0712.2975/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[manuscript]{aastex}
23: 
24: %% preprint2 produces a double-column, single-spaced document:
25: % \documentclass[preprint2,longabstract]{aastex}
26: 
27: \documentclass[preprint2]{aastex}
28: \usepackage{natbib}
29: %\usepackage{psfig}
30: %\usepackage{longtable}
31: \usepackage{lscape}
32: %\usepackage{deluxetable}
33: %\usepackage{amssymb}
34: \bibpunct{(}{)}{;}{a}{}{,}
35: %\usepackage{times}
36: %% Sometimes a paper's abstract is too long to fit on the
37: %% title page in preprint2 mode. When that is the case,
38: %% use the longabstract style option.
39: 
40: %% \documentclass[preprint2,longabstract]{aastex}
41: 
42: %% If you want to create your own macros, you can do so
43: %% using \newcommand. Your macros should appear before
44: %% the \begin{document} command.
45: %%
46: %% If you are submitting to a journal that translates manuscripts
47: %% into SGML, you need to follow certain guidelines when preparing
48: %% your macros. See the AASTeX v5.x Author Guide
49: %% for information.
50: 
51: \newcommand{\vdag}{(v)^\dagger}
52: \newcommand{\myemail}{skywalker@galaxy.far.far.away}
53: \def\Deg{${}^\circ$\llap{.}}
54: \def\Min{${}^{\prime}$\llap{.}}
55: \def\Sec{${}^{\prime\prime}$\llap{.}}
56: \def\timesec{${}^{s}$\llap{.}}
57: \def\deg{${}^\circ$}
58: \def\min{${}^{\prime}$}
59: \def\sec{${}^{\prime\prime}$}
60: %% You can insert a short comment on the title page using the command below.
61: %\slugcomment{Based on observations obtained with the NASA   
62: %{\it Hubble} Space Telescope 
63: %and   {\it Spitzer} Space Telescope , 2.2m@ESO and  Blanco\,4m@CTIO telescopes}
64: 
65: 
66: 
67: %% If you wish, you may supply running head information, although
68: %% this information may be modified by the editorial offices.
69: %% The left head contains a list of authors,
70: %% usually a maximum of three (otherwise use et al.).  The right
71: %% head is a modified title of up to roughly 44 characters.
72: %% Running heads will not print in the manuscript style.
73: 
74: \shorttitle{X-ray properties of protostars in Orion}
75: \shortauthors{Prisinzano et al.}
76: 
77: %% This is the end of the preamble.  Indicate the beginning of the
78: %% paper itself with \begin{document}.
79: 
80: \begin{document}
81: 
82: %% LaTeX will automatically break titles if they run longer than
83: %% one line. However, you may use \\ to force a line break if
84: %% you desire.
85: 
86: \title{X-ray properties of protostars in the  Orion Nebula}
87: %% Use \author, \affil, and the \and command to format
88: %% author and affiliation information.
89: %% Note that \email has replaced the old \authoremail command
90: %% from AASTeX v4.0. You can use \email to mark an email address
91: %% anywhere in the paper, not just in the front matter.
92: %% As in the title, use \\ to force line breaks.
93: 
94: \author{L. Prisinzano\altaffilmark{1}, 
95: 	G. Micela\altaffilmark{1},
96: 	E. Flaccomio\altaffilmark{1},
97: 	J. R. Stauffer\altaffilmark{2},
98: 	T. Megeath\altaffilmark{3},
99: 	L. Rebull\altaffilmark{2},
100: 	M. Robberto\altaffilmark{4},
101: 	K. Smith\altaffilmark{4},
102: 	E. D. Feigelson\altaffilmark{5},
103: 	N. Grosso\altaffilmark{6} and
104: 	S. Wolk\altaffilmark{7}}
105: 
106: %\email{loredana@astropa.inaf.it}
107: 
108: %\author{J. R. Stauffer\altaffilmark{2} and L. Rebull\altaffilmark{2}}
109: 
110: %\author{T. Megeath\altaffilmark{3}}
111: 
112: %\author{M. Robberto\altaffilmark{4} and K. Smith\altaffilmark{4}}
113: 	  
114: %\author{S. Wolk\altaffilmark{5}}
115: 	  
116: %\author{E. D. Feigelson\altaffilmark{6}}
117: 
118: %\and
119: 
120: %\author{N. Grosso\altaffilmark{7}}
121: 
122: \altaffiltext{1}{INAF - Osservatorio Astronomico di Palermo,
123:    		Piazza del Parlamento 1,  
124: 		I-90134 Palermo, Italy, loredana@astropa.inaf.it}
125: \altaffiltext{2}{Spitzer Science Center, California Institute of Technology, 
126: 	 1200 East  
127: 	 California Boulevard, Pasadena, CA 91125}
128: \altaffiltext{3}{Department of Physics, University of Toledo, 2801 W. 
129: 	Bancroft Ave, 
130: 	 Toledo OH 43606}
131: \altaffiltext{4}{Space Telescope Science Institute, 3700 San Martin Drive, 
132: Baltimore,  MD 21218}
133: \altaffiltext{5}{Department of Astronomy and Astrophysics, Penn State
134: University,  
135: 525	 Davey Lab, University Park, PA 16802, USA}
136: \altaffiltext{6}{Observatoire astronomique de Strasbourg, Universit\'e 
137: 	Louis-Pasteur, 
138: 	 CNRS, INSU, 11 rue de l'Universit\'e, 67000 Strasbourg, France}
139: \altaffiltext{7}{Harvard Smithsonian Center for Astrophysics, MS 65, 60 
140: 	Garden Street, 
141: 	  Cambridge, MA 02138}
142: 
143: 
144: %% Notice that each of these authors has alternate affiliations, which
145: %% are identified by the \altaffilmark after each name.  Specify alternate
146: %% affiliation information with \altaffiltext, with one command per each
147: %% affiliation.
148: 
149: %\altaffiltext{1}{Visiting Astronomer, Cerro Tololo Inter-American Observatory.
150: %CTIO is operated by AURA, Inc.\ under contract to the National Science
151: %Foundation.}
152: %\altaffiltext{2}{Society of Fellows, Harvard University.}
153: %\altaffiltext{3}{present address: Center for Astrophysics,
154: %    60 Garden Street, Cambridge, MA 02138}
155: %\altaffiltext{1}{Visiting Programmer, Space Telescope Science Institute}
156: %\altaffiltext{5}{Patron, Alonso's Bar and Grill}
157: 
158: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
159: %% style, abstract will output a Received/Accepted line after the
160: %% title and affiliation information. No date will appear since the author
161: %% does not have this information. The dates will be filled in by the
162: %% editorial office after submission.
163: 
164: \begin{abstract}
165: The origin and evolution of the X-ray emission in very young stellar objects (YSOs)
166: are not yet well understood because it is very hard to observe YSOs in the
167: protostellar phase.
168: We study the X-ray properties of Class\,0-I objects in the Orion Nebula
169:  Cluster (ONC) and compare them with those of the more evolved 
170:  Class\,II and III members. Using Chandra Orion Ultradeep Project (COUP) data,
171:    we study the X-ray properties of stars in different evolutionary 
172:    classes: 
173:   luminosities, hydrogen column densities N$_{\rm H}$, average 
174:   plasma temperatures and  time variability are  compared
175:      in order  to understand if the interaction between the circumstellar
176:   material and the central object can influence the X-ray
177:   emission. 
178: We have assembled the deepest and most complete photometric catalog 
179: of objects in the ONC region from the UV 
180: to 8\,$\mu$m using data from
181:  the HST Treasury Program, deep and almost simultaneous UBVI
182: %($I_{\rm lim}\sim 21.7$) 
183: and JHK 
184: %($K_{\rm lim}\sim 18.7$) 
185: images  taken, respectively,
186: with WFI@2.2m ESO and ISPI@4m CTIO telescopes, and   {\it  Spitzer}  IRAC imaging.
187: We select high probability candidate 
188: Class\,0-I
189: protostars, distinguishing between  those
190: having a spectral energy distribution  which rises from K up to 8\,$\mu$m
191: (Class\,0-Ia) from those  where the SED rises from K up to 4.5\,$\mu$m 
192: and decreasing afterwards (Class\,0-Ib).
193: In addition, we select a sample of ``bona fide"
194: Class\,II stars 
195:   and a set of Class\,III stars with IR emission
196: consistent with normal photospheres.
197:   
198: Our principal result is that  Class\,0-Ia objects are significantly less luminous 
199: in X-rays, both in the total and hard bands,
200:  than  
201:  the more evolved Class\,II stars with mass larger than 0.5\,M$_\odot$;
202:  these latter show X-ray luminosities similar to those of Class\,0-Ib stars.
203: This result supports the hypothesis that the onset of X-ray emission occurs at a very 
204: early stage of star formation and is in agreement with the result found in \citet{giar07}.
205:  Spectral properties of Class\,0-I stars are similar to those of the more
206:  evolved Class\,II and III objects, except for a larger absorption likely due to
207:  gas in the envelope or disk of  the protostellar objects. Our data
208:  suggest that the three different classes have similar X-ray temporal
209:  variability. 
210: \end{abstract}
211: 
212: %% Keywords should appear after the \end{abstract} command. The uncommented
213: %% example has been keyed in ApJ style. See the instructions to authors
214: %% for the journal to which you are submitting your paper to determine
215: %% what keyword punctuation is appropriate.
216: 
217: \keywords{open clusters: general --- open clusters: individual (Orion Nebula
218: Cluster) --- X-rays: stars --- stars: pre-main sequence --- stars: activity}
219: 
220: %% From the front matter, we move on to the body of the paper.
221: %% In the first two sections, notice the use of the natbib \citep
222: %% and \citet commands to identify citations.  The citations are
223: %% tied to the reference list via symbolic KEYs. The KEY corresponds
224: %% to the KEY in the \bibitem in the reference list below. We have
225: %% chosen the first three characters of the first author's name plus
226: %% the last two numeral of the year of publication as our KEY for
227: %% each reference.
228: 
229: 
230: %% Authors who wish to have the most important objects in their paper
231: %% linked in the electronic edition to a data center may do so by tagging
232: %% their objects with \objectname{} or \object{}.  Each macro takes the
233: %% object name as its required argument. The optional, square-bracket 
234: %% argument should be used in cases where the data center identification
235: %% differs from what is to be printed in the paper.  The text appearing 
236: %% in curly braces is what will appear in print in the published paper. 
237: %% If the object name is recognized by the data centers, it will be linked
238: %% in the electronic edition to the object data available at the data centers  
239: %%
240: %% Note that for sources with brackets in their names, e.g. [WEG2004] 14h-090,
241: %% the brackets must be escaped with backslashes when used in the first
242: %% square-bracket argument, for instance, \object[\[WEG2004\] 14h-090]{90}).
243: %%  Otherwise, LaTeX will issue an error. 
244: 
245: \section{INTRODUCTION}
246:  Young stellar objects (YSOs)  exhibit an X-ray luminosity
247: significantly higher than similar stars in the main sequence phase. However,
248: it is still unclear at what stage of their formation process YSOs begin to emit
249: X-rays  and how this X-ray emission evolves 
250: with time \citep{fava03,feig07}. 
251: 
252: Evidence has accumulated over the past few years that among the 
253: YSOs, the X-ray luminosity
254: function (XLF) of Classical T Tauri stars (CTTS) is different from that of 
255: Weak T Tauri stars (WTTS).
256: Such a difference has been investigated by some 
257: because
258: the accretion and outflow processes, characteristic of the CTTS  
259: phase, involve the interaction of
260: ionized material with the star and disk magnetic field, and
261: these media play a 
262: fundamental role in the coronal X-ray emission of young stars.  
263:   Using the {\it Chandra} Orion Ultradeep Project (COUP)  data
264:  \citep{getm05a}, \citet{prei05b} find that the accreting stars,
265:  as discriminated using the Ca\,{\footnotesize II} 8542\AA\ IR triplet line,
266:   are less  X-ray luminous than
267:  non-accretors, while no differences are found for stars with $(K-L)$~
268:  IR excesses,  indicative of 
269:  the presence of a (possibly passive) disk,
270:  with respect to stars without IR excesses. 
271:  These results are in agreement with those found  in the previous work 
272:  in the  ONC by \citet{flac03b,flac03c} 
273:  who found a strong difference in the X-ray
274:  luminosity of accreting and non-accreting stars.
275: % However, the results of such comparisons strongly depend on the star samples 
276: % used in the analysis that can be affected by different selection criteria if, 
277: % for example,
278: % WTTS are selected by their X-ray detection, while CTTS are selected by the
279: % emission of their H$\alpha$ line. 
280:  
281:   In order to understand the evolution of the X-ray emission during the star
282:   formation process, \citet{prei05a} looked
283:  for a possible relation between X-ray emission and stellar age
284:   in the ONC. They found, in the  age range $\sim0.1-10$\,Myr, 
285:   a slow decay in the X-ray luminosity, L$_{\rm X}$. 
286: Because, as the star contracts, 
287:   the bolometric luminosity, L$_{\rm bol}$,
288: decreases faster than L$_{\rm X}$, 
289:   on average L$_{\rm X}$/L$_{\rm bol}$ and the X-ray surface flux, F$_{\rm X}$,
290:    increase during this age range.
291: 
292:   In this work, we want to investigate the onset and the evolution 
293:   of X-ray emission during the early phase of the star formation by
294:   statistically comparing the X-ray properties of the stars in different evolutionary phases
295:   rather than considering 
296: individual stellar ages, usually derived from theoretical
297: isochrones. We hope in this way to avoid some of the problems introduced
298: when using inferred ages (due to their often large random and
299: systematic uncertainties; see, e.g., \citeauthor{hill07} 2007).
300: %
301: 
302:  YSOs in their early phases  are characterized
303:  by IR excesses due to  the presence of a circumstellar envelope and/or a disk
304:  that contains warm dust. The first IR classification in Class\,I, II and III 
305:  sources, which is based on the IR spectral index, was proposed by 
306:  \citet{lada87}.  The discovery in the sub-mm of cold objects, younger than the
307:  IR sources, leads to the introduction of the Class\,0 sources \citep{andr93}. 
308:  Based on their mid-IR properties, YSOs
309: are usually classified as follows 
310:  \citep[e.g.][]{alle04,mege04,lada06,alle07}: 
311:  Class\,0 objects 
312: are protostars surrounded by the initial 
313:  infalling envelope, typically detected in the
314:  sub-millimeter and far-IR; Class\,I objects are evolved protostars
315:  surrounded  by an envelope and a circumstellar
316:  disk; Class\,II stars are
317:  objects with a circumstellar disk but lacking significant
318: evidence for a dense envelope and finally
319:  Class\,III stars
320:  are  objects that have lost their initial envelope and disk and show
321:  optical and near-IR colors
322:  consistent with normal photospheres.  
323:  %However, in the last few years,
324:  %with the advent of the {\it Spitzer} data,  has it been 
325:  %possible to  classify many YSOs in their different
326:  %evolutionary stages based on their IR properties for entire star
327:  %forming regions.
328:  By providing very sensitive 3.6-70\,$\mu$m data with good angular
329: resolution, {\it Spitzer} observations have provided
330: a much better census of the stellar populations in a number
331: of star-forming regions.  However, {\it Spitzer} data have also
332: complicated the YSO classification system, allowing detection
333: of some Class 0 objects at wavelengths as blue as 3.6\,$\mu$m.
334: 
335:  Our intent is  to compare the
336:  X-ray properties of protostars - where to date there are relatively
337:  few studies - with the X-ray properties of the more evolved
338:  Class\,II and III stars, for which the
339:  X-ray emission is well studied.  Such a comparison is fundamental in order 
340:  to understand the onset of 
341:  X-ray emission in YSOs and its evolution in the early phases of
342:  star formation.
343:  
344: Previous studies of X-ray emission from Class 0-I objects include \citet{ozaw05},
345:  who studied core F of the $\rho$\,Ophiuchi cloud and found
346: a high X-ray detection rate (64\%) of Class\,I sources. They  suggest that an
347: evolutionary effect in temperature and extinction 
348:  is present from Class\,I to Class\,III stars, these latter
349: having lower coronal temperatures and extinction. 
350: 
351: A similar result has also been found by \citet{giar07}. 
352: \citet{getm07} studied the IC\,1396N cometary globule where they found 6 X-ray
353: emitting protostars, including the Class\,0-I IRAS 21391+5802 which they claim to
354: be one of the youngest objects ever detected in X-rays. 
355: The comparison of X-ray emission from different evolutionary classes is, 
356: however, hampered by the difficulty of detecting X-rays from Class\,0 protostars.
357: For example, no X-ray emission has been found from Class\,0 stars
358: in the Serpens, one of the most sensitive surveys \citep{giar07}, even if
359: important detections have been found in other surveys
360: \citep[e.g.][]{tsub01,hama05,tsuj05,getm07}.
361: 
362: %A list of Class\,0 and Class\,I single 
363: %protostars in a variety of star
364: %forming regions observed with the Chandra and/or XMM Observatory, is given in
365: %Table\,5 of \citet{getm07} where the protostars detected in X-rays are indicated.
366: 
367: The ONC is a prime target for this kind of investigation. It is the nearest
368: concentrated site of star formation with a very large number of known members
369: covering all evolutionary stages.
370: It is the best studied region at all wavelengths, in particular, in the X-ray
371: band.
372:  Many interesting results have been obtained with the extraordinarily long 
373: {\it Chandra} observation performed by the COUP project. 
374:  To understand the evolution of X-ray activity from 
375: the very first stages of the star formation process,  it is necessary to
376: distinguish Class\,0-I objects from Class\,II stars. Such a classification,
377: based on the shape of the spectral energy distribution (SED) is not
378: available in the literature and 
379: requires complete and deep near and mid IR data.
380: 
381: We present here the results of the X-ray study of YSOs
382:  in the ONC classified using the {\it Spitzer} IRAC catalog \citep{mege05},
383:   groundbased JHK data (Robberto et al. 2007, in preparation)
384:  and the deep photometric catalog obtained within 
385: the HST Treasury Program \citep{robb05b}. 
386: Using the COUP data, we have studied the X-ray properties of YSOs
387: from their initial protostellar phase
388: (Class\,0-I objects) to the more evolved pre-main sequence (PMS) stage 
389: of no longer accreting and disk free stars (Class\,III objects). 
390: We have classified the ONC members using conservative criteria in 
391: order to minimize the contamination of the resulting samples.
392: Because of the large area coverage and the richness of the ONC, 
393: we have obtained an homogeneous and statistically significant sample
394:  of X-ray detected Class\,0-I 
395: objects. 
396: We are thus able, for the first time, to compute the X-ray luminosity function 
397: (XLF) of Class\,0-I stars and to compare it with the XLF of 
398: Class\,II and III stars.
399: 
400: 
401: 
402: \section{OBSERVATIONS}
403:   We have assembled a deep  photometric catalog
404:    of the ONC  from the UV
405: to  8\,$\mu$m obtained using new optical/NIR observations taken within 
406: the HST Treasury Program \citep{robb05b}, using a variety of telescopes and
407: detectors, and {\it Spitzer} IRAC \citep{wern04,fazi04} imaging.
408: The  optical photometry was derived  
409: using images taken with the  Wide-Field Planetary Camera 2
410: (WFPC2) of HST  \citep{robb05b}
411: and with the  Wide Field Imager (WFI) at the ESO 2.2\,m
412: Telescope (Robberto et al. 2007, in preparation).
413:  $JHK$ photometry was obtained from images taken    
414: with the Infrared Side Port Imager (ISPI) camera at the CTIO Blanco\,4m
415: telescope (Robberto et al. 2007, in preparation).
416: {\it Spitzer} IRAC photometry in the four standard filters centered on 3.6, 4.5, 5.8 and 
417: 8.0\,$\mu$m was obtained by \citet{mege05}.
418: A detailed description of the optical-IR observations is 
419: given in Table\,\ref{observations}. We note that the magnitude limits given in
420: this table are typical values for each survey. However,
421: due to the complex structure of the nebulosity in the ONC (and due to
422: varying integration times for different parts of the region), the 
423: sensitivity is not uniform. 
424: For example the limiting magnitude at 3.6\,$\mu$m is about 16 within
425: most of the COUP
426: FOV, while it is about 10 within a region of 1\min\ in radius around the
427: brightest stars of the Trapezium cluster.
428: 
429: %\clearpage
430: %\thispagestyle{empty}
431: %\setlength{\voffset}{30mm}
432: {\rotate
433: \tabcolsep 0.15truecm
434: \begin{table*}
435: \caption{Log of the new optical-IR observations used in this work.}
436: \label{observations}
437: \centering
438: \begin{tabular}{cccccccc}
439: \hline\hline
440: Nights&Telescope&Camera&Filters&Exp. time\tablenotemark{a}&Mag. Lim.\tablenotemark{b}& Field of & Reference\\
441:       & 	&      &       &	 [sec]         &	 &  view    & \\
442: \hline
443: 2006 May      & HST	   &WFPC2& F336W\tablenotemark{c}&2$\times$400  &F336W$\simeq$20.0&30\min$\times$ 30\min& 1\\ 
444: 2006 May      & HST	   &WFPC2& F439W\tablenotemark{c}&80  &F439$\simeq$20.7 &30\min$\times$ 30\min& 1\\ 
445: 2006 May      & HST	   &WFPC2& F656N\tablenotemark{c}&400  &F656N$\simeq$18.0&30\min$\times$ 30\min& 1\\ 
446: 2006 May      & HST	   &WFPC2& F814W\tablenotemark{c}&10  &F814$\simeq$20.0 &30\min$\times$ 30\min& 1\\ 
447: 2005 Jan. 1-2 &ESO\,2.2\,m&  WFI   &U, B, V, I$_c$&3, 30, 280&I$_c\simeq$21.7&35\min$\times$ 34\min&2   \\
448: %2005 Jan. 1-2 &ESO\,2.2\,m&  WFI   &U, B, V, I$_c$&30	 &	  5	 &35\min$\times$ 34\min&2  \\
449: %2005 Jan. 1-2 &ESO\,2.2\,m&  WFI   &U, B, V, I$_c$&3	   &	  1	 &35\min$\times$ 34\min&2  \\
450: 2005 Jan. 1   &CTIO\,Blanco\,4m&ISPI&J, H, K$_s$  &3, 30  &K$_s\simeq$18.7&6$\times$(10\Min25$\times$10\Min25)&2 \\
451: 2005 Jan. 2   &CTIO\,Blanco\,4m&ISPI&J, H, K$_s$  &3, 30  &K$_s\simeq$18.7&8$\times$(10\Min25$\times$10\Min25)&2 \\
452:   & {\it Spitzer}\tablenotemark{d}       &IRAC & (3.6, 4.5, 5.8, 8.0)\,$\mu$m& &[3.6]$\simeq$16.8&5.6 sq. degrees&3  \\
453: \hline
454: \end{tabular}
455: \renewcommand{\footnoterule}{}  % to avoid a line before footnotes
456: \tablenotetext{a}{Exposure time per filter}
457: \tablenotetext{b}{Magnitude limit computed for stars with photometric errors smaller than 0.1}
458: \tablenotetext{c}{WFPC2 filters F336W, F439W, F656N, F814W corresponds to the standard filters U, B, H$\alpha$
459: and I, respectively}
460: \tablenotetext{d} {Data part of the {\it Spitzer}/GTO Programs 43 and 50 
461: taken in high dynamic range mode using 0.6 and 12 sec at each position.}
462: \tablerefs{(1) \citet{robb05b}; (2) Robberto et al. (2007) in preparation; (3) \citet{mege05}}
463: \end{table*}
464: }
465: %\clearpage
466: \setlength{\voffset}{0mm}
467: 
468: \begin{figure} 
469: \centerline{\includegraphics[width=8cm]{f1.ps}}
470: \caption{The  field of view (FOV) of the catalogs considered in this work
471:  are overplotted  on the
472:  spatial distribution of the
473: stars detected with IRAC. Note that the FOV 
474: of the IRAC (dashed line), \citet{rebu00} and 2MASS data 
475: (not indicated here) are larger than the field shown in this figure.}
476: \label{f01}
477: \end{figure}
478: %\clearpage
479: 
480: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
481:  \subsection{X-ray data}
482: X-ray data used in this work were obtained with
483: the $\sim$838\,ks COUP observation of the ONC, taken with the ACIS camera that
484: covers a FOV of 17\min $\times$17\min. Details of data reduction and the
485: derived catalog of 1616 sources can be found
486: in \citet{getm05a}.
487: \subsection{Optical data}
488: New optical photometric data taken with the WFPC2 camera of HST
489: were used  by \citet{robb05b} to derive the optical photometric catalog 
490: used in this work. This catalog contains 1754 
491: selected point sources with at least a
492: detection in the F814 band.
493: 
494: The recent HST survey has been complemented with UBVI images taken 
495: with the WFI camera of the ESO\,2.2\,m Telescope. Details on data reduction and
496: photometric catalog derivation will be described in Robberto et al. (2007, in
497: preparation). This catalog contains 2744 selected point sources with at least a
498: detection in the I band.
499: 
500: Additional optical data were retrieved from literature including the catalogs of 
501: \citet[][ hereafter H97]{hill97}, \citet{rebu00} and \citet{herb02}.
502: \subsection{$JHK$ data}
503: The $JHK$ photometric catalog has been obtained using images
504:  taken with the ISPI camera at the CTIO Blanco\,4\,m
505: Telescope. Details on data reduction and photometry will be given in 
506: Robberto et al. (2007, in preparation). 
507: The survey includes a mosaic of 11  fields, each of
508: 10\Min25$\times$10\Min25, centered 
509: around the Trapezium cluster.
510: % and contains a total of 47\,612 detections,
511: % including multiple items related to stars observed more than once. 
512: All fields were observed once using a 5 point dithering pattern. 
513: %except the three central fields that were observed twice. 
514: %A total of   47\,612 point sources were detected, including multiple detections 
515: %of the same object. 
516: From the source catalog
517: we excluded  saturated objects, that is, those with a peak flux in each band
518:  higher than 37\,000 counts and entries with errors in magnitudes larger than 
519:   0.2. After this selection, we 
520:  were left with 17\,824 entries in the table.
521: 
522: Finally, we constructed a consolidated catalog of individual objects by 
523: merging entries with positions within 0\Sec6, the radius at which
524: spurious identifications start to dominate.
525: For each object, we computed the mean coordinates 
526: and magnitudes. 
527:  Our final $JHK$ photometric catalog contains  13\,987 objects.
528: % but the objects with an error smaller than 0.1 mag in at
529: %least one of the three JHK magnitudes are 6523.
530: 
531: Additional $JHK$ data were retrieved from the  2MASS catalog
532: \citep{cutr03}. 
533: 
534: \subsection{{\it Spitzer} IRAC data}
535: A description of the data, the reduction procedure and the derived 
536: photometric catalog obtained from the {\it Spitzer} IRAC images is given in
537: \citet{mege05}.  The survey covers the Orion A and B molecular clouds for a
538: total of 5.6 sq. degrees with a sensitivity of 17.0, 16.0, 14.5 and 13.5 mag in
539: the 3.6\,$\mu$m, 4.5\,$\mu$m, 5.8\,$\mu$m and 8.0\,$\mu$m  bands, respectively.
540: The uncertainties in the calibration are about 5\%, but for a few stars are
541: significantly worse due to the bright nebulosity in 
542:  the 5.8 and 8.0\,$\mu$m bands. 
543:  Within the region  studied in this work, i.e. the COUP FOV,
544: the catalog includes 
545: 1244 sources 
546: %IDL> RESTORE,'../SAVEFILES/catalog.save'
547: %IDL> RESTORE,'../SAVEFILES/coup_fov.save'
548: %IDL> cat=catalog[coup_fov]
549: %IDL> tmp=where(FINITE(cat.MA36_SP) AND FINITE(cat.MA45_SP))
550: %IDL> help,tmp
551: %TMP		 LONG	   = Array[1244]
552:  detected both at 
553: 3.6\,$\mu$m, 4.5\,$\mu$m and 334 
554: %IDL> tmp=where(FINITE(cat.MA58_SP) AND FINITE(cat.MA80_SP))
555: %IDL> help,tmp
556: %TMP		 LONG	   = Array[334]
557: sources detected both at 
558:  5.8\,$\mu$m, 8.0\,$\mu$m. 
559:  The {\it Spitzer} Orion Survey also included  MIPS
560:  observations at 24\,$\mu$m that we do not use in the following analysis 
561:  since they are saturated in the COUP region due to the high
562:  background from the HII region.
563: 
564: \subsection{Cross identifications of sources}
565: All the photometric catalogs obtained from the observations
566: described in Table\,\ref{observations} were assembled in a single
567: database.
568: % containing all the objects detected at least in one of
569: %the considered catalogs. 
570: Common objects between two catalogs
571: were found  using the nearest neighbour method by first considering
572: two sources as the same object if their positions agree to within
573: a specified amount -
574: 1\sec-3\sec -  depending on the
575: absolute astrometric accuracy of the catalogs.
576: Matched sources  were then used to compute the median and the
577: standard deviation of the offsets.
578: These median values were used to remove
579: the systematic offsets between each
580: pair of matched catalogs,
581: and new matches were performed, for all pairs of catalogs, using new tolerances  
582: computed using the standard deviation of the offsets. 
583: 
584: Using the same matching procedure we have cross-correlated 
585: these data with the  
586: literature catalogs of \citet{hill97},
587:  \citet{rebu00}, 
588:  2MASS \citep{cutr03}, \citet{herb02}, \citet{muen02} and
589:  the X-ray source list from the COUP Survey \citep{getm05a}.
590: 
591: Finally we collected
592: all data from all the cross-matched catalogs, including all objects detected in
593: at least one catalog.
594:   
595: The  field of view (FOV) covered by these  surveys
596: are overplotted in Fig.\,\ref{f01} on the
597:  spatial distribution of the
598: stars detected with IRAC around the ONC. 
599: Note that the FOV of the IRAC, \citet{rebu00} and
600: 2MASS data (not indicated)
601:  are larger than the field shown in the figure.
602: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
603: %\subsection {Sample selection}
604: 
605: Since this work is focused on the X-ray properties of
606: the young stars in the ONC, only stars within the COUP FOV will be considered 
607: in the following  analysis. In addition, 
608: photometric data   with errors smaller 0.2 mag will be considered.
609: %% In a manner similar to \objectname authors can provide links to dataset
610: %% hosted at participating data centers via the \dataset{} command.  The
611: %% second curly bracket argument is printed in the text while the first
612: %% parentheses argument serves as the valid data set identifier.  Large
613: %% lists of data set are best provided in a table (see Table 3 for an example).
614: %% Valid data set identifiers should be obtained from the data center that
615: %% is currently hosting the data.
616: %%
617: %% Note that AASTeX interprets everything between the curly braces in the 
618: %% macro as regular text, so any special characters, e.g. "#" or "_," must be 
619: %% preceded by a backslash. Otherwise, you will get a LaTeX error when you 
620: %% compile your manuscript.  Special characters do not 
621: %% need to be escaped in the optional, square-bracket argument.
622: 
623: 
624: 
625: 
626: 
627: %% The displaymath environment will produce the same sort of equation as
628: %% the equation environment, except that the equation will not be numbered
629: %% by LaTeX.
630: 
631: 
632: %% If you wish to include an acknowledgments section in your paper,
633: %% separate it off from the body of the text using the \acknowledgments
634: %% command.
635: 
636: %% Included in this acknowledgments section are examples of the
637: %% AASTeX hypertext markup commands. Use \url without the optional [HREF]
638: %% argument when you want to print the url directly in the text. Otherwise,
639: %% use either \url or \anchor, with the HREF as the first argument and the
640: %% text to be printed in the second.
641: \section{YSO CLASSIFICATION \label{classification}}
642: In this Section we describe how we select ONC stars belonging to the three
643: different evolutionary classes 0-I, II and III. Our approach is to select
644: conservative, even if incomplete, samples in order to be able to
645: compare the X-ray properties
646: of the stars in different evolutionary stages using samples
647: that are as ``uncontaminated" as possible.
648:  
649: %Using these data we are able to select stars of different evolutionary classes in the ONC
650: %and compare their fundamental properties. 
651: %To this aim, we use a selection criterion as conservative as possible, even if incomplete,
652: %in order to avoid the contamination of highly reddened different class objects
653:  %in the selected samples. This approach leaves a number of unclassified
654:  %objects but  allows us to compare the X-ray properties
655: %of the stars in different evolutionary stages from samples "uncontaminated"
656: % as possible.
657: 
658: \subsection{Class\,0-I protostars}
659: According to \citet{lada84}  and \citet{adam87}, the first stage of YSO 
660: evolution is the infall  phase around a central protostar undergoing
661: accretion, characterized by
662: strong emission in the sub-millimeter and far-infrared and weak emission
663: shortward of 24\,$\mu$m. These objects, indicated as Class\,0, 
664: can have outflows and are defined as 
665:  protostars with
666: half or more of their mass in the envelope. More evolved protostars are the
667: Class\,I objects
668: with both a disk and envelope surrounding the central star.
669: Both these classes of protostars are characterized by
670:  a steeply rising SED in the infrared range,
671:  although pure Class\,0 objects are expected to be   
672:   fainter at near and mid-IR bands \citep{alle07}.
673: 
674: Since our data extend only to 8\,$\mu$m,
675:  we cannot distinguish between Class\,0 and Class\,I stars and we will
676:  combine them into one group as Class\,0-I protostars;
677: we classify  them using the following criterion:
678: we computed ${\rm log}(\lambda F_\lambda)$ 
679: over the wavelengths
680: $[\lambda_1,...\lambda_5]=[2.2,3.6,4.5,5.8,8.0]\,\mu$m,
681:  i.e. the K-band and the four IRAC bands with
682: the \citet{lada06} flux conversion factors that are in agreement with 
683: the \citet{reac05} values; 
684: for each star we then computed the
685: slopes for each pair of adjacent bands
686: %$$\frac{\Delta{\rm log}(\lambda F_{\lambda})}{\Delta {\rm log}\lambda}$$
687: $$\frac{{\rm log}(\lambda_{i+1} F_{\lambda_{i+1}})-
688: 	{\rm log}(\lambda_{i}   F_{\lambda_{i}})}
689: 	{{\rm log}\lambda_{i+1}-{\rm log}\lambda_{i}}~~~ {\rm for}~~i=1,4;$$
690:  we consider Class\,0-I objects those for which the slopes are all
691: larger than 0, i.e. objects with rising SEDs. This 
692:  criterion
693: is more conservative
694: than using the single slope (usually indicated as spectral index)
695: computed from the linear fit of the 
696: ${\rm log}(\lambda F_\lambda)$ function over the whole wavelength range,
697: because it only selects monotonically increasing 
698: SEDs.
699: Moreover,  it is more robust
700: than those based on only a
701: color-color diagram (CCD)  since it is based
702: simultaneously on all the
703:  known magnitudes. 
704: %in addition, it does not depend on the limits adopted to distinguish different class objects
705: % in the CCD \citep{alle04,wins07}.
706: % Therefore our Class\,0-I sample are subsamples of those selected with other methods
707: % (\citet[e.g.][]{alle04,mege04,hart05}).
708: 
709: Our sample of Class\,0-I objects includes all the objects for which we have at least
710: three of the five magnitudes and therefore at least two
711: slopes. This allows us to include objects detected only in the K, 3.6\,$\mu$m 
712: and 4.5\,$\mu$m
713: bands but not at 5.8 \& 8.0\,$\mu$m; because of the bright nebular background 
714: and the lower sensitivity 
715:  at 8\,$\mu$m,
716:  many objects
717: are not detected in these two bands (roughly 44\% %(18/41 see class0I_def_1.pro)
718: of the Class 0-I sample).
719: 
720:  Using this criterion, we 
721:  end up with 41  potential Class 0-I objects in the COUP field of view
722:  \citep{getm05a}.
723: %restore,'../SAVEFILES/protostar_1.save'
724: %help,protostar_1
725: %PROTOSTAR_1     LONG      = Array[41]
726:   Eighteen
727: %IDL> help,CL0I_COMPLEX_1
728: %CL0I_COMPLEX_1  LONG      = Array[18]
729:  of them have a detection
730:   in at least one of the optical 
731:  UBVI bands. Since protostars are expected
732:   to be heavily embedded and therefore highly reddened objects,
733:   we do not expect to detect them 
734:    in the visible.
735:   Of these 18  peculiar objects,
736: 5 (COUP \# 693,     747,     826,     948,    1011) % see kastner05.pro
737: are in the list of the X-ray detected proplyds given in
738: \citet{kast05}. From a visual inspection of ACS/HST images, we conclude that
739: all 18 of these objects have  a counterpart in the HST optical
740: bandpasses, while the other 23 objects are not detected at these bands.
741: We eliminate these 18 objects from our list of candidate Class\,0-I stars
742:  and we consider  
743: the remaining 23 stars  as Class\,0-Ia objects to distinguish them from the
744: additional sample  of candidate Class\,0-I described below.
745:  The SEDs of the 23 selected Class\,0-Ia stars are plotted  in 
746: Fig.\,\ref{f02}.  The photometry is given in 
747: Table\,\ref{cl0I_1} where Column\,1 gives the sequential number in our catalog,
748: Column\,2 is the COUP identification number, Columns\,3 and 4 are the J2000
749: coordinates, Columns\,5 to 11 give the JHK and IRAC magnitudes.
750:  Of these 23 objects, 
751: %IDL> help,CL0I_ONLYIR_X_1
752: %CL0I_ONLYIR_X_1 LONG      = Array[10]
753: 10 ($\sim$40\%) have an X-ray counterparts
754:  in the COUP data. 
755: % Three of these (COUP \# 360, 611, 702)  are in the \citet{getm05b} 
756: One of these (COUP \# 702) is in the \citet{getm05b} 
757:  list of nonflaring 
758:  COUP X-ray sources without optical and NIR counterpart  and was  
759:   flagged by them as a
760: probable extragalactic object (EG).
761:  %None of our objects is, instead, in the list of
762:  %flaring COUP X-ray sources without optical and NIR counterparts 
763:  %\citep[Table\,1][]{getm05b}.
764: 
765: %\clearpage
766: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
767: \begin{figure*} 
768: \centerline{\includegraphics[width=18cm]{f2.ps}}
769: \caption{SED of the 23 Class\,0-Ia candidate objects with a rising SED from K to
770: 8.0\,$\mu$m. The sequential number in our catalog and the spectral index
771: computed using the available IRAC magnitudes  are also indicated.}
772: \label{f02}
773: \end{figure*}
774:  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
775: %\clearpage
776: 
777: Fig.\,\ref{f03} shows the color-color diagrams
778: (CCDs) obtained using the K and IRAC magnitudes 
779: of all the stars in the COUP field of view.
780: %\clearpage
781: \begin{figure*} 
782: \centerline{\includegraphics[width=19cm]{f3.ps}}
783: \caption{CCD of the stars in the COUP field of view (dots)
784: created using the $K$ and the
785: {\it Spitzer} IRAC magnitudes. 
786: The total number of objects in each diagram is reported in the three panels.
787: Squares and  large filled circles represent 
788: the   Class\,0-Ia and 0-Ib objects, respectively, 
789:  while
790:  {\tt X} symbols indicate
791:   those with an X-ray counterpart; asterisks
792: indicate objects rejected because they have an optical counterpart.
793: % {\it Dashed lines} indicate the
794: %\citet{hart05} limits adopted to select Class\,0-I objects. 
795: The reddening vectors computed using the IR reddening law 
796: given in \citet{flah07} for Orion A are also plotted.}
797: \label{f03}
798: \end{figure*}
799: %\clearpage
800: %{\it Squares} and {\it large filled circles}
801: % represent Class\,0-Ia and \,0-Ib objects,
802: %respectively,
803: %and {\tt X} symbols indicate
804: %those with an X-ray counterpart; 
805: %{\it asterisks} indicate objects
806: %rejected because they have an optical counterpart. 
807: %We report in each panel the total number of plotted objects
808: %and those of X-ray detected and undetected Class\,0-I stars.  
809: %>>>REMOVE (ovvio, direi)
810:  The plot on the left panel includes more
811: stars than those  on the central and right ones 
812: because it does not require detections at 5.8 and 8.0\,$\mu$m. 
813: %<<<
814: The limited sensitivity of the 8.0\,$\mu$m channel means that few
815: objects with zero color (consistent with photospheric emission) appear
816: in the right panel. 
817: The selected  Class\,0-Ia objects with increasing SED,
818: indicated in Fig.\,\ref{f03} by squares, show [3.6]-[4.5]$>$0.7
819: and [5.8]-[8.0]$\gtrsim$1.1. The objects with [3.6]-[4.5]$>$ 0.7
820: and [5.8]-[8.0]$\lesssim$1.1 that are usually also classified as Class\,0-I 
821: stars \citep{alle04,mege04}, show a decreasing or flat SED at wavelengths longer
822: than 4.5\,$\mu$m. 
823: The spectral energy distributions of protostars in this
824: spectral regime are complex as
825: rotation \citep{keny93} and outflow cavities
826: \citep{whit03} reduce the
827: extinction toward the poles of the protostellar envelopes allowing
828: radiation to leak  and scatter
829: out.  Consequently, protostars
830: can also exhibit decreasing slopes, particularly between 4.5 and 8.0~$
831: \mu$m, depending on their
832: inclination \citep{whit03}.  Furthermore, highly
833: reddened Class II
834: objects can have increasing slopes; however, these slopes will
835: generally flatten or decline for
836: $\lambda \ge 4.5 \mu$m due to the flattening of the reddening law
837: and the broad silicate feature
838: at 9.7~$\mu$m \citep{flah07}.  Thus, although the IRAC data
839: can enhance the identification
840: of protostars by extending the SEDs to 8 microns; establishing a
841: sample of protostars uncontaminated by
842: reddened Class II objects is still difficult  without photometry at
843: wavelengths $\lambda > 8$~$\mu$m.
844: 
845: To take into account these cases, we also include the objects with  a 
846: decreasing SED for $\lambda > 4.5 \mu$m
847: in our Class\,0-I sample.
848: %IDL> help,PROTOSTAR
849: %PROTOSTAR       LONG      = Array[69]
850: %IDL> help,PROTOSTAR_1
851: %PROTOSTAR_1     LONG      = Array[41]
852: %PROTOSTAR_2=69-41=28
853: There are 28  such objects in total. %Six
854: %IDL> help,CL0I_ONLYIR_2
855: %CL0I_ONLYIR_2   LONG      = Array[22]
856: % CL0I_COMPLEX_2=28-22=6
857: % of these have optical
858: % counterparts and  as before, we discard them from
859: % further consideration, leaving 22 objects. 
860:  %In our investigation of their 
861:  %X-ray properties, we will refer to these 22 objects as 
862: % the Class\,0-Ib sample, thus distinguishing them from the previously selected 
863: % sample of 23  candidate protostars with continually rising SEDs 
864: % (the Class\,0-Ia objects).
865: After rejecting the six objects with optical counterparts in the HST imaging, we 
866: are left with 
867: 22 objects,  which we
868: refer to hereafter as Class\,0-Ib sources. 
869:  The  SEDs of the 
870:  Class\,0-Ib objects are shown in
871:  Fig.\,\ref{f04}, while the photometry is given in 
872: Table\,\ref{cl0I_2}, 
873: analog
874:  to Table\,\ref{cl0I_1}.
875:  Among these objects,   18 
876: ($\sim$80\%) are in the COUP list of
877:  X-ray detection, including  COUP source 1565, which was classified as an 
878:  extragalactic object by \citet{getm05b}. 
879: %\clearpage
880:  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
881: \begin{figure*} 
882: \centerline{\includegraphics[width=18cm]{f4.ps}}
883: \caption{SED of the 22 Class\,0-Ib candidate objects with a rising SED
884: from K to 4.5\,$\mu$m. The sequential number in our catalog and the 
885: spectral index
886: computed using the available IRAC magnitudes  are also indicated.}
887: \label{f04}
888: \end{figure*}
889:  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
890: %\clearpage
891:  
892: If we consider both 
893: Class\,0-Ia and Class\,0-Ib samples,
894: our list of X-ray detected Class\,0-I objects (10+18)
895:   is one of the largest  sample of Class\,0-I stars with 
896:  evidence of X-ray emission \citep{getm07,giar07}.
897:  
898: Our selected  Class\,0-I
899: objects fall in the region
900: of the envelope models indicated
901: in \citet{alle04} and \citet{mege04}. 
902: There are
903: a small number 
904:  of objects falling in the Class\,0-I 
905: regions of the CCD which were
906: not selected with our criterion because either
907: they have an optical counterpart (18+6 in total) or 
908: their SED band-to-band slopes do not increase as required by our
909: criterion.
910: Whereas some of these rejected 
911: objects could be Class\,0-I stars, by adopting a conservative selection 
912: criterion, we 
913: have minimized
914: the adverse effect of sample contamination 
915:  for our comparison of X-ray properties.
916: 
917: A careful examination of the near-IR and IRAC
918: colors of our Class\,0-I candidates (see Figs.\,\ref{f03}
919:  and \ref{f05})
920: suggests to us that the Class\,0-Ia's (on average) are
921: inconsistent with their being heavily reddened Class II's.
922: In particular, the bottom-right panel of Figure 5 shows
923: that, using the Orion A reddening law of Flaherty, the
924: Class\,0-Ia's are too red in [4.5]-[5.8] to be heavily
925: reddened Class\,II's - and in any case would require
926: reddenings of A$_{\rm V}\geq60$.  The bottom left panel of
927: Fig.\,\ref{f05} also indicates that the Class\,0-Ia's, on average,
928: would need very large reddenings (A$_{\rm V}\sim50$) for them to
929: be Class\,II sources.  Such heavily embedded objects would
930: likely be very young in any case - and we therefore believe
931: it is more likely that they are, indeed, Class\,0 or Class\,I
932: objects.   By contrast, the Class\,0-Ib objects are located
933: in all of the diagrams in Figs.\,\ref{f03}
934:  and \ref{f05}  in positions
935: consistent with being moderately reddened (A$_{\rm V}\sim30$
936: on average) Class\,II's.   We suspect the Class\,0-Ib's
937: are a mixture of true Class0 or I objects, and heavily
938: reddened Class\,II's - without more data we cannot
939: clean the sample any further than we have.
940: 
941:  Another possible worry is that our Class\,0-I candidate sample includes 
942:  systems seen edge-on in or around the HII region. In these cases, 
943:  relatively evolved Class\,II objects may display a SED peaking at 
944:  mid-IR and far-IR wavelengths, as suggested by \citet{robb02}.
945:   However, since edge-on
946:  disks should be randomly distributed in the nebula, 
947:  clustering in the dark regions
948:  of the selected objects (see Fig.\,\ref{f06}a) strongly suggests
949:   that they are
950:  real protostars.
951:  
952: %since the [5.8-8.0] color is particularly insensitive to the reddening
953: %\citep{inde05,flah07} and the Class\,II objects have $[5.8-8.0]$ colors 
954: %smaller than 1.1 (see Section \ref{classII_sec}).
955: 
956: 
957:  
958: \subsection{Class\,II stars \label{classII_sec}}
959: Following \citet{hart05}, Class\,II objects within the COUP field of view
960: were selected using the condition $0.2<[3.6]-[4.5]<0.7$ and
961:  $0.6<[5.8]-[8.0]<1.1$. These criteria are slightly more restrictive than 
962:  those adopted by \citet{mege04} in order to attempt to create
963:  as "pure" a set of Class II's as possible.
964:  With our selection criteria,  
965:  we may lose reddened Class\,II
966:  objects, but assuming that X-ray properties of Class\,II
967:   stars are independent of their reddening, our sample should be representative
968:   of the Class\,II population. 
969:      
970:   Using this criterion, we find 148 
971: %IDL> RESTORE,'../SAVEFILES/classII.save',/v
972: %IDL> help,clII_sub
973: %CLII_SUB        LONG      = Array[148]
974: %IDL> help,clII_x_sub
975: %CLII_X_SUB      LONG      = Array[130]
976:  Class\,II stars with NIR photometry
977: given in Table\,\ref{clII},
978:  analogous to Table\,\ref{cl0I_1}; 130 of these objects (about 
979: 88$\%$) are X-ray  detected. One of them, the COUP source 1401, is listed by
980: \citet{getm05b} as a  flaring source and  
981: classified as an Orion Molecular Core (OMC) member, 
982: whereas another Class\,II object,
983:  COUP  source % 151 and 
984:  751, was classified by the same work as  extragalactic object. 
985:  Among the 130  % see classII_def.pro
986: %XOPT_CLII       LONG      = Array[130]
987: %EWCA_ALL        LONG      = Array[537]
988: %EWCA_ALL_CLII   LONG      = Array[83]
989: %EWCA_ACCR_CLII  LONG      = Array[28]
990: X-ray detected Class\,II stars
991: with a counterpart in the \citet{hill97} catalog,
992: 83  have a measure of the equivalent width of the 
993: Ca\,{\small II} 8542\AA\ line $W$(Ca\,{\small II}). Among these,
994: 28 have $W$(Ca\,{\small II})$<-$1, 
995: which is the limit used by \citet{flac03a} to classify strong accretors.
996: Therefore we have about 55 %(83-28)/83 
997: objects (i.e. $\sim$66\%) 
998: that have a near-IR bright inner disk but relatively low 
999: accretion rate.   
1000: Of the 148 Class\,II stars we have identified, we will consider 
1001:  in the following analysis only 
1002: the 87 X-ray detected  and the 9 X-ray undetected stars for which 
1003:   masses are known \citep{hill97,getm05a}.
1004: %help,CLII_lum_x
1005: %CLII_LUM_X      STRUCT    = -> <Anonymous> Array[96] 87 x+9 no X 
1006: We retain only these
1007: stars  in order to study the
1008: X-ray properties as a function of stellar mass.
1009: 
1010: \subsection{Class\,III stars \label{classIII_sec}}
1011: 
1012:  Class\,III objects are PMS stars that do not show IR
1013: excesses and therefore 
1014: were selected as those having IRAC colors near zero.
1015: In order to include all the stars not detected in the longer
1016:  wavelength IRAC
1017: bands, we selected all the stars with $K-[3.6]<0.5$ and
1018:  rejected those with $[3.6]-[4.5]>0.2$ or $[4.5]-[5.8]>0.2$ or
1019: $[5.8]-[8.0]>0.2$, when these latter colors were available,
1020:  i.e. objects with IR excesses. Stars 
1021:  undetected in the IRAC bands at wavelengths
1022: longer than  3.6\,$\mu$m were included since 
1023: their non-detection in these  bands
1024: is compatible with purely photospheric SEDs.
1025: 
1026: Using these criteria, we selected 205 objects,
1027:  with photometry 
1028: given in Table\,\ref{clIII}; 160 of  them
1029:  (about 80\%)
1030: have an X-ray counterpart.
1031:  Among these latter, 150
1032: are in the list   of
1033: \citet{hill97}. In particular, 130 of them have
1034:  a proper motion membership
1035: probability larger than 90\%, 6 have a membership probability equal to 0 and 
1036: the other 14 have no  membership probabilities. 
1037: For the remaining 10 stars without
1038: a counterpart in the list  of
1039:  \citet{hill97}, 
1040:  X-ray detection is the only known 
1041: membership criterion. 
1042: We note that none of these 10 objects is either in the list of flaring
1043: or  in that of non-flaring sources of \citet{getm05a}.
1044:  Therefore
1045:  the membership class for these objects 
1046:  remains uncertain.
1047:  
1048:  Of the 45 objects without an X-ray detection, 18 have
1049: a counterpart in the \citet{hill97} catalog but only 5 of these have a proper
1050: motion membership probability larger than 90\%. 
1051: 
1052: The remaining 
1053: 40 objects without a membership criterion
1054: could be  either Class\,III stars without X-ray activity, or
1055:  field stars. 
1056: Of the 205 Class\,III stars initially selected, we 
1057: will consider the 103 %help,coup_clIII
1058: X-ray detected stars for which the mass is known 
1059: \citep{getm05a} and with \citet{hill97} membership probability larger than 90\%;
1060: in addition, from the sample of X-ray
1061:  undetected Class\,III candidates, we will retain
1062: only the 2 objects with membership probability larger than 90\%
1063: and for which  a mass has been derived by \citet{hill97}.
1064:  
1065: Fig.\,\ref{f05} shows the selected stars
1066: in the IRAC CCD and color-magnitude diagrams. 
1067: %{\it Dots} indicate
1068: %all the stars in the COUP field of view with detections at the bands used in
1069: %each panel. {\it Squares} and {\it large filled circles} 
1070: %are the Class\,0-Ia and 0-Ib objects, respectively,
1071: % {\it plus symbols  } are the stars classified  as Class\,II and 
1072: % {\it empty diamonds} are
1073: %the Class\,III stars. The total number of detections and the total for each
1074: %subclass are given in each panel. 
1075: These diagrams show that our samples of YSOs most likely do not include AGN or 
1076: star-forming galaxies;
1077: that is, they do not include objects falling in the wedge-shaped
1078: regions outlined in the upper panels of Fig.\,\ref{f05}.
1079: Star-forming galaxies are usually dominated by PAH  features in
1080: the 5.8 and 8.0\,$\mu$m  and therefore
1081: have $[5.8]-[8.0]>1.0$ but $[3.6]-[4.5]\lesssim0.3$, that corresponds to 
1082: the regions indicated
1083: by solid lines 
1084: in the two upper panels of Fig.\,\ref{f05} \citep{gute07}.
1085: AGN, instead, have 
1086: $[3.6]-[4.5]$ and $[5.8]-[8.0]$ colors very similar to the Class\,II YSOs,
1087: but are typically much fainter than YSOs, having, for example,
1088: [4.5] magnitudes  \citep{mege05},
1089:  much fainter than those of
1090: the selected YSOs (see  bottom-left panel in
1091: Fig.\,\ref{f05}).
1092: 
1093: %\clearpage
1094: \begin{figure*} 
1095: \centerline{\includegraphics[width=15cm]{f5.ps}}
1096: \caption{IRAC CCD and color-magnitude diagrams of the  different
1097: populations in the ONC.  Dots indicate
1098: all the stars in the COUP field of view with detections at the bands used in
1099: each panel. Squares  and large filled circles
1100: are the Class\,0-Ia and 0-Ib objects, respectively,  plus symbols 
1101:  are the stars classified  as Class\,II and  diamonds are
1102: the Class\,III stars.  The reddening vector corresponding to A$_{\rm V}=30$,
1103: computed using the \citet{flah07} reddening law, is plotted on each panel.
1104:  The solid lines 
1105: in the  above panels separate the region dominated by YSO's
1106: from the region expected to be dominated by galaxies.
1107: %The total number of detections and the total for each
1108: %subclass are given in each panel.
1109: }
1110: \label{f05}
1111: \end{figure*}
1112: %\clearpage
1113: 
1114: \section{SPATIAL DISTRIBUTION}
1115: The spatial distribution of YSOs within the molecular cloud from which
1116: they have formed
1117: may give  important clues for understanding the star formation
1118: process. Using the classification 
1119:  defined in the previous section, we have analyzed the spatial distribution
1120: of the YSOs in the ONC, as shown in 
1121: Fig.\,\ref{f06}a where we plot   the 
1122: position of all  2590  objects in
1123: our catalog within the COUP field of view that have been detected in 
1124: at least two of the UBVIJHK and IRAC bands and the Class\,0-I candidates.
1125:  Stars of Class\,II and III  
1126: are plotted in Fig.\,\ref{f06}b and c, respectively.
1127: %\clearpage
1128:  \begin{figure*} 
1129: \centerline{\includegraphics[width=18cm]{f6.ps}}
1130: \caption{Spatial distribution of all the objects in
1131: our catalog within the COUP field of view that have been detected in at least
1132: two of the UBVIJHK and IRAC bands. Squares  and  large filled 
1133: circles indicate
1134: Class\,0-Ia and 0-Ib objects,  plus symbols  
1135: the   Class\,II stars 
1136: and  diamonds  the   Class\,III objects. 
1137: In all the panels, X-ray detected objects
1138: are indicated with an {\tt X} symbol. }
1139: \label{f06}
1140: \end{figure*}
1141: %\clearpage
1142: 
1143: Figure\,\ref{f07} shows the velocity integrated $^{13}$CO map obtained by
1144: \citet{ball87}   and the {\it Spitzer} IRAC image at 3.6\,$\mu$m 
1145:   where the
1146:  different classes selected in this work have been overplotted.
1147: %  In particular, yellow squares and  cyan circles indicate Class\,0-Ia and 
1148: %0-Ib objects, green plus symbols
1149: %are the Class\,II stars and  red diamonds are the Class\,III stars.
1150: The  upper green box indicates a 40\sec$\times$50\sec\ region around 
1151: the Becklin-Neugebauer object  and the
1152: Kleinmann nebula, also indicated as BN-KL 
1153: %(RA(2000)=05$^h$35$^m$14.16$^s$,
1154: (RA$_{2000}$=05$^h$35$^m$14\timesec16,
1155: Dec$_{2000}$=-05\deg22\min21.5\sec). %, from SIMBAD). 
1156: The bottom green box
1157:  indicates a  60\sec$\times$75\sec\ region 
1158: around the Orion Molecular Cloud 1 (OMC-1S) 
1159: (RA$_{2000}$=05$^h$35$^m$14\timesec50,
1160: Dec$_{2000}$=-05\deg25\min49.0\sec).% , from SIMBAD
1161: These boxes show the limits of the areas studied in \citet{gros05}. 
1162: 
1163: Most of the  Class\,0-I stars are  
1164:    concentrated in the central and north-east parts
1165: of the COUP field of view  
1166:   along a region running north-south from the
1167: Trapezium that 
1168: follows the north-south distribution of the $^{13}$CO emission, i.e. of the 
1169: dense star-forming gas. The distribution of our selected Class\,0-I objects
1170:  is biased by our NIR observations which are more sensitive
1171: towards the north of the molecular ridge where the nebular
1172: contribution is lower. The lack of Class\,0-I objects in
1173: the BN-KL and OMC-1s regions is due to the bright nebulosity, as 
1174: is evident from the image at 3.6\,$\mu$m, that causes a decrease of sensitivity 
1175: in the central region. For this reason, the spatial
1176:  distribution of our Class\,0-I objects 
1177:  cannot be directly compared with that obtained by 
1178:  \citet{lada00,lada04}, although in fact the two spatial 
1179:  distributions are quite similar.
1180: %\clearpage  
1181: \begin{figure*}[t] 
1182: %\centerline{\includegraphics[width=14cm]{f7.eps}}
1183: \centerline{\includegraphics[width=18cm]{f7.eps}}
1184: \caption{{\it (Left)} Velocity integrated $^{13}$CO emissivity from
1185: \citet{ball87}  in the region around the Trapezium cluster and
1186: {\it (right)} {\it Spitzer} IRAC image at 3.6\,$\mu$m in the same region. 
1187: Both images are shown with inverted color map. Yellow squares and  light blue
1188:  circles
1189:  indicate Class\,0-Ia and 0-Ib objects,  green plus symbols
1190: are Class\,II stars while red diamonds are Class\,III stars.
1191: The 40\sec$\times$50\sec upper green box indicates  
1192: the BN-KL  region while the 60\sec$\times$75\sec   bottom green box 
1193: indicates  the Orion Molecular Cloud 1 South (OMC-1S).}
1194: \label{f07}
1195: \end{figure*}
1196: %\clearpage
1197: 
1198: %Panel (a) of Fig.\,\ref{sectors_2} shows
1199: % the percentage of 
1200: %objects belonging  to
1201: % Class\,0-I (first number), Class\,II (second number) and Class\,III
1202: %(third number) in each of the sectors delimited by the four  line segments,
1203: %  computed with respect to the total number of selected objects in each sector, while
1204: %panel (b) shows
1205: % the percentage of 
1206: %objects belonging  to
1207: % Class\,0-I (first number), Class\,II (second number) and Class\,III
1208: %(third number) in each sector,
1209: %  computed with respect to the total number of
1210: %stars of each class in the whole COUP FOV. 
1211: 
1212: 
1213: %Most of them, about 40\%, are however concentrated
1214: %in  the central sector. 
1215: Figures\,\ref{f06} and \ref{f07} show the same structure already
1216: found in \citet{feig05} using  the lightly X-ray 
1217: absorbed stars: they interpreted this distribution as due to a process of 
1218: violent relaxation.  
1219: Note that, the latter study examined the spatial distributions
1220: stratified by absorption, while here the stars are stratified by individual YSO
1221: evolutionary classification. The features of the spatial distribution
1222: visible in Figs.\,\ref{f06} and \ref{f07} include: the OMC-1S core,
1223: BN/KL core and northern extension which are part of the larger filament along
1224: the middle of the Orion A cloud and finally the East-West asymmetry that in
1225: \citet{feig05} was
1226: attributed to non-equilibrium dynamics. The latter only shows up in Class\,III
1227: systems.
1228: 
1229:  The spatial distributions of  
1230: Class\,II and III stars are also 
1231: in agreement with the spatial distribution found by 
1232: \citet{hill97} who,  using \citet{dant94} isochrone ages, 
1233:  found that youngest stars are clustered toward the cluster
1234: center  while slightly older stars are more widely distributed. 
1235: The  elongated asymmetric 
1236: distribution of our candidate Class\,0-I stars and the large-scale EW
1237: asymmetry are consistent with the
1238: distribution derived by \citet{hill98} who concluded
1239:  that the structure of the ONC is more the result 
1240: of the different environmental conditions rather than the result of dynamical
1241: evolution.  
1242: 
1243:  There is
1244: a lower percentage of Class\,II objects in the central region,
1245: but this could be due to the reliance on the 8.0\,$\mu$m detections for
1246:  the selection of Class\,II stars.
1247: Our spatial distribution are, therefore, strongly dependent on the 
1248: sensitivity of our NIR data which is quite limited toward the central region,
1249: where the contamination by the bright 
1250: OB stars and nebular emission is very high.
1251: 
1252: %
1253: % \begin{figure}[!hb]
1254: %\centerline{\psfig{figure=sectors_2.ps}}
1255: %\caption{Panel (a) shows
1256: % the percentage of 
1257: %objects belonging  to
1258: % Class\,0-I (first number), Class\,II (second number) and Class\,III
1259: %(third number) in each of the sectors, delimited by the four  drawn segments,
1260: %  computed with respect to the total number of selected objects in each sectors.
1261: %Panel (b) shows
1262: % the percentage of 
1263: %objects of Class\,0-I (first number), Class\,II (second number) and Class\,III
1264: %(third number) with respect to the total number of
1265: %stars of each class in the whole COUP FOV. }
1266: %\label{sectors_2}
1267: %\end{figure}
1268: 
1269: \section{EVOLUTION OF X-RAY ACTIVITY}
1270: 
1271: The very deep X-ray COUP data
1272: together with the relatively large sample of Class\,0-I objects selected with
1273: the {\it Spitzer} observations 
1274:  allow us, for the first time, to analyze and compare, 
1275: in a statistically significant way, the X-ray properties of the 
1276: YSOs in  their initial protostellar phase 
1277: with those of more evolved PMS stars. 
1278: %\citet{prei05a} shows that the X-ray
1279: %luminosities remain roughly constant from 10$^5$ to 10$^7$\,yrs, possibly with a
1280: %slight decline in L$_{\rm X}$ but smaller than the scatter.
1281: 
1282: As already mentioned in Section\,\ref{classification}, we find that
1283: the fraction of Class\,0-Ia objects detected in the X-rays
1284: is almost 40\% (10/23),  that of 
1285:  Class\,0-Ib  and   Class\,III  
1286:     is about 80\%  (18/22 and 160/205, respectively) while that of
1287: Class\,II objects
1288: is  about 88\%( 130/148).  The analysis described in this section aims at
1289: understanding whether
1290:  the X-ray emission can be related to the IR properties of YSOs, which we
1291:  take as signposts for 
1292: the different phases of protostellar and PMS evolution.   
1293: 
1294: \subsection{X-ray luminosities \label{xluminosity}}
1295: We use the results of the spectral analysis presented by \citet{getm05a} to
1296: compare the X-ray properties of the  different populations of YSOs in the
1297: ONC.
1298: For the X-ray luminosities of
1299: detected stars we adopt  
1300: the absorption corrected luminosities in
1301:  the total 0.5--8.0\,keV band, L$_{t,c}$, from Table 8 of \citet{getm05a}.
1302: 
1303: In order to compare unbiased luminosity functions  of the  
1304: different classes, we derive  upper limits to
1305: the X-ray luminosities of the X-ray undetected objects.
1306: Upper limits to the photon count rates of the 13+4 undetected Class\,0-Ia 
1307: and Class\,0-Ib objects were calculated 
1308: with PWDetect \citep{dami97b,dami97a} 
1309: consistently with the source
1310:  detection procedure of the other COUP sources \citep{getm05a}, adopting a
1311:  detection threshold significance   of 5\,$\sigma$.
1312:  The upper limits for the 9 undetected Class\,II stars and 
1313:  for the 2 undetected Class\,III stars
1314:   were taken from Table\,11 of \citet{getm05a}.
1315:  
1316: %we considered from the  and III samples only stars for which the mass
1317: %has been estimated using the \citet{sies00} PMS model \citep{getm05a}. 
1318: %With this constraint, we have 83 X-ray detected Class\,II stars.
1319: % For the Class\,III samples, instead, we have
1320: %53 X-ray detected and 2 X-ray undetected.
1321: % Of these latter 2 objects, we consider the
1322: %only star with an high proper motion membership probability and exclude the
1323: %other for which such probability is unknown.
1324: 
1325: In order to convert upper limits to the count rates into unabsorbed 
1326:  X-ray luminosities, we  derive
1327: an extinction dependent conversion factor. 
1328: The CCDs in Fig.\,\ref{f03} indicate that
1329:  X-ray detected and undetected Class\,0-I objects have similar reddening.
1330:  We therefore computed the conversion factors of the X-ray detected 
1331:  Class\,0-I objects as the ratio between the unabsorbed X-ray luminosities 
1332:  (L$_{t,c}$) and the count rates  taking into
1333:  account the PSF fraction within the extracted area and the effective exposure
1334:  time at the source position \citep{getm05a}, and we plotted these values as a
1335:  function of the N$_{\rm H}$\footnote{The  H column densities were derived
1336:   by \citet{getm05a}
1337:  from the spectral analysis of the COUP sources.} values.
1338:  We  assume for the X-ray undetected Class\,0-I objects a conversion
1339:  factor computed as the
1340:  median value of the conversion factors of the 
1341:  X-ray detected Class\,0-I objects, which corresponds for these  stars to 
1342:  $N_{\rm H}\simeq4.5\times10^{22}$cm$^{-2}$. 
1343:  However,
1344:  in order to check whether  our results are robust with respect to the assumed 
1345:  N$_{\rm H}$ of undetected objects, 
1346:  we also considered the conversion
1347:  factors corresponding to the highest and minimum 
1348:  conversion factors of  
1349:  the X-ray detected 
1350:  Class\,0-I objects.\footnote{We  used the median conversion factor 
1351:   to derive the X-ray luminosity 
1352: (log $L_{t,c}=28.52$\,erg/s)
1353: of the COUP source 1197 for which \citet{getm05a} did not perform
1354: a spectral fit  because of the low counting statistics.} 
1355: % ($N_{\rm H} \sim 3.63 \times 10^{23}$  cm$^{-2}$) and with the minimum 
1356: % value
1357: % ($N_{\rm H} \sim 10^{23}$cm$^{-2}$) of the N$_{\rm H}$ 
1358: % distribution. 
1359: We used the same procedure to estimate the upper limits to the X-ray
1360: luminosities in the 2.0--8.0\,keV hard band.
1361: 
1362: The X-ray properties of the Class\,0-I  COUP X-ray sources, taken from 
1363: \citet{getm05a}, are given in Table\,\ref{coup_cl0I}. We provide
1364: sequential numbers from our catalog, sequential COUP
1365: numbers,  background-corrected counts in the 0.5--8.0\,keV band,
1366: the fractions of the PSF within the source extraction areas, 
1367: effective exposure times,
1368: the hydrogen column densities
1369:  obtained from the spectral fit,  plasma temperatures, 
1370:   emission measures and finally the  absorption-corrected  X-ray 
1371:  luminosities in the 2.0--8.0\,keV hard band, L$_{h,c}$, and 
1372:  in the 0.5--8.0\,keV total band, L$_{t,c}$.
1373: 
1374:   Values
1375: computed for the X-ray undetected Class\,0-I stars are given in 
1376: Table\,\ref{cl0I_up_lim}. We list
1377: sequential numbers from our catalog, the
1378: celestial coordinates from the {\it Spitzer} catalog, 
1379: upper limits to the source counts computed
1380: with PWDetect in the hard and total bands,
1381:  the effective {\it Chandra} exposure time at the object positions and,
1382:  finally, the upper limits to the unabsorbed X-ray luminosities
1383:  in the hard and total energy bands.
1384: 
1385: X-ray properties of X-ray detected Class\,II and III stars, 
1386: taken from \citet{getm05a},
1387: are given in Tables \ref{coup_clII_1}   
1388: and \ref{coup_clIII_1} where column names are analogous to those given in
1389: Table\,\ref{coup_cl0I}.
1390: 
1391: The X-ray luminosities of 
1392: X-ray undetected 
1393: Class\,II and III objects were derived, as for Class\,0-I stars, 
1394: using the median conversion factor 
1395: of the X-ray detected stars in the same class.
1396: This assumption is based on the fact that the NIR colors of the
1397:  X-ray undetected Class\,II stars  indicate
1398:  that  these
1399:  objects are not more absorbed than those detected with COUP.
1400:  In addition, the conversion factors
1401: corresponding to the lowest and highest  observed N$_{\rm H}$ values 
1402:  were also  considered
1403: ($10^{21}$ and 4.0 $\times 10^{22}$  cm$^{-2}$  for the Class\,II stars and
1404:  0.5$\times 10^{21}$ and $2.0\times 10^{22}$  cm$^{-2}$  for the Class\,III stars). 
1405:  Values
1406: computed for the X-ray undetected Class\,II and III stars are given in 
1407: Table\,\ref{clII_III_up_lim}, analogous to 
1408: Table\,\ref{cl0I_up_lim}.
1409: 
1410: \subsection{X-ray luminosity functions\label{xlf_section}}
1411: %\clearpage
1412: \begin{figure*}[!ht]
1413: \centerline{\includegraphics[width=18cm]{f8.ps}}
1414: \caption{Maximum likelihood cumulative XLF in the total (0.2--8.0)\,keV 
1415: energy band
1416: of the four NIR classes. The three panels refer to Class\,II and Class\,III
1417: stars in three different
1418:  mass ranges. The red and green lines 
1419: indicate the XLF of Class\,III and II objects, respectively, with masses in the
1420: given  range, while  the 
1421: black lines are the XLF of Class\,0-I stars, for which masses are unknown
1422: and are therefore the same in the three panels. Lines  of the same color 
1423: in each panel show the XLF computed using the three different luminosity 
1424: conversion factors for the X-ray non detections. 
1425: Thick lines indicate the XLF used to perform the statistical tests  (see text).
1426: Note that the
1427: distributions do not reach the f=100\% because of the presence of upper limits.}
1428: \label{f08}
1429: \end{figure*}
1430: %
1431: \begin{figure*}[!ht]
1432: \centerline{\includegraphics[width=18cm]{f9.ps}}
1433: \caption{Maximum likelihood cumulative XLF in the hard (2.0--8.0)\,keV 
1434: energy band
1435: of the four NIR classes. The figure is analogous to Fig.\,\ref{f08}.}
1436: \label{f09}
1437: \end{figure*}
1438: %\clearpage
1439: 
1440: We now compare the X-ray luminosity functions  of the stars of the four
1441: considered NIR classes. In order to 
1442:  take into account 
1443: nondetections, XLFs were derived using the
1444: Kaplan-Meier maximum-likelihood estimator.
1445: For Class\,II and III sources, we considered separately stars in the three mass
1446: ranges used by \citet{prei05a}.
1447: Fig.\,\ref{f08} shows the comparison of the  XLFs
1448: of the four populations. The  red and green lines 
1449: indicate the XLF of Class\,III and II objects, respectively, with masses in the
1450: given range, while  the 
1451:  black and  blue lines are the XLFs of Class\,0-Ia and 0-Ib stars,
1452:  for which masses are unknown
1453: and are therefore the same in the three panels. Lines of the same color 
1454:  in each
1455: panel show the XLF obtained 
1456: using the three different count rate to luminosity conversion factors
1457: adopted for the X-ray non detections of each subsample and give an idea of the
1458: uncertainty of the XLFs. 
1459: We removed from this analysis the four objects of 
1460: our sample of Class\,0-Ia protostars  without detections
1461: at 4.5\,$\mu$m. Since this band is
1462: one of the most sensitive bands and the least contaminated by nebulosity,
1463: these objects could be PAH  contaminated sources or even knots of nebulosity. 
1464: By discarding these objects, we are more confident to consider a sample of 
1465: protostar candidates as less contaminated as possible.
1466: The figure shows that 
1467:  for the  two highest mass
1468: ranges,  the X-ray
1469: luminosities of Class\,III stars are highly elevated with respect to those of 
1470: Class\,II stars and 0-Ib objects. These latter two classes have very similar XLFs and
1471:  are significantly brighter than Class\,0-Ia stars. 
1472: 
1473:   Because the Class\,0-Ia and Class\,0-Ib stars in our sample are heavily
1474:   absorbed, 
1475:   a note of caution is needed. Since absorption precludes the observation
1476:   of the soft part of the intrinsic source X-ray spectra, we 
1477:   cannot rule out the presence of soft, completely absorbed  emission
1478: from Class\,0-I stars. 
1479: Nevertheless,
1480: we know from previous studies of  CTTS stars that a very soft
1481:  component, kT$<$0.3\,keV, in the few cases in which it is observed,
1482:   is mainly due to accretion \citep{kast02,schm05,gude07,argi07,huen07,gros07},
1483: while  harder components are due to coronal-like emission.
1484: This is in agreement with the previous
1485: finding for Class\,II stars in other regions \citep[e.g.][]{stel04},
1486: where the X-ray spectra of Class\,II's were found to include an
1487: additional softer emission component compared to WTT's.  Our
1488: Orion stars are too heavily absorbed to detect this soft component.
1489: 
1490: Coronal emission, as analyzed with  resolution spectra from {\it 
1491: Charge Coupled Devices}, 
1492: can  
1493: usually be decomposed into softer $\sim$0.8--1.0\,keV and harder 
1494: $\sim$2\,keV components \citep{prei05b}. The soft coronal component, like 
1495: the accretion component, might remain unobserved in heavily extincted sources
1496: such as our Class\,0-I stars and  
1497: thus the X-ray luminosities in the total band can underestimate the true
1498: emission in such objects. 
1499: In addition, magnetic flare components can be strong around 0.5--2.0\,keV,
1500: and dominate the X-ray flux. 
1501: For these reasons, we also considered the comparison of the XLFs of different
1502: classes,
1503: by restricting the analysis to the X-ray
1504:  luminosities in the hard (2.0--8.0)\,keV energy band. 
1505:  We note that we recomputed with PWDetect the upper limits in the hard band for a 
1506:  total of 31 undetected objects, assuming a threshold of 4.8$\sigma$ estimated
1507: from the background in this band. Among these objects, 
1508:  12 are flagged by the PWDetect code
1509: as affected by a X-ray detected source nearby, which therefore raises the upper
1510: limit count rate. In two cases out of 12, however, the asymmetry of the Chandra
1511: PSF at large off-axis angles is such as to make PWDetect clearly overestimate 
1512: the contribution of the detected source to the computed upper limit at the
1513: desired position, which is instead dominated by the background; in these
1514: instances we computed the upper limit from the background alone. 
1515: The results of this analysis are shown in Fig.\,\ref{f09},
1516: analogous to Fig.\,\ref{f08} but for the hard energy band.  
1517: 
1518: To check if these results are statistically 
1519: significant, we performed statistical tests  developed for ``survival
1520: analysis", suited for the analysis of censored datasets. The
1521: probabilities that the X-ray luminosities of the stars in two samples are drawn
1522: from the same parent distribution were computed with the
1523: ASURV\footnote{Astronomy Survival Analysis  available from the
1524: StatCodes http://www.astro.psu.edu/statcodes} package \citep{feig85}.
1525:  The results of this
1526: analysis are given in Tables\,\ref{xlf_prob} and \ref{xlf_prob_hard} 
1527: where, for each pair of samples,
1528:  we list the probabilities
1529: computed using the five tests of the ASURV package.
1530:  We note that the tests were
1531: performed in the most conservative way, i.e. by considering for each pair
1532: of samples the nearest XLFs among those computed with different conversion factors
1533: for the upper limits. 
1534: %\clearpage
1535: \input{tab2}
1536: \input{tab3}
1537: %\clearpage
1538: The tests show  that Class\,0-Ia and 0-Ib stars  
1539: have statistically different XLFs both in the hard and total bands, 
1540: at a significance level larger than 99\%;
1541: Class\,0-Ia and II are different 
1542: at a significance level of 99\%  for the [0.9-1.2] and [0.5-0.9]\,M$_\odot$ 
1543: mass ranges. The difference is marginal, both
1544: in the total and hard band,  
1545: for the [0.1-0.5]\,M$_\odot$ 
1546: mass range.
1547:  The difference between Class\,II and III objects
1548:  is  quite marginal (between 85\% and 95\%) for the highest mass range, 
1549:  while for lower mass stars, Class\,II and III stars do not show any
1550: significant difference.
1551: %
1552: %
1553: 
1554: These results  clearly indicate an evolution of the X-ray activity
1555: from Class\,0-Ia, the least X-ray luminous objects, up to Class\,III, 
1556: which show the highest X-ray luminosities, at least for stars with mass larger
1557: than 0.9\,M$_\odot$.  Although   
1558: the difference in the intrinsic
1559: X-ray activity between Class\,II and III stars is quite marginal,
1560: it is in agreement with 
1561:  previous results indicating that the X-ray luminosities of accretors in the ONC are smaller
1562:  than those of non accretors \citep{flac03b,prei05b}. Note, however, that these 
1563:   results cannot be compared directly   because of the different 
1564:   criteria adopted to select the two populations.
1565: %\clearpage
1566:  \begin{figure}[!h]
1567: \centerline{\includegraphics[width=9cm]{f10.ps}}
1568: \caption{Absorption corrected
1569: X-ray luminosity  of  Class\,0-Ia stars (squares) and 
1570:  Class\,0-Ib (circles) as a function of the [5.8]-[8.0] (panel a) and 
1571:  [3.6]-[4.5] (panel b) colors.
1572:  Upper limits are indicated by arrows.}
1573: \label{f10}
1574: \end{figure}
1575: %\clearpage
1576: 
1577: The most significant and novel result is the difference in X-ray activity level 
1578: between Class\,0-Ia and Class\,II and 
1579: between Class\,0-Ia and Class\,0-Ib sources.
1580: Our results are even more robust if we include the four objects without 
1581: detections
1582: at 4.5\,$\mu$m.
1583: 
1584:  Based on our selection criterion, Class\,0-Ia objects are protostars with SEDs
1585: increasing up to 8.0\,$\mu$m, while Class\,0-Ib are protostars with SEDs
1586: increasing up to 4.5\,$\mu$m and decreasing at longer wavelengths. 
1587: %Following
1588: %\citet{alle04}, we ascribe
1589: %the decrease of the SED 
1590: %to a different density of the circumstellar envelope.  In this
1591: %sense, Class\,0-Ib stars may be in a different evolutionary phase  
1592: %with respect to Class\,0-Ia stars.
1593: %
1594:  Fig.\,\ref{f10}a shows the absorption corrected
1595: X-ray luminosities plotted vs. the [5.8]-[8.0] color, that 
1596: discriminates Class\,0-Ia stars (squares) 
1597: from Class\,0-Ib ones (filled circles), the  former
1598: being redder than [5.8]-[8.0]$\simeq$1.1 and the latter bluer than this value.
1599: This plot indicates a trend of increasing X-ray luminosity with decreasing  
1600: [5.8]-[8.0] color. %, i.e., in our above interpretation,
1601: %with a different envelope density. 
1602: Fig.\,\ref{f10}b shows the same luminosities as a function of the [3.6]-[4.5]
1603: color.
1604:  While this latter color does not discriminate between Class\,0-Ia and
1605: 0-Ib objects, it clearly confirms that, from the X-ray point of view,
1606: Class\,0-Ia and 0-Ib  objects belong to two different populations. This again
1607: suggests us
1608: that the contamination by reddened Class\,II objects can be very high 
1609: for Class\,0-Ib
1610: objects, while the Class\,0-Ia sample is dominate by true protostars. 
1611: We could then be witnessing the onset
1612: of X-ray activity in very young YSOs.
1613: 
1614: Class\,0-I objects are affected by larger absorptions with respect to the other
1615: classes as we will confirm in Sect.\,\ref{gas_dust} %(Fig.\,\ref{f12}) 
1616: using the X-ray derived
1617: N$_{\rm H}$ values. Can this fact results in a systematic underestimation of 
1618:  L$_{\rm X}$ values for Class 0-I stars? Although we do apply absorption 
1619: corrections  to
1620: the observed X-ray fluxes when deriving luminosities both in the 0.5-8.0\,keV
1621: and 2.0-8.0\,keV bands, these are corrections to the {\em observed} X-ray
1622: spectra and emission from relatively cool plasma would remain unobserved, and
1623: therefore unaccounted for, in heavily observed sources.
1624: 
1625: In order to verify that the X-ray luminosities of heavily absorbed Class\,0-I
1626: objects are accurately corrected for absorption, we assumed that these sources
1627: have X-ray spectra similar to  Class\,II and III objects, and computed through
1628: extensive simulations the effect of varying the absorbing hydrogen column
1629: density on the determination of X-ray luminosities.
1630: 
1631: For this purpose we considered the best fit spectra of a subsample of $\sim$200
1632: Class\,II and III objects with small absorption 
1633: (Log\,N$_{\rm H} <21.5$cm$^{-2}$), consisting of either one or two thermal
1634:  components.  For each
1635: of these spectra, using XPSEC and the {\sc fakeit} command, we computed 10
1636: simulated spectra for each of five different  N$_{\rm H}$ values in the range
1637: $10^{21}$cm$^{-2}<$N$_{\rm H} <10^{23}$cm$^{-2}$. We then analyzed the
1638: simulated spectra with a procedure analogous to that used  by \citet{getm05a}
1639: for the analysis of the observed spectra. Each spectrum was rebinned so to have
1640: a minimum of 15 photon per channel and was then fit with both an isothermal and
1641: a two component model. After adopting the simplest of the two models that gave
1642: a statistically acceptable fit, we computed the unabsorbed flux in the
1643: 0.5-8.0\,keV and 2.0-8.0\,keV bands. We then computed the absorption corrected
1644: XLFs for each of the five absorption values. Most of the simulated sources did
1645: not have enough photons when absorbed by N$_{\rm H} =10^{23}$cm$^{-2}$. In
1646: order to avoid selection effects, since we are not accounting for upper
1647: limits,  we could only compare the XLFs for the 115 sources whose spectra were
1648: bright enough to be analyzed even when considering the highest absorption
1649: value. The result is that the 5  XLFs for the 0.5-8.0\,keV luminosities have
1650: median within $\sim$0.1\,dex but tend to become wider as the $N_H$ is
1651: increased. The five XLFs for 2.0-8.0\,keV luminosities are instead very similar
1652: in shape and with medians within 0.03\,dex, except for the one relative to the
1653: highest absorption values which is {\em brighter} by $\sim$0.1 dex. We conclude
1654: that, in the assumption that Class 0-I sources have the same spectra of Class
1655: II-III ones, the X-ray luminosities we derive are not underestimated.
1656: 
1657: We cannot 
1658: exclude that some of the
1659:   Class\,0-Ib stars are actually heavily absorbed Class\,II stars
1660:   (see Fig.\,\ref{f03}).  However,
1661: even in this case, 
1662: our result would 
1663:   plausibly indicate a significant
1664:   evolution of X-ray activity in the early stages of YSO formation.
1665:   
1666: Due to various IR sensitivity limitations at very high
1667: obscurations, the Class\,0-I sampling  might be restricted to higher masses
1668: (with stronger IR bolometric luminosities), and through the L$_{\rm X}$-mass
1669: correlation, would have higher L$_{\rm X}$ values introducing a bias. But we find
1670: that these heavily obscured stars have lower, not higher, L$_{\rm X}$ values, further
1671: supporting our result.
1672: 
1673: Another bias may be introduced by the requirement that all stars considered have
1674: IRAC detections in one or more bands. Therefore, it is possible that lower-mass
1675: (i.e. bolometrically weaker) Class\,III stars have been systematically omitted
1676: from the sample, leaving behind too many higher mass, i.e. stronger L$_{\rm X}$ stars.
1677: But this is not relevant as the XLF comparisons in Figs.\,\ref{f08}
1678: and \ref{f09}
1679: are binned by mass. We might have few 0.1--0.5\,M$_\odot$ Class\,III stars, but
1680: those shown should be  unbiased in L$_{\rm X}$.
1681:   
1682: 
1683: %In our case, however,
1684: %the  ratio of the emission measures
1685: % of the hard and soft  components for Class\,II stars suggests that  the
1686: %hard component dominates.
1687: %Moreover we will see in the next section that the intrinsic source spectra,
1688: %as determined from the fits, do not seem to vary from Class\,0-I
1689: %to Class\,III, in spite of the large differences in average H column densities
1690: %among these  classes. 
1691: %When we compare the X-ray luminosity of Class\,II stars with those of
1692: %Class\,0-I, we are thus comparing in a consistent way the same
1693: %(hard) contribution  of the X-ray emission.
1694: %
1695: %
1696: 
1697: 
1698: Another possible worry is that flares could dominate the flux from some of our 
1699: objects, and our X-ray spectra could pertain to the flaring state rather than
1700: the quiescent state.  We have verified  that this is not the case by using the
1701: technique described in \citet{wolk05} to recompute X-ray spectra and fluxes for
1702: our stars using photons detected only during non-flaring intervals. Fits to
1703: these new spectra yield XLFs that do not differ significantly  from those shown
1704: in Fig.\,\ref{f08}. We have also verified that using these new spectra
1705: the results discussed in the next section do not change significantly.
1706: 
1707: We note that our results are consistent with the non-detection of class\,0
1708: sources of Serpens \citep{giar07}. In fact, their limiting sensitivity, 
1709: (4.0$\times 10^{29}$\,erg/s in the hard band) estimated
1710: assuming temperature and absorption similar to those we found for our X-ray
1711: detected Class\,0-I sources, corresponds to the highest tenth percentile of the
1712: XLF of our Class\,0-Ia stars. Considering the small size of the Serpens 
1713: Class\,0 star sample, their non-detection is not surprising.
1714: %
1715: \subsection{Gas and dust absorption\label{gas_dust}}
1716: Figure\,\ref{f11} shows the
1717:  {\it Spitzer}/IRAC [3.6]-[4.5] color vs. {\it Chandra} X-ray median
1718: energy of all selected sources. The typical uncertainty  in the median energy 
1719: has been estimated using the results of the MARX simulations made in 
1720: \citet{getm06},
1721: where the median energy errors were computed as a function of the source net 
1722: counts. We used the value of 0.3\,keV, corresponding to objects with 
1723: MedE$\gtrsim$2\,keV and with about 30 net counts which is  the 
1724: minimum of net counts we have for Class\,0-I objects. The typical 
1725: uncertainty  in the IRAC
1726: color  has been estimated assuming a 5\% uncertainty for the IRAC magnitudes.
1727: %\clearpage
1728: \begin{figure}[!h] 
1729: \centerline{\includegraphics[width=9cm]{f11.ps}}
1730: \caption{{\it Spitzer}/IRAC [3.6]-[4.5] color vs. {\it Chandra} X-ray median
1731: energy of all selected sources. The different classes ONC objects
1732: are indicated using the same symbols of Fig.\,\ref{f05}, while filled
1733: squares are the values of the YSO in IC\,1396N taken from \citet{getm07}.
1734: Typical error bars in the median energy 
1735: and in the IRAC colors are also shown.} 
1736: \label{f11}
1737: \end{figure}
1738: %\clearpage
1739: 
1740: As discussed in \citet{getm07}, high X-ray median energy 
1741: (MedE$\gtrsim1.7$\,keV) can be considered as
1742: an indicator of absorbing column density that arises from gas,
1743: primarily metallic elements
1744:  (e.g. N, O, Ne, Mg, Si, Ar, Fe, etc.). On the other hand, 
1745: if the object is known to be a protostar, the [3.6]-[4.5] color
1746: is primarily  a measure of the density of grains in the local infalling envelope.
1747: Thus, the correlation between these two quantities, first found in 
1748: \citet{getm07}, has been interpreted by these authors as evidence of absorption
1749: from the protostellar envelope rather than from the 
1750: ambient molecular cloud material. 
1751:  Although our data do not cover the whole  X-ray
1752: median energy, they show a trend  consistent with that found using the IC\,1396N
1753: data, also shown in Fig.\,\ref{f11}.  Nevertheless, the 
1754: interpretation of our data is not as clear as it was in IC1396N,
1755: since the ONC background is quite complex while
1756:  the maximum line-of-sight column is A$_V\sim10$\,mag.
1757: 
1758: Fig.\,\ref{f12}    
1759:  compares the absorbing column density N$_{\rm H}$
1760: for the  different populations. 
1761: In the left panel, log\,N$_{\rm H}$  
1762: is plotted for the X-ray detected objects as a function of the unabsorbed X-ray
1763: luminosity L$_{t,c}$.
1764: The right panel shows, for each
1765: class, the median values of the absorbing column density distributions,
1766: values between the first and third quartiles and 
1767: values between the minimum and the maximum of the distributions of the 
1768: log\,N$_{\rm H}$ values. Note that only objects with   
1769: errors in log\,N$_{\rm H}$ smaller than 0.5 have been considered. However, the
1770: median values of the distribution do not change significantly if all data are
1771: considered. 
1772: 
1773: A clear trend of increasing absorbing column density from Class\,III to
1774:  Class\,0-I can be seen both  in the left and 
1775: right panels. This trend is
1776: expected since Class\,0-I objects are characterized by a large amount of 
1777: infalling material and are also typically embedded in the molecular cloud; 
1778: a smaller amount of obscuring material is instead present for Class\,II
1779: objects that have shed their envelope and 
1780: formed a circumstellar disk, while the extinction suffered
1781: from Class\,III objects is   due to  interstellar material only
1782: (both within the cloud  and on the line of sight to the cloud). 
1783:  %
1784: %\clearpage
1785: \begin{figure*} 
1786: \centerline{\includegraphics[width=16cm]{f12.ps}}
1787: %\centerline{\includegraphics[width=16cm]{f12_av.ps}}
1788: %\centerline{\includegraphics[width=16cm]{f12_ch.ps}}
1789: \caption{Left panel shows the log\,N$_{\rm H}$  
1790: values as a function of the unabsorbed X-ray
1791: luminosity L$_{t,c}$ for the X-ray detected objects. 
1792: Symbols are as in Fig.\,\ref{f05}.
1793: Right panel shows, for each
1794: class, the median values of the absorbing column density distributions
1795: indicated by the big black dots.   
1796: Solid bars indicate the values of log\,N$_{\rm H}$
1797: included between the first and third quartiles, 
1798: while the dashed lines (partially hidden by solid bars) 
1799: indicate the values included between the minimum and
1800: the maximum of the distributions of the 
1801: log\,N$_{\rm H}$ values.
1802:  Note that only objects with errors in log\,N$_{\rm H}$ smaller
1803: than 0.5 were considered.} 
1804: \label{f12}
1805: \end{figure*}
1806: %\clearpage
1807: 
1808: We  note that the estimate of the N$_{\rm H}$  values of Class\,0-I
1809: stars is based primarily on 
1810: the hard component of the
1811:  X-ray spectrum.   That is, at the temperatures of the protostars 
1812:  (kT$_{\rm av}\sim2$\,keV)
1813:   and with log N$_{\rm H}\sim23$\,cm$^{-2}$,
1814:  the observed flux is almost entirely due to  the hard component 
1815:  of the X-ray spectrum.
1816:  At the same temperatures, if the N$_{\rm H}$  was
1817:  smaller, the total flux
1818:  would be distributed  between the  soft and hard component and the shape of
1819:  the spectrum would be different.
1820:  Our technique should thus be the most accurate for objects with high
1821:   N$_{\rm H}$ values -- thus yielding particularly good  N$_{\rm H}$ values
1822:   for the, on average, heavily embedded protostars.
1823:  Therefore, we argue that the low XLFs for protostars we have derived
1824:  (see Figure 9) are unlikely to be due to errors in our estimates
1825:  of the X-ray absorption towards these objects, as confirmed from the analysis
1826:  of the hard XLFs and the simulated spectra of Class\,II and III (see
1827:  Sect.\,\ref{xlf_section}).
1828: 
1829: % KT
1830: %
1831: \subsection{Evolution of YSO X-ray spectra and temporal characteristics}
1832: % NH
1833: Fig.\,\ref{f13} shows 
1834: the results of the comparison  of the average 
1835: plasma temperature $k$T$_{\rm av}$ amongst the
1836: different classes.  For X-ray spectra fit with two isothermal
1837: components, $k$T$_{\rm av}$ was 
1838: computed as the mean of the soft and hard $k$T spectral components,
1839: weighted by their respective emission  measures; 
1840: for spectra fit with a single component $k$T$_{\rm av}$=$k$T.
1841: The  $k$T$_{\rm av}$ values for the X-ray detected objects
1842: of the different populations
1843: are plotted as a function of the unabsorbed X-ray luminosity L$_{t,c}$.
1844:  Note that only objects with  
1845: errors on $k$T$_{\rm av}$ smaller than 1\,keV have been considered.
1846: However the
1847:  distributions do not change significantly if all data are
1848: considered.  
1849:  The bottom-right panel of Fig.\,\ref{f13} shows, for each
1850: class, the median values of the average plasma temperatures, 
1851: values between 
1852: the first and third quartiles and values between the minimum and the 
1853: maximum of the distributions of the 
1854: $k$T$_{\rm av}$ values. No trend in plasma energy is found as a function of
1855: evolutionary stage from Class\,0-Ia to Class\,III systems.
1856: 
1857: A trend of increasing of the average plasma temperature with 
1858: increasing unabsorbed
1859: X-ray luminosity is clearly evident for all the samples, as has often been
1860: reported for other types of coronal sources
1861: \citep[e.g.][]{schm90}. A similar result is obtained 
1862: by considering only the subsample of stars with X-ray spectra fit with 
1863: two temperature components:
1864: high luminosity stars tend to have high values of the ratio
1865: between the high and low temperature emission measures (EM$_2$/EM$_1\sim 2.5$),
1866:  suggesting that the hot plasma component dominates over the cold one
1867: for  the high luminosity stars. 
1868: No significant difference between the  classes is observed in 
1869: Fig.\,\ref{f13}.
1870: 
1871: Finally we checked if ONC stars of different classes  show different
1872:   X-ray temporal behavior. To do this, we used the results of
1873:   the nonparametric one-sample Kolmogorov-Smirnov (K-S) test performed by
1874:   \citet{getm05a}. The results of the K-S test, P$_{\rm KS}$, strongly depend on the source
1875:   count-statistics and therefore on their count rate. 
1876:   For this
1877:   reason, we compared the probabilities of variability P$_{\rm var}$=(1-P$_{\rm KS}$)
1878:   for the  different
1879:   classes of stars with count rates in definite ranges. 
1880:  The results are given in Table\,\ref{probks_var} where, for each
1881: class and for a given range of Count Rate (CR), we report  the number of stars with 
1882: P$_{\rm var}>99$\%  with respect to the total number of stars of each
1883: class in the corresponding range of count rates. The percentage of variable
1884: stars is also indicated within brackets. Even if the statistics for each sample
1885: is quite poor, the percentages indicate that the  different classes  
1886: have very similar fractions of variable stars.
1887: In practice, when observed with high statistics, all the sources are
1888: variable, independently  on their class.
1889: %\clearpage
1890: \input{tab4}
1891: %\clearpage
1892: %
1893: \begin{figure*} 
1894: \centerline{\includegraphics[width=16cm]{f13.ps}}
1895: \caption{ kT$_{\rm av}$
1896: values as a function of the unabsorbed X-ray
1897: luminosity L$_{t,c}$ for the X-ray detected objects, 
1898:  plotted separately for Class\,0-I, Class\,II and Class\,III stars,
1899:   in the first three panels.
1900: Symbols are as in Fig.\,\ref{f05}.
1901: The bottom-right panel shows, for each
1902: class, the median values of  the average plasma temperature distributions
1903:  indicated by the
1904: big black dots. 
1905: Solid bars indicate values of the temperature distributions between 
1906: the first and third quartiles, while  dashed lines  (partially hidden by 
1907: solid bars)
1908:  indicate values between the minimum and the maximum 
1909:  of the distributions of the
1910:  kT$_{\rm av}$ values. 
1911:  Note that only objects with errors 
1912: in kT$_{\rm av}$ smaller than 1\,keV were considered.} 
1913: \label{f13}
1914: \end{figure*}
1915: %\clearpage
1916: %
1917: \section{CLASS\,0-I: COUNTERPARTS AT OTHER WAVELENGTHS\label{discussion}}
1918: We list in Table\,\ref{classI_matches}  the counterparts of the  Class\,0-I
1919: objects identified in this work that have been detected in 
1920: X-rays (the COUP survey), JHKL bands
1921:  \citep[][ hereafter LMH00 and MLL02, respectively]{lada00,muen02},
1922: 10 and 20\,$\mu$m \citep[][ hereafter RBP5]{robb05a} and in the 
1923: 11.7\,$\mu$m  \citep[][ hereafter SBS05]{smit05}.
1924: None of our candidate Class\,0-I sources
1925: are in the list of 1.3\,cm radio sources of \citet{zapa04b}.
1926: 
1927: The JHKL photometric survey from \citet{lada00}  included 391 stars
1928:  in the central 6\Min55$\times$6\Min55
1929: of the Trapezium cluster. By selecting stars with $K-L>$1.5, they obtained a list
1930: of 78 candidate protostars. We correlated this list with our catalog using a matching
1931: radius of 1\Sec2. As expected, the \citet{lada00}   sample, selected on the basis of
1932: the $K-L$ color, is larger
1933: than our sample for which the selection requires at least two colors computed using the K 
1934: and the four IRAC magnitudes. For this
1935: reason, we find that only 4  of the 78 \citet{lada00} candidate protostars 
1936: are in common with our Class\,0-Ia stars and 9 with our Class\,0-Ib stars.
1937:  Among the remaining 65 objects, 16 cannot be
1938: Class\,0-I objects since they have an
1939: optical counterpart in our catalog and 27 have
1940: [3.6]-[4.5]$<$0.7 and are therefore consistent with star-disk systems (Class\,II)
1941: rather than protostars; for the remaining 22 \citet{lada00} candidate protostars,
1942: the available magnitudes do not
1943: satisfy our selection criteria.
1944: 
1945: Among our Class\,0-I objects, only 5 (Class\,0-Ib) have been detected at 
1946:  10 and 20\,$\mu$m by \citet{robb05a}; of these 5, 2 are in the list of 
1947: the   embedded objects previously studied in \citet[][ hereafter SBS04]{smit04}  
1948: in the OMC-1S region  at  8.8, 11.7 and 18.8\,$\mu$m.
1949: 
1950: %\clearpage 
1951: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1952: \begin{figure*} 
1953: \centerline{\includegraphics[width=16cm]{f14.ps}}
1954: \caption{SED of Class\,0-I candidate objects in the \citet{robb05a} list
1955: with photometry at 10 and/or 20\,$\mu$m. Dots indicate fluxes computed
1956: used the magnitudes from our catalog,   plus symbols correspond to
1957:  the magnitudes at 10 and 20\,$\mu$m from the \citet{robb05a} catalog and
1958:  {\tt X} correspond to the magnitudes at 8.8, 11.7 and 18.8\,$\mu$m from the
1959:  \citet{smit04} catalog. Multiple values at 10\,$\mu$m are 
1960:  relative to the two different nights of observation in \citet{robb05a}.}
1961: \label{f14}
1962: \end{figure*}
1963:  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1964: %\clearpage
1965: 
1966: We interpolated the flux conversion coefficients given in 
1967: \citet{lada06} to  10 and 20\,$\mu$m and we computed the corresponding
1968: fluxes. SEDs with all the available fluxes for these 5 objects are shown in 
1969: Fig.\,\ref{f14}.  
1970: %Most of  the matched objects
1971: %show a depression in the SED around 10\,$\mu$m that can be ascribed 
1972: %to the silicate absorption feature, 
1973: %while fluxes at 18.8 and 20\,$\mu$m, when known, correspond
1974: %to their having a rising SED, consistent with that expected
1975: %for Class\,0-I objects. 
1976: 
1977: In order to understand the properties of our sample of Class\,0-I objects,
1978: we also considered the compact radio sources detected by \citet{zapa04b} and 
1979: \citet{zapa04a} 
1980: using, respectively,  1.3\,cm (in a region of 30\sec$\times$30\sec around
1981:  OMC-1S) and 3.6\,cm observations 
1982:  (in a region of 4\min$\times$4\min\ around the Trapezium cluster) 
1983:  performed with the Very Large Array. 
1984: With a matching radius of
1985: 1\sec, we retrieved in our catalog  7 of the 11 radio sources at  1.3\,cm
1986: and 67 of the 77 compact     3.6\,cm  sources. %Of these latter,
1987:  %62 have an optical/NIR counterpart and 6 have only X-ray detection; 
1988: %some of these objects have colors consistent with those of Class\,II objects.
1989: %but there are also objects with significant  excesses in the [4.5]-[5.8] color.
1990: %  Among  the 9 radio sources that \citet{zapa04a} indicate as 
1991: %not having a counterpart from  the
1992: %literature, we have matched with our catalog  the  \citet{zapa04a} sources
1993: %1, 15, 25, 26, 35. 
1994: 
1995: None of the \citet{zapa04a,zapa04b}  matched sources is in our list of Class\,0-I 
1996: objects. The absence of 1\,cm radio counterparts in our protostar sample
1997: might imply that 
1998: the radio sources are mainly magnetically active non-thermal Class\,III stars
1999: rather than thermal protostellar outflows.
2000: %Fig.\,\ref{trapezium_ch3} shows a portion of the OMC-1s IRAC image at 
2001: %5.8\,$\mu$m where the Class\,0-I stars falling in this region are indicated
2002: %by {\it red circles}; {\it green circles} indicate those common with the
2003: %\citet{robb05a} catalog, while the {\it red X symbols} indicate the 3 Class\,0-I
2004: %sources detected also at 1.3\,cm by \citet{zapa04b}.  
2005: 
2006: %We note that our Class\,0-Ia source %3023 old id
2007: % corresponds to 
2008: %the ``optical outflow source"  
2009: %\citep{odel03}  while the Class\,0-Ib source
2010: % 241 has a high velocity (100\,km s$^{-1}$) 
2011: % CO outflow \citep{zapa04a}. This latter star,    
2012: % corresponding to the COUP source  554, has been
2013: % interpreted in \citet{gros05} as embedded T Tauri stars rather than
2014: % protostar.  The SED of this star 
2015: % (see Fig.\,\ref{robb05a}), 
2016: % extended up to 20\,$\mu$m, does not allow us to exclude its being
2017: %   a Class\,0-I object.
2018:  %For the same reason, we believe
2019: % that all the other Class\,0-I candidates, having SEDs similar, are objects in
2020: % a protostellar phase or  embedded Class\,II  stars. 
2021: %Our list includes
2022: % the COUP sources 555 and 632, corresponding to our sources 12 and 44,
2023: % respectively, with the former object identified as an embedded 
2024: %Class\,II star, and the latter 
2025: % as an evolved protostar in \citet{gros05}.
2026: Note that the north-east region of the COUP FOV, where most of our candidate
2027:  Class\,0-I are concentrated, is not included in the FOV of the literature sources
2028:  mentioned in this section. For this reason, the SEDs of Class\,0-I objects in
2029:  this region cannot yet be extended to wavelengths longer than 8\,$\mu$m, which
2030:  is crucial to confirm (or refute) their protostellar nature.
2031: 
2032: Finally, our list of candidate Class\,0-I protostars includes only two objects
2033: (COUP sources 420 and 484) of the list of stars  studied in \citet{gros05}, 
2034: located at the bottom edge of the OMC-1S region, i.e., far from
2035: the density peak and its associated nebular emission.
2036:  We know that the absence of 
2037: IRAC protostars near %$\theta^1$C \,Ori 
2038: OMC-1s and in the BN/KL region is due to nebular contamination and crowding, 
2039: not to the real absence of protostars in these  two high density regions. In
2040: addition, we cannot classify
2041:  the COUP sources 582, 594, 615, 633, 641, 659 and 667, indicated in
2042:  \citet{gros05} as embedded YSOs in a subcluster, since they do not have any
2043:  counterpart    at the IRAC bands.
2044:  
2045:  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2046: %\begin{figure} 
2047: %\centerline{\includegraphics[width=9cm]{trapezium_ch3.ps}}
2048: %\caption{Portion of the OMC-1s IRAC image at 5.8\,$\mu$m; .}
2049: %\label{trapezium_ch3}
2050: %\end{figure}
2051:  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2052: 
2053: %
2054: \section{SUMMARY AND CONCLUSIONS}
2055: 
2056: We have selected a sample of 45 candidate
2057:  Class\,0-I objects in the ONC field observed in the X-rays by COUP
2058: using new deep $JHK$, [3.6], [4.5], [5.8] and [8.0] photometry obtained at
2059: the ISPI@4m CTIO telescope and from {\it Spitzer} IRAC observations. New
2060: deep optical observations taken with the WFPC2 camera of HST and the WFI camera 
2061: of ESO\,2.2\,m
2062: were also used to reject candidate Class\,0-I objects with optical counterparts.
2063: 
2064: We distinguish between candidate protostars 
2065: with SEDs increasing monotonically from K up to 8\,$\mu$m, indicated as 
2066: Class\,0-Ia, and 
2067: candidate protostars with SED increasing from K up to 4.5\,$\mu$m and 
2068: decreasing at longer
2069:  wavelengths, indicated as Class\,0-Ib, since  the different slope of the SED
2070:  at wavelengths larger than 4.5\,$\mu$m could be an
2071: indicator of a different density in the circumstellar material. 
2072:  
2073: Out of the 23 Class\,0-Ia stars, 10 have been detected in the X-rays with COUP,
2074:  while out the 22 Class\,0-Ib stars, 18 have been  detected; 
2075: we are thus in the position to study, for the first time,
2076:  the X-ray properties of a significant and homogeneous sample of  candidate
2077: protostars. 
2078: 
2079: Samples of bona fide more evolved 
2080: Class\,II and III stars in the same region
2081:   have also been selected in order to study the time and mass evolution of
2082:     X-ray activity.
2083: 
2084: Our principal result is that
2085: Class\,0-Ia stars are significantly less luminous in X-rays, both in the total
2086: and hard bands, than more
2087: evolved Class\,II stars with mass larger than 0.5\,M$_\odot$, 
2088: while Class\,0-Ib stars show X-ray luminosities similar
2089: to Class\,II stars; finally, our data confirm previous determinations 
2090: \citep{neuh97,flac03b,prei05b} that 
2091:  Class\,III stars with masses 
2092: $>$0.3\,M$_\odot$
2093:  are more X-ray luminous than Class\,II objects, even if the result
2094: is quite marginal with our data.
2095:   
2096: Our result that Class\,0-Ia objects are less luminous in X-rays than objects of
2097: other classes, support the hypothesis that the onset of X-ray emission occurs at
2098: a very early stage, when the objects show an increasing  SED at least up to 
2099: 8.0\,$\mu$m. 
2100: The lack of a detailed theory of protostellar evolution does not
2101: allow us to interpret the  difference in the evolutionary stage
2102:  between Class\,0-Ia and 0-Ib protostars.
2103:  Although Class\,0-Ib objects are selected as candidate protostars, our data
2104: suggest that this sample could be dominated by reddened Class\,II stars.  
2105:  If we consider the  X-ray limiting sensitivity in \citet{giar07}, we deduce
2106:   that our results are in agreement with the non-detection of
2107:  Class\,0 protostars in Serpens. 
2108: %is density of the envelope is low enough
2109: %(log\,$\rho \lesssim -13.5 {\rm g\,cm}^{-3}$) that the SED . 
2110:   
2111: The X-ray spectral properties of Class\,0-Ia and 0-Ib 
2112: are similar to those of the more evolved
2113: Class\,II and III objects except for a larger
2114: absorption likely due to the enhanced local density surrounding protostellar
2115: objects.
2116: The  different evolutionary classes also show  similar X-ray variability 
2117: characteristics and the protostellar variability is indistinguishable from
2118: T\,Tauri variability. This result supports the XEST determination that
2119: variability of Class\,II and III stars are the same \citep{stel07} and the COUP
2120: determination that T\,Tauri M stars have the same variability as more massive
2121: stars \citep{cara07}.
2122: Also related is the COUP result that YSOs with high X-ray luminosities have the
2123: same first ionization potential-related elemental abundances as older flaring
2124: stars \citep{magg07}.
2125: 
2126:  All of these results support the general theme that there
2127: is only one mechanism in YSO X-ray production at energies $\gtrsim$1\,keV, 
2128: and that this mechanism arises
2129: from stellar magnetic activity which does not depend on accretion or disks.
2130: The various recent arguments 
2131: \citep{kast02,schm05,gude07,argi07,huen07,gros07} that accretion accounts for some
2132: X-ray emission and line ratios, probably only refers to soft components which
2133: are not accessible to our study \citep{feig07}.
2134: \acknowledgments
2135: This work is based on observations made with {\it Hubble} Space Telescope and 
2136: the {\it Spitzer} Space Telescope,
2137: which is operated by the Jet Propulsion Laboratory (JPL), California Institute
2138: of Technology under a contract with NASA.
2139: Partial support for this work was provided by NASA through an award issued
2140: by JPL/Caltech, by ASI and INAF. EDF is supported by NASA contract NAS8-38252.
2141: This research has made use of NASA's Astrophysics Data System (ADS)
2142: Abstract Service, and of the SIMBAD database, operated at CDS,
2143: Strasbourg, France.
2144: This research has also made use of data products from the Two Micron
2145: All-Sky Survey (2MASS), which is a joint project of the University of
2146: Massachusetts and the Infrared Processing and Analysis Center, funded by
2147: the National Aeronautics and Space Administration and the National Science
2148: Foundation.  These data were served by the NASA/IPAC Infrared Science
2149: Archive, which is operated by the Jet Propulsion Laboratory, California
2150: Institute of Technology, under contract with the National Aeronautics and
2151: Space Administration.
2152: We wish to thank Francesco Damiani and Konstantin Getman for helpfull
2153: discussions. We also thank the anonymous referee for suggestions that helped
2154: improve this paper. 
2155: \bibliographystyle{aa} 
2156: \bibliography{orion}
2157: 
2158: %\clearpage
2159: %\pagestyle{empty}
2160: %\setlength{\voffset}{48mm}
2161: \input{tab5}
2162: \input{tab6}
2163: \input{tab7}
2164: \input{tab8}
2165: %\clearpage
2166: %\pagestyle{plaintop}
2167: %\setlength{\voffset}{0mm}
2168: 
2169: %\input{../TABLES/classIa_matches}
2170: %\input{../TABLES/classIb_matches}
2171: %
2172: %\input{../TABLES/coup_cl0I_1}
2173: %\input{../TABLES/coup_cl0I_2}
2174: \input{tab9}
2175: %\clearpage
2176: %\thispagestyle{empty}
2177: \input{tab10}
2178: %\clearpage
2179: \input{tab11}
2180: \input{tab12}
2181: %\clearpage
2182: %\thispagestyle{empty}
2183: \input{tab13}
2184: %\clearpage
2185: \input{tab14}
2186: 
2187: \end{document}
2188: 
2189: %%
2190: %% End of file `sample.tex'.
2191: