Revealing the "missing" low-mass stars in the S254-S258 star forming region by deep X-ray imaging
aa r X i v : . [ a s t r o - ph . GA ] J un Astronomy&Astrophysicsmanuscript no. 17074 c (cid:13)
ESO 2018July 25, 2018
Revealing the “missing” low-mass stars in the S254-S258 starforming region by deep X-ray imaging ⋆ P. Mucciarelli , , T. Preibisch , and H. Zinnecker , , Universit¨ats-Sternwarte M¨unchen, Ludwig-Maximilians-Universit¨at, Scheinerstr. 1, 81679 M¨unchen, Germanye-mail: [email protected],[email protected] Exzellenzcluster Universe, Boltzmannstr. 2, 85748 Garching, Germany Astrophysikalisches Institut Potsdam, An der Sternwarte 16, 14482 Potsdam, Germany Deutsches SOFIA Institut, Universit¨at Stuttgart, Pfa ff enwaldring 31, 70569 Stuttgart, Germany NASA-Ames Research Center, MS 211-3, Mo ff ett Field, CA 94035, USAReceived 13 April 2011; accepted 4 June 2011 ABSTRACT
Context.
X-ray observations provide a very good way to reveal the population of young stars in star forming regions avoiding thebiases introduced when selecting samples based on infrared excess.
Aims.
The aim of this study was to find an explanation for the remarkable morphology of the central part of the S254–S258 starforming complex, where a dense embedded cluster of very young stellar objects (S255-IR) is sandwiched between the two H IIregions S255 and S257. This interesting configuration had led to di ff erent speculations such as dynamical ejection of the B-starsfrom the central cluster or triggered star formation in a cloud that was swept up in the collision zone between the two expandingH II regions. The presence or absence, and the spatial distribution of low-mass stars associated with these B-stars can discriminatebetween the possible scenarios. Methods.
We performed a deep
Chandra
X-ray observation of the S254–S258 region in order to e ffi ciently discriminate young stars(with and without circumstellar matter) from the numerous older field stars in the area. Results.
We detected 364 X-ray point sources in a 17 ′ × ′ field ( ≈ × ∼ . M ⊙ . A clustering analysis identifies three significant clusters:the central embedded cluster S255-IR and two smaller clusterings in S256 and S258. Sixty-four X-ray sources can be classified asmembers in one of these clusters. After accounting for X-ray background contaminants, this implies that about 250 X-ray sourcesconstitute a widely scattered population of young stars, distributed over the full field-of-view of our X-ray image. This distributedyoung stellar population is considerably larger than the previously known number of non-clustered young stars selected by infraredexcesses. Comparison of the X-ray luminosity function with that of the Orion Nebula Cluster suggests a total population of ∼ Conclusions.
The observed number of ∼
250 X-ray detected distributed young stars agrees well with the expectation for the low-mass population associated to the B-stars in S255 and S257 as predicted by an IMF extrapolation. These results are consistent with thescenario that these two B-stars represent an earlier stellar population and that their expanding H II regions have swept up the centralcloud and trigger star formation (i.e. the central embedded cluster S255-IR) therein.
Key words.
Stars: formation, low-mass, pre-main-sequence – X-ray: stars – Galaxy: clusters: individual: S254-S258
1. Introduction
The south-eastern part of the molecular cloud complex in theGem OB1 association contains an embedded star-forming re-gion with several di ff use H II regions (S254–S258, Sharpless1959, see Fig. 1). The most prominent of these H II regions,S255 and S257 (Chopinet et al. 1974), are both powered byB0 stars, have diameters of ∼ ′ , and a projected separa-tion of ∼ ′ . Sandwiched right between them is a densedusty molecular cloud filament (S255-IR or S255-2; Heyer et al.1989; Di Francesco et al. 2008) that contains numerous em-bedded infrared sources (Zinnecker et al. 1993; Howard et al.1997; Itoh et al. 2001; Longmore et al. 2006; Ojha et al. 2006;Chavarr´ıa et al. 2008). Masers, HH-objects, jets, and molecular ⋆ Tables 2, 3, and 5 are only available in electronic form at theCDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or viahttp: // cdsweb.u-strasbg.fr / cgi-bin / qcat?J / A + A / outflows (Snell & Bally 1986; Miralles et al. 1997; Minier et al.2005, 2007; Goddi et al. 2007; Jiang et al. 2008; Wang et al.2011) provide clear evidence of very recent and ongoing starformation activity in this cloud. The combination of Spitzer mid-infrared observations (Allen et al. 2005) with near-infrared im-ages of a 26 ′ × ′ region led to the detection of 510 sourceswith near- or mid-IR excess (Chavarr´ıa et al. 2008), 87 and 165of which were classified as Class I and Class II sources, respec-tively. The large majority (80%) of these infrared excess-selectedyoung stellar objects (YSOs) were found to be clustered. Thecentral cluster S255-IR is the richest of these. It contains at least
140 infrared excess sources, among them 23 Class I sources.Another large fraction of the known YSO population is locatedin an elongated cluster at the southern edge of S256, and thereare several smaller clusters at di ff erent locations (see Fig. 12 inChavarr´ıa et al. 2008). The census is incomplete because the
Spitzer images of the densecluster su ff er from source crowding and saturation e ff ects. Mucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region About 1 ′ north of the center of S255-IR, a strong far-infrared source, S255-N, was detected by Ja ff e et al. (1984).This object is also associated with massive star formationand it is believed to be at an even earlier evolutionary stagethan S255-IR (Kurtz et al. 1994, 2004; Cyganowski et al. 2007;Wang et al. 2011). Another deeply embedded region, S255-S(Wang et al. 2011), is located about 1 ′ south-west from S255-IR. It exhibits strong mm continuum emission (Wang et al. 2011;Di Francesco et al. 2008) but no other sign of active star forma-tion in near- and mid-IR observations. Minier et al. (2007) sug-gested that this sub-region is in a very early pre-stellar phaseof evolution. Scenarios for the spatial and temporal sequenceof star formation in the S254-S258 complex have been recentlydiscussed in Chavarr´ıa et al. (2008), Bieging et al. (2009), andWang et al. (2011).The distance of the S254-S258 complex was only poorlyknown until recently. Pismis & Hasse (1976) and Mo ff at et al.(1979) derived a value of 2.5 kpc; similar values (e.g.,Chavarr´ıa et al. 2008), but also lower values down to 1.5 kpcwere used in later studies. Rygl et al. (2010) performed high-precision astrometry using the 6.7 GHz methanol maser emis-sion from the source J0613 + . ± .
07) kpc for S255.We will use this new and reliable distance for our present study.
The most remarkable part of the S254-S258 complex and the fo-cus of the study presented here is the central region around thetwo H II regions S255 and S257. A very interesting, as yet un-explained, feature of this region is that the two B0 stars excitingthese H II regions, ALS 19 and HD 253327, appear to be moreor less isolated and do not have any obvious co-spatial low-massclusters (Zinnecker et al. 1993). This is remarkable because, ac-cording to the standard field star IMF (e.g., Kroupa 2001) eachB0 star ( M ∗ ≈ M ⊙ ) should be accompanied by about 300lower-mass stars. Fundamentally di ff erent possible explanationsfor the apparent absence of low-mass stars around these B0 starshave been proposed over the years.One scenario is based on the assumption of bimodal starformation . It assumes that the two high-mass B0 stars formedindependently from the low-mass young stars in the centralcluster and in a fundamentally di ff erent processes. The lack ofclusters of low-mass stars around these two B0 stars wouldthen imply that these high-mass stars formed in isolation (seeZinnecker et al. 1993). One problem with this scenario is that inbasically all other well investigated star forming regions, high-mass stars are always associated with large numbers of low-massstars (Testi et al. 1999; Brice˜no et al. 2007). The case of S255and S257 would then represent a quite unique exception if theabsence of low-mass stars was confirmed.The second scenario assumes that the two B0 stars formedtogether with the low-mass stars in the dense central clus-ter S255-IR, but were dynamically ejected , e.g. by means ofclose stellar encounters and N-body interactions. This modelpredicts that both B0 stars should move away from the cen-tral clusters with substantial velocities. Unfortunately, the avail-able HIPPARCOS proper-motions for the stars are not accurateenough to either support or rule out this scenario.A third scenario assumes multiple stellar generations andtriggered star formation . Here, the two B0 stars belong to anearlier generation of stars that formed several Myr ago in thisarea. The expanding H II regions swept up di ff use gas and dustin their surroundings into shells, and formed the dense cloud in the interaction zone between them. This process of creating newclouds at the edges of shells or bubbles driven by high-mass starsis well established and observed at many locations (see, e.g.,Brand et al. 2011; Zavagno et al. 2010; Deharveng et al. 2009).The particularly strong compression of the cloud at the intersec-tion of the two shells, caused by the ongoing expansion of theH II regions, may have triggered the formation of a new genera-tion of stars, i.e. the embedded cluster of young stellar objects.The ongoing expansion of the interacting bubbles would alsoprovide a natural explanation why the youngest regions S255-Nand S255-S are found just above and below the central youngcluster S255-IR at the intersection of the shells. A discriminant between the di ff erent evolution scenarios forthe S255 /
257 region is the presence or absence of low-massstars associated with the two B0 stars. While no low-mass starsshould be present in the case of the bimodal star formationmodel or the dynamical ejection model, the multi-generationmodel predicts the presence of several hundred low-mass starsnear these B0 stars, since the stellar populations in basically allwell-investigated OB associations follow the standard field starIMF (Brice˜no et al. 2007). A handful emission line stars and afew dozen infrared excess objects (see Chavarr´ıa et al. 2008) areknown inside or near the two H II regions, but their numbersare far to small for the expected low-mass population associatedwith the massive B0 stars.The apparent lack of associated low-mass stars may, how-ever, just be a result of the sensitivity limits of the existing ob-servations: a population of several Myr old low-mass stars wouldbe quite hard to identify in the present optical and infrared im-ages, for several reasons. First, the low-mass stars would not bedensely clustered around the B0 stars but scattered over a ratherwide area, up to ∼
10 pc away from the massive stars, as typicalfor subgroups in OB associations. Second, these low-mass starswould be quite hard to see in most existing optical or infraredimages of the region: since a ∼
10 Myr old 1 M ⊙ [0.2 M ⊙ ] starsshould have magnitudes of V ≥ ≥ ff use infrared emission in this region. Third, eventhe availability of Spitzer observations is of limited use here: al-though Spitzer data are sensitive enough to detect a good frac-tion of the low-mass stars, the Spitzer images of the region aredominated by unrelated field stars (note that this region lies veryclose to the galactic plane). The usual approach to identify youngstars by their infrared excesses is not feasible here, because atan age of more than a few Myr, most of the low-mass associ-ation members have already lost their circumstellar disks andthus should not exhibit infrared excesses (Brice˜no et al. 2007). It is thus impossible to identify and distinguish a population ofseveral ( ∼ − ) Myr old low-mass association members fromunrelated field stars with optical or infrared photometry alone. Sensitive X-ray observations can provide a very good solu-tion of this problem, since they allow to detect the young starsby their strong X-ray emission (e.g., Feigelson et al. 2007) ande ffi ciently discriminate them from the numerous older field starsin the survey area. The median X-ray luminosity of < ∼
10 Myrold solar-mass stars is ≈ . erg s − ; this is nearly 1000 timeshigher than for solar-mass field stars (see Preibisch & Feigelson2005), and makes these young stars relatively easily detectableX-ray sources. Another very important aspect is that X-ray ob-servations trace magnetic activity rather than photospheric or cir-cumstellar disk emission from young stars, and are thus comple- ucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region 3 mentary to the available optical and infrared data of the region.The X-ray selected sample of low-mass stars will be not biasedtoward stars with circumstellar disks identified in the Spitzerdata. Furthermore, an X-ray image is not subject to confusionfrom bright di ff use emission by heated gas and dust. X-rayscan penetrate deeply into obscuring material and are very e ff ec-tive in detecting embedded YSOs (Getman et al. 2005a). ManyX-ray studies of star forming regions have demonstrated thesuccess of this method (see, e.g., Preibisch & Zinnecker 2002;Broos et al. 2007; Forbrich & Preibisch 2007; Townsley et al.2011). Also, the relations between the X-ray properties and ba-sic stellar properties in young stellar populations are now verywell established from very deep X-ray observations such as the Chandra
Orion Ultradeep Project (COUP) (see Getman et al.2005b; Preibisch et al. 2005). To summarize, a deep X-ray im-age of the S254-S258 complex can reveal the full young stellarpopulations in the area and provide essential information aboutthe star formation history.At distances beyond 1 kpc, very good angular resolution isrequired to resolve the individual sources in the dense youngclusters and to allow a reliable identification of the X-ray sourceswith the numerous infrared sources (note that the complex isalmost exactly on the galactic plane, b = − . ◦ ). The Chandra
X-ray observatory, that provides an on-axis PSF of ≤ ′′ , is theonly currently active X-ray mission that has su ffi cient angularresolution for this purpose.We have therefore performed a deep Chandra
X-ray ob-servation of this extraordinary star forming region in order touncover the population of low-mass association members. Ourstudy focuses on the central region of the S254-S258 complex,i.e. the two H II regions S255 and S257 and the embedded clusterS255-IR between them. A characterization of the size, the spatialdistribution, and the properties of the low-mass population canprovide important information on the star formation history anddiscern between the di ff erent models for the relation betweenS255 / S257 and the embedded cluster S255-IR. In Section 2 wedescribe the
Chandra observations and data reduction. Section3 presents the basic X-ray properties of the detected sources.Section 4 analyzes the X-ray population of the S254-S258 starforming complex, and Section 5 discusses the spatial distributionof the X-ray sources and the implications on the star formationprocess in S254-S258. A more detailed analysis of the opticaland infrared properties of the individual X-ray detected youngstars (that can provide direct information on the ages, masses,and the circumstellar disks around these stars) will be presentedin a forthcoming paper.
2. Observations and data reduction
The S254-S258 complex was observed (PI: Th. Preibisch)in November 2009 with the Imaging Array of the
Chandra
Advanced CCD Imaging Spectrometer (ACIS-I). ACIS-I pro-vides a field of view of 17 ′ × ′ on the sky. At the 1.6 kpcdistance of S254-S258 this corresponds to 7 . × . α (J2000) = h m . s , δ = + ◦ ′ ′′ . The observation was performed in the stan-dard “Timed Event, Faint” mode (with 3 × ff -axis angles at these detectors is strongly de-graded, their point-source sensitivity is reduced; only six X-raysources are detected in the field of these ACIS-S chips.The basic data products of our observation are the two Level2 processed event list provided by the pipeline processing atthe Chandra
X-ray Center, that list the arrival time, locationon the detector and energy for each of the 626 140 detected X-ray photons. We combined the two pointings with the merge all script, a
Chandra contributed software that make use of standard
CIAO tools. The mean background count rate in our mergedimage, determined from several large source-free regions, is2 . × − counts s − pixel − , corresponding to a mean back-ground level of 0.02 counts pixel − .At a distance of 1.6 kpc, the expected ACIS point sourcesensitivity limit for a 5-count detection on-axis in a 75 ks ob-servation is L X , min ∼ . erg s − , assuming an extinction of A V ≤ . N H ≤ × cm − ) as typical for the starsin the H II regions, and a thermal plasma with kT = Chandra
Orion Ultradeep Project(Preibisch et al. 2005; Preibisch & Feigelson 2005), we can ex-pect to detect almost all stars in S254-S258 with masses greaterthan 0 . M ⊙ and about half of the 0 . − . M ⊙ stars. The ex-pected level of detection completeness is > ∼
90% for stars with M ∗ ≥ . M ⊙ (corresponding approximately to spectral typesearlier than M1) and drops below 50% at M ∗ ≤ . M ⊙ (spec-tral types < ∼ M5). Note that these values are valid for the centralpart of the observed field; sensitivity is ∼ − The source detection was performed in a two-step process. Thefirst detection step was performed in a rather aggressive man-ner in order to find even the weakest possible sources, deliber-ately accepting some degree of false detections. In the secondstep, this list of potential sources was then cleaned from spuri-ous detections by a detailed individual analysis. We employedthe wavdetect algorithm (Freeman et al. 2002, a
CIAO mexican-hat wavelet source detection tool) for locating X-ray sources inour merged image, and used a rather low detection threshold of10 − . This step was performed in three di ff erent energy bands,the total band [0.5-8.0] keV, the soft band [0.5-2.0] keV, and thehard band [2.0-8.0] keV, and with wavelet scales between 1 and16 pixels. We also performed a visual inspection of the imagesand added some 30 additional candidates to the merged catalogfrom the wavelet analysis, resulting in a final catalog of 511 po-tential X-ray sources.To clean this catalog from spurious sources, we thenperformed a detailed analysis of each individual candidatesource with the ACIS Extract (AE hereafter) software package (Broos et al. 2010). A full description of the procedures usedin AE can be found in Getman et al. (2005b), Townsley et al.(2003) and Broos et al. (2007). The following three steps were Chandra Interactive Analysis of Observations, version 4.2:http: // cxc.harvard.edu / ciao / index.html http: // / xray / docs / TARA / ae users guide.html Mucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region Fig. 1. Top:
Negative grayscale representationof the optical image of the S254-S258 complexfrom the Digitized Sky Survey. The black el-lipses represent the five H II regions that definethe complex. The location of the central, op-tically invisible embedded cluster S255-IR ismarked by the arrow.
Center:
Spitzer
IRAC 4 image of the cen-tral part of the S254-S258 complex. This im-age was created from the basic calibrated dataproducts for the programs 201 and 30784 re-trieved from the
Spitzer archive and mosaickedwith the MOPEX software available from theSpitzer Science Center. Note that parts of thebright emission from the central embeddedcluster S255-IR is saturated in these data.
Bottom:
Chandra
ACIS-I image of S254-S258in the [0.5–8.0] keV band. Blue ellipsoids rep-resent extraction regions for the individual de-tected X-ray sources based on a model of thelocal PSF that encircles 90% of total energy.ucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region 5
Table 1.
Chandra observation log
Obs.Id. Date Start – End Time [UT] Exposure time Level 2 events10983 2009-11-16 11:40:48 – 23:35:08 40 570 s 340 41112022 2009-11-20 05:14:15 – 15:18:20 34 155 s 285 729 performed by AE in order to prune our input catalog from spuri-ous detection (including afterglows):1. Extraction regions were defined as the 90% contours of thelocal PSF (or smaller in the case of other nearby sources),and source events were extracted. Energy dependent correc-tions for the finite extraction regions were applied;2. Local background events were extracted after masking allthe sources in the catalog;3. The Poisson probability ( P B ) associated with the “null hy-pothesis”, i.e. that no source exist and the extracted eventsare solely due to Poisson fluctuations in the local back-ground, is computed for each source.All candidate sources with P B > .
01 were rejected as back-ground fluctuations. After 8 iterations of this pruning procedureour final catalog consisted of 364 sources. It contains 344 pri-mary sources with P B < . . < P B < .
01. The extraction regions for the sources in ourfinal catalog are plotted on the
Chandra image in Fig. 1.
The AE software also determines basic properties for each of thedetected sources, such as the net (i.e., background-subtracted)counts in various energy bands, the median photon energy, sta-tistical test for variability, and a measure of the incident pho-ton flux. These properties are reported in Table 2 (available inthe electronic edition). Sources are sorted by increasing rightascension and identified by their sequence number (Col. 1) ortheir IAU designation (Col. 2). While the general X-ray proper-ties were determined from the merged data set (the AE softwareis well suited for this purpose), we note that the spectra (seeSect. 3.1.3) were extracted from the individual observations.
As in any X-ray observation, there must be some degree ofcontamination by galactic field stars as well as extragalacticsources. To quantify the expected level of this contamination, weconsider the results from the recent
Chandra
Carina ComplexProject (CCCP; see Townsley et al. 2011), for which the indi-vidual pointings had very similar exposure times ( ≈ −
80 ks)as our S254-S258 pointing. Furthermore, S254-S258 is at nearlythe same galactic latitude as the Carina Nebula, suggesting thatthe background contamination should be very similar in thesetwo regions.For the CCCP data set, the classification study of Broos et al.(2011a), which considered the X-ray, optical, and infrared prop-erties of the sources (that di ff er for the di ff erent contaminantclasses), found that 716 X-ray sources in the 1.46 square-degreeCCCP survey are are foreground stars, 16 are background stars,and 877 are extragalactic (AGN) contaminants. Scaling thesenumbers to the field-of-view of our S255 pointing gives 39 fore-ground stars, 1 background star, and 48 extragalactic (AGN)contaminants. However, since S254-S258 is considerably closer(1.6 kpc) than the Carina Nebula (2.3 kpc), the number of fore-ground stars should be accordingly smaller, approximately by a factor of (1 . / . ≈ .
34. Furthermore, the number of fore-ground stars in the Carina Nebula is particularly high since thisdirection is close to the tangent point of the Carina spiral arm.These considerations imply that the contamination in our S254-S258 field should be clearly dominated by ∼
48 expected extra-galactic sources (AGNs).A characteristic of extragalactic contaminants is that theiroptical and infrared counterparts should be very faint, in factmostly undetected in the available optical and infrared images.As we describe in more detail below, our search for counterpartsof the X-ray sources left 46 X-ray sources outside the centralembedded cluster S255-IR without optical or infrared counter-parts. This number agrees well with the expected number of ex-tragalactic contaminants.Assuming at most 10 contaminating foreground / backgroundstars, the total expected number of contaminants would be < ∼ < ∼
3. Properties of the X-ray source in S254-S258
An accurate determination of the intrinsic X-ray source lu-minosities requires good knowledge of the X-ray spectrum.However, for the majority of the X-ray sources the number ofdetected photons is too low for a detailed spectral analysis. Only25 sources in our catalog have more than 80 net counts, thepractical lower limit for meaningful spectral analysis. For thesebright sources we performed a detailed spectral fitting analysisto derive the plasma temperature and the extinction, and fromthese quantities we can calculate the intrinsic (i.e. , extinction-corrected) X-ray luminosities, as described in detail in Section3.1.3.For the weaker X-ray sources, for which a meaningful spec-tral analysis is not feasible, one cannot determine intrinsic X-rayluminosities without knowledge of the extinction. This is a sub-stantial problem because the young stars in the S254-S258 com-plex show a very wide range of extinctions. There are numerousoptically visible stars with low obscuration (at most a few mag-nitudes of visual extinction), while other stars su ff er from cloudextinction up to about A V ∼
20 mag, and embedded YSOs showadditional circumstellar extinctions up to A V ∼
50 mag and be-yond (Chavarr´ıa et al. 2008). This implies that we cannot simplyuse a common count-rate to flux conversion factor to determineintrinsic X-ray luminosities but have to consider each source in-dividually.
An estimate of the observed (i.e. not the intrinsic) X-ray fluxis computed by AE. This quantity, called
FLUX
2, is calculatedfrom the number of detected photons and using a mean value ofthe instrumental e ff ective area (through the Ancillary ResponseFunction, ARF) over energy. The FLUX . −
8] keV range), are reported in column (3) inTable 3 (available in the electronic edition). It should be noted
Mucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region
Fig. 2.
The blue histogram shows the distribution of FLUX2 val-ues for the X-ray sources in S254-S258. For comparison, thedistribution of fluxes from the COUP data from Getman et al.(2005b) is shown by the black histogram, scaled to the distanceof S254-S258, i.e. 1.6 kpc.that this
FLUX ff er from a systematic error with re-spect to the true incident flux, because the use of a mean ARFis only correct in the hypothetical case of a flat incident spec-trum, an assumption that probably not fulfilled. Nevertheless, the FLUX
FLUX
Chandra
Orion UltradeepProject (COUP, see Getman et al. 2005b). Note that the COUPfluxes were scaled to the 1.6 kpc distance of S254-S258. Thetwo distribution show a very similar shape in the range between ≈ − . photons cm − s − and ≈ − . photons cm − s − . Thedi ff erences between the two distributions can be explained asfollows:First, the COUP sample shows a few stars with fluxes of > − . photons cm − s − , while no such very bright sourcesare seen in S254-S258. These very bright sources are the highlyX-ray luminous O-type stars in the Orion Nebula Cluster. Thefact that the S254-S258 does not contain such high-mass starsexplains the absence of similarly high values in the observeddistribution of incident fluxes for S254-S258.Second, the peak and turn-over of the S254-S258 distribu-tion at ≈ − . photons cm − s − is a direct consequence ofthe higher sensitivity limit of our S254-S258 X-ray observation.As S254-S258 is about 4 times more distant than the ONC, andsince the exposure time of our S254-S258 Chandra observationis less than one tenth of the 840 ks COUP observation, the ex-pected sensitivity limit should be about 150 times higher.Third, the number of sources per bin is always lower forS254-S258 compared to the ONC. This suggests that the totalstellar population in the observed part of S254-S258 is smallerthan in the ONC as observed in the COUP.
An estimate of the intrinsic, i.e. extinction corrected, X-ray lu-minosity for sources that are too weak for a detailed spectralanalysis can be obtained with the
XPHOT software , developedby Getman et al. (2010). XPHOT is based on a non-parametricmethod for the calculation of fluxes and absorbing X-ray col-umn densities of weak X-ray sources. X-ray extinction and in-trinsic flux are estimated from the comparison of the apparentmedian energy of the source photons and apparent source fluxwith those of high signal-to-noise spectra that were simulatedusing spectral models characteristic of much brighter sources ofsimilar class previously studied in detail. This method requiresat least 4 net counts per source (in order to determine a meaning-ful value for the median energy) and can thus be applied to 255of our 364 sources. Columns (4) to (7) of Table 2 report appar-ent and intrinsic (corrected for absorption, noted with subscript c ) luminosities in the hard and total band, assuming a distanceof 1.6 kpc. The resulting intrinsic X-ray luminosities range from10 . to 10 . erg s − .Figure 3 shows the distribution of median photon ener-gies and the deduced hydrogen column densities estimated by XPHOT . The median value of the derived hydrogen colum den-sities is log (cid:16) N H [cm − ] (cid:17) = .
04, corresponding to a visual ab-sorption of A V ∼ (cid:16) N H [cm − ] (cid:17) = .
21 ( A V ∼ For the 25 sources in our sample with more than 80 net countswe performed a spectral fitting analysis using AE and the
XSPEC software v12.5 (Arnaud 1996). We used models with one-or two-temperature thermal
VAPEC components (Smith et al.2001) and the
TBABS multiplicative model to describe the ef-fect of extinction by interstellar (and circumstellar) material (asmeasured by the hydrogen column density N H ). The plasmaabundances for the VAPEC components were fixed at the val-ues adopted by the XEST study (G¨udel et al. 2007) to be typi-cal for pre-main sequence stars . For the extinction we used thestandard interstellar abundances in the TBABS model as listedin Wilms et al. (2000). In order to evaluate the goodness ofour fits we choose to apply the C-statistic (a maximum likeli-hood method; Cash (1979); Wachter et al. (1979)) which is bet-ter suited than the classic χ statistic for low-count data.For two sources, (cid:16) N H [cm − ] (cid:17) =
20 to log (cid:16) N H [cm − ] (cid:17) = .
08, cor-responding to visual absorptions between A V < ∼ . A V ∼
65 mag. These values are in good agreement with the es-timates derived with
XPHOT . The median value is 22.04, corre-sponding to A V ∼ ≈ . ∼
15 keV (170 MK). In Table 4 we alsoreport luminosities derived from the spectral fit, assuming a dis- http: // / users / gkosta / XPHOT / The adopted abundances, relative to the solar photospheric abun-dances given by Anders & Grevesse (1989), are: C = = = = = = = = = = = = Fig. 3.
Distributions of median photon energies (left) and hydrogen column densities (right) determined by
XPHOT (right) for thewhole sample (black histograms). The blue dashed histograms show the distributions restricted to the sub-sample of sources locatedin the 1 arcmin radius region centered on the embedded cluster S255-IR.tance of 1.6 kpc. The intrinsic luminosities are calculated fromthe spectral fit parameters, setting extinction to zero. The rangeof extinction-corrected intrinsic X-ray luminosities spans fromlog L t , c = .
56 erg s − to 32.11 erg s − for the full [0.5-8.0] keVband.The X-ray properties can give us some clues about the natureof the sources. The majority of the X-ray sources has plasmatemperatures and X-ray luminosities in the typical ranges foundfor YSOs in other star forming regions (see, e.g., Preibisch et al.2005); together with the fact that most X-ray sources haveclear optical / infrared counterparts, this suggests that these X-ray sources actually are young stars in the S254-S258 complex.However, the sources for which the spectral fit yields extremelyhigh plasma temperatures of kT > . / infrared counterpart, and, third, it islocated at the periphery of the S254-S258 region, well outsidethe boundaries of the molecular clouds. Furthermore, its X-rayspectrum can also be well reproduced with a power-law model,as typical for AGN X-ray sources. Similar arguments apply tosource 30.The other X-ray sources with extremely high plasma tem-peratures are located in dense clouds, have infrared counterparts,and are thus probably deeply embedded very young stellar ob-jects.It is interesting to compare the X-ray luminosities from thespectral fits to those derived with XPHOT . Figure 5 shows thatthe results from these two di ff erent methods agree quite wellfor the majority of cases. Only for four of the highest luminos-ity objects we find that XPHOT seems to systematically over-
Fig. 5.
Comparison of the intrinsic full band [0.5-8.0] keV lu-minosities derived from the spectral fits to the 25 bright sources(Table 4) to the intrinsic full band [0.5-8.0] keV luminosities de-termined with
XPHOT .estimate the X-ray luminosities by ≈ . − . XPHOT are reliable. We therefore will use the
XPHOT re-sults for our analysis of the X-ray luminosity function presentedbelow.
Mucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region −5 −4 −3 C oun t s s ec − k e V − H =0.356 kT=0.601,3.43 Cstat=724 NET_CNTS=175−0.410.5 2 5−5×10 −4 −4 r e s i du a l s Energy (keV) 10 −5 −4 −3 C oun t s s ec − k e V − H =9.69 kT=5.44 Cstat=916 NET_CNTS=374−0.810.5 2 5−5×10 −4 −4 r e s i du a l s Energy (keV)10 −5 −4 −3 C oun t s s ec − k e V − H =2 kT=0.86,8.73 Cstat=953 NET_CNTS=251−0.510.5 2 505×10 −4 r e s i du a l s Energy (keV) 10 −5 −4 −3 C oun t s s ec − k e V − H =3.65 kT=7.91 Cstat=934 NET_CNTS=396−16.310.5 2 505×10 −4 r e s i du a l s Energy (keV)
Fig. 4.
Chandra
X-ray spectra and best-fit models of four bright X-ray sources in S254-258. The crosses show the measured spectra,the solid lines show the best-fit models; in the two cases where a two temperature model was required, the dotted lines show the twoindividual spectral model components. CXOU J061244.17 + + + + A first analysis of the time variability of individual X-ray sourcesis performed by AE by comparing the arrival times of the indi-vidual source photons in each extraction region to a model as-suming temporal uniform count rates. The statistical significancefor variability is quantified computing the 1-sided Kolmogorov-Smirnov statistic (Col. 15 of Table 2). In our sample, 23 sourcesshow significant X-ray variability (probability of being con-stant P const < . . < P const < . Chandra pointings, i.e. at a time di ff erence of about 4 days (see Figure 7). One can seethat the majority of sources show changes in the count rates byless than a factor of 2. Only for 21 sources the count rates dif-fer by more than 3 σ between the two observations. The resultis consistent with the assumption that the X-ray emission fromyoung stellar objects is a superposition of many flares of di ff er-ent amplitude, where weak flares are very frequent while verystrong flares occur more rarely, at rates of about one such eventper week (e.g., Getman et al. 2008).
4. Characteristics of the X-ray stellar population ofS254-S258
In order to identify counterparts of the X-ray sources in otherwavelengths, we used the optical images from the DigitizedSky Survey (DSS), the Two Micron All Sky Survey (2MASS)point source catalog, and the
Spitzer -IRAC catalog fromChavarr´ıa et al. (2008). The results of the cross-correlation arereported in Table 5 (available in the electronic edition). Ourvisual inspection of the red and blue DSS images gave opti-cal counterparts to 95 X-ray sources (i.e. 26% of all 364 X-ray ucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region 9
Fig. 6.
Lightcurves for six significantly variable sources. The solid dots show the arrival time (measured from the start of theobservation) and the energy of each of the detected source photons. The histograms show the corresponding binned lightcurves.
Table 4.
Spectral parameters of brighter sources: the spectral fit was performed with an absorbed thermal plasma model with one( tbabs * vapec ) or two components ( tbabs *( vapec+vapec )). Source CP / CXOU J Net Counts log N H kT kT log L s log L h log L h , c log L t log L t , c No. IR class (counts) (cm − ) (keV) (keV) (erg s − ) (erg s − ) (erg s − ) (erg s − ) (erg s − )(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)1 yes 061147.81 + ± ± + ± ± + ± ± / III 061206.65 + ± ± + ± ± + ± ∗ – 30.62 31.46 31.49 31.52 31.6532 yes 061231.17 + ± ± / III 061244.17 + ± ± ± + ± ± + ∗ ± / III 061246.71 + ± ± / III 061251.04 + ± ± + ± ±
724 – 28.43 31.13 31.35 31.13 31.48190 yes 061253.73 + ± ± + ± ± + ± ± ± / I 061258.21 + ± ± + ± ± + ± ± + ± ± / II 061325.57 + ± ± / III 061325.94 + ∗ ± + ± ± + ± ± + ± ± ∗ indicates frozen parameters in the fit.Col. (2): presence of an optical or infrared counterpart, infrared class from Spitzer photometry (if available).Col. (3): sources flagged with “F” showed flare-like variability during our observations; lightcurves are shown in Fig. 6.Col. (4): absorbing hydrogen column density of the best-fit.Cols. (5) and (6): plasma temperature(s) of the best-fit.Cols. (7) to (11): X-ray luminosities (for an assumed distance of 1.6 kpc) in the soft ( s , [0 . − .
0] keV) band, the hard ( h [2 . − .
0] keV) band, and the total ( t [0 . − .
0] keV) band.Absorption-corrected luminosities are denoted with the subscript c . sources). Our cross-correlation with the 2MASS Point SourceCatalog lead to 231 near-infrared counterparts (i.e. a counterpartfraction of 63%). Our cross-correlation with the infrared catalogfrom Chavarr´ıa et al. (2008) yielded 293 infrared counterparts,i.e. 80% counterpart fraction. For 58 X-ray sources we did notfind a counterpart in any of the inspected optical and infrared im-ages. Twelve of these sources are located in or very close to thecentral embedded cluster S255-IR. These X-ray sources may bevery deeply embedded protostars or young stellar objects locatedbehind the dense molecular cloud clumps; the non-detection ofoptical / infrared counterparts would then be related to very highextinction. The remaining 46 X-ray sources without known opti-cal / infrared counterpart outside this cluster show a rather homo-geneous spatial distribution, as expected for (mostly extragalac-tic) contaminants.It is interesting to consider the infrared classification of thesesources based on the IRAC spectral energy distribution (SED)slope determined by Chavarr´ıa et al. (2008). Unfortunately, thematching of our X-ray source-list with this infrared catalog is notstraightforward. The catalog contains 26 821 infrared sources.However, most of these are only detected in the deep near-infrared images, and just about 6400 of these are detected in the Spitzer data. Infrared classifications are only available for the462 infrared sources that are detected in all four IRAC bands.The majority of the catalog entries are very faint NIR sources,and many of these are probably background objects rather thanyoung stars in S254-S258. Furthermore, the fact that only 4225of the 26 821 sources are detected in all three of the J -, H -, and K -bands also suggest that there may well be a significant numberof spurious detections among the faint sources detected in onlyone band. This very high number of faint infrared sources pro-duces serious problems in any attempt to find the correct infraredcounterparts for our X-ray sources, since many Chandra sources have more than one possible counterparts within the X-ray errorradius. In these cases, the closest positional match is not nec-essarily the true counterpart. Due to the increasing number ofinfrared sources at fainter magnitudes, good positional matcheswith very faint infrared sources may in fact be just chance super-positions of physically unrelated sources, and one of the otherpossible matches may be the true counterpart . A reliable iden-tification of the infrared counterparts requires a sophisticatedapproach and will be addressed in the next step of our study.Nevertheless, we can mention here the results of a preliminarysource matching, where we only considered the spatially clos-est match to each X-ray source. We find that 8 X-ray sourceshave closest matches classified as Class I YSOs (embedded veryyoung stellar objects with infalling envelopes), 50 X-ray sourceshave closest matches classified as Class II YSOs (Classical T-Tauri stars, CTTs), and 8 X-ray sources have closest matchesclassified as Class III YSOs (“Weak line” T-Tauri stars, WTTs)in the infrared catalog. The X-ray luminosity function (XLF) is the product of thedistribution of X-ray luminosities of stars with a given massand the number of stars per mass interval, i.e. the initialmass function (IMF). Although the correlation between stel-lar mass and X-ray luminosity shows a considerable scatter(see, e.g., Preibisch et al. 2005), X-ray studies of a large num-ber of young stellar clusters have shown that the XLF appearsto be rather universal and constant in di ff erent environments We note that similar problems were encountered in an X-ray andinfrared study of the Carina Nebula; see Preibisch et al. (2011) for amore detailed discussion.ucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region 11
Fig. 7.
This plot shows for each X-ray source the net counts inthe first pointing compared to the net counts in the second point-ing (obtained about 4 days later). The sequence of blue errorbars at the upper and right edge show the Poisson statistical un-certainties for di ff erent numbers of net counts. The solid lineindicates the expected relation for sources with constant countrates in the two pointings; the dotted lines are o ff set by factorsof 2. Fig. 8.
Comparison of the XLF of S254-S258 (thick grey line)to the XLF of the Orion Nebula Cluster (from the COUP data;Getman et al. 2005b, black line). The straight lines show the re-sults of the power-law fits to the distributions in the luminosityrange L X = . − . erg s − .(see Feigelson & Getman 2005; Getman et al. 2006; Wang et al.2007).To construct the XLF of S254-S258, we use the intrinsic fullband [0 . − .
0] keV luminosities calculated by
XPHOT (Table 3column (7)). Our XLF of S254-S258 is shown in Figure 8 andcompared to the XLF for the stars in the Orion Nebula Clusterfrom the COUP (Getman et al. 2005b). Obviously, the S254-S258 XLF peaks and turns over at a higher luminosity (near L X ≈ . erg s − ) than the COUP XLF, because the X-ray detected sample for S254-S258 is incomplete for low-mass starsdue to the lower sensitivity as discussed above. The slopes of thebright parts of these two distributions, however, can be seen toagree well.For a more quantitative analysis, we performed power-lawfits of the form dN / d (log L X ) ∝ α × log L X for the observed dis-tribution of luminosities. We used the maximum-likelihood tech-nique described by Maschberger & Kroupa (2009), that yieldsan estimate for the exponent from the observed distribution func-tion (i.e. not a fit of the histogram). The resulting power-law ex-ponents for the distribution of X-ray luminosities in the range L X = [10 . − . ] erg / s are α = − . ± .
09 for the OrionCOUP data, and α = − . ± .
10 for our S254-S258 data. Thisconsistency confirms the results from comparisons of other re-gions (e.g., the CCCP; Feigelson et al. 2011).This result shows that it is reasonable to assume that theXLF of S254-S258 has a similar shape as the ONC XLF. Wecan therefore make a quantitative estimate of the size of the to-tal young stellar population in the observed part of S254-S258by determining the vertical o ff set between the two distributions.We find that the total population in the observed part of theS254-S258 complex is ≈ . × of that in the ONC. Since thetotal population of the Orion Nebula Cluster (within 2.06 pc,Hillenbrand & Hartmann 1998) is about 2800 stars, the observedregion of the S254-S258 should contain ∼ The spatial distribution of the 364 X-ray sources in the
Chandra field shows a complex pattern. Besides the prominent and densecentral cluster S255-IR a few further apparent clusterings aswell as a widely distributed population of X-ray sources thatare spread homogeneously over the entire field-of-view of our
Chandra observation can be seen. For a quantitative charac-terization of the spatial distribution we performed a nearestneighbor analysis (see Casertano & Hut 1985) to identify sta-tistically significant clusterings in an objective way. The surfacedensity estimator at the position of each source i is given by µ j ( i ) = ( j − . (cid:16) π D j ( i ) (cid:17) , where D j ( i ) is the angular distancefrom source i to its j th nearest neighbor. We used j = ∼ ff ects such as mirror vignetting and the increasing width of thepoint-spread function, it gets several times worse near the edgesof the ACIS field-of-view. We therefore plot in Fig. 9 the sourcedensity µ as a function of the o ff axis-angle. The general trendof decreasing source density with increasing o ff axis-angle canbe clearly seen.Clusters can be defined as spatially confined groups ofsources for which the local surface density clearly exceeds thevalues found at other locations in the image at similar o ff axis-angles. The obvious dense central cluster S255-IR appears (asexpected) as a very prominent peak at low o ff axis-angles inFig. 9. Using a density threshold of µ ≥
22 arcmin − , we findthat 45 X-ray sources can be considered as members of this clus-ter. Two further prominent peaks can be seen in the plot: onepeak consisting of 12 sources near o ff axis-angle 5 ′ is caused For comparison, we note that the diameter of the Orion NebulaCluster ( ≈ ′′ ) would be ≈ ′ at the distance of 1.6 kpc.2 Mucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region Fig. 9.
Top: Nearest neighbor analysis surface density at the lo-cation of each X-ray source plotted against the o ff axis-angle.The general decrease of density with increasing o ff axis-angleis related to instrumental e ff ects. The dot-lined boxes mark themembers of the three identified clusters.Bottom: Spatial distribution of the X-ray sources. The membersof the three clusters marked in the top plot are show by the col-ored asterisks.by a cluster of sources in the S256 region, while another peakaround o ff axis-angle 9 ′ with 7 sources represents a clustering inthe S258 region.Our clustering analysis thus reveals three significant clusters,that contain a total population of 64 X-ray sources. The remain-ing 300 X-ray sources are thus in a distributed, non-clusteredspatial configuration. As discussed above, up to ≈
50 X-raysources may be unrelated contaminants. This leads to a popu-lation of ∼
250 widely distributed X-ray detected young stars inthe observed area.
5. The size of the distributed X-ray population
In order to check whether the distributed X-ray sources may rep-resent the low-mass stellar population associated to the B0 starsin S255 and S257, we compare the size of the distributed X-raypopulation with the expected number of based on IMF extrapo-lations. According to the Kroupa (2001) field IMF model, eachB0 star ( M ≈ M ⊙ ; see Martins et al. 2005) should be associ-ated by ≈
320 low-mass stars ([0 . − M ⊙ ). However, not all ofthese low-mass stars will be detected in our X-ray observation,as the X-ray luminosities of young stars are related to the stellarmass. The expected number of X-ray sources can be found bycomparing the X-ray detection limit to the typical X-ray lumi-nosity functions for young stars in specific mass ranges.As discussed above, the X-ray detection limit of our datais L X , min ∼ . erg s − in the central part of the observedfield. However, this limit shows considerable systematic vari-ations as a function of the o ff -axis angle and gets several timesworse near the edges of the ACIS field. If we consider the widelydistributed population of X-ray sources, we have to take intoaccount that most of these sources are located outside the cen-tral few arcmininute region of maximum sensitivity. Broos et al.(2011b) performed a detailed analysis of the spatial sensitivityvariations over the ACIS field. From the values for the com-pleteness limits in di ff erent o ff -axis angle slices in their Table8 we find that the area-weighted average of the completenesslimit over the full field-of-view is about 0.5 dex higher than theon-axis value. This implies that the average X-ray completenesslimit of our Chandra data for the widely distributed populationis thus log L X ∼ . erg s − .Since the X-ray luminosity functions for young stars are verysimilar for most studied regions (see Feigelson & Getman 2005;Getman et al. 2006) we can assume that the young stars in S254-S258 follow the same relations between stellar mass and X-rayluminosity as established by the data from the Chandra
OrionUltradeep Project (see Preibisch & Feigelson 2005). This im-plies that we should detect ≈
70% of the young stars in the massrange [0 . − M ⊙ and ≈
30% of the stars in the mass range[0 . − . M ⊙ .Since the Kroupa (2001) IMF predicts ≈
80 stars with[0 . − M ⊙ and ≈
250 stars with [0 . − . M ⊙ for each B0star, the expected number of X-ray detected stars associated tothe two B0 stars is ∼ × (0 . × + . × = ≈
250 distributed X-ray sources (after correction forbackground contamination) is actually quite close to this expec-tation value and consistent with the assumption that these X-raysources trace the low-mass stellar population associated to thetwo B-stars.
6. Conclusions
Our deep
Chandra observation of the S254-S258 complex led tothe detection of 364 X-ray sources, about 50 of which are ex-pected to be background contaminants. The X-ray properties ofmost sources (luminosity, plasma temperature, and variability)are in the typical ranges found for young stellar objects. Thissupports the assumption that these X-ray sources objects repre-sent the population of young low-mass stars in the S254-S258complex. Our analysis of the spatial distribution of the X-raysources with a nearest neighbor method reveals three signifi-cant clusters: the dense central cluster S255-IR, and two smallerclusterings related to S256 and S258. About 20% of the X-ray sources are members of one of these clusters, whereas the ucciarelli, Preibisch, & Zinnecker: Revealing the “missing” low-mass stars in the S254-S258 star forming region 13 large majority ( ∼ not expect to see this distributed pop-ulation in the context of the models that two B0 stars have eitherformed in isolation or were ejected from the central embeddedcluster. Our results therefore suggest that the two B-stars and theassociated distributed low-mass stars represent a stellar popula-tion that is distinct from the embedded cluster of YSOs in S255-IR. This is in agreement with the model scenario in which theobserved star formation activity in the dense embedded clusterlocated in the interaction zone between the S255 and S257 H IIregions has been triggered by the compression of the cloud dueto the expansion of the H II regions.A detailed analysis of the optical and infrared properties ofthe individual X-ray detected young stars that can provide di-rect information on the ages, masses, and the circumstellar disksaround these stars will be presented in a upcoming study. Acknowledgements.
This work is based on observations obtained withthe
Chandra
X-ray Observatory, which is operated by the SmithsonianAstrophysical Observatory for and on behalf of the National AeronauticsSpace Administration (NASA) under contract NAS8-03060. Our analysis pre-sented in this paper was supported by the Munich Cluster of Excellence:“Origin and Structure of the Universe”. This publication makes use of dataproducts from the Two Micron All Sky Survey, which is a joint project ofthe University of Massachusetts and the Infrared Processing and AnalysisCenter / California Institute of Technology, funded by the National Aeronauticsand Space Administration and the National Science Foundation, and of obser-vations made with the Spitzer Space Telescope, which is operated by the JetPropulsion Laboratory, California Institute of Technology under a contract withNASA.
References
Allen, L. E., Hora, J. L., Megeath, S. T., et al. 2005, in IAU Symposium, Vol. 227,Massive Star Birth: A Crossroads of Astrophysics, ed. R. Cesaroni, M. Felli,E. Churchwell, & M. Walmsley, 352–357Anders, E. & Grevesse, N. 1989, Geochim. Cosmochim. Acta, 53, 197Arnaud, K. A. 1996, in Astronomical Society of the Pacific ConferenceSeries, Vol. 101, Astronomical Data Analysis Software and Systems V, ed.G. H. Jacoby & J. Barnes, 17– + Bieging, J. H., Peters, W. L., Vila Vilaro, B., Schlottman, K., & Kulesa, C. 2009,AJ, 138, 975Brand, J., Massi, F., Zavagno, A., Deharveng, L., & Lefloch, B. 2011, A&A,527, A62 + Brice˜no, C., Preibisch, T., Sherry, W. H., et al. 2007, Protostars and Planets V,345Broos, P. S., Feigelson, E. D., Townsley, L. K., et al. 2007, ApJS, 169, 353Broos, P. S., Getman, K. V., Povich, M. S., et al. 2011a, ApJS, 194, 4Broos, P. S., Townsley, L. K., Feigelson, E. D., et al. 2010, ApJ, 714, 1582Broos, P. S., Townsley, L. K., Feigelson, E. D., et al. 2011b, ApJS, 194, 2Casertano, S. & Hut, P. 1985, ApJ, 298, 80Cash, W. 1979, ApJ, 228, 939Chavarr´ıa, L. A., Allen, L. E., Hora, J. L., Brunt, C. M., & Fazio, G. G. 2008,ApJ, 682, 445Chopinet, M., Deharveng-Baudel, L., & Lortet-Zuckermann, M. C. 1974, A&A,30, 233Cyganowski, C. J., Brogan, C. L., & Hunter, T. R. 2007, AJ, 134, 346Deharveng, L., Zavagno, A., Schuller, F., et al. 2009, A&A, 496, 177Di Francesco, J., Johnstone, D., Kirk, H., MacKenzie, T., & Ledwosinska, E.2008, ApJS, 175, 277Favata, F., Flaccomio, E., Reale, F., et al. 2005, ApJS, 160, 469Feigelson, E., Townsley, L., G¨udel, M., & Stassun, K. 2007, Protostars andPlanets V, 313Feigelson, E. D. & Getman, K. V. 2005, ArXiv Astrophysics e-printsFeigelson, E. D., Getman, K. V., Townsley, L. K., et al. 2011, ApJS, 194, 9Forbrich, J. & Preibisch, T. 2007, A&A, 475, 959Freeman, P. E., Kashyap, V., Rosner, R., & Lamb, D. Q. 2002, ApJS, 138, 185 Getman, K. V., Feigelson, E. D., Broos, P. S., Micela, G., & Garmire, G. P. 2008,ApJ, 688, 418Getman, K. V., Feigelson, E. D., Broos, P. S., Townsley, L. K., & Garmire, G. P.2010, ApJ, 708, 1760Getman, K. V., Feigelson, E. D., Grosso, N., et al. 2005a, ApJS, 160, 353Getman, K. V., Feigelson, E. D., Townsley, L., et al. 2006, ApJS, 163, 306Getman, K. V., Flaccomio, E., Broos, P. S., et al. 2005b, ApJS, 160, 319Goddi, C., Moscadelli, L., Sanna, A., Cesaroni, R., & Minier, V. 2007, A&A,461, 1027G¨udel, M., Briggs, K. R., Arzner, K., et al. 2007, A&A, 468, 353Heyer, M. H., Snell, R. L., Morgan, J., & Schloerb, F. P. 1989, ApJ, 346, 220Hillenbrand, L. A. & Hartmann, L. W. 1998, ApJ, 492, 540Howard, E. M., Pipher, J. L., & Forrest, W. J. 1997, ApJ, 481, 327Imanishi, K., Koyama, K., & Tsuboi, Y. 2001, ApJ, 557, 747Itoh, Y., Tamura, M., Suto, H., et al. 2001, PASJ, 53, 495Ja ff e, D. T., Davidson, J. A., Dragovan, M., & Hildebrand, R. H. 1984, ApJ, 284,637Jiang, Z., Tamura, M., Hoare, M. G., et al. 2008, ApJ, 673, L175Kroupa, P. 2001, MNRAS, 322, 231Kurtz, S., Churchwell, E., & Wood, D. O. S. 1994, ApJS, 91, 659Kurtz, S., Hofner, P., & ´Alvarez, C. V. 2004, ApJS, 155, 149Longmore, S. N., Burton, M. G., Minier, V., & Walsh, A. J. 2006, MNRAS, 369,1196Martins, F., Schaerer, D., & Hillier, D. J. 2005, A&A, 436, 1049Maschberger, T. & Kroupa, P. 2009, MNRAS, 395, 931Minier, V., Burton, M. G., Hill, T., et al. 2005, A&A, 429, 945Minier, V., Peretto, N., Longmore, S. N., et al. 2007, in IAU Symposium, Vol.237, IAU Symposium, ed. B. G. Elmegreen & J. Palous, 160–164Miralles, M. P., Salas, L., Cruz-Gonzalez, I., & Kurtz, S. 1997, ApJ, 488, 749Mo ff at, A. F. J., Jackson, P. D., & Fitzgerald, M. P. 1979, A&AS, 38, 197Ojha, D., Tamura, M., & Sirius Team. 2006, Bulletin of the Astronomical Societyof India, 34, 119Pismis, P. & Hasse, I. 1976, Ap&SS, 45, 79Preibisch, T. & Feigelson, E. D. 2005, ApJS, 160, 390Preibisch, T., Hodgkin, S., Irwin, M., et al. 2011, ApJS, 194, 10Preibisch, T., Kim, Y., Favata, F., et al. 2005, ApJS, 160, 401Preibisch, T. & Zinnecker, H. 2002, AJ, 123, 1613Preibisch, T., Zinnecker, H., & Schmitt, J. H. M. M. 1993, A&A, 279, L33Rygl, K. L. J., Brunthaler, A., Reid, M. J., et al. 2010, A&A, 511, A2 + Sharpless, S. 1959, ApJS, 4, 257Smith, R. K., Brickhouse, N. S., Liedahl, D. A., & Raymond, J. C. 2001, ApJ,556, L91Snell, R. L. & Bally, J. 1986, ApJ, 303, 683Stassun, K. G., van den Berg, M., Feigelson, E., & Flaccomio, E. 2006, ApJ,649, 914Testi, L., Palla, F., & Natta, A. 1999, A&A, 342, 515Townsley, L. K., Broos, P. S., Corcoran, M. F., et al. 2011, ApJS, 194, 1Townsley, L. K., Feigelson, E. D., Montmerle, T., et al. 2003, ApJ, 593, 874Wachter, K., Leach, R., & Kellogg, E. 1979, ApJ, 230, 274Wang, J., Townsley, L. K., Feigelson, E. D., et al. 2007, ApJS, 168, 100Wang, Y., Beuther, H., Bik, A., et al. 2011, A&A, 527, A32 + Wilms, J., Allen, A., & McCray, R. 2000, ApJ, 542, 914Wolk, S. J., Harnden, Jr., F. R., Flaccomio, E., et al. 2005, ApJS, 160, 423Zavagno, A., Russeil, D., Motte, F., et al. 2010, A&A, 518, L81 ++