\circ$
529: closer to the Galactic plane than the on-cloud fields on average, we
530: expect larger differences between the counts of individual regions
531: compared to the other c2d clouds. Figures \ref{diffcount1} and
532: \ref{diffcount2} therefore separate the Lupus I cloud from the III and
533: IV clouds. The observed differential source counts are compared to the
534: predictions of the Galactic source counts by \cite{Wainscoat1992}. In
535: all plots a steady increase is seen in number of sources per magnitude
536: bin up to a break at about 16--16.5~magnitudes in IRAC bands 1 and 2
537: and 15--15.5 in bands 3 and 4 corresponding roughly to the sensitivity
538: levels of the surveys.
539: 
540: \begin{table}
541:   \caption{Ecliptic and galactic coordinates for the centers of the mapped IRAC regions and their areas.}\label{cloudinfo}
542: \begin{tabular}{lccc}\hline\hline
543:   & $(\alpha,\delta)$ & $(l,b)$ & area  \\
544:   &  deg              & deg     & deg$^2$ \\\hline
545:   Lupus I    & 235.4 ; $-$34.2   & 338.6 ; 16.7 & 1.391 \\
546:   Lupus III  & 242.7 ; $-$38.6   & 340.2 ;  9.4 & 1.336 \\
547:   Lupus IV   & 240.4 ; $-$41.8   & 336.7 ;  8.2 & 0.374 \\
548:   OC2        & 243.2 ; $-$34.1   & 343.7 ; 12.4 & 0.051 \\
549:   OC6        & 250.0 ; $-$40.6   & 342.7 ;  4.0 & 0.051 \\ \hline
550: \end{tabular}
551: \end{table}
552: 
553: Some interesting differences are seen between the shorter wavelength
554: IRAC bands 1 and 2 compared to the longer wavelength bands 3 and 4. In
555: the shorter bands the match between the on- and off-cloud counts is
556: poor with the off-cloud fields showing significantly more sources in
557: given magnitude bins than do the on-cloud fields; consistent with
558: their proximity to the Galactic plane. The on-cloud source counts are
559: clearly higher in the Lupus III and IV regions than the Lupus I
560: region. In each plot the observed on-cloud source counts at IRAC bands
561: 1 and 2 are traced reasonably well by the Wainscoat model
562: predictions. These effects all suggest that the source counts in the
563: shorter IRAC bands are dominated by background stars as expected. In
564: the longer IRAC bands the situation is slightly different: both the
565: off-cloud and on-cloud source counts are here seen to exceed the
566: predictions of the Wainscoat models at faint magnitudes. Also the
567: difference between the different on-cloud regions internally or
568: compared to the off-cloud regions are found to be smaller than in IRAC
569: bands 1 and 2. This is likely due to the longer wavelength source
570: counts starting to be dominated by the extra-galactic background,
571: which naturally is independent of the Galactic location.
572: 
573: Compared to other c2d clouds, the differential source counts towards
574: the Lupus clouds show a similar behaviour to that seen in Serpens,
575: where the source counts exceed the galactic models in IRAC bands 3 and
576: 4 and match better in IRAC bands 1 and 2 \citep{Harvey2006}. Both
577: clouds are the closest to the direction of the galactic centre, with
578: the Lupus off-cloud fields being the closest of the whole c2d data set
579: (Table \ref{cloudinfo}). The source counts in the Chamaleon II (Cha II
580: hereafter) and Perseus clouds, well away from the Galactic plane, show
581: good matches with the Wainscoat models at all bands
582: (\citealt{Porras2007} and \citealt{Jorgensen2006}, respectively). This
583: illustrates the strong dependence of the source counts on galactic
584: coordinates and suggests the presence of a larger number of mid-IR
585: sources close to the Galacic center than those predicted by the
586: Wainscoat models. Of course, the excess of mid-IR source counts
587: towards the Lupus clouds could also be partially attributed to the
588: presence of YSO's in these clouds in cases where the excesses are
589: larger in the on-cloud than in the off-cloud regions. This would also
590: imply a larger number of YSO's in Lupus III and IV than in Lupus I.
591: 
592: 
593: \begin{figure}
594: \resizebox{\hsize}{!}{\includegraphics{fig4bm.eps}}
595: \caption{Differential source counts for the on- and off-cloud regions
596:   (grey and black, respectively). The predictions from the Wainscoat
597:   models are shown with the dashed lines for each of the three parts
598:   of the cloud (Lupus IV in dot-dashed line, Lupus III in dotted line
599:   and Lupus I with dashed line).} \label{diffcount}
600: \end{figure}
601: \begin{figure}
602: \resizebox{\hsize}{!}{\includegraphics{fig5bm.eps}}
603: \caption{As in Fig.~\ref{diffcount} but here only for the Lupus I
604:   cloud.} \label{diffcount1}
605: \end{figure}
606: \begin{figure}
607: \resizebox{\hsize}{!}{\includegraphics{fig6bm.eps}}
608: \caption{As in Fig.~\ref{diffcount} but here only for the Lupus III
609:   and IV clouds.} \label{diffcount2}
610: \end{figure}
611: 
612: \section{Young stellar objects and pre-main-sequence stars in Lupus}
613: \label{YSOc_list}
614: 
615: This section presents and discusses the complete list of Young Stellar
616: Objects (YSOs hereafter) and Pre-Main Sequence (PMS) stars in the
617: three Lupus clouds, obtained from the new Spitzer observations. The
618: final list of objects is obtained by merging the IR-excess sources
619: from Spitzer (YSOs) with the list of all other known young stars in
620: the clouds from previous optical and near-IR surveys inside the
621: Spitzer covered area (PMS stars). These two c2d standards were
622: introduced in  \citet{Harvey2007a} and \citet{Alcala2008}. The full 
623: list of YSOs and PMS stars is given in Table \ref{YSO_list1}.
624: 
625: 
626: %The following sections describe in detail how our complete list of
627: %objects was selected and fixed: the IRAC and MIPS data were used to
628: %establish a list of YSOc and these were hand-checked to make the YSO
629: %list.   in Table \ref{YSO_list1}.  All other
630: %spectroscopically confirmed members (PMS) of the clouds found in other
631: %studies which escaped detection with the c2d SED filter either because
632: %they did not show high quality flux detections or substantial excess
633: %in the IRAC1 to 4 and MIPS1 bands are also included in the list.
634: %Finally, all the candidate PMS stars in the IRAC observed area from¤
635: %previous optical and near-IR photometric surveys but without
636: %spectroscopic confirmation were also included. The detection of clear
637: %excesses at mid-IR wavelengths of some of these objects partially
638: %confirms their previous selection as cloud members and helps us also
639: %assessing the quality of the different selection criteria used to
640: %select them . If the PMS candidate has been
641: %spectroscopically confirmed, it is listed in Table \ref{YSO_list1} as
642: %`PMS' in the column for PMS Status, otherwise is listed as a candidate
643: %or `Cnd'. 
644: 
645: %To provide a definitive confirmation of the youth and membership to
646: %the cloud of the new YSO candidates from Spitzer reported in this work
647: %for the first time, an optical spectroscopy program is on-going. The
648: %optical spectroscopy will be used to determine their spectral types
649: %and search for the Li~I absorption and H$\alpha$ emission lines. The
650: %results of these observations will be presented elsewhere and here we
651: %use all currently available spectroscopic information from the literature.
652: 
653: \subsection{Selection of YSOs with IRAC and MIPS data}
654: \label{yso_sel}
655: 
656: We call an object a Young Stellar Object candidate
657: (YSOc) if it appears in the catalog of such objects delivered in the
658: c2d delivery \citep{Evans2007}. This
659: sample of objects was also visually inspected to subtract suspect
660: galaxies and to add obvious YSOs missed by the c2d criteria either
661: because they were too faint at short wavelengths, but associated with
662: millimeter emission, or because they were saturated in the Spitzer
663: images.  The objects in the resulting list are called simply YSOs;
664: however some caution should still be exercised.  
665: 
666: %One of the main goals of the c2d project is to provide a comprehensive
667: %list of all the YSOs detected by Spitzer in the clouds. For that we
668: %will use the IRAC and MIPS colors of the sources and a detailed method
669: %developed by the team to minimize the presence of extragalactic
670: %sources in the YSO catalogs. 
671: 
672: Complete details on the YSO selection method can be found in the c2d
673: Delivery Documentation \citep{Evans2007} and in \citet{Harvey2007a},
674: but in short, it consists in the definition of an empirical
675: probability function which depends on the relative position of any
676: given source in several color-color and color-magnitude diagrams where
677: diffuse boundaries have been determined to obtain an optimal
678: separation between young stars and galaxies. For that comparison, the
679: SWIRE catalog \citep{Surace2004} was ``extincted'' and resampled to
680: match as accurately as possible the c2d sensitivity limits for each
681: star forming cloud and to provide the statistical color distributions
682: expected for the background galaxies in our fields. The filter also
683: includes a flux cut-off to exclude bright galactic post-AGB stars in
684: the background which resemble Class III objects in the cloud. Figure 
685: \ref{YSO_sel} shows the Spitzer color-color
686: and color-magnitude diagrams used to select the YSOs in Table
687: \ref{YSO_list1}.
688: 
689: 
690: %The method requires detection in all IRAC bands and in MIPS1 with S/Ns
691: %$>$ 3 (quality classes `A', `B', `C' or `K' in the delivered catalogs,
692: %`K' meaning confused) to classify an object as either a YSO or a
693: %background galaxy. 
694: 
695: 
696: \begin{figure}
697: \centering \includegraphics[width=12cm]{fig7bm.ps}
698: \caption{Color-magnitude and color-color diagrams for the Lupus I
699:   (left), III (center) and IV (right) clouds. The black dot-dashed
700:   lines show the ``fuzzy'' color-magnitude cuts that define the YSO
701:   candidate criterion \citep{Harvey2007a} in the various
702:   color-magnitude spaces. The red dot-dashed lines show hard limits,
703:   fainter than which objects are excluded from the YSO
704:   category.} \label{YSO_sel}
705: \end{figure}
706: 
707: The application of this method in the Lupus
708: catalogs yields 18, 69 and 12 YSO candidates in Lupus I, III and IV
709: respectively (shown in Figure \ref{YSO_sel} as red dots and crosses
710: for point- and extended sources, respectively). Out of these, only 4
711: (22 \%), 26 (38 \%) and 2 (16 \%) were already known from previous
712: ground-based optical and near-IR observations. This shows that the
713: space observatory, with its highly sensitive detectors, has multiplied
714: by 5, 4, and 6 the number of known YSOs in the three clouds
715: respectively, capturing very low-mass young objects which escaped
716: detection due to low sensitivities and also getting the stars with
717: moderate excess only in the mid-IR which were not identified as excess
718: sources in near-IR surveys. 
719: 
720: The optical, IRAC and MIPS images of all the objects in the sample
721: were examined to confirm the point-source nature and to search for
722: blended features or long-period visual binaries. In Lupus I we
723: identified two galaxies, clearly resolved in the optical data, which
724: corresponded to two sources each in the Spitzer bands. Interestingly,
725: all of them were classified as 'YSOc\_PAH\_em' in the Spitzer catalog
726: of the region, due to their strong 8 $\mu$m emission due to PAHs (see
727: the delivery document for a description of these classes). No other
728: resolved galaxies were found in the optical images of all the other
729: YSO candidates in Lupus III and IV. However, a fake source was found
730: in the extended IRAC emission around the embedded object IRAS
731: 15356-3430 (see also \S~\ref{outflow_sources}). These sources were taken
732: off our YSO list and reduced the list of YSOs in Lupus I from 18 to
733: 13. 
734: 
735: \subsection{Sample of PMS and PMS candidate stars in Lupus}
736: \label{pms_sel}
737: 
738: We will use the term Pre-Main Sequence (PMS) star for other objects 
739: added to the list whose youth had already been confirmed using other 
740: observational techniques, mostly optical spectroscopy. 
741: If an object has not been spectroscopically confirmed as young but it was 
742: selected by its optical and near-IR photometry as such it will be labeled as 
743: a PMS candidate. 
744: 
745: We have added to our list the PMS candidates found by
746: \citet{Chapman2007} in the MIPS-only covered area. Because the MIPS coverage was much more extensive than the IRAC
747: coverage, there are sources that have clear excesses over photospheres
748: based on 2MASS and MIPS data, but for which we have no IRAC data
749: \citep{Chapman2007}. They would have been classified as YSOs, but
750:  since they cannot be tested against the galaxy
751: filter, we do not include them as YSOs or even YSOc. They are listed
752: as PMS candidates. The embedded class 0 
753: source Lupus3MMS, detected by \cite{Tachihara2007} at 1.2 mm, was 
754: added to the YSO sample in Lupus III. This object was detected at all 
755: IRAC and MIPS bands, but with too
756: poor S/N in the shorter wavelengths to be classified as a YSO
757: automatically. 
758: 
759: Also, we added all
760: previously known members of Lupus cited in Comer\'on (2008, in prep.),
761: which come mostly from the list of H$\alpha$-selected PMS stars in
762: Lupus of \citet{Schwartz1977} after the revisions of
763: \citet{Krautter1991} , \citet{Hughes1994} and \citet{Comeron2003}. This only added to our YSO
764: list the PMS stars without any detectable IR excess with good enough S/N.
765: It must be noted that the multi-color
766: criteria described in the previous section already recovered the 28
767: Classical T Tauri stars in the cloud listed in Comer\'on 2008 (in
768: prep.)  with clear IR excess in the Spitzer wavelength range and
769: good detections in all IRAC and MIPS1 bands and did not select the
770: other 13 from that list that did not show detectable IR
771: excesses. Interestingly, the only two bona-fide CTTs with IR excess
772: but not selected as YSOs by our criteria are in high background
773: emission regions and therefore have bad quality detection in one or
774: more Spitzer bands.
775: 
776: The candidate PMS stars from the previous optical and near- and mid-IR
777: surveys of \citet{Nakajima2000}, \citet{LopezMarti2005},
778: \citet{Allers2006}, and \citet{Allen2007} were also added to our list
779: of objects. These sources are labeled as PMS candidates in the 
780: ``PMS'' column of Table \ref{YSO_list1} and their respective 
781: original references are given in column ``Sel''. The
782: percentages of candidates from these studies that show IR excess in
783: the Spitzer bands are 6/18 (33 \%), 5/15 (33 \%), 1/3 (33 \%), and
784: 16/16 (100 \%), respectively. According to this, the H$\alpha$
785: narrow-band imaging criteria by \citet{LopezMarti2005} have a success
786: rate equal to that of the optical and near-IR deep photometric surveys
787: of \citet{Nakajima2000} and \citet{Allers2006}. Overall, the small abundance
788: of mid-IR excess sources in these samples is surprising
789: since the all three estimators used to select them are traditional proxies for 
790: disk mass accretion towards the central star and should be correlated
791: with the presence of a disk (see \S~\ref{disks} for a discussion on the 
792: disks properties). Finally, all the candidates selected
793: by \citet{Allen2007} in the Lupus III core show mid-IR excess since
794: IRAC data from the GTO observations were used to select them but those
795: not classified as YSOs by our criteria fall in the galaxy areas of the
796: color magnitude diagrams and have SEDs difficult to explain with star
797: plus disk models. This comparison with other
798: samples of YSO candidates gives us confidence on the robustness of the
799: c2d color criteria for selecting a reliable set of YSOs in the
800: clouds. In the following sections, we will analyze separately the YSO
801: sample and the total list objects. 
802: 
803: 
804: 
805: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%   Table   %%%%%%%%%%%%%%%%%%%%%%%%%%%
806: \begin{table}[h]
807: \begin{center}
808: \center{\caption{Total Number of Stars and YSOs in the
809:        Lupus Clouds Organized by SED Class.
810:        \label{YSO_classes}}}
811: 
812: \begin{tabular}{lcccccc}
813: \tableline  \tableline
814: Lada Class  & \multicolumn{3}{c}{YSOs}          & \multicolumn{3}{c}{Total} \\
815:            & Lupus I  &  Lupus III  &  Lupus IV   & Lupus I  & Lupus III   & Lupus IV \\
816: \tableline
817: I   &  2 (15 \%)  &  2 (3 \%)  &  1 (8 \%)  &  2 (12 \%)  &  5 (4 \%)  &  1 (6 \%) \\
818: Flat   &  3 (23 \%)  &  6 (9 \%)  &  1 (8 \%)  &  3 (18 \%)  &  8 (6 \%)  &  1 (6 \%) \\
819: II   &  6 (47 \%)  & 41 (59 \%)  &  5 (42 \%)  &  8 (47 \%)  & 56 (45 \%)  & 11 (61 \%) \\
820: III   &  2 (15 \%)  & 20 (29 \%)  &  5 (42 \%)  &  4 (23 \%)  & 55 (44 \%)  &  5 (28 \%) \\
821: Total & 13 &  69 &  12 &  17 & 124 &  18 \\
822: \tableline
823: \end{tabular}
824: \end{center}
825: \tablecomments{The percentages for each class in the table are calculated with respect to
826: the total numbers at the bottom of each column. The `Total' population consists of the YSOs 
827: plus the PMS stars and candidates (\S~\ref{YSOc_list}).}
828: \end{table}
829: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
830: 
831: \subsection{Class distribution of the sample}
832: 
833: The complete sample contains 17, 124
834: and 18 objects in Lupus I, III and IV, respectively. Table
835: \ref{YSO_classes} shows the number of objects per cloud and
836: their respective SED classes, as defined by \cite{Lada1984} and
837: extended by \cite{Greene1994} (i.e. Class I for objects with
838: $\alpha_{(K-24 \mu m)} > 0.3$, Class Flat for $0.3 > \alpha_{(K-24 \mu
839:   m)} \geq -0.3$, Class II for $-0.3 > \alpha_{(K-24 \mu m)} \geq
840: -1.6$, and finally Class III for $\alpha_{(K-24 \mu m)} < -1.6$). The
841: table also shows the total number of YSOs, as defined above, for each
842: cloud to facilitate comparison to other c2d clouds where the PMS
843: sample will be obviously different. The number of sources shows that Lupus 
844: in general
845: presents a moderate star formation activity compared with the other
846: clouds surveyed by the Spitzer c2d program, only larger than that
847: measured in Cha II \citep{Alcala2008}. It also shows that amongst
848: them, the most active regions are Lupus III, followed by Lupus
849: I and IV, which present a lower amount of YSOs.
850: 
851:    \begin{figure}
852:    \centering
853:    \includegraphics[angle=90,width=15cm]{fig8bm.ps}
854:    \caption{Distribution of objects in $\alpha_{(K-24 \mu m)}$ for
855:    Lupus I, III and IV. Thin lines show the YSO population and dotted
856:    lines the total sample of stars. The intervals defining the Lada
857:    classes are marked with vertical dashed lines. The comparison illustrates
858:    that Lupus I is a region with a small number of YSOs, many of them
859:    being early classes, Lupus III contains all kinds of objects, with a
860:    great number of Class III sources while Lupus IV contains almost no
861:    early class sources and relatively more late classes.}
862:    \label{alpha_hist}
863:    \end{figure}
864: 
865: 
866:    Figure \ref{alpha_hist} shows the distribution of $\alpha_{(K-24
867:      \mu m)}$ values in the three clouds. It illustrates that Lupus
868:    III shows a larger variety of objects, from Class I to Class III,
869:    while Lupus I has a much smaller number of YSOs relatively rich in
870:    Class I and Flat spectrum objects, and Lupus IV shows a clear lack of early class
871:    objects. The figure also shows that the addition of spectroscopic
872:    or photometric member candidates to the YSO samples mostly adds
873:    Class III objects in Lupus III, whose association with the cloud
874:    cannot be confirmed until optical spectroscopy reveals youth
875:    signatures. This is a selection effect, since Lupus III is the most
876:    studied region of the three and illustrates that the difference in
877:    percentages between the YSO and the total sample is only driven 
878:    by different available data and not based on physical characteristics
879:    of the stars. For this reason, we will concentrate
880:    in the YSO sample on the following discussion, which has been
881:    selected as genuine IR-excess objects and therefore likely belong
882:    to the star-forming clouds. The analysis below is only valid as a rough indicator
883:    of the relative abundances of objects within Lupus. A more solid
884:    statistical analysis which combines all the YSO samples in all c2d
885:    clouds will be presented elsewhere (N. Evans et al., in
886:    preparation).
887: 
888: One interesting difference is that the percentage of Class I and Flat
889: sources compared to the number of Class II and III sources is
890: particularly high in Lupus I, compared to Lupus III and IV. Assuming
891: that the classes correspond to a succession of evolutionary phases
892: from the cores to the disks, this suggests that Lupus I, the least
893: active cloud of the three in terms of YSO abundance, is in a less 
894: evolved phase of evolution, while Lupus III and Lupus IV are respectively 
895: more evolved than Lupus I.
896: 
897: It is also possible to study the later phases of disk evolution: the
898: percentage of Class III vs Class II plus III (i.e. N(III)/(N(II) +
899: N(III))) YSOs in the clouds (50$\pm$25\% in Lupus IV, 32$\pm$16\% in
900: Lupus III and 25$\pm$12\% in Lupus I) suggests that star formation
901: started longer ago in Lupus IV than in any of the other clouds. The
902: errors are dominated by the actual number of Class III objects, or
903: alternatively, the completeness of the sample for this kind of
904: objects. Given our selection criteria based on presence of IR excess,
905: the sample will be biased against the Class III objects with little
906: excesses. However, we can estimate the Class III completeness level to
907: be of the order of 50\% from the number of spectroscopically confirmed
908: young stars from the `Sz' catalog which were observed to have
909: detectable IR excess in the Spitzer bands (namely, 13 out of 28). This
910: uncertainty dominates the total error budget of these ratios and let
911: us only conclude significatively that Lupus IV has a larger percentage of
912: Class III objects than any of the other two clouds. Actually, this is
913: the largest number of Class III vs Class II+III sources in any of the
914: c2d studied star-forming clouds
915: (e.g. \citealt{Harvey2007a,Alcala2008}). Comparing Lupus III and IV,
916: which have a similar density of YSOs, these numbers also suggest that
917: the major star-formation activity already took place (and is about to
918: start again, see \S~\ref{extinction}) in Lupus IV, while it is currently
919: taking place in Lupus III, mostly in its very dense star-forming
920: core. 
921: 
922:  We can confirm our Class III completeness ratio of 50\% given
923: above with the list of XMM detections in 
924: the Lupus III core by \cite{Gondoin2006}, which is deeper than
925: the ROSAT All Sky Survey by \cite{Krautter1997}. That survey 
926: covers a circular area of 30$'$ in diameter around the Lupus III core and found 
927: 102 X-ray source detections, out of which 25 are associated with 
928: optical and IR counterparts. A cross-match of
929: our total list of objects and the list of X-ray detections with a match
930: range of 4 arcsec yielded 24 matches, 13 of which have SEDs indicating
931: the presence of a disk and the remaining 11 showing no IR excess. 
932: All stars with associated X-ray emission were labeled in Table \ref{YSO_list1}
933: with a reference to \cite{Gondoin2006}.
934: Assuming that the presence of X-ray emission is a signature of youth, this 
935: yields an X-ray disk fraction in the core in Lupus III of 54\% which is of the order of the 
936: Class III completeness we derived from the independent measurement
937: with the H$\alpha$-selected `Sz' catalog above.
938: 
939: The full list of PMS stars is given in Table \ref{YSO_list1}: columns
940: 2 to 5 give their c2d and previous names of the objects and their
941: coordinates in the Spitzer catalog, column 6 gives the selection
942: source as described above, a flag determining the status of membership
943: spectroscopic confirmation (PMS status) of the object is given in
944: column 7, the Spitzer-derived class based on the SED slope
945: $\alpha_{(K-24 \mu m)}$ is given in column 8 and the references to the
946: objects are given in column 9.
947: 
948: %The following sections \ref{disks} and \ref{clouds} describe the
949: %properties of the individual sources and global properties of the
950: %clouds and clusters, respectively.
951: 
952: %\section{Evolution: Classes and Stages}
953: \section{Individual sources properties}
954: 
955: \label{disks}
956: 
957: The properties of the circumstellar material surrounding the young
958: stars in Lupus can be studied with the optical to mid-infrared
959: emission in several different ways. We are specifically interested in
960: estimating the overall disk fraction, the amount of circumstellar dust, 
961: and the morphology of the circumstellar disks (flared versus flat) which
962: is likely to represent some stage in the
963: formation of planetary systems. The special interest that the new Spitzer data bring to the
964:   analysis comes from the fact that the IRAC and MIPS data, due to
965:   their wavelength coverage, probe the region in the disks around 
966:   low-mass stars between $\sim$ 0.1 and 5 AU, where planet formation
967:   may take place. Therefore, large samples of Spitzer SEDs of
968:   star-forming regions with ages in the critical time-frame between 1
969:   and 10 Myr can be used to study statistically the initial conditions
970:   for planet formation.
971: 
972: \subsection{Color-Color and Color-Magnitude diagrams}
973: \label{colorcolor}
974: 
975: Infrared color-color (CC) and color-magnitude (CM) diagrams are good
976: diagnostic tools for the investigation of circumstellar matter around
977: YSOs in a statistical way \citep[][and references
978: therein]{Hartmann2005,Lada2006}. In recent years, several authors have
979: produced grids of YSO models and computed their colors in the Spitzer
980: bands to allow direct comparison with the observations
981: \citep{Whitney2003,Allen2004,Robitaille2006}.  Both previous data sets
982: and the models roughly agree in the spatial distributions of the
983: objects of different Lada classes in the different CC and CM diagrams
984: although the interplay between the presence or absence of envelopes
985: and disks together with the whole possible range of system
986: inclinations and total interstellar extinctions makes the problem
987: highly degenerate. Therefore it is impossible to associate a position
988: in those diagrams with a physical configuration for individual
989: objects. \cite{Robitaille2006} computed a very large grid of YSOs
990: with a broad range of physical parameters, and inclination angles that
991: can be used to make statistical analyses. In discussing the
992: evolutionary stages of our models, they adopt a "Stage" classification
993: analogous to the Class scheme, but referring to the actual
994: evolutionary stage of the object, based on its physical properties
995: (e.g., disk mass or envelope accretion rate) rather than properties of
996: its SED (e.g., slope). Stage 0 and I objects have significant
997: infalling envelopes and possibly disks, Stage II objects have
998: optically thick disks (and possible remains of a tenuous infalling
999: envelope), and Stage III objects have optically thin disks. Figures
1000: \ref{cccmd1} and \ref{cccmd2} compare three CC and CM diagrams of the
1001: Lupus YSO sample with the synthetic colors of a model cluster from
1002: \cite{Robitaille2006}.  To make these comparisons, we have scaled the
1003: models to 150 pc in the first case to compare them with the YSOs in
1004: Lupus I and IV and to 200 pc to compare them with those in Lupus III.
1005: 
1006: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_9 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1007: \begin{figure}[!ht]
1008: \epsscale{1.0}
1009: \plotone{fig9bm.ps}
1010: \caption{The Color-color and color-magnitude diagrams for the
1011:  spectroscopically confirmed PMS objects (gray filled diamonds) and
1012:  candidates (open diamonds) with Spitzer and 2MASS data reported in
1013:  Table~\ref{YSO_list1} in the Lupus I and IV clouds are plotted in the
1014:  left panels of each diagram. The colors derived from the SED models
1015:  by \citet{Robitaille2006} are plotted (in gray-scale intensity
1016:  representing the density of points) in the right panels of each
1017:  diagram.  The areas corresponding to the Stages I, II, and III as
1018:  defined by \citet{Robitaille2006} are also indicated in each
1019:  diagram. The label 'ALL' mark the regions where models of all
1020:  evolutionary stages can be present. The green circle represents
1021:  normal reddening free photospheres. The PMS stars in
1022:  Lupus I and IV fall in regions corresponding to Stage I to III sources.
1023:  \label{cccmd1}}
1024: \end{figure}
1025: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1026: 
1027: 
1028: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_10 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1029: \begin{figure}[!ht]
1030: \epsscale{1.0}
1031: \plotone{fig10bm.ps}
1032: \caption{Same as figure \ref{cccmd1} but with the YSOs in 
1033: Lupus III. \label{cccmd2}}
1034: \end{figure}
1035: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1036: 
1037: The comparison between the position of the Lupus YSOs and those of the
1038: model grids in the [3.6]-[4.5] vs [5.8]-[8.0] CC diagram shows several
1039: early Class~I and Flat source SEDs in Lupus I and IV (Figure
1040: \ref{cccmd1}), redder in [3.6]-[4.5] than the Stage II objects from
1041: Robitaille et al. and a large majority of Class~II sources in Lupus
1042: III (Figure \ref{cccmd2}). The position of the maximum density of such
1043: kind of objects is also consistent with that predicted by
1044: \citet{Allen2004} with a grid of physical disk models of
1045: \citet{Dalessio2005}. The [8.0] vs [3.6]-[8.0] CM diagram of Lupus III
1046: (Figure \ref{cccmd2}) shows a bimodal distribution with a large number
1047: of objects with photospheric [3.6]-[8.0] colors and a range of 8.0
1048: \micron\ magnitudes and another one with a similar range of
1049: luminosities and a range of color excesses in the IRAC bands
1050: compatible with the presence of moderate to large mid-IR excess.  No
1051: special difference is found between the distributions of
1052: spectroscopically confirmed PMS objects (grey diamonds) and candidate
1053: YSOs in any of the diagrams, which suggests that both populations
1054: contain indeed the same kind of objects.
1055: 
1056: \subsection{Spectral Energy Distributions}
1057: \label{SEDs}
1058: 
1059: We have constructed SEDs for each source, similar to what has been
1060: done for other star-forming regions \citep[e.g.,
1061: ][]{Hartmann2005,Lada2006,Sicilia-Aguilar2006}. These allow the full
1062: set of multiwavelegth data to be displayed for each source. Figure 
1063: \ref{seds_I} shows the SEDs of all Class I and `Flat' sources
1064: in the sample. For all objects, the c2d catalog provides the 2MASS
1065: near-IR magnitudes and the IRAC1-4 and MIPS1-3 fluxes. We have added
1066: all the complementary data described in \S~\ref{binaries} plus IRAS
1067: fluxes from the PSC and 1.3 mm fluxes from \cite{Nuernberger1997}. The
1068: open dots are the observed fluxes and the Spitzer data have been highlighted
1069: in grey to compare with previous observations. 
1070: 
1071: Figure \ref{seds_II} shows all the SEDs for the Class II and III
1072: objects in the Lupus sample for which there was sufficient data to
1073: make a good fit to a stellar SED and characterize the stars and disks
1074: separately. Two different
1075: procedures were applied for the SED fits: for the objects with known
1076: spectral types, marked as `PMS' in Table \ref{YSO_list1}, we obtained
1077: the best-fit visual extinction $A_V$ by fitting all photometry between
1078: $V$ and $J$ to the stellar NEXTGEN models \citep{Hauschildt1999}. This
1079: method provides very good agreement with the traditional use of the
1080: $R_C-I_C$ color, as compared with the results by \citet{Alcala2008}.
1081: The $\chi^2$ minimization is extended to try all spectral types
1082: between A0 and M9 and all visual extinctions $A_V$ between 0 and 30
1083: magnitudes for the new YSOs for which there is no spectral type
1084: available in the literature. This technique is similar to that used by
1085: Spezzi et al. (2007b, in preparation) with the c2d Cha II sample, and
1086: provides agreement within $\pm$ 200 K in the effective temperature
1087: when applied to objects with known spectral types and moderate
1088: extinctions. Once the spectral type and extinction are computed for
1089: the sample, the observed fluxes are dereddened with the extinction law
1090: by \citet{Weingartner2001} with $R_V = 5.5$ and fitted to the
1091: appropriate NEXTGEN
1092: stellar atmosphere models.  The sample of SEDs of stars from the
1093: Schwartz catalog (Sz) was also compared with the SED fits made by
1094: \citet{Hughes1994} and there is in general good agreement in the
1095: resulting dereddened stellar and disk energy distributions. The
1096: resulting spectral types and extinctions, together with their
1097: corresponding references are given in Table \ref{sed_results}.
1098: 
1099:   The SEDs show considerable variety. A useful classification
1100:   system has to rely on disk models which are again degenerate in several
1101:   parameters as e.g. the inclination angle of the disk or the degree
1102:   of dust processing of the disk. Therefore, we will not attempt a
1103:   full characterization of the SED types in this section but rather
1104:   define four types of objects and give their frequencies in the
1105:   sample. To help with the classification, we have plotted with a
1106:   dashed line the median SED of the Classical T Tauri stars (CTTs)
1107:   from Taurus \citep{Dalessio1999} normalized to the optical
1108:   dereddened fluxes of all the low-mass stars. Based on this
1109:   comparison benchmark, we can identify the systems whose colors from
1110:   optical to the millimeter follow closely the decaying slope of a
1111:   classical accreting optically thick disk around a low-mass star (e.g.  Sz100 or
1112:   SSTc2d J160901.4-392512). We will call these systems `T'-type, for T
1113:   Tauri. We then call `L'-type (from `low' IR excess) all those
1114:   objects where the IR excess is clearly smaller than the median SED of
1115:   a CTTs (e.g. ACK2006-19). These are what \citet{Lada2006} call the `anemic' disks. We
1116:   call `H'-type objects, those systems with higher IR fluxes than
1117:   those of the median CTTs SED (e.g. 2MASS J16081497-3857145). 
1118:   Finally, we will call `E'-type from
1119:   {\sl empty} the spectrally confirmed young stars in the clouds where
1120:   no IR excess at all is detected in the Spitzer bands (e.g. Sz 67). These are
1121:   systems with a very little amount of cold dust or ``debris''-like
1122:   disks (see also some in \citealt{Sicilia-Aguilar2006} and
1123:   \citealt{Lada2006}).
1124: 
1125:   These classifications are given in Table \ref{sed_results} for
1126:   each system. Considering the whole sample of YSOs and PMS stars and
1127:   candidates in Lupus, 22$\pm$10\% of them are
1128:   `T'-Type, 39$\pm$18\% are `L'-type, 6$\pm$2\% are `H'-type, and
1129:   19$\pm$8\% are `E'-type. The remaining 20\% are the Class I
1130:   and Flat SED sources. The large errors in this percentages come
1131:   from the difficulty in classifying the borderline cases if we
1132:   take into account the implicit uncertainties in the fitting process
1133:   of the dereddened photometry to the stellar
1134:   photospheres and from the completeness estimation for disk-less members
1135:   presented in \S~\ref{pms_sel}. We can consider a negligible amount of
1136:   `H'-type objects, likely due to source variability at different
1137:   wavelegths and problems with the stellar normalization. From the
1138:   remaining sample, approximately 40\% show `L'-type SEDs, 20\%
1139:   show `T'-type SEDs and another 20\% show `E'-type SEDs. The large
1140:   pertentage of `L'-type objects in Lupus is consistent with that
1141:   found in Serpens \citep{Harvey2007a} and suggests more evolved inner
1142:   disks in these objects compared to the initial inner disk
1143:   confguration in the T Tauri stars. This percentages contrasts interestingly with 
1144:   the 30\% of `T'-type and 20\% of `L'-type SEDs found in the older 3 Myr star-forming 
1145:   cluster IC 348 by \cite{Lada2006}. These different disk populations in different
1146:   clouds could be related to different environmental conditions (e.g. clustered
1147:   vs extended star-formation), to a different stellar population (abundance of low-mass vs 
1148:   high-mass stars) or alternatively to a faster evolutionary time-scales for
1149:   the less flared `L'-type disks. More samples of disks will be needed to
1150:   pursue this research, which is outside the scope of this paper.
1151:   
1152:   One fifth of the sample are
1153:   `E'-type stars and this gives an overall disk fraction for the Lupus clouds 
1154:   of 68\% to 81\%, depending on whether the assume a 50\% completeness
1155:   ratio for diskless stars of or not. This range of values matches
1156:   well with other disk fractions calculated for 1 to 2 Myr-old  star forming regions
1157:   with near and mid-IR data \citep{Haisch2001,Sicilia-Aguilar2006}, 
1158:   including Cha II \citep{Alcala2008}.  The `E'-type objects in Lupus show 
1159:   photospheric IRAC and MIPS fluxes which 
1160:   indicate that the stars do not have any circumstellar dust out
1161:   to a distance of at least 50 AU, depending on the stellar luminosity
1162:   (see e.g. HR 6000, Sz 119 or Sz 124). Once spectroscopically confirmed 
1163:   as young members of the clouds and therefore with ages $<$ 3 Myrs, they 
1164:   will provide very valuable constraints to the disk dissipation time-scales and
1165:   mechanisms and will be the subject of a separate study.
1166: 
1167:   Also interestingly, the Spitzer data of a subsample of the `L'-type
1168:   objects show a very peculiar SED, with almost photospheric IRAC
1169:   and MIPS1 fluxes typical of Class III sources, and a MIPS2 excess
1170:   flux at 70 $\mu$m comparable to those of the CTTs. This subgroup has
1171:   been labeled `LU' in Table \ref{sed_results} and contains the
1172:   objects SSTc2d J154240.3-341343, Sz 91, and Sz 111. It represents
1173:   a 2$\pm$1\% of the young population in Lupus. Similar objects
1174:   have been found in Serpens \citep{Harvey2007a} and Cha II
1175:   \citep{Alcala2008} and seem to be extremely rare. These objects,
1176:   called cold disks \citep{Calvet2005,Brown2007}, are interpreted as
1177:   optically thick disks with large inner holes of several to tens of
1178:   AUs, where potentially planets could be currently forming. However,
1179:   both the objects in Lupus and the star ISO-ChaII 29 in Cha II
1180:   \citep{Alcala2008} are different from previously known systems of
1181:   this kind in that the long wavelength at which they show the sizable
1182:   excess (70 $\mu$m) implies larger inner holes of up to 70 AU, while
1183:   previously known objects show excess already at 24 $\mu$m, implying
1184:   inner holes smaller than 50 AU.
1185: 
1186: 
1187: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_11_a %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1188: \begin{figure}[!ht]
1189: \epsscale{0.9}
1190: \plotone{fig11bma.ps}
1191: \caption{Spectral energy distributions of the Class~I and Flat SED PMS
1192:        objects and YSOs ordered by RA and labeled with their
1193:        name and ID as in Table \ref{YSO_list1}. Only the observed fluxes are
1194:        plotted with open dots and upper limits with arrows. The new Spitzer data are shown in grey
1195:        to distinguish them from previous data.
1196:        \label{seds_I}}
1197: \end{figure}
1198: 
1199: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1200: 
1201: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_11_b %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1202: 
1203: \begin{figure}[!ht]
1204: \epsscale{0.9}
1205: \plotone{fig11bmb.ps}\\
1206: Fig.~\ref{seds_I} - Continued
1207: \end{figure}
1208: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1209: 
1210: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_12_a %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1211: \begin{figure}[!ht]
1212: \epsscale{0.9}
1213: \plotone{fig12bma.ps}\\
1214: \caption{Spectral energy distributions of the Class~II and Class~III
1215:        PMS objects and candidates ordered by RA and labeled with their
1216:        names and ID as in Table \ref{YSO_list1}. The observed, dereddened
1217:        fluxes and upper limits are represented with open and solid dots and arrows,
1218:        respectively. The plotted error bars are usually smaller than
1219:        the symbols. NEXTGEN stellar models for the spectral type of
1220:        each star are shown in dotted lines. For all low-mass objects,
1221:        the median SED of the T Tauri stars in Taurus from
1222:        \cite{Dalessio1999} is shown normalized to the stellar
1223:        photosphere in dashed line for comparison.  \label{seds_II}}
1224: \end{figure}
1225: 
1226: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_12_b %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1227: 
1228: \begin{figure}[!ht]
1229: \epsscale{0.9}
1230: \plotone{fig12bmb.ps}\\
1231: Fig.~\ref{seds_II} - Continued. 
1232: \end{figure}
1233: 
1234: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_12_c %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1235: 
1236: \begin{figure}[!ht]
1237: \epsscale{0.9}
1238: \plotone{fig12bmc.ps}\\
1239: Fig.~\ref{seds_II} - Continued.
1240: \end{figure}
1241: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1242: 
1243: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_12_d %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1244: 
1245: \begin{figure}[!ht]
1246: \epsscale{0.9}
1247: \plotone{fig12bmd.ps}\\
1248: Fig.~\ref{seds_II} - Continued.
1249: \end{figure}
1250: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1251: 
1252: 
1253: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_12_e %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1254: 
1255: \begin{figure}[!ht]
1256: \epsscale{0.9}
1257: \plotone{fig12bme.ps}\\
1258: Fig.~\ref{seds_II} - Continued.
1259: \end{figure}
1260: 
1261: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_12_f %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1262: 
1263: \begin{figure}[!ht]
1264: \epsscale{0.9}
1265: \plotone{fig12bmf.ps}\\
1266: Fig.~\ref{seds_II} - Continued.
1267: \end{figure}
1268: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1269: 
1270: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_12_g %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1271: 
1272: \begin{figure}[!ht]
1273: \epsscale{0.9}
1274: \plotone{fig12bmg.ps}\\
1275: Fig.~\ref{seds_II} - Continued.
1276: \end{figure}
1277: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1278: 
1279: 
1280: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_12h %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1281: 
1282: \begin{figure}[!ht]
1283: \epsscale{0.9}
1284: \plotone{fig12bmh.ps}\\
1285: Fig.~\ref{seds_II} - Continued.
1286: \end{figure}
1287: 
1288: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_12i %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1289: 
1290: \begin{figure}[!ht]
1291: \epsscale{0.9}
1292: \plotone{fig12bmi.ps}\\
1293: Fig.~\ref{seds_II} - Continued.
1294: \end{figure}
1295: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1296: 
1297: \clearpage
1298: 
1299: \subsection{The luminosity function in Lupus}
1300: \label{star_luminosities}
1301: 
1302: %This section analyse the luminosity function of the complete Lupus
1303: %sample. In order to do this, we will consider both the YSOs and full
1304: %list of PMS stars. The latter category contains the least biased
1305: %sample of the PMS population in the clouds since the IR-excess YSOs
1306: %are complemented by all the spectroscopically confirmed PMS in the
1307: %clouds without measurable IR excess. In this section we will
1308: %concentrate on the analysis of this sample of ``bona-fide'' members of
1309: %the Lupus clouds.
1310: 
1311: A basic step in the characterization of the PMS population is the
1312: determination of its degree of completeness in the effective
1313: temperature and luminosity scales, which might be correlated in those
1314: objects that formed more or less simultaneously.  Both parameters can
1315: be accurately determined from the spectral type and the SEDs shown in
1316: the previous section for Class II and III objects. In order to calculate the best possible stellar
1317: luminosities for the sample, we integrate the NEXTGEN stellar model
1318: normalized to the dereddened optical fluxes (dotted curves in Figure
1319: \ref{seds_II}), and assume distances of 150 pc for Lupus I and IV and 
1320: 200 pc for Lupus III. The computed stellar luminosities can be found in Table
1321: \ref{sed_results}. The complete emission of the objects is obtained in
1322: the bolometric luminosity, where we integrate under all the observed
1323: fluxes of the SED, the corresponding bolometric flux is then converted
1324: to luminosities with the same distances quoted above for the different
1325: clouds.
1326: 
1327: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_18 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1328: \begin{figure}[!ht]
1329: \epsscale{0.7}
1330: \includegraphics[angle=90,scale=.60]{fig13bm.ps}
1331: \caption{Bolometric luminosity function for the Lupus young stellar population ({\sl solid histogram})
1332: and estimated correction for completeness effects ({\sl dashed
1333: histogram}). \label{lum_hist}}
1334: \end{figure}
1335: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_18 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1336: 
1337: Figure \ref{lum_hist} shows a histogram of the total bolometric
1338: luminosities for all YSOs and PMS stars in Lupus as a solid line. Compared to
1339: the same figure in Cha II \citep{Alcala2008} and in Serpens
1340: \citep{Harvey2007b}, it shows a larger population of low and very-low
1341: luminosity objects, which was already noted by previous surveys of the
1342: region (see F. Comer\'on 2008, in prep. and references therein), and a
1343: large number of objects with luminosities between 0.3 and 0.1
1344: L$_\odot$. Interestingly, the lowest luminosity members of the
1345: population are Lupus3MMS, the Class 0 object in Lupus III
1346: \citep{Tachihara2007}, SSTc2d J161027.4-390230, another flat SED
1347: object in the same cloud followed by late M-type brown dwarfs from
1348: \cite{Comeron2003}, \cite{Allers2006} and \cite{Allen2007}.
1349: 
1350: The peak of the luminosity function appears at 0.2 L$_\odot$, which
1351: corresponds to a 0.2 M$_\odot$ star (spectral type M5 at an age of 1
1352: Myr) and up to 1.0 M$_\odot$ for objects of 5 Myrs according to the
1353: PMS tracks by \citet{Baraffe1998} plus a tail at lower luminosites. 
1354: \cite{Harvey2007b} made an estimate of the completeness of the c2d
1355: catalogs by comparing a trimmed version of the deeper SWIRE catalog of
1356: extragalactic sources (\citealt{Surace2004}), taken to represent 100\%
1357: completeness by c2d standards, with the number counts of the c2d
1358: catalog in Serpens per luminosity bin. Their estimations suggest a 100
1359: \% completeness at luminosities $\geq$ 0.3 L$_\odot$ in the c2d
1360: catalogs, which corresponds to a 0.3 M$_\odot$ star (spectral type
1361: M4.5 at an age of 1 Myr).  The dashed line in Figure 13 is the
1362: luminosity histogram corrected for completeness effects in each
1363: luminosity bin, which suggests that up to 15 additional very
1364: low-luminosity objects ($\log L/L_\odot < -1.7$) were missed below the
1365: noise level of our c2d observations or were confused in the galactic
1366: region of the color-magnitude diagrams. The comparison of both
1367: luminosity histograms yields a completeness level in luminosity for
1368: our disk study of $\log L/L_\odot \approx -1.0$, which corresponds to
1369: a mass of $M \sim 0.1 M_\odot$ for a star of $\sim$ 1 Myr
1370: \citep{Baraffe1998}. The lower luminosity limit with respect to that
1371: in Serpens stems from the different distances to both clouds and is
1372: roughly consistent with the distance squared ratio.
1373: 
1374: \subsection{Disk fractional luminosities}
1375: \label{fractional_luminosities}
1376: 
1377: One way to characterize the circumstellar accretion disks around young
1378: stars is to compute the disk fractional luminosities $L_{\rm
1379: disk}/L_{\rm star}$ and compare them with that of passive reprocessing flaring
1380: disks \citep{Kenyon1987}. We computed those ratios by subtracting the
1381: stellar fluxes from the observed dereddened fluxes in the SEDs and
1382: integrating the IR excess throughout the SED. That was done for all
1383: the objects with sufficient data to allow a good fit of the stellar
1384: model to the SED, crucial to get a sensible disk luminosity. The
1385: resulting flux ratios are also given in Table \ref{sed_results}. 
1386: 
1387: The upper panel of Figure \ref{lstar_ldisk} shows the stellar
1388: luminosity versus the disk fractional luminosity for the sample of YSOs
1389: and PMS objects in Lupus. The grey dots correspond to visual binary
1390: systems, as reported in \S~\ref{binaries}, while the black dots
1391: represent single stars. In the lower planel, we have collapsed the
1392: stellar luminosity axis to get a histogram on disk fractional
1393: luminosities for the entire sample. This diagram offers a complete
1394: view of the star plus disk population in Lupus and shows several
1395: interesting results and trends. The $\log L_{\rm disk}/L_{\rm star}$
1396: ratio of objects in which we could not detect any sizable IR excess
1397: were set to $-3$ so that they appear to the left of both panels. The
1398: disk luminosity ratios show a strong dependence on the band used to
1399: normalize the dereddened fluxes to the stellar photospheric model and
1400: the extinction determination. These are only accurate in those objects
1401: with known spectral types, but the deviations are small enough to
1402: extract statistical trends from the diagram.
1403: 
1404: \begin{figure}[!ht]
1405: \epsscale{1.0}
1406: \includegraphics[scale=0.8]{fig14bm.ps}
1407: \caption{{\sl Top}: Correlation plot between the stellar luminosities and 
1408: the disk fractional luminosities for the PMS stars in Lupus. Solid and
1409: grey dots are single and binary stars, respectively and the dot-dashed
1410: line marks the approximate 24 $\mu$m detection limit. {\sl Bottom}:
1411: distribution of disk fractional luminosities for the Lupus sample.
1412: The solid and dashed lines show the single and binary stars
1413: distributions, respectively. Objects with negligible IR excess are
1414: shown in the $-3$ abscisa for completeness. \label{lstar_ldisk}}
1415: \end{figure}
1416: 
1417: Horizontal dashed lines in the upper panel separate the regions where
1418: the luminosities correspond to those of Herbig Ae stars (HAeBe's,
1419: spectral types earlier than F0), T Tauri stars (down to M7) and Brown
1420: dwarfs (BDs, below M7) of 1 Myr according to \cite{Baraffe1998}. The
1421: only two known HAeBe's in Lupus, HR 5999 and HR 6000, share the upper
1422: region with SSTc2d J161000.1-385401, a intermediate-mass HAeBe
1423: candidate in Lupus with small IR excess, which could be a background
1424: post-AGB star, given the resemblance of its SED to those of the old
1425: field objects found in Serpens \citep{Harvey2007b}. It can be also seen
1426: here, as well as in figure \ref{lum_hist}, that the great majority of
1427: the objects in Lupus are M-type T Tauri and very low mass stars.
1428: 
1429: The vertical lines separate the regimes where the luminosity ratios
1430: can be explained by different mechanisms: accretion disks ($L_{\rm
1431:   disk}/L_{\rm star}$ $>$ 0.1), and passive reprocessing disks (0.02
1432: $<$ $L_{\rm disk}/L_{\rm star}$ $<$ 0.08 \citealt{Kenyon1987}), and
1433: finally ``debris''-like disks ($L_{\rm disk}/L_{\rm star}$ $<$ 0.02,
1434: which is a conservative upper limit for the fractional disk luminosity
1435: of a transitional or young debris disk, \citealt{Currie2007}). The
1436: distribution of points and the histogram in the bottom panel show that
1437: the great majority of the disks found in Lupus have luminosities which
1438: imply some degree of accretion, similar to what was found in the Cha
1439: II cloud \citep{Alcala2008}. This result contrasts with the 40\%
1440: of disks with small IR excesses found in \S~\ref{SEDs} and shows
1441: that detailed Spitzer SED morphology studies are needed to provide a better
1442: description of the inner disks, which are greatly degenerate with the flux
1443: ratios in the T Tauri phase.
1444: 
1445: The grey dots and dashed histogram show the distribution of those YSOs
1446: in the plot which were found to be visual binaries in the optical
1447: images by Comer\'on et al (in prep.). It has been suggested that close
1448: binaries ($<$ 20 AU) play a central role in the evolution of the disks
1449: by dramatically reducing the incidence of disks around multiple stars
1450: \citep{Bouwman2006}, while wide ($>$ 100 AU) binaries do not affect
1451: significatively the presence and evolution of the circumstellar disks
1452: \citep{Pascucci2007}. Here we present further evidence for this lack
1453: of correlation between disks and wide orbit companions in Lupus, where
1454: binaries and single stars show the same distribution in terms of
1455: luminosity or disk fractional luminosity.
1456: 
1457: Finally, there seems to be a trend by which lower luminosity objects show
1458: higher disk to star luminosity ratios than higher luminosity
1459: stars. However, the luminosity
1460: distribution of the `H' and `T'-type objects only, the least affected by the 
1461: detection limits in the IR,
1462: is statistically equal to the luminosity distribution for the total sample. 
1463: Therefore, the trend reflects an observational bias caused by the
1464: limited detection capability of excesses at longer wavelengths for
1465: the faintest sources. The flux sensitivity limit for 24 $\mu$m detection 
1466: with S/N $>$ 3 is $\approx$ 1 mJy in our observations. This corresponds to photospheric
1467: emission of a $\approx$ 0.2 $L_\odot$ ($\log L_{\rm star} \approx
1468: -0.7$) star in our sample, which corresponds to a 0.2 $M_\odot$ (M5.5)
1469: star of 1 Myr. The star Sz 92 is close to the border with 0.47
1470: $L_\odot$ ($\log L_{\rm star} \approx -0.3$), a very small excess at
1471: 24 $\mu$m ($L_{\rm disk}/L_{\rm star}$ = 0.06, $\log L_{\rm
1472:   disk}/L_{\rm star}$ = -1.2) and a flux density at 24 $\mu$m of 1.5
1473: mJy. Fainter objects with larger excesses will be detected as soon as
1474: their 24 $\mu$m is larger than the limiting flux and this number will
1475: obviously depend on the level of background emission and crowdedness
1476: of the area where a given object is. This limit, shown with a
1477: dot-dashed line in the figure, explains the skewness of the
1478: data points towards the bottom right of the diagram and illustrates
1479: the disk parameter space probed with these observations. The survey is not
1480: sensitive to ``debris''-like disks around stars less massive than 0.2
1481: $M_\odot$ but detects all actively accreting disks down to the
1482: substellar regime since those should present detectable IR excess at
1483: all IRAC bands which are much more sensitive.
1484: 
1485: \subsection{A Two-Dimensional Classification System}
1486: \label{2D_param}
1487: 
1488: Finally, to explain the diversity in the SEDs, we developed a {\sl
1489:   second-order} set of parameters to classify the diversity of 2MASS +
1490: IRAC + MIPS SEDs that are found in the star-forming regions:
1491: $\lambda_{\rm turn-off}$ and $\alpha_{\rm excess}$  
1492: \citep{Cieza2007}. In short, $\lambda_{\rm excess}$ is the last
1493: wavelength in microns where the observed flux is photospheric and
1494: $\alpha_{\rm excess}$ is the slope computed as ${\rm dlog}(\lambda
1495: F_\lambda)/{\rm dlog}(\lambda)$ starting from $\lambda_{\rm turn-off}$
1496: and up to 70 $\mu$m when available. The first parameter gives us an
1497: indication of how close the circumstellar matter is to the central
1498: object and the second one is a measure of how optically thick and flared
1499: it is. Both numbers can be found for all the stars with disks in the
1500: sample in Table \ref{sed_results}.
1501: 
1502: \begin{figure}[!ht]
1503: \includegraphics[angle=90.,width=14cm]{fig15bm.ps}
1504: \caption{Distribution of excess slopes $\alpha_{\rm
1505:    excess}$ with respect to the wavelength at which the infrared
1506:    excess begins $\lambda_{\rm turn-off}$ for the Lupus sample (open
1507:    circles). Solid dots are objects which were detected at 1.3 mm by
1508:    \citet{Nuernberger1997} and grey dots are objects for which only
1509:    upper limits were obtained at 1.3 mm. CoKu Tau/4 shows the position
1510:    of the cold disks in this diagram for comparison. \label{lto_ae}}
1511: \end{figure}
1512: 
1513: Figure \ref{lto_ae} shows both values for the sample of YSOs in Lupus
1514: for which we could perform a sensible SED fit. Three objects (namely
1515: Par-Lup3-3, Par-Lup3-4 and Sz 99) were not included in this analysis
1516: since their SEDs did not allow for a good determination of
1517: $\lambda_{\rm turn-off}$ due to too much IR excess at short
1518: wavelengths in the first two cases and lack of IRAC data in the third
1519: one. \citet{Cieza2007} showed that $\lambda_{\rm turn-off}$ is a good
1520: ``evolutionary'' parameter, since it separates efficiently the
1521: Classical T Tauri stars from the Weak Line T Tauri stars. They also
1522: showed that the range of $\alpha_{\rm excess}$ values, which scale
1523: with the amount of emitting material in the disk for central objects
1524: of similar luminosity, grows with $\lambda_{\rm turn-off}$, which was
1525: interpreted by those authors as a signature of multiple ``evolutionary
1526: paths'' of the inner disks of T Tauri stars from the actively
1527: accreting phase (with typically $\lambda_{\rm turn-off} < 2\mu$m) to
1528: progressively more settled and optically thin disks with larger values
1529: of $\lambda_{\rm turn-off}$. The Lupus sample shows a remarkably
1530: similar distribution in Figure \ref{lto_ae}, as also do the Serpens and 
1531: Cha II samples, suggesting that the result is a common feature of 
1532: young stellar populations.
1533: 
1534: The diagram is also useful for identifying transitional disks, or cold
1535: disks \citep{Calvet2005,Brown2007}, i.e. systems with an optically
1536: thick outer disk with large inner holes (several AU). These objects
1537: appear in the right-upper part of this diagram. For comparison, we
1538: show the position of CoKu Tau/4, a well known T Tauri disk with a
1539: inner hole of $\sim$ 10 AU devoid of matter \citep{Dalessio2005}. We
1540: also labeled one of the new cold disks found in Lupus III, Sz 91,
1541: which is indeed is the most extreme upper-right object of the whole
1542: distribution. The determination of robust cut-offs in this diagram for
1543: identifying ``bona-fide'' disks with holes and their frequency in the
1544: full c2d data set is underway and will be presented elsewhere.
1545: 
1546: In order to compare the inner disk architectures of our sample
1547: of Lupus PMS stars with their outer disks, we have marked the `Sz'
1548: stars detected in 1.3 mm continuum observations by
1549: \cite{Nuernberger1997} with black dots, and the upper limits with grey
1550: dots in Figure \ref{lto_mm}.  As expected, all the millimeter 
1551: detections and upper limits in the diagram
1552: appear at $\alpha_{\rm excess} \geq -1.5$, which imply relatively
1553: optically thick inner disks for all values of $\lambda_{\rm
1554:   turn-off}$.
1555: 
1556: \begin{figure}[!ht]
1557: \includegraphics[angle=90.,width=15cm]{fig16bm.ps}
1558: \caption{Relationship between the continuum flux at 1.3 mm
1559:   from \citet{Nuernberger1997} and the wavelength $\lambda_{\rm
1560:     turn-off}$ at which the infrared excess begins for the Classical T
1561:   Tauri stars in Lupus. The arrow show the positions of the upper
1562:   limits, including that for the cold disk Sz 91. \label{lto_mm}}
1563: \end{figure}
1564: 
1565:   However, the really interesting result comes when comparing the
1566:   millimeter fluxes, which are a proxy for disk total mass
1567:   \citep{Beckwith1991,Andrews2005}, with $\lambda_{\rm turn-off}$, which is an
1568:   indirect indicator of the degree of dust depletion and clearing of
1569:   the inner disk. Figure \ref{lto_mm} shows the 1.3 mm continuum
1570:   fluxes and upper limits for the Lupus sample normalized to 200
1571:   pc compared to the $\lambda_{\rm turn-off}$. The horizontal line shows the 
1572:   average of the normalized
1573:   millimeter fluxes and the dotted lines the 1$\sigma$ error bar of
1574:   the mean. With the exception of Sz 98, disks with a range of
1575:   $\lambda_{\rm turn-off}$ or inner disk clearings appear to show
1576:   quite similar total disk masses, being all well inside the standard
1577:   deviation of the mean. The case of Sz 98 could be different, since
1578:   its SED suggests a possible contribution of a remnant envelope,
1579:   which would contribute to the total millimeter flux. In spite of
1580:   the low number of detections, the general result is that objects
1581:   with a range of inner disk configurations show total disk masses in
1582:   the same order of magnitude. This suggests an inside-out process
1583:   which empties the inner disk while the outer disk remain unchanged. 
1584:   \cite{Alexander2007} show that
1585:   disk clearing by photoevaporation is more efficient for smaller disk
1586:   masses while e.g. \cite{Edgar2007} argue that the efficiency of
1587:   giant planet formation and migration, the main competing mechanism
1588:   for inner disk dissipation, is proportional to the disk mass. The apparent
1589:   lack of correlation between inner disk clearing and total disk mass
1590:   is not compatible with the photoevaporation scenario while it is
1591:   with the planet formation one. However, the current
1592:   sample only contains 8 good detections, and several caveats should be
1593:   taken into account in the interpretation of this figure: apart from
1594:   the low number statistics, the $\lambda_{\rm turn-off}$ value will
1595:   be sensitive to the inclination angle, while the millimeter flux is
1596:   considered roughly insensitive to it (e.g. \citealt{Dalessio1999}).
1597:   More work will be done to quantify the extent of this relation with
1598:   larger data sets and will be presented elsewhere.
1599: 
1600: \clearpage
1601: 
1602: \subsection{Outflow sources}
1603: \label{outflow_sources}
1604: 
1605: Several previous studies have shown the suitability of the IRAC bands,
1606: and more specifically IRAC band 2 at 4.5 $\mu$m for detecting new
1607: high-velocity shocked outflows via the H$_2$ emission line found in
1608: that band (e.g. \citealt{Noriega-Crespo2004},
1609: \citealt{Jorgensen2006}). In this section we present the Spitzer data
1610: for the known outflow sources in the surveyed Lupus clouds (from
1611: the review by Comer\'on 2008, in prep.) and new emission nebulae found in the IRAC
1612: mosaics which could be high velocity outflows and shocks. Table
1613: \ref{outflows} summarizes the results of this effort and Figures
1614: \ref{hh_objects} and \ref{irasHH} show them.
1615: 
1616: 
1617: \begin{figure}[!ht]
1618: \epsscale{0.9}
1619: \plotone{fig17bm.ps}
1620: \caption{Known HH objects detected in Lupus with IRAC, together with a
1621: candidate new outflow source in Lupus III and a nebulous object in a
1622: ring of stars in Lupus IV. All images have a size of 2$'$x2$'$,
1623: linear stretch and are from the IRAC1 band at 3.6 $\mu$m, except
1624: the one showing HH 185, which is IRAC4 at 8.0 $\mu$m. North is up and
1625: East is to the left. Coordinates are given in Table
1626: \ref{outflows}. \label{hh_objects}}
1627: \end{figure}
1628: 
1629: 
1630: The HH 185 object, reported by \cite{Heyer1989}, shows an ellipsoidal
1631: shape close to the Flat SED object IRAS 15398-3359 in all IRAC
1632: bands.  HH 187 \citep{Heyer1989} is a faint nebulosity seen at IRAC1
1633: and 2 bands only, and brighter in the latter. HH 228 and HH 78 also
1634: appear in the Spitzer bands as small ($\sim$ 4$''$) faint nebulosities
1635: around their driving sources. The small [SII] outflow in Par-Lup3-4
1636: reported by \cite{Fernandez2005} and the HH 186 36$''$ [SII] jet
1637: around Sz 68 are not detected by the Spitzer observations. Only
1638: nebulosities appearing consistently in both epoch images were included
1639: to avoid artifacts. All the IRAC images were searched for nebulosities
1640: and we report here the detection a jet-like nebulous structure found
1641: at all IRAC bands in Lupus III and a nebulous object in a ring of
1642: point sources found in Lupus IV. The driving sources of these two
1643: objects were not identified.
1644: 
1645: 
1646: \begin{figure}[!ht]
1647: \epsscale{0.9}
1648: \plotone{fig18bm.ps}
1649: \caption{IRAC and MIPS images of the extended nebula around the 
1650: source IRAS 15356-3430. All images have a size of 4$'$x5$'$ and
1651: linear stretch. North is up and East is to the left. Coordinates
1652: are given in Table \ref{outflows}. \label{irasHH}}
1653: \end{figure}
1654: 
1655: 
1656: Figure \ref{irasHH} shows the remarkable emission around the the
1657: bright class I source IRAS 15356-3430. A 40$''$ elliptical nebulosity
1658: is observed in all IRAC bands and also detected in MIPS 24 and 70
1659: $\mu$m bands. The nebulosity is also found with a slightly smaller
1660: extension in the optical and 2MASS images. \cite{Carballo1992}
1661: classified it as a possible YSO or a galaxy based on its IRAS colors
1662: and \cite{Strauss1992} confirmed its YSO nature with optical
1663: spectroscopy. Its 70 $\mu$m flux is as high as that of Lupus3MMS but it
1664: has never been observed in the millimeter wavelengths. The SED slope
1665: of the object is smaller than that of Lupus3MMS or IRAS 15398-3359 and
1666: that makes it detectable in the near-IR bands. In any case, the
1667: Spitzer SEDs of all these sources with extended emission are clearly
1668: associated with extremelly embedded and actively accreting objects.
1669: 
1670: Interestingly, one class II and one class I YSO (Sz 98, Lupus3MMS)
1671: happen to have detection quality flag of `K' only in IRAC2 band at 4.5
1672: $\mu$m in the c2d catalog, which means that these sources could not be
1673: fitted by a stellar PSF at that wavelength. This suggests the possible
1674: presence of unresolved ($< 2''$) outflows detected preferentially
1675: through the shocked H$_2$ line in IRAC band 2 and is consistent with
1676: the fact that all three of them are extremely active accretors with
1677: large $\alpha_{(K-24 \mu m)}$ values and excesses well above the median SED of the T
1678: Tauri stars in Taurus. Similarly, in the case of the highly accreting
1679: HAeBe star HR 5999, a point source could not be extracted for IRAC
1680: bands 1 and 2 so the corresponding fluxes were removed for the SED
1681: analysis of that object. This again hints to bright extended emission
1682: around 4.5 $\mu$m on a scale smaller than the large PSF of the star.
1683: 
1684: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%   Table   %%%%%%%%%%%%%%%%%%%%%%%%%%%
1685: \begin{table}[!h]
1686: \begin{center}
1687: \caption{Probable high-velocity outflows and nebulae in Lupus\label{outflows}}
1688: \begin{tabular}{lccccc}
1689: \tableline  \tableline
1690: Assoc.   & R.A.  &  Dec.   &  & & \\
1691: YSO ID   & (J2000)  &  (J2000)   & HH identification / driving source   & Ref. & Notes \\
1692: \tableline
1693: \multicolumn{6}{c}{\bf Lupus I} \\
1694:  2  & 15 38 48.36 &  -34 40 38.2   & IRAS 15356-3430              & 4 & a \\
1695:  10  & 15 43 01.29 &  -34 09 15.42  & HH 185 / IRAS 15398-3359     & 1 & b \\
1696: 15, 16  & 15 45 19.03 &  -34 17 32.43  & HH 187 / Sz 68 / Sz 69        & 1 & c \\
1697: \multicolumn{6}{c}{\bf Lupus III} \\
1698: 47  &  16 08 32.11 &  -39 03 18.23 & HH 228E / Sz 102            & 2 & d \\
1699: 47  &  16 08 27.18 &  -39 03 00.68 & HH 228W / Sz 102            & 2 & d \\
1700: 87  &  16 09 12.38 &  -39 05 00.61 & HH 78 / Lupus3 MMS       & 3 & e \\
1701: \ldots & 16 10 57.95  &  -38 04 37.90 & \ldots                     & 4 & f \\
1702: \multicolumn{6}{c}{\bf Lupus IV} \\
1703: \ldots & 16 00 39.04  &  -42 06 51.51 & \ldots                   & 4 & h \\
1704: \tableline
1705: \end{tabular}
1706: 
1707: \tablenotetext{}{a~~Bright 40 $''$ nebula towards the NE in all IRAC and MIPS bands.}
1708: \tablenotetext{}{b~~Bright 12 $''$ nebula towards the NE in all IRAC bands.}
1709: \tablenotetext{}{c~~Faint 3$''$ nebula knot at 3.6 and 4.5 $\mu$m.}
1710: \tablenotetext{}{d~~Faint knots at 3.6 and 4.5 $\mu$m.}
1711: \tablenotetext{}{e~~Faint knots mostly at 3.6 and 4.5 $\mu$m.}
1712: \tablenotetext{}{f~~8$''$ jet in all IRAC bands.}
1713: \tablenotetext{}{h~~9$''$ knots at all IRAC bands in a ring of 15 point sources.}
1714: 
1715: \tablerefs{1) \citet{Heyer1989}; 2) \citet{Krautter1986} ; 3) \citet{Reipurth1988}; 
1716:            4) This work}	
1717: \end{center}
1718: \end{table}
1719: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1720: 
1721: 
1722: 
1723: \newpage
1724: 
1725: \section{Clouds and Cluster properties}
1726: \label{clouds}
1727: 
1728: We can use the Spitzer data to also study the star formation history in the Lupus
1729: clouds. In particular, we want to estimate the star-formation activity and the 
1730: clustering properties of the young stellar population in relation to the current 
1731: cloud structure.
1732: 
1733: %The general properties of the clouds and stellar clusters found in the Spitzer
1734: %data are now studied with the help of the Spitzer-derived extinction maps
1735: %(\S~\ref{extinction}) and the spatial distribution of the YSOs 
1736: %(\S~\ref{spatial_distribution}). The Star Formation Rates for the different
1737: %regions are calculated and compared with the predictions
1738: %for different star-formation modes in \S~\ref{SFR}. Finally, in \S~\ref{SFE} the enclosed
1739: %mass and star formation efficiencies of the different regions are computed and 
1740: %discussed.
1741: 
1742: \subsection{Extinction maps in Lupus}
1743: \label{extinction}
1744: 
1745: The structure of the interstellar medium in the direction of the Lupus
1746: clouds has been extensively studied with (sub)millimeter molecular
1747: line observations \citep[e.g.,
1748: ][]{Murphy1986,Hara1999,Vilas-Boas2000,Tachihara2001} as well as with
1749: optical and infrared star-counts (\citealt{Cambresy1997},
1750: \citeyear{Cambresy1999}) on spatial scales of several
1751: arcminutes. These studies reveal a large clumpy structure with a
1752: substantial number of overdensities, dominated by the ``classical''
1753: Lupus I to IV clouds, where star-formation has also been detected. See
1754: Comer\'on 2008 (in prep.) for a complete review of these observations.
1755: 
1756: The c2d data offer an exceptional tool for producing extinction maps
1757: in all the imaged clouds. This is done by estimating the visual
1758: extinction towards each source classified as a background star, based
1759: on their SED from 1.25 $\mu$m to 24 $\mu$m. This provides multiple
1760: line-of-sight measures, which are then convolved with Gaussian beams
1761: of 90$''$ to 300$''$ to produce homogeneous extinction maps. These
1762: maps are part of the c2d data delivery. In the case of Lupus I, the
1763: low number of background stars only allowed reliable construction of
1764: an extinction map with a minimum FWHM of 120$''$. For more information
1765: about these maps, see \cite{Evans2007}. 
1766: 
1767: The maps with the largest beams (300$''$) are used to estimate the
1768: enclosed cloud masses (\S~\ref{SFE}) and to compare
1769: them with the position of the different YSOs (\S~\ref{spatial_distribution}).
1770: They can be seen in Figures \ref{spa_distr_I} to 
1771: \ref{spa_distr_IV}. The extinction traces the northern cloud
1772: in the Lupus I map with a peak of $A_V \sim$ 23 mag and shows two large
1773: clumps in Lupus III and IV, with peak extinctions of 33 and 36
1774: mag, respectively. The maximum in Lupus III coincides with the
1775: rich star-forming cluster, while it roughly coincides with the Flat
1776: SED source SSTc2d J160115.6-415235 in Lupus IV. This is the first
1777: report of such a high extinction inactive core in Lupus IV to our knowledge,
1778: which was previously unnoticed due to its current lack of activity.
1779: 
1780: \subsection{Spatial distribution and clustering of YSOs in Lupus}
1781: \label{spatial_distribution}
1782: 
1783: Figures \ref{spa_distr_I} to \ref{spa_distr_IV} show the spatial
1784: distribution of the PMS stars in the three Lupus clouds compared with
1785: the c2d extinction maps: In Lupus I, the most striking result is that the 
1786: majority of the YSOs
1787: fall in a ridge of high extinction which extends in the North of the
1788: cloud with NE-SW direction.  A southern high extinction region
1789: contains the pair of T Tauri stars Sz 68 and Sz 69. The majority of
1790: the objects, of any SED class but with larger abundance of Class I
1791: and Flat SED sources, appear very close to high
1792: extinction regions, with the exception of the Class III sources Sz 67
1793: and SSTc2d J153803.1-331358 and the Flat spectrum source IRAS
1794: 15398-3359. The close match between the YSO distribution
1795: and extinction map in the region and the early class of
1796: most of the objects suggests that we are seeing the objects in the
1797: places where they were just born.
1798: 
1799: The sources in Lupus III are dramatically concentrated in the dense
1800: star-forming cluster at RA $\sim$ 16$^h$09$^m$ and DEC $\sim$
1801: -39$^{\rm o}$10$'$, which contains the two bright intermediate mass Herbig
1802: Ae/Be stars HR 5999 and HR 6000 in the centre.  In this case, objects of
1803: all classes are found in the vicinity of the dense cluster and at large
1804: distances from it. Figure~\ref{spa_distr_III_zoom} zooms into the cluster, showing
1805: the extraordinary density of PMS objects at the highest extinction
1806: area in the whole cloud, where star formation is extremely
1807: active. Amongst many Class II and III sources, the cluster contains
1808: four Class I sources, including Lupus3MMS, claimed to be the only
1809: Class 0 object in the Lupus clouds \citep{Tachihara2007}. Note that
1810: only 3 of them appear in Figure~\ref{spa_distr_III_zoom}, while the
1811: forth is visible in Figure ~\ref{spa_distr_III} to the South of the
1812: cluster.
1813: 
1814: Figure~\ref{spa_distr_IV} shows the distribution of YSOs in Lupus IV
1815: around a very dense, yet quite unpopulated, extinction peak at RA
1816: $\sim$ 16$^h$02$m$ and DEC $\sim$ -41$^{\rm o}$50$'$ which contains
1817: the only Flat spectrum source in the cloud:
1818: SSTc2d J160115.6-415235. The rest of the Class II and III sources,
1819: together with the Class I source IRAS 15589-4134 are found surrounding
1820: the core, but following a different spatial pattern, which suggests that
1821: they might have been formed in a different structure. As the 
1822: contour plot nicely shows, the peak extinction at this clump is higher than that of
1823: the star-forming cluster in Lupus III (see also \S~\ref{extinction}).
1824: This fact together with the relative Class III predominance in Lupus IV 
1825: suggests that star-formation may have taken place before in other
1826: regions of this cloud and may be about to start in this core. We
1827: must note, however, that an object like Lupus3 MMS would have not
1828: been detected in this core, since it was only thanks to its millimeter detection
1829: that it was identified as a cloud member in Lupus III
1830: and such observations are so far not available for Lupus IV.
1831: 
1832: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_13 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1833: \begin{figure}[!ht]
1834: \plotone{fig19bm.ps}
1835: \caption{Spatial distribution of the spectrally confirmed PMS objects
1836:   (black symbols) and candidates (white symbols) in Lupus~I as
1837:   function of Lada Class, over-plotted on the contours from the c2d
1838:   extinction map (continuous lines).  The contour levels of extinction
1839:   are from 2~mag to 20~mag, in steps of 2~mag. The shaded areas, from
1840:   light to dark-gray, display the regions observed with MIPS, WFI and
1841:   IRAC, respectively.  The dashed lines outside the IRAC area are the
1842:   contour levels of extinction from \citet{Cambresy1999}, from 1~mag
1843:   to 6~mag in steps of 0.35~mag. The higher resolution and sensitivity
1844:   to higher $A_V$ of the c2d extinction map with respect to that of
1845:   \citet{Cambresy1999} can be appreciated.  
1846: \label{spa_distr_I}}
1847: \end{figure}
1848: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1849: 
1850: 
1851: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_14 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1852: \begin{figure}[!ht]
1853: \plotone{fig20bm.ps}
1854: \caption{Spatial distribution of the PMS objects (filled symbols) and 
1855:        candidates (open symbols) in Lupus~III as function of Lada
1856:        Class, over-plotted on the contours from the c2d extinction map
1857:        (continuous lines).  Symbols as in Figure~\ref{spa_distr_I}.
1858:        \label{spa_distr_III}}
1859: \end{figure}
1860: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1861: 
1862: 
1863: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_15 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1864: \begin{figure}[!ht]
1865: \plotone{fig21bm.ps}
1866: \caption{Zoom of the dense star-forming cluster in Lupus III. Symbols
1867:          as in Figure~\ref{spa_distr_I}, except the dashed lines, 
1868:          which are the 2 to 14 mag contours on the c2d extinction map in
1869:         steps of 2 magnitudes.
1870: 	\label{spa_distr_III_zoom}}
1871: \end{figure}
1872: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1873: 
1874: 
1875: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_16 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1876: \begin{figure}[!ht]
1877: \plotone{fig22bm.ps}
1878: \caption{Spatial distribution of the PMS objects (filled symbols) and 
1879:        candidates (open symbols) in Lupus~IV as function of Lada
1880:        Class, over-plotted on the contours from the c2d extinction map
1881:        (continuous lines).  Symbols as in Figure~\ref{spa_distr_I}.
1882:        \label{spa_distr_IV}}
1883: \end{figure}
1884: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1885: 
1886: The observation of these complex extinction structures and YSO
1887: distributions has stimulated for many years the discussion on how
1888: star-formation takes place, whether it is a hierarchical process or
1889: whether it takes place in a centrally-condensed way (see \citealt{Lada2003}, LL03
1890: hereafter, and references therein).  In order to study the subject with
1891: the new Spitzer data, we need to set some standards
1892: that will be applied across the c2d observational set. Similar to the
1893: other c2d clouds (see, e.g., \citealt{Alcala2008}) we identified
1894: substructure in the Lupus clouds using the nearest neighbor algorithm
1895: applied by \cite{Gutermuth2005} (see J{\o}rgensen et al. 2008, in prep. for
1896: details). Following the discussion by J{\o}rgensen et al. we divide
1897: concentrations of PMS objects up into ``clusters'' or ``groups''
1898: depending on whether they have more or less than 35 members at a given
1899: volume density level. We furthermore identify structures with volume
1900: densities higher than 1~$M_\odot$~pc$^{-3}$ (the criterion for a
1901: cluster by LL03) or 25~$M_\odot$~pc$^{-3}$ as being ``loose'' or
1902: ``tight'', respectively.
1903: 
1904: Figure \ref{vol_density} shows the result of the nearest neighbor
1905: analysis for the whole Lupus sample. When applied to Lupus this
1906: algorithm breaks the complex into three separate entities: two 
1907: tight concentrations of PMS
1908: stars at the 25~M$_\odot$~pc$^{-3}$ level and a loose one at the
1909: 1~M$_\odot$~pc$^{-3}$ level (taking into account the difference in
1910: distance between Lupus I and IV on one hand and Lupus III on the other).
1911: The Lupus IV cloud has one loose group and Lupus I a tight group.  The
1912: Lupus III cloud has one tight cluster containing 118 members. Table
1913: \ref{clustering_results} shows the number of PMS objects of the
1914: different classes in the regions identified above. This analysis
1915: allows ordering the objects by degree of clustering, with Lupus III
1916: being the most clustered region, followed by Lupus I and Lupus IV.
1917: 
1918: Overall, both the cloud density structure and the distribution of stars in the
1919: three regions suggest a centrally-condensed structure dominating
1920: the star-formation process only in Lupus III and a more disperse distribution
1921: of volume density enhancements in the other two clouds, mostly at
1922: Lupus IV, where star-forming cores have formed along a cloud filament
1923: in the NE-SW direction. As a whole,
1924: the three clouds clearly show a hierarchical structure with separate
1925: gas concentrations that evolve independently. This scenario argues
1926: in favor of a larger presence of turbulence over gravitational forces
1927: (LL03), which is consistent with the Lupus clouds being
1928: flanked by the large Scorpius-Centaurus OB associations 
1929: and therefore subject to their strong high-energy radiation 
1930: \citep{Comeron2007a}.
1931: 
1932: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fig_17 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1933: \begin{figure}
1934: \plotone{fig23bm.ps}
1935: \caption{Volume density plot for the three Lupus clouds as determined
1936:   with the nearest-neighbor algorithm by J. J{\o}rgensen et al. (2007,
1937:   in preparation). The derived contours are compared with extinction
1938:   map from \cite{Cambresy1999} in gray scale from 2 to 4.1 in steps
1939:   of 0.7 mag. The blue contour
1940:   corresponds to the 1$\times$LL03 level, the yellow to the
1941:   25$\times$LL03 level, and the green ones correspond to levels of
1942:   0.125, 0.25, 0.5, 2, 4 and 8 times that level. The two groups and
1943:   the cluster in Lupus III are identified with labels.\label{vol_density}}
1944: \end{figure}
1945: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1946: 
1947: \clearpage
1948: 
1949: \subsection{Star Formation Rates}
1950: \label{SFR}
1951: 
1952: %\subsection{The star formation in the Lupus dark clouds}
1953: %\label{star_formation}
1954: 
1955: %One of the goals of the c2d project was the compilation of consistent
1956: %and homogeneous star formation rates (SFR) across all five observed
1957: %star-forming regions. Using our complete samples of YSOs in the three
1958: %Lupus clouds, the cloud areas and the cloud masses estimated from the
1959: %extinction maps, the Star Formation Rates and the Star Formation
1960: %Efficiencies (SFE) of the Lupus clouds can be calculated.
1961: 
1962: We can also use our complete samples to estimate the star formation
1963: rates. In section \ref{YSOc_list} we computed the total numbers of YSOs and
1964: PMS objects for each of the three clouds. The IRAC+MIPS coverage areas
1965: reported in \S~\ref{IRAC_BDP} were used for the YSOs and the
1966: overlapping MIPS areas as given by \cite{Chapman2007} for the PMS
1967: stars. Using the typical average mass of 0.5~$M_\odot$ for consistency
1968: with the other c2d clouds, we find total masses in YSOs of $\sim$ 6,
1969: 34, and 6 $M_\odot$ in Lupus I, III and IV, respectively. A more
1970: appropriate average value for the stellar mass in Lupus of
1971: 0.2~$M_\odot$ (\S~\ref{star_luminosities}) would yield total masses 
1972: in YSOs of $\sim$ 2, 14, and 2
1973: $M_\odot$ in Lupus I, III and IV, respectively and therefore it would
1974: scale all the results down by $\sim$ 40 \% of the quoted values
1975: accordingly. These numbers could be slightly underestimated since
1976: very embedded objects in the extinction peaks could have not 
1977: been detected, as it was illustrated by our inability to select Lupus3 MMS
1978: based on Spitzer data only. However, given the relative small areas
1979: of the very embedded regions, the number of hidden Class 0
1980: of I objects should not be large compared to the total number of stars
1981: in the cloud.
1982: 
1983: Age estimates for the YSOs in Lupus suffer from many the uncertainties
1984: in assigning PMS tracks to positions on CMDs (e.g.,
1985: \citealt{Mayne2007}), including the uncertain distances to the Lupus
1986: clouds. \cite{Hughes1994} use 150 pc for all clouds and obtained an
1987: age range from less than 1 to 10 Myr with a peak at 3.2 Myr, however,
1988: \cite{Comeron2003} found ages between 1 and 1.5 Myr for objects in
1989: Lupus III using a distance of 200 pc (see also the discussion on the
1990: ages by \citealt{Comeron2007a}). Detailed 
1991: determinations of the individual stellar ages for the entired sample will be 
1992: performed once an optical spectroscopy survey of the sample is 
1993: completed. This section assumes the average
1994: value of 2 Myr for all the stars in Lupus. Table \ref{starform} shows
1995: the results of these calculations: star formation rates of 3, 17, and
1996: 3 M$_{\odot}$ Myr$^{-1}$ are found, or up to two times these values if
1997: we include the PMS stars and candidate PMS stars, in the
1998: Lupus I, III and IV clouds, respectively.
1999: 
2000: The numbers are consistent with those found in another low-star
2001: forming regions and assume that all stars were formed approximately
2002: at the same time. However, it is know that some of these young stellar populations
2003: might have appeared at different times, producing a spread in ages 
2004: \citep[e.g.][]{Hughes1994}. The clusters are clearly young
2005: so it is reasonable to assume that new-born stars should be close
2006: to the places were they were born. Indeed, early Class objects were
2007: found systematically close to the high extinction regions in Lupus I and many 
2008: Class II objects cluster around the two bright Herbig Ae/Be stars HR 5999 and HR 6000
2009: in the center of the Lupus III cluster. The formation of these latter objects could have 
2010: been triggered by the strong winds of the two intermediate-mass stars but we found
2011: no other evidence for triggered-star formation like clustered
2012: populations of objects with the same class in our mosaics. 
2013: The ages calculated for the PMS stars in the clouds
2014: are also in agreement with the expected cloud removal time-scales of $<$ 5 Myrs 
2015: after the onset of the star-formation process \citep{Ballesteros2007a}.
2016: 
2017: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%   Table   %%%%%%%%%%%%%%%%%%%%%%%%%%%
2018: \begin{center}
2019: \begin{table}[!h]
2020: \caption{Numbers, Densities and Star Formation Rates in Lupus \label{starform}}
2021: \begin{tabular}{lcccccccc}
2022: \tableline \tableline 
2023:  Type     & $\Omega$ & Area   & N   & N/$\Omega$ & N/Area & N/Vol & SFR & SFR/Area \\
2024:           & deg$^{-2}$& pc$^2$&     & deg$^{-2}$  & pc$^{-2}$ & pc$^{-3}$ & M$_{\odot}$ Myr$^{-1}$ & M$_{\odot}$ Myr$^{-1}$ pc$^{-2}$ \\
2025: \tableline
2026: \multicolumn{9}{c}{\bf Lupus I}\\
2027:  YSOs  &      1.39   & 9.49   & 13  &	9.4  & 1.4  & 0.43  & 3.3  & 0.35 \\
2028:  Total &      3.82   & 26.07  & 17  &	4.5  & 0.6  & 0.13  & 4.3  & 0.16 \\
2029: \multicolumn{9}{c}{\bf Lupus III}\\
2030:  YSOs  &      1.34   & 16.26  & 69  &	51.5 & 4.2  & 1.05  & 17.3 & 1.06 \\
2031:  Total &      2.88   & 34.94  & 124 &	43.0 & 3.5  & 0.60  & 31.0 & 0.89 \\
2032: \multicolumn{9}{c}{\bf Lupus IV}\\
2033:  YSOs  &      0.37   & 2.52   & 12  &	32.4 & 4.8  & 3.00  & 3.0  & 1.19 \\
2034:  Total &      1.08   & 7.37   & 18  &	16.7 & 2.4  & 0.90  & 4.5  & 0.61 \\
2035: \tableline
2036: \end{tabular}
2037: 
2038: \tablecomments{$\Omega$ is the solid angle subtended by the cloud, `Area'
2039:   is the equivalent in square parsec, `N' is the number of objects, and `SFR' stands
2040:   for Star Forming Rate.}
2041: 
2042: \end{table}
2043: \end{center}
2044: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2045: 
2046: 
2047: \subsection{Cloud Masses and Star Formation Efficiencies}
2048: \label{SFE}
2049: 
2050: Table \ref{cloud_mass} summarizes the cloud mass measurements
2051: from the literature as cited in the Comer\'on (2008, in
2052: prep.). To those, we add the c2d-derived masses of the clouds from the
2053: extinction maps with the lowest spatial resolution (300$''$) as
2054: explained in the Delivery Documentation \citep{Evans2007}. The table
2055: shows large differences in these mass estimates, most of which are due
2056: to the fact that the areas used to compute the total cloud masses
2057: differ from the c2d area, whereas both the number of YSOs and c2d cloud mass
2058: estimates were obtained from the same area. For this reason, we will
2059: use the c2d values (marked with an `e' in the table) in the
2060: following. 
2061: 
2062: %%%%%%%%%%%%%   Table   %%%%%%%%%%%%%%%%%%%%%%%%%%%
2063: \begin{table}[!h]
2064: \begin{center}
2065:   \caption{Estimates of the cloud masses and Star Formation
2066:     Efficiencies (SFE) in Lupus \label{cloud_mass}}
2067: \begin{tabular}{llcc}
2068: \tableline \tableline
2069: Tracer     & Mass$^\dagger$ &  \multicolumn{2}{c}{SFE (\%)} \\
2070:            & $M_{\odot}$    & YSOs & Total  \\
2071: \tableline
2072: \multicolumn{4}{c}{\bf Lupus I}\\
2073:       $^{13}$CO &   878$^a$ &   0.7 &   0.9 \\
2074:       C$^{18}$O &   326$^b$ &   1.8 &   2.4 \\
2075:      Extinction & 22851$^c$ &   0.0 &   0.0 \\
2076:      Extinction &   654$^d$ &   0.9 &   1.2 \\
2077:      Extinction &   479$^e$ &   1.2 &   1.6 \\
2078: \multicolumn{4}{c}{\bf Lupus III}\\
2079:       $^{13}$CO &  1195$^a$ &   2.8 &   4.9 \\
2080:      C$^{18}$O  &   418$^b$ &   7.6 &  12.9 \\
2081:      Extinction & 10547$^c$ &   0.3 &   0.6 \\
2082:      Extinction &  1666$^d$ &   2.0 &   3.6 \\
2083:      Extinction &   846$^e$ &   3.9 &   6.8 \\
2084: \multicolumn{4}{c}{\bf Lupus IV}\\
2085:      C$^{18}$O  &   216$^b$ &   2.7 &   4.0 \\
2086:      Extinction &  1406$^c$ &   0.4 &   0.6 \\
2087:      Extinction &   225$^d$ &   2.6 &   3.8 \\
2088:      Extinction &   196$^e$ &   3.0 &   4.4 \\
2089: 
2090: \tableline
2091: 
2092: \end{tabular}
2093: 
2094: 
2095: \tablenotetext{^\dagger}{All the masses have been normalized to the distances of 150 pc for Lupus I and IV \\ and 200 pc for Lupus III.}
2096: \tablenotetext{^a}{Total mass of H$_2$ derived from the $^{13}$CO luminosity \citep{Tachihara1996}.}
2097: \tablenotetext{^b}{Total mass of H$_2$ derived from the C$^{18}$O luminosity \citep{Hara1999}.}
2098: \tablenotetext{^c}{Cloud mass for A$_{\rm V}$ above 2 \citep{Cambresy1999}.}
2099: \tablenotetext{^d}{Mass in dense condensations as derived from extinction \citep{Andreazza1996}.}
2100: \tablenotetext{^e}{Cloud mass for Av $>$ 2 for Lupus III and IV and for Av $>$ 3 for Lupus I using the c2d extinction  \\ maps.}
2101: 
2102: \end{center}
2103: \end{table}
2104: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2105: 
2106: 
2107: \begin{table}[h]
2108: \caption{Star Formation efficiencies (SFE) and class numbers by region in Lupus. 
2109: \label{clustering_results}}
2110: \begin{tabular}{lllllllllll} \hline\hline
2111:                 &        &  Total &    I        &  Flat  &   II   &  III  &$\langle A_V \rangle$&  Mass  &  Volume  &  SFE   \\ \hline
2112: Total           &        &  159  &  8 (4.4\%)  &  12    &   75   &  64   &     $\ldots$        & 1501.4 & $\ldots$ &  5.0\% \\
2113: Extended        &        &   23  &  3 (9.1\%)  &   2    &   11   &  20   &     $\ldots$        & 386.1  & $\ldots$ &  2.8\% \\\hline
2114: I               &   LG   &    6  &  0 (0.0\%)  &   1    &    5   &   0   &        2.75         &  99.19 &  1.69    &  2.9\% \\        
2115: $\ldots$tight   &   TG   &    5  &  0 (0.0\%)  &   1    &    4   &   0   &        6.06         &  15.72 &  0.0326  & 13.7\% \\        
2116: III             &   LC   &  118  &  4 (3.4\%)  &   8    &   53   &  45   &        1.47         &  793.6 &  98.0    &  6.9\% \\        
2117: $\ldots$tight   &   TC   &   79  &  4 (5.1\%)  &   5    &   37   &  25   &        3.85         &  318.9 &  5.69    & 11.0\% \\        
2118: IV              &   LG   &   12  &  1 (8.3\%)  &   1    &    6   &   4   &        2.96         &  169.6 &  3.39    &  3.4\% \\ \hline 
2119: \end{tabular}
2120: 
2121: \tablecomments{ The SFEs and number of objects of per class are presented for different sub-samples according to their spatial distribution: the first two rows show the Total and Extended populations, `LG' stands for Loose Group, `TG' is Tight Group, `LC' is Loose Cluster, and `TC' is Tight Cluster, defined at \S~\ref{spatial_distribution} and shown in Fig. \ref{vol_density}. The $\langle A_V \rangle$, the enclosed masses and volumes were calculated for each of the regions following J{\o}rgensen et al.}
2122: 
2123: \end{table}
2124: 
2125: Based on these mass estimates and following \cite{Alcala2008}, we
2126: derived the Star Formation Efficiency (SFE) as
2127: $SFE=\frac{M_{star}}{M_{cloud}+M_{star}}$ where $M_{cloud}$ is the
2128: cloud mass and $M_{star}$ is the total mass in YSOs objects, i.e. the
2129: sum of the masses over the entire stellar populations, which range
2130: from 6 to 34 M$_\odot$ for the YSOs and from 8 to 62 M$_\odot$ if we
2131: include the PMS stars (c.f. \S~\ref{SFR}). The average stellar mass
2132: (0.5~$M_\odot$) was used here to derive $M_{star}$. The results on the
2133: efficiency of star formation are provided in
2134: Table~\ref{cloud_mass}. We estimate the SFE in the Lupus clouds to be
2135: 0.7 -- 2.4 \%, 2.0 -- 12.9 \% and 2.6 -- 4.4 \% for the three clouds,
2136: respectively. The large cloud masses from \cite{Cambresy1999} were
2137: derived for much larger areas and therefore were not considered. In
2138: spite of the large differences in total cloud mass estimates, these
2139: numbers are consistent the previous estimates in the literature for
2140: Lupus (0.4 -- 3.8\%, \citealt{Tachihara1996}) and somewhat higher than
2141: estimates for other T associations like e.g. Taurus (1--2\%,
2142: \citealt{Mizuno1999}).
2143: 
2144: We can also calculate the same numbers for the different
2145: sub-structures identified in the distribution of the PMS
2146: objects. Table \ref{clustering_results} gives the number of objects of
2147: the different classes in the different groups and clusters described
2148: in \S~\ref{spatial_distribution} along with the enclosed cloud masses
2149: calculated in the areas subtended by each of the groups and clusters 
2150: and the c2d extinction maps. The resulting SFE values for each
2151: region are also given in the table. All SFEs are larger for the 
2152: groups and clusters than for the extended population or the global cloud
2153: SFEs, as to be expected. The SFEs for the Loose Group (LG) and Tight Group (TG) in
2154: Lupus I (see \S~\ref{spatial_distribution}) are 2 and 8 times the
2155: cloud average SFE in Table \ref{cloud_mass}, while the SFE in the 
2156: Loose Group in Lupus IV almost matches the global SFE in that cloud. 
2157: The large numbers in the tight group and cluster in Lupus I and
2158: III are due to the relatively small areas containing a moderate
2159: numbers of forming stars and seems to suggest again that the star-formation
2160: process is more clustered in both clouds, with bigger differences
2161: in the SFEs between the clustered and extended population than in
2162: Lupus IV, where there is no substantial difference.
2163: 
2164: The caveat for all these SFE analyses is that the
2165: definition of the cloud area used for the cloud mass calculations 
2166: depends on the surface density of PMS stars and therefore assumes that
2167: the remaining cloud outside the chosen area is not related to the formation
2168: of the stars inside of it, which is probably correct, but also that a given amount
2169: of cloud gas density will always produce a given amount of stars, i.e., it 
2170: assumes that the star-formation process is universal and not affected by
2171: other characteristics of the cloud like its composition or its initial angular
2172: momentum. With this caveat in mind, it is however worth noticing that the 
2173: SFEs calculated for the bulk of the PMS populations in the three Lupus
2174: clouds (i.e. the loose cluster and groups) scale linearly with the enclosed 
2175: masses in the areas defining the loose clusters or groups. This correlation 
2176: between SFE and cloud mass is suggestive and was never found when a 
2177: cutoff in $A_V$ was used to define the cloud boundaries and masses. The 
2178: result is also interesting since it matches well the dynamical and fast 
2179: star-formation scenario recently proposed (LL03, \citealt{Bate2003}): it might help 
2180: explaining the presence of objects of very different evolutionary stages at small distances
2181: as the result of several mild star-formation events in small clouds before a big
2182: star-formation event contributes to the bulk of the coeval PMS stars in each of the 
2183: regions \citep{Ballesteros2007b}.
2184: 
2185: %\clearpage
2186: 
2187: \section{Summary}
2188: \label{summary}
2189: 
2190: We present observations of the Lupus I, III and IV dark clouds at 3.6,
2191: 4.5, 5.8 and 8.0 $\mu$m made with the Spitzer Space Telescope Infrared
2192: Array Camera (IRAC) and discuss them along with optical,
2193: near-infrared, observations made with the Multiband Imaging
2194: Photometer for Spitzer (MIPS) onboard Spitzer and millimeter flux from
2195: the literature to provide a complete description of the three clouds and
2196: their young stellar populations. The main results can be summarized
2197: as follows:
2198: 
2199: \begin{itemize}
2200: 
2201: 
2202: \item{We performed a census of the Young Stellar Objects (YSO)
2203:     applying the c2d Spitzer color criteria in the three clouds and
2204:     increase the number of cloud members by more than a factor of 4.
2205:     The PMS population consists of 159 stars in the three clouds
2206:   with infrared (IR) excess or spectroscopically determined
2207:   membership, mostly found in the high density regions of the clouds
2208:   and greatly dominated by low and very low-mass
2209:   objects. The sample is complete down to M $\approx$ 0.1 M$_\odot$ 
2210:   and probes well down into the sub-stellar regime.}
2211: 
2212: \item{30 -- 40\% of sources are multiple with binary separations between
2213:   0.7 and 10$''$ (100 to 2000 AU). These long period binaries 
2214:   appear not to affect the disk properties.}
2215: 
2216: \item{A large majority of the YSOs in Lupus are Class II or Class III
2217:   objects, with only 20 (12\%) of Class I or Flat spectrum
2218:   sources. Objects of all classes appear equally distributed in the
2219:   clouds and tend to cluster around the cloud high density peaks,
2220:   except in Lupus IV where they do not follow the extinction distribution.}
2221: 
2222: \item{The disk survey is complete down to ``debris''-like systems in stars as
2223:     small as M $\approx$ 0.2 M$_\odot$ and includes sub-stellar objects with
2224:     larger IR excesses. The disk fraction in Lupus is 70 -- 80\%, consistent
2225:     with an age of 1 -- 2 Myr. However, the young population contains 20\% optically
2226:     thick accretion disks and 40\% relatively less flared disks regarding their
2227:     Spitzer SEDs. }
2228: 
2229: \item{A larger variety of inner disk structures is found for larger
2230: inner disk clearings, suggesting several possible evolutionary
2231: paths for the primordial disks. Similar disk masses are
2232: found for a range of inner disk clearings, which provides evidence
2233: against a clearing of the inner disks by photoevaporation.}
2234: 
2235: \item{All previously known Herbig-Haro objects with sizes larger than 3$''$ were
2236: found in the Spitzer images of the clouds and two new sources are 
2237: reported: a jet-like structure in Lupus III and a nebulous object
2238: in Lupus IV.}
2239: 
2240: \item{Lupus I consist of a filamentary cloud structure with three density
2241: enhancements closely followed by early class objects.  Lupus III 
2242: contains a very active star-forming cluster with a very large number of
2243: objects of all classes. Lupus IV shows the highest extinction
2244: peak in Lupus with few late class objects away from the density peak.}
2245: 
2246: \item{Clustering analysis of the PMS distribution recovers separate
2247: structures in the three clouds, with Lupus III being the most centrally
2248: populated and rich, followed by Lupus I and Lupus IV. Overall, the
2249: cloud structures are compatible with predictions from the hierarchical 
2250: star formation scenario.}
2251: 
2252: \item{We estimate star formation efficiencies of a few percent and a star
2253: formation rate of 2 -- 10 M$_\odot$ Myr$^{-1}$ in the Lupus clouds. We 
2254: also find a tentative linear correlation between the star formation efficiencies and 
2255: the enclosed cloud masses of the three main stellar groups in Lupus.}
2256: 
2257: \end{itemize}
2258: 
2259: \input{tab9.tex}
2260: 
2261: \input{tab10.tex}
2262: 
2263: \input{tab11.tex}
2264: 
2265: \input{tab12.tex}
2266: \input{tab13.tex}
2267: \input{tab14.tex}
2268: 
2269: 
2270: \begin{acknowledgements}
2271:      
2272:   Support for this work, part of the Spitzer Space Telescope Legacy
2273:   Science Program, was provided by NASA through Contract Numbers
2274:   1256316, 1224608 and 1230780 issued by the Jet Propulsion
2275:   Laboratory, California Institute of Technology under NASA contract
2276:   1407. Astrochemistry at Leiden is supported by a NWO Spinoza and
2277:   NOVA grant, and by the European Research Training Network ``The
2278:   Origin of Planetary Systems'' (PLANETS, contract number
2279:   HPRN-CT-2002-00308). B. M. thanks the Fundaci\'on Ram\'on Areces for
2280:   early financial support. The authors would like to thank the referee
2281:   for very good suggestions on the structure and contents of the paper,
2282:   to E. Solano and R. Gutierrez from the Spanish Virtual Observatory
2283:   at LAEFF for providing easy automatic access to ancillary
2284:   Vizier data for the sample, and to Jennifer Hatchell for providing very detailed
2285:   comments.
2286:  
2287: \end{acknowledgements}
2288: 
2289: 
2290: %\bibliographystyle{aa}
2291: %\bibliography{biblio}
2292: \input{ms.bbl}
2293: 
2294: 
2295: \end{document}
2296: 
2297: %%
2298: %% End of file `sample.tex'.
2299: