arXiv:1307.5105v1 [astro-ph.SR] 19 Jul 2013

31 Pages·2013·1.04 MB·English

The Gould’s Belt Very Large Array Survey I:
The Ophiuchus complex
Sergio Dzib1, Laurent Loinard1,2, Amy J. Mioduszewski3, Luis F. Rodr´ıguez1,4, Gisela N.
Ortiz-Leo´n1, Gerardo Pech1, Juana L. Rivera1, Rosa M. Torres5, Andrew F. Boden6, Lee
3 Hartmann7, Neal J. Evans II8, Cesar Bricen˜o9 and John Tobin10
1
0
2

l
u ABSTRACT
J
We present large-scale (∼ 2000 square arcminutes), deep (∼ 20 µJy), high-resolution (∼ 1(cid:48)(cid:48))
9
1 radio observations of the Ophiuchus star-forming complex obtained with the Karl G. Jansky
Very Large Array at λ = 4 and 6 cm. In total, 189 sources were detected, 56 of them associated

] with known young stellar sources, and 4 with known extragalactic objects; the other 129 remain
R
unclassified, but most of them are most probably background quasars. The vast majority of the
S
young stars detected at radio wavelengths have spectral types K or M, although we also detect
.
h 4 objects of A/F/B types and 2 brown dwarf candidates. At least half of these young stars
p
are non-thermal (gyrosynchrotron) sources, with active coronas characterized by high levels of
-
o variability, negative spectral indices, and (in some cases) significant circular polarization. As
r expected, there is a clear tendency for the fraction of non-thermal sources to increase from the
t
s younger (Class 0/I or flat spectrum) to the more evolved (Class III or weak line T Tauri) stars.
a
The young stars detected both in X-rays and at radio wavelengths broadly follow a Gu¨del-Benz
[
relation, butwithadifferentnormalizationthanthemostradio-activetypesofstars. Finally, we
1
detect a ∼ 70 mJy compact extragalactic source near the center of the Ophiuchus core, which
v
should be used as gain calibrator for any future radio observations of this region.
5
0
Subject headings: astrometry —magnetic fields — radiation mechanisms: non–thermal — radio contin-
1
uum: stars — techniques: interferometric
5
.
7 1. Introduction
0 1CentrodeRadioastronom´ıayAstrof´ısica,Universidad
3 NacionalAut´onomadeM´exico Gould’s Belt (see Poppel 1997 for a recent re-
1 ApartadoPostal3-72, 58090, Morelia, Michoac´an, Mexico view)isaflattenedlocalGalacticstructure,about
: ([email protected])
v 1kpcinsize,thatcontainsmuchoftherecentstar-
i 2Max Planck Institut fu¨r Radioastronomie, Auf dem forming activity in the Solar neighborhood. In
X Hu¨gel69,53121Bonn,Germany
r 3NationalRadioAstronomyObservatory,DomeniciSci- particular, it comprises the nearest sites of active
a enceOperationsCenter, star-formation – i.e. the Taurus, Ophiuchus, and
1003LopezvilleRoad,Socorro,NM87801,USA Orionmolecularcomplexes. Thispaperisthefirst
4King Abdulaziz University, P.O. Box 80203, Jeddah in a series that will describe the results of deep,
21589,SaudiArabia
large-scale radio observations of these nearby re-
5PaulHarris9065,LasCondes,Santiago,Chile
gions of star-formation; it will focus on the Ophi-
6Division of Physics, Math, and Astronomy, California
uchus complex.
InstituteofTechnology,1200ECaliforniaBlvd.,Pasadena,
CA91125,USA Acombinationofseveralpropertiesconspireto
7DepartmentofAstronomy,UniversityofMichigan,500
ChurchStreet,AnnArbor,MI48105,USA
5101-A,Venezuela
8DepartmentofAstronomy,TheUniversityofTexasat
10National Radio Astronomy Observatory, Char-
Austin, 1 University Station, C1400, Austin, TX 78712,
lottesville,VA22903
USA
9Centro de Investigaciones de Astronom´ıa, M´erida,
1
make Ophiuchus one of the most interesting tar- twospecificfields,butreachedanoiselevelseveral
gets for star-formation studies (see Wilking et al. times better than that of previous observations.
2008 for a recent review). Its proximity (120 pc; A total of 35 compact radio sources were iden-
Loinardetal.2008)ensureshighlinearspatialres- tified, most of them associated with young stel-
olution, and enables the detection of fainter ob- lar objects (Wilking et al. 2001). More recently,
jects (e.g. brown dwarfs) than in other regions. Gagn´e et al. (2004) reported the detection at 6
Inaddition,thestellarpopulationassociatedwith cm of a dozen additional compact sources, but it
thecentraldarkcloudLynds1688(oftencalledthe is not clear what fraction of these sources might
Ophiuchus core; see Figure 1) provides us with an be associated with young stars. Finally, a num-
example of a young (τ < 0.5 Myr; Wilking et al. berofobservations(withbettersensitivityand/or
2008) stellar cluster that probes a mode of star- higher angular resolution) have been obtained to-
formation intermediate between the more isolated wardspecificobjects(particularlytheprototypical
situation exemplified by Taurus, and the more Class 0/I sources IRAS 16293–2422, VLA 1623,
clustered mode typically found in higher mass re- andYLW15–Wootten1989; Chandleretal.2005;
gions. As a consequence, Ophiuchus has been one Loinardetal.2007,2013;Andr´eetal.1993;Ward-
of the best-studied regions of star-formation. In Thompson et al. 2011; Girart et al. 2000, 2004).
the last decade alone, very detailed surveys of its The analysis of the radio properties of young
stellar population have been obtained in X-rays stars in Ophiuchus has helped clarify the origin
(Imanishi et al. 2001; Gagn´e et al. 2004; Ozawa of the radio emission produced by young stellar
et al. 2005; Sciortino et al. 2006), near-infrared objects in general. For sources with flat, Class I
(Alves de Oliveira et al. 2010; Barsony et al. or Class II infrared spectral energy distributions,
2012),andmid-infrared(Padgettetal.2008). The the radio emission (at ν (cid:46) 30 GHz) tends to be of
properties of the circumstellar disks around these thermal bremsstrahlung (free-free) origin,1 and to
young stars have been studied in the mid-infrared trace the dense central base of ionized winds (e.g.
(Kessler-Silacci et al. 2006; Geers et al. 2006; Rodr´ıguez 1999). This type of emission is char-
Lahuis et al. 2007) and millimeter/sub-millimeter acterized by a spectral index α (defined such that
regimes (Andrews & Williams 2007). Finally, the thefluxdependsonfrequencyasS ∝να)between
ν
distribution of the interstellar material has also –0.1 and +2 (α = +0.6 for the classical case of a
been studied in detail thanks to millimeter spec- partially optically thick isotropic wind; Panagia
troscopy (Ridge et al. 2006), sub-millimeter con- 1973), and only presents moderate variability ((cid:46)
tinuumobservations(Motteetal.1998;Johnstone 20%). In contrast, for Class III sources, the emis-
et al. 2000; Stanke et al. 2006; Young et al. 2006), sion is generally non-thermal (gyrosynchrotron)
and far-infrared imaging (Padgett et al. 2008). and related to magnetic activity near the stellar
Numerousradioobservationshavealsobeenob- surface. This kind of emission often (but not al-
tained, but the characterization of the Ophiuchus ways) exhibits strong variability and some level
complexatradiowavelengthsremainssignificantly of circular polarization (Dulk 1985). The spec-
lesssystematicthaninotherfrequencybands. Fol- tral index for gyrosynchrotron emission depends
lowing the early work by Brown & Zuckerman ontheenergydistributionoftheelectrons,andcan
(1975) and Falgarone & Gilmore (1981), the first vary between −2 and +2 (Dulk 1985), although
large-scale observations were obtained with the a slightly negative index occurs in many cases.
Very Large Array (VLA) in its C and D config- Furthermore, non-thermal radiation mechanisms
urations at 1.4 and 5 GHz by Andr´e et al. (1987) produce high brightness emission (T (cid:38) 106−7 K)
b
and Stine et al. (1988). While most of the de- confined to a very compact region ((cid:46) 10 R ). As
(cid:12)
tected radio sources were background extragalac-
tic objects, roughly a dozen could be confidently 1The Class 0 source IRAS 16293–2422 is somewhat of an
identified with young stars in the Ophiuchus clus- odditywherethermaldustemissionappearstoremainthe
ter. The next major step came with the deep dominantemissionprocesswellintothecentimeterregime
(Loinardetal.2013andreferencetherein). This,however,
5 GHz VLA observations of the Ophiuchus core,
isaveryunusualsituation,presumablypartlyrelatedtothe
again in the C and D configuration, reported by large amount of dense gas and dust around IRAS 16293–
Leous et al. (1991). These observations targeted 2422; we do not expect to find any source of this type in
thepresentsurvey.
2
a consequence, it is detectable even in very long mapped this area, which contains over 400 young
baselineinterferometry(VLBI)radioexperiments, stellarobjects,usingamosaicof47VLApointings
as shown by Lestrade et al. (1991), Phillips et al. (see Figure 1). Ten additional pointings were se-
(1991) or Andr´e et al. (1992). It should be men- lected to cover regions associated with other dust
tioned that the separation between thermal (free- clouds(particularlyL1689,butalsoL1709,L1712,
free) and non-thermal(gyrosynchrotron) emission and L1729; see Figure 1) located to the east of
at the Class II/III boundary is not a sharp one the Ophiuchus core (these clouds are known col-
as a number of Class II and even a few Class I lectively as the Ophiuchus streamers).
sources have been found to emit non-thermal ra- The FWHM of the primary beam (i.e. the field
diation(e.g.Forbrichetal.2007; Dzibetal.2010; of view) of the VLA has a diameter of 10(cid:48) at 4.5
Deller et al. 2013). GHz, and 6(cid:48) at 7.5 GHz. As a consequence, the
In the present article, we report on new ten individual fields targeting the streamers cover
radio observations of the Ophiuchus complex an area of 785 square arcminutes at 4.5 GHz and
which largely surpass previously published stud- 283 square arcminutes at 7.5 GHz. The spac-
ies thanks to a combination of high sensitivity, ing between the individual pointings of the mo-
sub-arcsecond angular resolution, and large field saic observed toward L1688 follows a somewhat
of view. These data will be used both to discuss irregular pattern chosen to optimize the compro-
thepopulationofradiosourcesinOphiuchus, and mise between uniform sensitivity and inclusion of
the relation between radio properties of young thelargestpossiblenumberofknownyoungstars.
stellar objects and their characteristics at other The total area covered by these fields is 1185
wavelengths. The observations are described in square arcminutes at 4.5 GHz and 878 square ar-
Section 2; the results are presented in Section 3, cminutes at 7.5 GHz.
andanalyzedinSections4and5. Ourconclusions Each observing session was organized as fol-
are provided in Section 6. lows. The standard flux calibrator 3C 286 was
first observed for ∼10 minutes. We subsequently
2. Observations spent one minute on the phase calibrator J1626-
2951followedbyaseriesofthreetargetpointings,
The observations were obtained with the Karl
spending three minutes on each. This phase cali-
G. Jansky Very Large Array (VLA) of the Na-
brator/target sequence was repeated until all tar-
tional Radio Astronomy Observatory (NRAO) in
getfieldswereobserved. Thus,threeminuteswere
itsCnBandBconfigurations. Twofrequencysub-
spentoneachtargetfieldforeachepoch. Thedata
bands, each 1 GHz wide, and centered at 4.5 and
were edited, calibrated, and imaged in a standard
7.5 GHz, respectively, were recorded simultane-
fashion using the Common Astronomy Software
ously. The observations were obtained on three
Applications package (CASA).
different epochs (February 17/19; April 3/4, and
In the 10 individual fields associated with the
May 4/6, 2011) typically separated from one an-
streamers, the noise level reached for each epoch
other by a month. This dual frequency, multi-
was 61 and 52 µJy beam−1 at 4.5 GHz and 7.5
epoch strategy was chosen to enable the charac-
GHz, respectively. For the mosaicked region, on
terization of the spectral index and variability of
the other hand, a nearly uniform noise level of 26
the detected sources, and to help in the identi-
µJy beam−1 is obtained at both frequencies. The
fication of the emission mechanisms (thermal vs.
frequency independence of the noise is the result
non-thermal).
of two effects that compensate each other: While
Since our aim was to examine the distribution
thenoiseinindividualfieldsissomewhatbetterat
and properties of radio sources in the Ophiuchus
7.5GHz,thelargerfieldofviewatlowerfrequency
complex in a statistically meaningful fashion, it
resultsinalargeroverlapbetweenfieldswhichre-
was important to systematically map a large area
duces the noise in the final mosaic. To produce
of the complex, including those regions known to
imageswithimprovedsensitivity,thethreeepochs
harbor a high density of young stars. The region
were combined, resulting in noise levels of 30 and
of highest stellar density is associated with the
25 µJy beam−1 at 4.5 GHz and 7.5 GHz, respec-
dark cloud Lynds 1688 (the Ophiuchus core). We
tively,fortheindividual(streamers)fields,and17
3
µJy beam−1 at both frequencies for the mosaic. measuredbycomparingthefluxesmeasuredatthe
The angular resolution of the observations is of three epochs. Specifically, we calculated, for each
order 1(cid:48)(cid:48). source and at each frequency, the difference be-
To test for circular polarization we produced tweenthehighestandlowestmeasuredfluxes,and
images of the V Stokes parameter in the inner normalized by the maximum flux. The resulting
quarter (in area) of the primary beam at each fre- values, expressed in percent, are given in columns
quency. At larger distances from the field center, 3 and 5 of Table 1. Circular polarization was con-
polarizationmeasurementsareunreliableasbeam fidently detected in 7 of our targets (Table 2).
squint (the separation of the R and L beams on Havingidentifiedtheradiosourcesintheregion
the sky) can create artificial circular polarization mapped, our next step was to try to determine
signals. which type of object they are associated with. In
our specific case, the two overwhelmingly domi-
3. Results nant possibilities are young stars and extragalac-
tic sources.3 We searched the literature for pre-
Thefirststepis,ofcourse,toidentifysourcesin
vious radio detections, and for counterparts at X-
our observations. This was done using the images
ray, optical, near- and mid-infrared wavelengths.
corresponding to the concatenation of the three
ThesearchwasdoneinSIMBAD,andaccessedall
epochs,whichprovidethehighestsensitivity. The
the major catalogs (listed explicitly in the foot-
criteria used to consider a detection as firm were:
note of Tab. 3). Note that the Spitzer c2d catalog
(i) sources with reported counterparts and a flux
includes cross-references to other major catalogs
larger than four times the σ noise of the area, or
which were taken into account in our counterpart
(ii) sources with a flux larger than five times the
search. We considered a radio source associated
σ noise of the area and without reported coun-
with a counterpart at another wavelengths if the
terparts. From this, a total of 189 sources were
separation between the two was below the com-
detected (see Table 1). To reflect the fact that
bined uncertainties of the two datasets. This was
these sources were found as part of the Gould’s
about 1.5 arcsec for the optical and infrared cat-
Belt Very Large Array Survey,asourcewithcoor-
alogs, but could be significantly larger for some
dinateshhmmss.ss−ddmmss.swillbenamedGBS-
of the radio catalogs (for instance, the NVSS has
VLA Jhhmmss.ss−ddmmss.s.
a positional uncertainty of about 5 arcsec). We
The flux of each source at 4.5 and 7.5 GHz are found that only 76 of the sources detected here
givenincolumns2and4ofTable1). Twosources hadpreviouslybeenreportedatradiowavelengths
of uncertainties on the fluxes are included: (i) the (column7ofTable3),whiletheother113arenew
error that results from the statistical noise in the radio detections. On the other hand, we found
images, and (ii) a systematic uncertainty of 5% a total of 100 counterparts at other wavelengths.
resulting from possible errors in the absolute flux Note that there is a significant number of sources
calibration. An estimation of the radio spectral that were previously known at radio wavelengths
index of each source (given in column 6 of Table and have known counterparts at other frequen-
1) was obtained from the fluxes measured in each cies. Asaconsequence,thenumberofsourcesthat
sub-band (at 4.5 and 7.5 GHz). To calculate the were previously known (at any frequency) is 134,
errors on the spectral indices, the two sources of while 55 of the sources in our sample are reported
errors (statistical and systematic) on the flux at here for the first time.
each frequency were added in quadrature and the
Based on their optical/infrared characteristics,
finalerrorwasobtainedusingstandarderrorprop-
2 of the 100 sources with counterparts are classi-
agation theory.
fiedintheliteratureasextragalacticsources,while
Oncethesourceswereidentifiedintheconcate- 55areclassifiedasyoungstellarobjects(YSO;see
nated images, we searched for them in the images
obtained from the individual epochs.2 An esti-
theaverageddata,butfoundnosuchobject.
mate of the level of variability of the sources was 3Althoughwecannotfullyruleoutthepossibilitythatother
objectsmightcontaminatethesample(aplanetarynebula,
2We also searched the individual epochs for objects that forinstance),theprobabilityofsuchanoccurrenceisvery
might be present there although they are not detected in low.
4
column8ofTable3). Oneadditionalobject(GBS- 4. Discussion
VLA J162626.31−242430.3, also known as VLA
4.1. Background Sources
1623B) is associated with the well-known Class 0
source VLA 1623, but it is still debated whether
It is clear from Section 3 that a large fraction
it is an outflow knot feature, or a protostellar ob-
of the radio sources detected here are background
ject (see the discussion by Ward-Thompson et al.
objects. To examine their statistics, we will con-
2011; and section 5.1 below). The remaining 42
centrate on the core region (for which we have a
radio sources with known counterparts at other
continuouscoverageatuniformsensitivity)andon
wavelengths are, to our knowledge, not classified
the observations at 4.5 GHz, which are more ap-
in the literature. On the other hand, two sources
propriate than those at 7.5 GHz for extragalactic
(GBS-VLA J162615.67-243421.2 and GBS-VLA
objects, since those usually have negative spec-
J162626.03-244923.7) have been classified in the
tral indices. Fomalont et al. (1991) showed that
literature as extragalactic, on the basis of their
the source counts at 5 GHz are appropriately de-
radio properties alone.
scribed by
To summarize, a total 134 of our 189 sources
were previously known (76 at radio wavelengths, (cid:18) N (cid:19) (cid:18) S (cid:19)−1.18±0.19
=0.42±0.05 .
and 100 at other frequencies, with some over- arcmin2 30µJy
lap between the two sub-samples). Of these 134
sources, 56 are classified as YSO and 4 as extra- According to those counts, the number of sources
galactic; theother74arenotclassified. Giventhe brighter than 150 µJy (the minimum flux of the
existing deep near and mid-infrared observations sourcesthatwedetected; seeTable1)expectedin
of Ophiuchus, it is unlikely that a major popula- the 1185 square arcminutes covered by our Ophi-
tion exists of unidentified young stars. In conse- uchus core observations is 75 ± 9. The number of
quence, we argue that most of these unclassified objects classified as extragalactic in the core is 3,
sources are extragalactic. For the same reason, andthereare79unclassifiedsourcesinthatregion.
most of the 55 objects detected here for the first Aswementionedearlier,mostoftheseunclassified
timearelikelybackgroundsources. Wenote,how- source are likely extragalactic, with only a small
ever, that 18 of the 129 unclassified objects (55 contribution(ofatmost15%)ofYSO.Thismeans
identified here for the first time and 74 previously that about 67 of the unclassified sources in the
known at radio wavelengths) are compact, have a core are extragalactic, and that the total number
positive spectral index, or exhibit high variabil- of observed extragalactic sources in that region is
ity (Table 4). Since these latter two properties about 70, in excellent agreement with the count
are not expected of quasars (which are certainly predicted by Fomalont et al. (1991). Note that
variable, but usually not strongly on such short the counts would in fact also be consistent with
timescale –e.g. Hovatta et al. 2008), but would be thepossibilitythatalltheunidentifiedsourcesare
natural characteristics of young stars, we argue extragalactic.
that a small population of YSO might be present
among the unclassified sources. This population 4.2. Radio properties of the YSO popula-
couldaccountfor,atmost,15%oftheunclassified tion
sources, and possibly significantly less. The dis-
In section 3, we mentioned that 56 of the radio
tribution of both the YSO and candidate YSO in
sources detected here are associated with young
our radio sources are shown in Figure 2.
stars. The spectral type and evolutionary status
It is interesting that 35 of the sources reported
for most of these objects are known (see Table 5)
here are only detected at radio frequencies and
and can be compared to their radio properties. In
withSpitzer(withnodetectionatanyotherwave-
Figure 3, the radio spectral index is plotted as a
lengths). These sources are usually not classified,
function of evolutionary status. There is a clear
and might be either young stellar sources, or fore-
tendencyformoreevolvedYSOstohaveasmaller
ground/background objects. It will be useful to
(i.e. more negative) spectral index. The younger
study this population further.
(Class 0, flat spectrum, and Class I) sources have
ameanspectralindexoforder0.5,indicatingthat
5
the dominant emission process is somewhat opti- or have high variability and a negative spectral
cally thick free-free emission. The older (Class II index. These source are almost certainly non-
andIII)sources,ontheotherhand,haveaslightly thermal and, as expected, they are mostly some-
negative mean spectral index, suggestive either what evolved –64% of them are Class III and/or
of optically thin free-free emission or of gyrosyn- WTTS.Theremightbeanevenlargerpopulation
chrotron radiation (see Section 1). In particular, of non-thermal YSOs in our sample since those
thereisasignificantpopulationofClassII/IIIob- sources which are either highly variable but with
jects with α < –0.1, and which are most likely positivespectralindex,orsteadybutwithanega-
non-thermal emitters. Note, however, that the tivespectralindexmightalsofallinthatcategory.
boundaries are not sharp since some very young Finally, if any of the unclassified sources given in
objects have slightly negative indices, while one Table4areindeedyoungstellarsources,theymust
of the more evolved YSOs reach a spectral index also be non-thermal.
above one. As expected, the extragalactic objects A final trend must be mentioned here. Out of
in our sample typically have negative spectral in- the 36 detected YSO with known spectral type,
dices, with a mean value of order –1. 31 fall in the range K0 to M5 (Figure 6), while no
The radio variability is shown as a function G type star, and only one F, two A, and one B
of evolutionary status in Figure 4. It is clear stars are detected. This points to a smaller num-
that younger sources are, on average, less variable ber of radio-bright early type stars compared to
than their more evolved counterparts. Since non- what would be expected on the basis of a typical
thermal emitters are often strongly variable, the stellar IMF alone. For instance, according to the
measured increase in variability confirms the con- Kroupa(2001)IMF,roughly50%ofallstarshave
clusiondrawnabovefromthespectralindexanal- a spectral type between K0 and M5. Could this
ysis that there is a significant population of non- be due to an IMF inherently deficient in higher-
thermal sources among the Class II/III sources in mass stars? Hsu et al. (2013) find that the pre-
our sample. A final piece of information supports mainsequencepopulationoftheL1641darkcloud
that conclusion: The radio flux appears to also in Orion, has a deficit of higher mass stars when
be, on average, higher for more evolved sources comparedtoacompactregionliketheOrionNeb-
(Figure 5), particularly from objects of Class III ulaCluster,andsuggestthattheenvironmentmay
as compared to those of Class II. Since younger playaroleindeterminingthehigh-massendofthe
sources experience more intense outflow activity, IMF.Wenote,however,thatEricksonetal.(2011)
they will be stronger thermal emitters than more constructedanIMFforanextinction-limitedsam-
evolved objects (this is consistent with the ob- pleof123YSOsinL1688andfoundittobeconsis-
served variability and spectral index trends de- tentwiththatoffieldstarforM>0.2M . Alter-
(cid:12)
scribed above). The stronger average radio flux natively, our finding might indicate that the frac-
for more evolved sources must, therefore, indicate tion of radio-bright young stars increases for later
that a different mechanism dominates the radio type stars, in agreement with the trend noticed
emission as young stars age. The most natural formoreevolvedstarsbyBergeretal. (2010). On
candidate is, again, gyrosynchrotron emission. It the other hand, the detected F/A/B stars are on
is important to point out, however, that young averagesignificantlybrighteratradiowavelengths
YSOs might well intrinsically emit as much non- thantheirMandKcounterparts, confirmingthat
thermal radiation as their older siblings, but be- (for earlier type stars), there is a good correlation
cause of their dense ionized winds, such emission between the bolometric and radio luminosities.
might be absorbed by the optically thick free-free It is also possible that the lower fraction of
emission (e.g. Forbrich et al. 2007; Deller et al. radio-bright early type stars could be the result
2013; see also Section 5.3). of an observational bias. The problem is that it is
The previous discussion shows that there ex- very difficult to identify YSOs with spectral types
ists a significant population of non-thermal radio G, F, A and B, unless they are actively accreting,
sources in our sample of detected YSOs. From as inferredfrom certainlines strongly inemission,
Table 5, we can see that 25 of the 55 young like Hα 6563˚A and He I 5875,6676˚A, and forbid-
stars in our sample (45%) either are polarized, den lines like [OI] or [SII], or have infrared excess
6
emission originating in a circumstellar disk.While relation were obtained mainly at 4.8 GHz and, in
for T Tauri stars we have well defined observables some cases, at 8.4 GHz.
toidentifythefullYSOpopulation,fortheearlier Of the 55 young stars in our sample, 35 have
type counterparts this is not the case. A strong known X-ray counterparts, and can be used to
Li I 6707˚A absorption line, in excess of what is study the L /L relation for young stellar ob-
X R
observed in young main sequence stars like the jects in a statistically significant fashion. Note
Pleiades(Bricen˜oetal.1997,1998)doesnotimply thatwecorrectedallX-rayluminositiestothedis-
youthinGandearliertypestars,becausetheydo tance of 120 pc adopted in this work. In Figure
not deplete Li I during their pre-main sequence 7, we show the X-ray luminosities of the young
phase. The strength of the Na I (8183,8195˚A) stellar objects as a function of their radio lumi-
doublet, which allows to distinguish low surface nosities at both radio frequencies observed in this
gravity stars still contracting toward the main se- work. In agreement with the results obtained by
quence (e.g. Martin et al. 1996; Slesnick et al. Gagn´e et al. (2004) and Forbrich et al. (2010),
2006; Downes et al. 2008; Lodieu et al. 2011; L /L ≤ 1015.5 for our sample. Indeed, a re-
X R
Schlieder et al. 2012), breaks down for spectral lation L /L ≈ 1014±1 provides a good match
X R
types earlier than about M0. Strong X-ray emis- to the distribution of points in this plot. This is
sion, characteristic of late F and G through M- equivalent,intermsoftheGu¨del-Benzrelation,to
type young stars, unless combined with other cri- κ = 0.03 for young stellar objects. We argue that
teria, cansufferfromsignificantcontaminationby this results provides an extension of the relation
young main sequence stars with ages up to ∼100 valid for YSOs.
Myr (Bricen˜o et al. 1997). So far, the best way
to determine the full membership of G through 5. Comments on some individual sources
B-type YSOs seems to be selecting as members
objectswhichsatisfyasprimarycriteriaradialve- 5.1. VLA 1623
locities and proper motions (if possible, e.g. Hsun
Atotaloffourradiosourcesassociatedwiththe
et al. 2013), combine these with other character-
Class 0 source VLA 1623 have been reported in
istics like X-ray emission, emission lines, infrared
the literature. Bontemps et al. (1997; see their
excesses,andplacetheminaH-Rdiagram,which
Figures 1 and 2) reported on the detection of
requiresareasonableknowledgeofthedistanceto
three roughly aligned objects that they called A,
eachandeveryobject,orassumptionofacommon
B, and C (from east to west). At higher an-
distance to a group or cluster. If, as suggested by
gular resolution, however, their source A breaks
the trend observed in Figure 5, Class III sources
down into two sub-condensation. They interpret
areindeed,onaverage,brighterradiosourcesthan
the easternmost (and weakest) of these two sub-
ClassIIYSOs,thefactthattheearly-typesample
condensations as the protostar VLA 1623 itself,
may be biased toward the later could be affecting
and the western sub-condensation, as well as the
our result.
sources B and C, as knots along a jet driven
by VLA 1623. In more recent publications, the
4.3. The radio – X-ray relation
two sub-condensations within source A of Bon-
Gu¨del&Benz(1993)andBenz&Gu¨del(1994) temps et al. (1995) have usually been referred
showedthattheradioandX-rayemissionsofmag- to as VLA 1623A and VLA 1623B (e.g. Ward-
netically active stars are correlated by a relation Thompson et al. 2011; Murillo & Lai 2013), while
of the form: source B of Bontemps et al. (1995) has been re-
named VLA 1623W; source C has, to our knowl-
L
X =κ·1015.5±1 [Hz]. edge, never been detected again. While there is
L
R general agreement that VLA 1623A is a proto-
star, the nature of VLA 1623B and VLA 1623W
where κ is unity for dMe and dK stars, BY Dra-
is still debated.4 Looney et al. (2000) obtained
type binaries and RS CVn binaries with two sub-
giants. For classical RS CVn binaries, Algol sys-
tems, FK Com stars and Post T Tauri stars κ ≈ 4Note,however,thatbothhavemeasurablepropermotions
similartothoseofothersourcesinOphiuchus(J.L.Rivera
0.17. Theradioobservationsusedtoestablishthis
7
dust observations at λ = 2.7 mm and concluded than with the higher mass members of the sys-
that VLA 1623B was, most likely, a stellar com- tem)issecuresincetheangularoffsetbetweenthe
panion of VLA 1623A. A similar conclusion was position of the radio source and the nominal posi-
reached by Murillo & Lai (2013) based on Sub- tion of the brown dwarf candidate is 0(cid:48).(cid:48)14, while
Millimeter Array (SMA) observations. The latter its separations from the other stars in the sys-
authors detected VLA 1623W as a bright point tem are 2(cid:48).(cid:48)42 and 3(cid:48).(cid:48)31. We also detected radio
sourceintheMIPS24microndata. Thisstrongly emission from two M stars near the brown dwarf
suggestthatitmightalsobeaprotostellarsource, boundary: GBS-VLA J162556.09-243015.3 is an
makingVLA1623atriplesystem. However,some M5 young stellar object of Class III, while GBS-
properties of the radio emission from VLA 1623B VLA J162759.95-244819.5 is an M4.75 weak line
are more easily explained if it is interpreted as a T Tauri star.
knot along the jet (Ward-Thompson et al. 2011). We will discuss these cases in more details in
VLA 1623 A, B, and W are detected in our a forthcoming dedicated publication, but would
observations as GBS-VLA J162626.39-242430.9, like to note here that in three of these four ob-
GBS-VLA J162626.31-242430.3, and GBS-VLA jects near the brown dwarf boundary, the radio
J162625.62-242429.2, respectively. Source C of emission shows clear indications (high levels of
Bontemps et al. (1997), however, is not seen in variability and/or a negative spectral index) of
our data. Both VLA 1623A and VLA 1623B have being non-thermal. Only in the case of GBS-
positive spectral indices of order 0.8, more typical VLA J162556.09-243015.3 is the emission more
of protostellar sources than of outflow features. likelythermal(free-free)sincethespectralindexis
VLA 1623W, on the other hand, has a spectral around zero and the radio flux quite steady. Our
index (–0.2 ± 0.6) consistent with the optically detections show that young stellar objects near
thin free-free emission expected from an outflow or beyond the brown dwarf boundary can be de-
knot (e.g. Pech et al. 2010). Interestingly, how- tectable radio emitters even at a distance of more
ever, the separations between VLA 1623A and than 100 pc. Like their older counterparts (e.g.
VLA1623B(1.2arcsec),andbetweenVLA1623A McLeanetal.2011;Ravietal.2011),theyappear
and VLA 1623W (10.5 arcsec) have not changed to have a large radio to bolometric luminosity ra-
appreciably during the ∼ 15 years separating the tio.
observations reported by Bontemps et al. (1997)
andthosedescribedhere. Thissetsanupperlimit 5.3. Non-thermal Radio Emission from
ofabout4kms−1 ontheirrelativemotion. While Class I Objects
thiswouldbeconsistentwiththeexpectedorbital
Protostellar objects (of Class 0 and I) are ex-
motioninasolarmassmultiplesystem,itisharder
pected to have strong winds producing thermal
to reconcile with sources B and W being outflow
bremsstrahlung (free-free) emission that is opti-
knots.
cally thick at least in the dense region immedi-
ately surrounding the protostar (e.g. Rodr´ıguez
5.2. Radio detections at the stellar-brown
1999). Inthissituation,eveniftheprotostaritself
dwarf boundary
emitted non-thermal (gyrosynchrotron) radiation,
Wedetectradioemissionfromtwobrowndwarf it ought to be absorbed by the ionized wind and
candidates: GBS-VLA J162722.96-242236.6 is as- should not reach the observer. It is noteworthy,
sociated with a brown dwarf reported and docu- however, that a small number of Class I sources
mented by Marsh et al. (2010), while GBS-VLA have been found to be non-thermal emitters. Per-
J162715.70-243845.6 is a brown dwarf candidate haps the most robust cases are IRS5 in Corona
(AlvesdeOliveiraetal.2010)locatedinthesouth- Australis (Feigelson et al. 1998; Deller, Forbrich
ernmost component in the triple system WL 20. & Loinard 2013), EC 95 in Serpens (Dzib et al.
We note that the association of the radio source 2010), and (more marginally) YLW 15 in Ophi-
with the brown dwarf candidate in WL 20 (rather uchus (Forbrich et al. 2007). The detectability of
non-thermal emission in these objects might be
etal.,inprep). Thus,theyareassociatedwiththatregion: due to a favorable geometry (e.g. when the proto-
neitherisanrelatedbackgroundorforegroundobjects. star is seen nearly pole-on or nearly edge-on, the
8
free-free opacity might be reduced; Forbrich et al. 6. Conclusions and perspectives
1997),ortotidalclearingofcircumstellarmaterial
In this paper, we have reported on radio ob-
in a tight binary system (Dzib et al. 2010).
servations of the Ophiuchus complex that largely
We find two possible non-thermal Class I
surpass all such previous observations thanks to a
sources in our sample. On the one hand, GBS-
combination of high sensitivity, good angular res-
VLA J162726.90-244050.8 corresponds to YLW
olution, and large field of view. A total of 189
15 which, as we just mentioned, is one of the
sources were detected, 56 of them associated with
known candidates non-thermal Class I sources.
knownyoungstellarsources,and4withknownex-
In our observations, its spectral index is negative
tragalacticobjects; theother129remainunclassi-
(suggesting a non-thermal process) but its flux is
fied, but most of them are certainly extragalactic
almostconstant. Forbrichetal.(2007)marginally
background sources. Most of the young stars de-
detected one of the components in YLW 15 dur-
tected at radio wavelengths have spectral types K
ing VLBI observations at 8.4 GHz, at a level of
orM,butwealsodetect2browndwarfcandidates.
145 µJy. An independent, more robust, VLBI
Interestingly, at least half of these young stars
detection would be necessary to confirm the non-
are non-thermal (gyrosynchrotron) sources, with
thermal nature of the source. The other target of
active magnetized coronas. These sources are
interest in this context is GBS-VLA J163200.97-
excellent targets for future astrometric observa-
245643.3 which is associated with the Class I
tionswithVLBIinstruments(similarto,butmuch
sourceWLY2-67. Whileithasapositivespectral
more extensive than those reported by Loinard
index, it exhibits significant variability and it is
et al. 2008) that would enable an accurate de-
found to be significantly circularly polarized (11
termination of the distance, kinematics, and in-
and 25 % at 4.5 and 7.5 GHz, respectively). This
ternal structure of the Ophiuchus region. Such
providesaclearindicationthattheradioemission
observations would be greatly aided by the detec-
is non-thermal. A VLBI detection should be at-
tion,reportedhere,ofanadequategaincalibrator
temptedofthistarget,asitwouldprovideadirect
(GBS-VLA J162700.00-242640.3) located toward
and independent confirmation of the non-thermal
the Ophiuchus core.
origin of the emission.
5.4. GBS-VLA J162700.00-242640.3: a L.L. is grateful to the von Humboldt Stiftung
new calibrator for Ophiuchus for financial support. S.D., L.L., L.F.R., G.N.O.,
G.P., and J.L.R. acknowledge the financial sup-
Oneoftheknownextragalactictargetsdetected
portofDGAPA,UNAMandCONACyT,M´exico.
here (GBS-VLA J162700.00-242640.3) is found to
The National Radio Astronomy Observatory is
have a flux of order 71 mJy both at 4.5 and 7.5
operated by Associated Universities Inc. under
GHz, toexhibitverylittlevariability(∼15%; Ta-
cooperative agreement with the National Science
ble 1) and to be unresolved in our VLA data. It
Foundation. CASA is developed by an interna-
has been detected in a number of previous radio
tional consortium of scientists based at the Na-
observations(Table3)aswellasinhigh-resolution
tional Radio Astronomical Observatory (NRAO),
Ka-band observations (at ν = 32.5 GHz) that
the European Southern Observatory (ESO), the
we obtained in 2011 in the BnA configuration of
National Astronomical Observatory of Japan
the VLA. The spectral energy distribution con-
(NAOJ),theCSIROAustraliaTelescopeNational
structed from all available radio data (Figure 8)
Facility(CSIRO/ATNF),andtheNetherlandsIn-
shows that it is a flat spectrum source. We have
stitute for Radio Astronomy (ASTRON) under
observed it with the Very Long Baseline Array
the guidance of NRAO. This research has made
(VLBA) at ν = 8.4 GHz, and detected it as
use of the SIMBAD database, operated at CDS,
an unresolved 70 mJy source. Since GBS-VLA
Strasbourg, France
J162700.00-242640.3 is located in the direction of
theOphiuchuscore(Figure2),itoughttobeused
REFERENCES
as main gain calibrator for any future interfero-
metric(conventionalorlongbaseline)observations Alves de Oliveira, C., Moraux, E., Bouvier, J., et
of the Ophiuchus complex. al. 2010, A&A, 515, A75
9
Andre,P.,Montmerle,T.,&Feigelson,E.D.1987, Dulk, G. A. 1985, ARA&A, 23, 169
AJ, 93, 1182
Dzib,S.,Loinard,L.,Mioduszewski,A.J.,Boden,
Andre, P., Deeney, B. D., Phillips, R. B., & A. F., Rodr´ıguez, L. F., & Torres, R. M. 2010,
Lestrade, J.-F. 1992, ApJ, 401, 667 ApJ, 718, 610
Andre, P., Ward-Thompson, D., & Barsony, M. Erickson, K. L., Wilking, B. A., Meyer, M. R.,
1993, ApJ, 406, 122 Robinson,J.G.,&Stephenson,L.N.2011,AJ,
142, 140
Andrews,S.M.,&Williams,J.P.2007,ApJ,671,
1800 Evans,N.J.,II,Dunham,M.M.,Jørgensen,J.K.,
et al. 2009, ApJS, 181, 321
Barsony, M., Haisch, K. E., Marsh, K. A., & Mc-
Carthy, C. 2012, ApJ, 751, 22 Falgarone, E., & Gilmore, W. 1981, A&A, 95, 32
Benz, A. O., & Guedel, M. 1994, A&A, 285, 621 Feigelson, E. D., Carkner, L., & Wilking, B. A.
1998, ApJ, 494, L215
Berger, E., Basri, G., Fleming, T. A., et al. 2010,
ApJ, 709, 332 Fomalont,E.B.,Windhorst,R.A.,Kristian,J.A.,
& Kellerman, K. I. 1991, AJ, 102, 1258
Bontemps, S., & Andre, P. 1997, Herbig-Haro
Flows and the Birth of Stars, 182, 63P Fomalont, E. B., Kellermann, K. I., Partridge,
R. B., Windhorst, R. A., & Richards, E. A.
Bouvier, J., & Appenzeller, I. 1992, A&AS, 92,
2002, AJ, 123, 2402
481
Forbrich, J., Massi, M., Ros, E., Brunthaler, A.,
Bricen˜o,C.,Hartmann,L.,Stauffer,J.,&Mart´ın,
& Menten, K. M. 2007, A&A, 469, 985
E. 1998, AJ, 115, 2074
Forbrich, J., Posselt, B., Covey, K. R., & Lada,
Briceno, C., Hartmann, L. W., Stauffer, J. R., et
C. J. 2010, ApJ, 719, 691
al. 1997, AJ, 113, 740
Gagn´e, M., Skinner, S. L., & Daniel, K. J. 2004,
Brown, R. L., & Zuckerman, B. 1975, ApJ, 202,
ApJ, 613, 393
L125
Geers, V. C., Augereau, J.-C., Pontoppidan,
Chandler, C. J., Brogan, C. L., Shirley, Y. L., &
K. M., et al. 2006, A&A, 459, 545
Loinard, L. 2005, ApJ, 632, 371
Girart, J.M., Rodr´ıguez, L. F., &Curiel, S. 2000,
Cieza,L.,Padgett,D.L.,Stapelfeldt,K.R.,etal.
ApJ, 544, L153
2007, ApJ, 667, 308
Girart, J. M., Curiel, S., Rodr´ıguez, L. F., et al.
Condon, J. J., Cotton, W. D., Greisen, E. W., et
2004, AJ, 127, 2969
al. 1998, AJ, 115, 1693
Grosso, N., Montmerle, T., Bontemps, S., Andr´e,
Cutri, R. M., Skrutskie, M. F., van Dyk, S., et
P., & Feigelson, E. D. 2000, A&A, 359, 113
al. 2003, ”The IRSA 2MASS All-Sky Point
Source Catalog, NASA/IPAC Infrared Science Guedel, M., & Benz, A. O. 1993, ApJ, 405, L63
Archive.href=”http://irsa.ipac.caltech.edu/applications/Gator/”
Hovatta, T., Lehto, H. J., & Tornikoski, M. 2008,
Cutri, R. M., & et al. 2012, VizieR Online Data A&A, 488, 897
Catalog, 2311, 0
Hsu, W.-H., Hartmann, L., Allen, L., et al. 2013,
Deller, A. T., Forbrich, J., & Loinard, L. 2013, ApJ, 764, 114
A&A, 552, A51
Imanishi, K., Koyama, K., & Tsuboi, Y. 2001,
Downes, J. J., Bricen˜o, C., Hern´andez, J., et al. ApJ, 557, 747
2008, AJ, 136, 51
10