Radio detections of southern ultracool dwarfs

1.6 MB·

by  C. Lynch

MNRAS000,1–10(2016) Preprint11January2016 CompiledusingMNRASLATEXstylefilev3.0
Radio detections of southern ultracool dwarfs
C. Lynch,1,2(cid:63) T. Murphy,1,2 V. Ravi,4 G. Hobbs,3 K. Lo,1,5 C. Ward1
1 Sydney Institute for Astronomy, School of Physics, The University of Sydney, NSW 2006, Australia
2 ARC Centre of Excellence for All-sky Astrophysics (CAASTRO)
3 Australia Telescope National Facility, CSIRO Astronomy and Space Science, PO Box 76, Epping, NSW 1710, Australia
4 Cahill Centre for Astronomy and Astrophysics, MC 249-17, California Institute of Technology, Pasadena, CA 91125, USA
5 University College London Genetics Institute, University College London, London WC1E 6BT, United Kingdom
6 Accepted2016January6.Received2016January5;inoriginalform2015September7
1
0
2 ABSTRACT
We report the results of a volume-limited survey using the Australia Telescope Com-
n
a pactArraytosearchfortransientandquiescentradioemissionfrom15southernhemi-
J sphere ultracool dwarfs. We detect radio emission from 2MASSW J0004348–404405
8 increasingthenumberofradioloudultracooldwarfsto22.Wealsoobserveradioemis-
sionfrom2MASSJ10481463–3956062and2MASSIJ0339352–352544,twosourceswith

] previous radio detections. The radio emission from the three detected sources shows
R
no variability or flare emission. Modelling this quiescent emission we find that it is
S consistent with optically thin gyrosynchrotron emission from a magnetosphere with
. anemittingregionradiusof(1–2)R ,magneticfieldinclination20◦–80◦,fieldstrength
h ∗
p ∼10–200G,andpower-lawelectrondensity∼104 –108 cm−3.Additionally,weplace
- upper limits on four ultracool dwarfs with no previous radio observations. This in-
o
creases the number of ultracool dwarfs studied at radio frequencies to 222. Analysing
r
t generaltrendsoftheradioemissionforthissampleof15sources,wefindthattheradio
s
activity increases for later spectral types and more rapidly rotating objects. Further-
a
[ more,comparingtheratiooftheradiotoX-rayluminositiesforthesesources,wefind
2MASS J10481463–3956062 and 2MASSI J0339352–352544 violate the Gu¨del-Benz
1
relation by more than two orders of magnitude.
v
9 Keywords: radiocontinuum:stars–stars:low-mass,browndwarfs–stars:magnetic
4 field – stars: activity
7
1
0
.
1
1 INTRODUCTION and X-ray emission does not imply a drop in the fraction
0
ofactivecoolstarsoverlaterspectraltypes.Thenumberof
6 Surveys of chromospheric Hα and coronal X-ray emission
1 active systems, as indicated by Hα emission, is observed to
fromlowmassstarsshowasteadydeclineinmagneticactiv-
: increase across later spectral types and peaks between M9
v itystrengthbeginninginlate-typeMdwarfs(e.g.Neuh¨auser
and L0 objects (Schmidt et al. 2015). Additionally neither
i et al. 1999; Gizis et al. 2000; West et al. 2004; Williams
X themagneticfieldstrengthorfillingfactorforlatetypeob-
et al. 2014; Schmidt et al. 2015). The strength of activity
jects is thought to decrease. In fact the detection of both
r in these two wavebands is frequently characterised by the
a quiescentandflaringnon-thermalradioemissionfromsome
ratio of the luminosity in the Hα/X-ray waveband to the
of the lowest mass stars and brown dwarfs (Berger et al.
bolometric luminosity (Hawley et al. 1996). The reduction
2001; Berger 2002, 2006; Berger et al. 2009; Burgasser &
in activity strength is thought to be associated with a de-
Putman 2005; Phan-Bao et al. 2007; Osten & Jayaward-
creaseinplasmaheatingthroughthedissipationofmagnetic
hana2006;McLeanetal.2012),collectivelycalledultracool
fields (Mohanty et al. 2002). However, recent atmospheric
dwarfs,confirmsthatatleastsomeoftheseobjectsarestill
modelling of late-type objects indicates that it is not un-
capable of generating strong magnetic fields.
reasonable to expect observable Hα emission for these ob-
jects.Therarefiedupperpartsofthestellaratmospheresare Mostradioloudultracooldwarfshaveaquiescentcom-
foundtobecapableofmagneticallycouplingdespitehaving ponent, and in some cases, this component is found to vary
a low levels of ionisation (Rodr´ıguez-Barrera et al. 2015). with the rotation of the star (e.g. McLean et al. 2011).
Thedecline inthe magneticactivitystrengthtracedbyHα There is still some debate over the nature of the quiescent
componentwherebothdepolarisedelectroncyclotronmaser
(ECM)(Hallinanetal.2007)andgyrosynchrotronemission
(cid:63) E-mail:[email protected] from a non-thermal population of electrons (Berger 2002;
(cid:13)c 2016TheAuthors
2 C. Lynch et al.
Burgasser&Putman2005;Ostenetal.2006a)areproposed sourcesconsistsof8Mdwarfsand7Ldwarfs,withspectral
sources for this emission. Furthermore, some radio loud ul- types ranging from M7.0 to L8.0. Additionally, this selec-
tracool dwarfs are observed to have strong radio flares that tionofsourcesincludes3knownradioloudultracooldwarfs:
canbeperiodic.TheECMmechanismisgenerallyaccepted 2MASS J1456-2809 (Burgasser & Putman 2005; Osten &
to be the source of the pulsed emission since it can account Wolk2009),2MASSJ10481463–3956062(Burgasser&Put-
forthehighbrightnesstemperature,directivity,andcircular man2005;Ravietal.2011),and2MASSIJ0339352–352544
polarisation of this emission (Hallinan et al. 2006). These (Berger et al. 2001). Further details about the selected sur-
radio flares are sometimes associated with periodic varia- vey targets are given in Table 1.
tions in the optical band (Berger et al. 2009; Williams & Theobservationsofthe15ultracooldwarfswerecarried
Berger 2015; Hallinan et al. 2015). Recent simultaneous ra- out with ATCA in its fully extended 6km configurations
dio and optical observations of a late-type M dwarf showed duringthetimeperiodbetweenAprilandJune2010.During
thattheobservedmodulationatbothwavelengthscouldbe our observations the two IF bands were centred on 5.5 and
accounted for by a propagating electron beam, powered by 9.0GHzsimultaneously.Alogoftheseobservationsisgiven
auroral currents, striking the stellar atmosphere (Hallinan in Table 2.
etal.2015).Thisresultssuggeststhatauroraemaybeubiq- Thevisibilitydatawerereducedusingthestandardrou-
uitous signatures of large-scale magnetospheres. tines in the miriad environment (Sault et al. 1995). The
Radiosurveysofultracooldwarfshavefoundthatabout flux scale and bandpass response were determined from the
9% of these system are radio luminous, with 21 currently ATCA primary calibrators, either PKS B1934-638 or PKS
known radio loud ultracool dwarfs (Berger 2006; McLean B0823-500.Observationsofabright,compactsecondarycal-
et al. 2012; Antonova et al. 2013; Route & Wolszczan 2013; ibrator was used to calibrate the complex gains and leak-
Kaoetal.2015).Correlationsbetweenthepresenceoftran- age between the orthogonal linear feeds in each antenna.
sientorquiescentradioemissionandotherdwarfproperties Observations of the secondary calibrators (listed in Table
such as rotation and tracers of magnetic activity at other 2) were carried out every 20 minutes, for 1.3 minutes. The
wavelengths (X-ray and Hα) are not well established. In flare emission observed from some ultracool dwarfs can be
fact,theradioluminosityofsomedetectedsystemsisfarin confused with low level radio frequency interference (RFI)
excessofthewell-knownGu¨del-Benzrelation(Gu¨del&Benz peaks. However, the flares from ultracool dwarfs are not
1993, GB), an empirically-derived ratio between radio and stronglylinearlypolarised.SotoavoidconfusingRFIwitha
X-ray luminosity that applies to magnetically active stars sourceflare,wefirstidentifiedandflaggedRFIintheStokes
overawiderangeofspectraltypes.Thedeviationfromthis Q and U polarisations and then extend these flags to the
relationobservedinsomeultracooldwarfssuggeststhatthe other polarisations. To carry out this flagging scheme we
chromospheric evaporation model usually applied to flare used the miriad flagging tools pgflag and blflag.
stars(Neupert1968;Machadoetal.1980;Allredetal.2006) Aftercalibration,thevisibilitydataforeachsourcewas
maynotapplytotheseobjects.Furthermore,littleisknown invertedandcleanedusingthemiriadtasksinvert,clean,
aboutthegeometryorstrengthofthemagneticfieldsinul- andrestor.Brightsourceslocatedinthesamefieldasthe
tracool dwarfs as well as the mechanism that populates the target source were identified and removed. This process in-
magnetospheres with non-thermal electrons. volved using the clean components for each of the bright
Toaddresstheseissues,wecarriedoutavolumelimited sources, while masking the location of the target ultracool
surveyofasampleof15late-typeMandLdwarfslocatedin dwarf, and subtracting the source from the visibility data
thesouthernhemisphereusingtheAustraliaTelescopeCom- using the miriad task uvmodel. The phase centre of the
pactArray(ATCA).TheCompactArrayBroadbandBack- resulting visibility data was shifted to the location of the
end(CABB;Wilsonetal.2011)allowsforabandwidthof2 target ultracool dwarf using the miriad task uvedit and
GHzperpolarisationineachoftwoindependentlytuneable then inverted and cleaned in the standard fashion.
intermediate frequency (IF) bands. These wideband capa- Tosearchforradioemissionfromeachofthesources,we
bilities of ATCA easily provide detailed information about madeimagesforboththe5.5and9.0GHzfrequencybands
how observed radio pulses and quiescent emission vary in where we averaged over the full ∼12 Hr of observation and
time and frequency. Such a characterisation is required if 2 GHz bandwidth, to ensure the best signal-to-noise ratio
we want to constrain the magnetospheric parameters and (SNR). We fit these images using the Common Astronomy
geometry of ultracool dwarfs (section 4). Additionally, to SoftwareApplication(casa)(McMullinetal.2007)toolim-
understand general trends of radio emission from ultracool stat, which reports the statistics for a supplied region of a
dwarfs with regards to their other physical properties, our image. To determine if a source has detectable radio emis-
observations are augmented with values for projected rota- sion, we compared the measured peak flux density at the
tional velocities (vsin(i)), Hα and X-ray luminosities from locationofthesourcetotheimageRMSdeterminedfroma
the literature (section 5). fittoaregioncentredonthesourcepositionwithdimensions
were 6 times the size of the restoring beam. We considered
peak flux densities greater than 3 times the image RMS as
detections. These fits were carried out in both the Stokes I
2 OBSERVATIONS AND DATA REDUCTION
and V images and the results are listed in Table 3.
The sample of 15 ultracool dwarfs were selected form the For sources observed with detectable radio emission in
all-sky-volume-limited compilations of late-M (Reiners & eitherofthesetwo2GHzimages,additionalIandVimages
Basri 2009) and L (Reid et al. 2008) dwarfs. From these using512MHzfrequencyaveragingweremadetoconstrain
two catalogs we selected sources with distances <10 pc the spectral index and polarisation frequency dependence.
and located in the Southern Hemisphere. This selection of These512MHzimageswerethenfitusingthesamemethod
MNRAS000,1–10(2016)
Radio detections of southern ultracool dwarfs 3
Table 1.PropertiesoftheSurveySources
2MASSNumber R.A. Decl. Spectral Distance vsin(i) Lbol Lx/Lbol LHα/Lbol Referencea
Type (pc) (kms−1) (L(cid:12))
10481258–1120082 104812.8 -112018.9 M7.0 4.5 3.0 -3.16 -4.43 -4.63 1,6
14563831–2809473 145638.1 -280953.3 M7.0 7.0 8.0 -3.29 -4.00 -4.02 1,5,7
11554286–2224586 115542.7 -222459.6 M7.5 9.7 33.0 -3.30 -4.40 -4.58 1,6,12
10481463–3956062 104813.5 -395617.0 M8.0 4.0 18.0 -3.39 -5.00 -5.15 1,5,8
00244419–2708242 002444.1 -270819.7 M8.5 7.71 9.0 -3.25 – -4.62 11
0339352–352544 033935.5 -352540.8 M9.0 5.0 26.0 -3.79 -3.70 -5.30 1,2
03341218–4953322 033413.3 -495328.6 M9.0 8.20 – – – <5.32 11
0853362–032932 085335.9 -032933.5 M9.0 9.0 13.5 -3.49 -3.70 -3.93 3
1507476–162738 150747.6 -162744.9 L5.0 7.3 32.0 -4.23 <-4.50 -8.18 1,4
08354256–0819237 083542.3 -081921.7 L5.0 9.0 23.0 -4.60 – -7.42 1,4,5
0004348-404405 000435.4 -404421.8 L5.0 10.0 42.0 -4.67 – -7.42 1
17502484–0016151 175024.6 +001613.7 L5.5 8.0 – – – – 1
0340094–672405 034009.3 -672408.7 L8.0 9.90 – – – – 10
02550357–4700509 025504.0 -470054.9 L8.0 4.97 67 -4.80 <-4.70 <-8.28 9,12
02572581–3105523 025726.1 -310550.0 L8.0 9.60 – -4.82 – – 1
aReferences:(1)Antonovaetal.(2013);(2)Bergeretal.(2001);(3)Berger(2002);(4)Berger(2006);(5)Burgasser&Putman(2005);
(6)McLeanetal.(2012);(7)Osten&Wolk(2009);(8)Ravietal.(2011);(9)Reidetal.(2008);(10)Reiners&Basri(2008);
(11)Reiners&Basri(2010);(12)Williamsetal.(2014)
Table 2.LogofObservations
2MASSNumber ObservationDate ATCAConfigurationa PrimaryCalibrator SecondaryCalibrator
10481258–1120082 17Apr201006:41–17Apr201016:39 6A PKSB0823-500 1045-188
14563831–2809473 12Apr201009:32–12Apr201021:59 6A PKSB1934-638 1519-273
11554286–2224586 19Apr201005:49–19Apr201016:22 6A PKSB0823-500 1143-245
10481463–3956062 18Apr201005:12–18Apr201015:43 6A PKSB0823-500 1104–445
00244419–2708242 24Apr201018:18–25Apr201006:11 6A PKSB1934-638 2357-318
0339352–352544 30May201018:55–31May201007:08 6C PKSB1934-638 0405-331
03341218–4953322 29May201018:43–30May201006:30 6C PKSB1934-638 0302-623
0853362–032932 30Apr201004:12–30Apr201013:36 6C PKSB1934-638 0906+015
1507476–162738 14Apr201010:13–14Apr201021:05 6A PKSB1934-638 1504-166
08354256–0819237 21Apr201003:16–21Apr201014:01 6A PKSB1934-638 0859-140
0004348–404405 25Apr201017:47–26Apr201005:31 6A PKSB1934-638 0022-423
217502484–0016151 13Apr201013:35–13Apr201022:44 6A PKSB1934-638 1741-038
0340094–672405 22Apr201022:03–23Apr201010:00 6A PKSB1934-638 0252-712
02550357–4700509 27Apr201017:17–28Apr201005:20 6C PKSB0823-500 0252-549
02572581–3105523 28May201018:11–29May201006:02 6C PKSB1934-638 0237-233
aThelabelsrefertodifferentvariantsoftheantennaspacings;thesearedefinedat:
http://www.narrabri.atnf.csiro.au/operations/array_configurations/configurations.html.
detailed above. By vector averaging the real components of predictedpositionsdeterminedfromthe2MASSastrometry
thevisibilitiesintimebinsof10s,30sand60soverthetwo (Cutri et al. 2003) and proper motion measurements from
2.0 GHz frequency bins, light curves were created to search the literature (Deacon et al. 2005; Schmidt et al. 2007; Fa-
for variability (section 3.1). herty et al. 2009).
Both 2MASS J1048-3956 and 2MASS J0339-3525 have
previous radio detections, while the detection of 2MASS
3 DETECTIONS
J0004-4044 is the first. Additionally, this survey includes
From fits to the 2 GHz frequency-averaged images, we find radio limits on 4 sources, 2MASS I J0340094–672405,
onlythreeofour15targetultracooldwarfshavedetectable 2MASS J02550357–4700509, 2MASS J03341218–4953322,
levels of Stokes I emission in at least one of the two ob- and2MASSJ00244419–2708242,withnopreviousradioob-
servingfrequencybands,2MASSJ10481463–3956062(here- servations. Combining our results with that of Antonova
after2MASSJ1048-3956),2MASSIJ0339352–352544(here- et al. (2013), Route & Wolszczan (2013), Burgasser et al.
after 2MASS J0339-3525), and 2MASSW J0004348–404405 (2013,2015),andKaoetal.(2015),thenumberofultracool
(hereafter 2MASS J0004-4044). The Stokes I and V images dwarfsstudiedatradioradiofrequenciesisnow222,with22
for these three sources at 5.5 GHz and 9.0 GHz are shown sources observed to have radio emission. From these num-
in Figures 2 and 3, respectively. The flux density peaks in bers, ∼10% of ultracool dwarfs are observed to have radio
each of these images lie within the mean beam size of the emission.ThisishigherthantheestimatebyAntonovaetal.
MNRAS000,1–10(2016)
4 C. Lynch et al.
6
5
9
3
8-
4
0
1
S J
S
A
M
2
5
2
5
3
9-
3
3
0
S J
S
A
M
2
4
4
0
4
4-
0
0
0
S J
S
A
M
2
Figure 1.Lomb-Scargleperiodogramofthe10stimeaveragedand2GHzfrequencyaveraged,StokesIfluxvaluesfor2MASSJ1048-
3956 (top), 2MASS J0339-3525 (middle), and 2MASS J0004-4044 (bottom). The periodograms are calculated for both the 5 GHz (left
column) and 9 GHz (right column) frequency bands. Dashed lines indicate false alarm probabilities of 0.01 (99%), 0.1 (90%), and 0.5
(61%).Wedidnotdetectsignificantvariabilityinthesethreesources.
(2013)whoget∼6%whentheyconsideronlytheirobserva- 4 CHARACTERISING THE QUIESCENT
tions and that of McLean et al. (2012). Additionally using EMISSION
thissetofobservations,Antonovaetal.(2013)notethatthe
2MASSJ1048-3956,2MASSJ0339-3525and2MASSJ0004-
majority of the ultracool dwarfs with observed radio emis-
4044 all have detectable levels of Stokes I emission at
sionhaveaspectraltypebetweenM7–L3.5.However,this
5.5 GHz. For both 2MASS J0339-3525 and 2MASS J0004-
result may be due to the small number of observations of
4044 the 5.5 GHz emission is also observed to be polarised,
objects with spectral types >L3.5 included in their sample.
with polarisation fractions of 0.53 for 2MASS J0339-3525
If we add to the Antonova et al. (2013) sample the results
and 0.44 for 2MASS J0004-4044. At 9 GHz the radio emis-
of Burgasser et al. (2013), Route & Wolszczan (2013), Kao
sion from 2MASS J0004-4044 is undetectable, however, for
et al. (2015), and our own radio detections, the fraction of
both2MASSJ0339-3525and2MASSJ1048-3956westillde-
ultracooldwarfswithobservableradioemissionremainscon-
tect Stokes Iemissionfrom these sources. We donot detect
stant (∼10%) across the spectral type range M7 – T8.
any Stokes V emission for any of the sources at this higher
frequency.
4.1 Spectral Indices
We did a least squares fit to the measured Stokes I values
from the 512 MHz images to constrain spectral indices for
3.1 Variability
the three detected sources (see Table 3) between 4.7 and
Tosearchforburstemissionandanypotentialperiodicityin 9.7 GHz. For all three sources the emission drops steeply
thedetectedsources,weconstructedlightcurvesofthereal withincreasingfrequencyandshowsnosignsofaturn-over
visibilitiesinStokesIandVfor2MASSJ1048-3956,2MASS in the lower observing frequency band.
J0339-3525,and2MASSJ0004-4044.Theselightcurveswere Comparingthespectralindiceswecalculatefor2MASS
made for 1 min, 0.5 min, and 0.1 min time averaged bins J0339-3525 and 2MASS J1048-3956 to previously cited val-
and 512 MHz, 1 GHz, and 2 GHz frequency averaged bins. ues, we find that they do not agree. For 2MASS J0339-
Inallcombinationsofaveragingwedonotdetectanyburst 3525, Berger et al. (2001) constrain the spectral index to
emission or variability in the quiescent component. be 2.1±0.3 between 4 and 8 GHz, indicating optically thick
We carried out a Lomb-Scargle analysis (Lomb 1976; emission. The discrepancy between our result and that of
Scargle 1982) using the astroML scientific python modules Berger et al. (2001) is most likely because Berger et al.
(VanderPlas et al. 2012) to test the significance of the non- (2001) include both flare and quiescent emission in their
variability.Theperiodogramsofthe3sourceswithobserved analysis while we only consider the quiescent component.
radioemissionareshowninFigure1.Thedashedlinesindi- YetnotethatthemeasuredfluxdensitiesfromBergeretal.
catefalsealarmprobabilitiesforthe99%,90%,and61%lev- (2001) vary by more than a factor of 2 over the 3 months
els.Forallthreesourcetherearenosignificantpeaksinthe covered by their observations. This flux density variation
Lomb-Scargle power spectrum for variability on timescales could also be related to a variation in the spectral index.
less than an hour up to 12 hours. The quiescent radio emission from 2MASS J1048-3956 has
MNRAS000,1–10(2016)
Radio detections of southern ultracool dwarfs 5
2MASS J1048-3956 2MASS J0339-3525 2MASS J0004-4044
Stokes I
Stokes V
Figure 2. Radio images for 2MASS J1048-3956 (left-column), 2MASS J0339-3525 (middle-column), and 2MASS J0004-4044 (right-
column)inStokesI(top-row)andStokesV(bottom-row)at5.5GHz.Additionally,contoursareoverlaidwithlevelsat3,9,18,27times
the Stokes I RMS value of 8.2µJy for 2MASS J1048-3956 and 3, 6, 9, 12 20 times the Stokes I RMS values of 10.5µJy and 8.3µJy for
2MASS J0339-3525 and 2MASS J0004-4044, respectively. For all sources the Stokes V contours are -10, -5, -3, 3, 5, 10 times the RMS
valuesof9.7µJyfor2MASSJ1048-3956,17.6µJyfor2MASSJ0339-3525,and13.3µJyfor2MASSJ0004-4044.
been studied over a wide range of radio frequencies (∼4.0- brightness temperature is given by,
20 GHz) by Ravi et al. (2011) who fit a spectral index of
α=1.71±0.09 to the observed radio emission. The smaller T =2.5×109(cid:18) Sν (cid:19)(cid:16) ν (cid:17)−2(cid:18) d (cid:19)2(cid:18) R (cid:19)−2 K, (1)
frequencycoverageofourobservationscouldbethecauseof b mJy GHz pc R
J
this difference. However, similar to 2MASS J0339-3525, the
whereR istheradiusofJupiterandisthetypicalradiusof
4-9GHzemissionof2MASSJ1048-3956isvariableonlong- J
verylowmassstarsandbrowndwarfs(Burrowsetal.2001).
timescaleswheretheradioemissionpreviouslyobservedby
Ravi et al. (2011) has a slightly higher flux density and is AssumingM-typestellarcoronaldimensionsof(1-2)R∗(Leto
etal.2000),themeasuredfluxdensitiesandappropriateup-
circularly polarised with a polarisation fraction of 0.25-0.4.
per limits for the non-detections imply brightness tempera-
Long-termvariabilityinthemeasuredfluxdensitiesandpo-
tures in the range of (0.5 - 6)×108 K at 5.5 GHz (see Table
larisation of UCDs has been observed in several other cases
3).
(McLean et al. 2012) and may indicate a significant change
inthephysicalcharacteristicsoftheemittingregionsinthese
sources.Suchvariabilityisalsowellknowninthecaseofra-
dio flares from close stellar binaries and are attributed to
4.3 Origins for the emission
changes in energisation (e.g., RS CVns; Mutel et al. 1998;
Richards et al. 2003). Thehighbrightnesstemperatures,combinedwiththespec-
tral indices, and the measured circular polarisation of
2MASS J0339-3525 and 2MASS J0004-4044, rule out ther-
malbremsstrahlungemissionandwefindtheemissiontobe
more consistent with gyrosynchrotron emission from a non-
4.2 Brightness Temperatures
thermalpopulationofacceleratedelectrons(Dulk1985).To
In order to assess the origin of the radio emission, we can modeltheobservedemissioncharacteristics,weusedtheex-
calculate the brightness temperature of the observed emis- pressions found in Robinson & Melrose (1984) for the ab-
sion. For a radio source at a distance d with an emitting sorption and emission coefficients of gyrosynchrotron emis-
volumeofradiusR,andfluxdensitySν atfrequencyν ,the sionfrommildlyrelativisticelectronswithapower-lawelec-
MNRAS000,1–10(2016)
6 C. Lynch et al.
2MASS J1048-3956 2MASS J0339-3525 2MASS J0004-4044
Stokes I
Stokes V
Figure 3. Radio image for 2MASS J1048-3956 (left-column), 2MASS J0339-3525 (middle-column), and 2MASS J0004-4044 (right-
column)inStokesI(top-row)andStokesV(bottom-row)at9.0GHz.Additionally,contoursareoverlaidwithlevelsat3,9,18,27times
theStokesIRMSvalueof10.6µJyfor2MASSJ1048-3956and3,6,9,1220timestheStokesIRMSvaluesof10.0µJyand8.2µJyfor
2MASS J0339-3525 and 2MASS J0004-4044, respectively. For all sources the Stokes V contours are -10, -5, -3, 3, 5, 10 times the RMS
valuesof10.6µJyfor2MASSJ1048-3956,19.1µJyfor2MASSJ0339-3525,and12.8µJyfor2MASSJ0004-4044.
Table 3.CharacteristicsoftheRadioEmission
2MASSNumber S5.5GHz(I)a S5.5GHz(V)a S9.0GHz(I)a S9.0GHz(V)a α4.7−9.7GHz TB
(µJy) (µJy) (µJy) (µJy) (K)
10481258–1120082 <44.7 <26.7 <34.5 <31.8 – <5.98×107
14563831–2809473 <29.9 <29.0 <37.4 <36.8 – <9.69×107
11554286–2224586 <31.2 <25.8 <34.8 <34.2 – <1.94×108
10481463–3956062 211.9±8.2 <29.1 131.4±10.8 <31.8 -1.1±0.11 2.23×108
00244419–2708242 <111.0 <37.8 <237.0 <48.9 – <4.36×108
0339352–352544 137.6±10.5 73.0±10.0 90.7±17.6 <57.3 -0.97±0.32 2.26×108
03341218–4953322 <29.4 <26.1 <37.8 <42.0 – <1.31×108
0853362–032932 <42.9 <35.4 <53.7 <50.4 – <2.30×108
1507476–162738 <37.6 <28.5 <36.6 <35.4 – <3.58×108
08354256–0819237 <32.0 <28.0 <42.0 <41.0 – <1.71×108
0004348–404405 100.0±8.3 44.3±8.2 <39.9 <38.4 <-0.74 6.61×108
17502484–0016151 <81 <36 <51 <39 – <3.43×108
0340094–672405 <27.0 <27.3 <39.0 <39.0 – <1.75×108
02550357–4700509 <30.9 <26.7 <34.6 <31.8 – <5.05×107
02572581–3105523 <66.0 <35.4 <63.0 <60.0 – <4.02×108
aUpperlimitslistedare3σ limitsbasedontheRMSvaluesmeasuredintheimagesforeachsource.
MNRAS000,1–10(2016)
Radio detections of southern ultracool dwarfs 7
Figure 4.ComparisonsbetweentworepresentativemodelcurvesandthemeasuredStokesI(top)andStokesV(bottom)fluxdensities
for2MASSJ1048-3956(left),2MASSJ0339-3525(middle),and2MASSJ0004-4044(right).Themeasuredvaluesresultfromfitstothe
512 MHz images for each source. The two different colours represent two different sets of model parameters that fall within the values
listedintable4.
Table 4.ModelParameters indexδ forthemodelusingtheapproximation(Dulk1985)
2MASSNumber R θB B log(Ne) δ =(1.22−α4.7−9.7GHz)/0.9
(R∗) (G) (cm−3) and the spectral indices given in Table 3. Additionally, the
optically thin assumption implies that the gyrosynchrotron
10481463–3956062 1.0–2.0 40◦ –80◦ 10–70 6.0–8.0
0339352–352544 1.0–2.0 20◦ –60◦ 20–223 4.1–6.8 turn over frequency is less than 4.7 GHz and further limits
0004348–404405 1.5–2.0 40◦ –60◦ 73–231 4.5–6.2 the range of suitable model parameters.
The model parameters that reproduce the observed
Stokes I and V emission for 2MASS J1048-3956, 2MASS
tron distribution given by, J0339-3525, and 2MASS J0004-4044 are listed in Table 4.
Generally the range of values for the densities, magnetic
Ne(E)=KE−δ field orientations and strengths, and the emission volumes
where K=N0(δ−1)E0δ−1 and is δ the energy index, and are consistent with previous constraints found for ultracool
with a low-energy cutoff E = 10 keV. dwarf magnetospheres (Burgasser & Putman 2005; Osten
0
Suchamodelrequiressomeknowledgeofthemagnetic et al. 2006b; Ravi et al. 2011; Lynch et al. 2015). Sample
geometryfortheultracooldwarf,specificallytheangleofin- SED’s are shown in Figure 4, along with the corresponding
clinationbetweenthelineofsightandthemagneticaxis,θB. observed fluxes in the 512 MHz images. The model param-
An estimate of this inclination angle can be obtained from eters for each curve are annotated in the figure, where the
observed variability in optical emission. However, 2MASS colour of the text corresponds to the colour of the appro-
J1048-3956, 2MASS J0339–3526, and 2MASS J0004-4044 priatemodelcurve.TheSEDfitsarequiteacceptable,with
are observed to have little variability at these wavelengths mostofthemodelfluxesagreeingwithupperlimitsandmea-
(Guenther et al. 2009; Schmidt et al. 2007; Stelzer et al. sured flux densities for both the Stokes I and V emission.
2012;Crossfield2014).Thustocharacterizetheplasmacon-
ditions responsible for this emission, we construct a sim-
ple coronal model consisting of a homogenous population
5 RADIO EMISSION TRENDS
ofmildly-relativisticpower-lawelectronsspirallinginauni-
form magnetic field. This simple model allows us to make To put the results of our survey into context, we compare
an order-of-magnitude estimate for the power-law electron ourradioresultswiththatofpreviousradiosurveysinFig-
densityandmagneticfieldstrengthwithoutmakingassump- ures 5 and 6. In these figures our measurements are the
tions concerning the geometry of the magnetic field. magentapointsandthegreypointsarefromRoute&Wol-
Forthismodelweassumedtheradiusoftheemittingre- szczan(2013),Antonovaetal.(2013),Burgasseretal.(2013,
giontorangefrom(1-2)R∗,consistentwithestimatesforthe 2015), and Kao et al. (2015), where appropriate. In these
emitting region dimension on M dwarfs (Leto et al. 2000), figures we include upper limits (open triangles), quiescent
and varied the emitting volume, the strength and orien- emission(squares)andflares(stars).Wefindthatourupper
tation of the magnetic field, and the non-thermal electron limits on the radio luminosity are lower than those placed
densitytobest-fitthemeasuredspectralenergydistribution by the previous surveys but are still comparable to the de-
andfractionalcircularpolarisationforeachsource.Sincethe tectedquiescentemissionfromtheleastbrightsources.The
spectralenergydistributionfor2MASSJ1048-3956,2MASS observed luminosities for 2MASS J1048-3956 and 2MASS
J0339-3525,and2MASSJ0004-4044isconsistentwithopti- J0339-3525areconsistentwiththosefromprevioussurveys.
callythingyrosynchrotronemission,weconstraintheenergy As a function of spectral type (Figure 5), we observe
MNRAS000,1–10(2016)
8 C. Lynch et al.
Figure5.(Top)Radioluminosityand(Bottom)ratiooftheradio Figure 6. (Top) Radio luminosity and (Bottom) ratio of the
tobolometricluminosityasafunctionofspectraltypeforultra- radio to bolometric luminosity as a function of projected rota-
cooldwarfs.Shownareflares(stars),quiescentemission(squares) tionalvelocityforultracooldwarfs.Upperlimits(opentriangles),
andupperlimits(opentriangles).Theresultsfromthissurveyare flares(stars)andquiescentemission(squares)areshownforboth
magentaandvaluesfromtheliterature(Route&Wolszczan2013; this survey (magenta) and the literature (grey; Route & Wol-
Antonovaetal.2013;Burgasseretal.2013,2015;Kaoetal.2015) szczan 2013; Antonova et al. 2013; Burgasser et al. 2013, 2015).
aregrey.Forobjectswithobservedquiescentandflareemission, Thedashedlineinthebottompanelindicatestheradioactivity-
thevaluesareconnectedbydashedlines. rotation saturation level for early type M dwarfs from McLean
etal.(2012).
the radio luminosity to be constant, agreeing with previous
(2012) these more rapidly rotating sources appear to tend
radiosurveyswhichfoundL ∼1023±0.5 ergs−1 forobjects
rad towardthehigherratioofradiotobolometricluminosityof
withspectraltypeM0-L5(Bergeretal.2010).Furthermore, L /L ∼10−6.5.
rad bol
whenwecomparetheratioofradiotobolometricluminosity
to spectral type, we observe the previously noted trend of
increasedactivitywithlaterspectraltype.Thisisincontrast
5.1 X-ray/Radio Correlation
with observations of other activity tracers, such as Hα and
X-rayemission,wherethisratiodecreasespastspectraltype As mentioned in section 1, there is a tight correlation
M7 (Berger et al. 2010). between the radio and X-ray emission for coronally ac-
Figure6showstheradioluminosityandratioofradioto tive stars ranging from spectral type F to mid-M, where
bolometricluminosityasafunctionofrotationrate.2MASS Lν/Lx∼10−15.5Hz−1 (Gu¨del & Benz 1993; Benz & Guedel
J1048−3956, 2MASS J0339-3525, and 2MASS J0004-4044 1994).Thefirstradioobservationofanultracooldwarffound
are all rapid rotators with vsin(i)(cid:38)20 km s−1. The radio Lν/Lx∼10−11.5Hz−1 for this source (Berger et al. 2001),
luminositiesforthesethreesourcesareobservedtobefairly suggesting that the GB relation may be severely violated
constant with rotation rate agreeing with the previous re- by these objects. Subsequent radio detections of ultracool
sults of McLean et al. (2012). However, note that rapid ro- dwarfs have found that these objects display a wide range
tation does not necessarily indicate a source will have ob- of behaviour with regard to the GB relation, where some
servable radio emission. In fact the most rapidly rotating sources are strongly radio overluminous varying from the
sourceinoursampleof15,2MASSJ02550357–4700509,has GB relation by several orders of magnitude, while others
anupperlimitofL ∼9×1011 ergs−1,whichislowerthan could be consistent with the this relation (Williams et al.
rad
the luminosity of the detected sources. 2014).
If we look at the ratio of the radio to bolometric lumi- Using the radio luminosities, L , for our 15 sources as
ν
nosityforthissampleofsourcesweseethat2MASSJ1048- well as X-ray luminosities, Lx, from the literature, we can
3956 and 2MASS J0339-3525 have ratios that lie along the determineifthesesourcesfallalongtheGBrelation.Figure
radio activity-rotation saturation level observed in early-M 7 shows this comparison, where we have also plotted the
dwarfs, Lrad/Lbol ∼10−7.5 (McLean et al. 2012). However, observed data from Gu¨del & Benz (1993) (grey points) and
note that sources with rotation rates (cid:38)30 km s−1, includ- the linear fit of log(Lν)=1.36[log(Lx)−18.9] to this data
ing our observation of 2MASS J0004-4044, all lie above the (Berger et al. 2010). The scatter of the data from Gu¨del &
early-M dwarf saturation level. As noted by McLean et al. Benz(1993)aroundthislineis0.6dexwhenwemeasurethe
MNRAS000,1–10(2016)
Radio detections of southern ultracool dwarfs 9
deviationatafixedLx.Therelativescatterofthedatatothe
best fit line (i.e. measured perpendicular to the line) is 0.2
dex. In order to be consistent with the analysis of Williams
etal.(2014),wedefinethedifferencebetweenthemeasured
ratio of radio to X-ray luminosity and the GB relation as
theperpendiculardistancebetweenthemeasuredvalueand
the best fit line.
Out of the three sources in our survey with detectable
levelsofradioemissiononly2MASSJ1048-3956and2MASS
J0339-3525 have measured X-ray luminosities in the litera-
ture. Comparing the ratio of the radio to X-ray luminos-
ity for these two sources we find that they differ from the
GB relation by 2.5 dex and 1.7 dex, for 2MASS J1048-3956
and2MASSJ0339-3525respectively.Thevariationfromthe
GB relation for these two sources, while significant, is not
nearly as extreme as the variation observed for other ultra-
cool dwarfs (e.g. TVLM-513; Berger et al. 2008; Williams
et al. 2014).
Mostofourmeasuredradioupperlimitsarewithin1.5
dex of the GB relation. Given these upper limits the actual
radioluminosityofthesesourceshasthepotentialtobecon-
Figure7.QuiescentradioluminosityasafunctionofX-raylumi-
sistent with the GB relation. However for 2 of our objects,
nosityforawiderangeofcoronallyactivestars.Thegreypoints
2MASS J1507476–162738 and 2MASS J02550357–4700509, are the results from Gu¨del & Benz (1993) and the red-line is a
theirmeasuredradioupperlimitsplacethem∼2.6dexaway fittothesedatafromBergeretal.(2010).Theresultsfromthis
fromtheGBrelation.Thisindicatesthatthemeasuredradio surveyaregivenbytheblack(5GHz)andmagenta(9GHz)data
luminosities for these two sources would have to be signif- points.Thesolidarrowsrepresenttheupperlimitsfromstacking
icantly less than the upper limits in order for them to be theobservationswithnon-detections.
consistent with the GB relation.
FollowingthemethodoutlinedinHancocketal.(2011), cm−3, spiralling in a magnetic field with a strength of 10 –
westacked11ofthe12observationswithnon-detections.We 300Gaussandorientationbetween40◦–80◦.Theseparame-
excludedtheobservationof2MASSJ00244419–2708242be-
ter ranges are still very large and could benefit from an ob-
cause we were unable to fully remove a bright field source
servation of the spectral turn-over frequency. In the case of
located at R.A. = 00:24:36.077 Decl.=-27.07.44.64. In the
gyrosynchrotronemissionthisfrequencyisdependentonthe
stacked image we do not detect any radio emission and
electrondensityandmagneticfieldorientationandstrength
measure upper limits of 12µJy (5.5 GHz) and 14.1µJy (9.0
(Dulk1985).Hereweassumethatthisfrequencyis<4GHz
GHz). Noting that the X-ray luminosities for these sources
based on the observed spectral energy distributions, how-
are Lx/Lbol>-5andcalculatingtheeffective distance for the ever an actual measurement will place stronger constraints
stakedimagetobe6.5pc,weplotthisupperlimitinFigure
on the model parameters.
7(solidarrows).Similartotheotherupperlimitsfromthis
We also compare the general emission trends of our
survey, this upper limit is 1.3 dex from the GB relation.
sample of 15 sources to the results of previous surveys of
ultracool dwarfs. As observed in previous radio studies of
ultracool dwarfs, we find the observed radio luminosities to
be constant with both spectral type and rotation rate. We
6 SUMMARY
also find that the ratio of radio to bolometric luminosity
From a sample of 15 late-M and L dwarfs, located within to increase towards later type objects and higher rotational
10pcoftheSun,wehavedetectedquiescentradioemission velocities. Additionally, using X-ray luminosities from the
at frequencies between 4.7 and 9.7 GHz for three sources, literaturewefindthattheratioofradiotoX-rayluminosity
2MASSJ1048-3956,2MASSJ0339-3525,and2MASSJ0004- for2MASSJ1048-3956and2MASSJ0339-3525varysignifi-
4044. While both 2MASS J1048-3956 and 2MASS J0339- cantlyfromtheGBrelation.Themajorityoftheradioupper
3525 have previous radio detections in the literature, this limits are (cid:46)1.5 dex from the GB relation, however for two
is the first detection of radio emission from 2MASS J0004- sourcetheupperlimitsdifferfromGBrelationbymorethan
4044.Additionally,weplacethefirstupperlimitsonthera- 2 dex. For these two sources the measured radio luminosity
dio emission from 2MASS J0024-2708, 2MASS J03341218– would have to be much lower than these upper limits to be
4953322, 2MASS J0340-6724, and 2MASS J02550357– consistent with the GB relation.
4700509.Thisincreasesthenumberofultracooldwarfsstud-
iedatradiofrequenciesto216,with17sourceswithobserved
radio emission.
ACKNOWLEDGEMENTS
We find that the observed Stokes I and V radio emis-
sion from 2MASS J1048-3956, 2MASS J0339-3525, and TheAustraliaTelescopeCompactArrayispartoftheAus-
2MASSW J0004348-404405 is well modelled by optically tralia Telescope which is funded by the Commonwealth
thin gyrosynchrotron emission from a homogenous popula- of Australia for operation as a National Facility managed
tion of power-law electrons, with density between 104 – 108 by CSIRO. Parts of this research were conducted by the
MNRAS000,1–10(2016)
10 C. Lynch et al.
Australian Research Council Centre of Excellence for All- Osten R. A., Hawley S. L., Allred J., Johns-Krull C. M., Brown
sky Astrophysics (CAASTRO), through project number A.,HarperG.M.,2006b,ApJ,647,1349
CE110001020. Phan-Bao N., Osten R. A., Lim J., Mart`ın E. L., Ho P. T. P.,
2007,ApJ,658,553
RaviV.,HallinanG.,HobbsG.,ChampionD.J.,2011,ApJ,735,
L2
Reid I. N., Cruz K. L., Kirkpatrick J. D., Allen P. R., Mungall
REFERENCES
F.,LiebertJ.,LowranceP.,SweetA.,2008,AJ,136,1290
AllredJ.C.,HawleyS.L.,AbbettW.P.,CarlssonM.,2006,ApJ, ReinersA.,BasriG.,2008,ApJ,684,1390
644,484 ReinersA.,BasriG.,2009,ApJ,705,1416
Antonova A., Hallinan G., Doyle J. G., Yu S., Kuznetsov A., ReinersA.,BasriG.,2010,ApJ,710,924
MetodievaY.,GoldenA.,CruzK.L.,2013,A&A,549,A131 RichardsM.T.,WaltmanE.B.,GhigoF.D.,RichardsD.S.P.,
BenzA.O.,GuedelM.,1994,A&A,285,621 2003,AstrophysicalJournalSupplementSeries,147,337
BergerE.,2002,ApJ,572,503 RobinsonP.A.,MelroseD.,1984,AustralianJournalofPhysics,
BergerE.,2006,ApJ,648,629 37,675
BergerE.,etal.,2001,Nature,410,338 Rodr´ıguez-Barrera M. I., Helling C., Stark C. R., Rice A. M.,
BergerE.,etal.,2008,ApJ,673,1080 2015,MNRAS,454,3977
BergerE.,etal.,2009,ApJ,695,310 RouteM.,WolszczanA.,2013,ApJ,773,18
BergerE.,etal.,2010,ApJ,709,332 SaultR.J.,TeubenP.J.,WrightM.C.H.,1995,inShawR.A.,
BurgasserA.J.,PutmanM.E.,2005,ApJ,626,486 PayneH.E.,HayesJ.J.E.,eds,AstronomicalSocietyofthe
BurgasserA.J.,MelisC.,ZaudererB.A.,BergerE.,2013,ApJ, PacificConferenceSeriesVol.77,AstronomicalDataAnalysis
762,L3 SoftwareandSystemsIV.p.433(arXiv:astro-ph/0612759)
Burgasser A. J., Melis C., Todd J., Gelino C. R., Hallinan G., ScargleJ.D.,1982,ApJ,263,835
BardalezGagliuffiD.,2015,AJ,150,180 SchmidtS.J.,CruzK.L.,BongiornoB.J.,LiebertJ.,ReidI.N.,
Burrows A., Hubbard W. B., Lunine J. I., Liebert J., 2001, Re- 2007,AJ,133,2258
viewsofModernPhysics,73,719 SchmidtS.J.,HawleyS.L.,WestA.A.,BochanskiJ.J.,Daven-
CrossfieldI.J.M.,2014,A&A,566,A130 portJ.R.A.,GeJ.,SchneiderD.P.,2015,AJ,149,158
CutriR.M.,etal.,2003,VizieROnlineDataCatalog,2246,0 StelzerB.,etal.,2012,A&A,537,A94
DeaconN.R.,HamblyN.C.,CookeJ.A.,2005,A&A,435,363 VanderPlas J., Connolly A. J., Ivezic Z., Gray A., 2012,
DulkG.A.,1985,ARA&A,23,169 in Proceedings of Conference on Intelligent Data Un-
FahertyJ.K.,BurgasserA.J.,CruzK.L.,SharaM.M.,Walter derstanding (CIDU). pp 47–54 (arXiv:1411.5039),
F.M.,GelinoC.R.,2009,AJ,137,1 doi:10.1109/CIDU.2012.6382200
Gizis J. E., Monet D. G., Reid I. N., Kirkpatrick J. D., Liebert WestA.A.,etal.,2004,AJ,128,426
J.,WilliamsR.J.,2000,AJ,120,1085 WilliamsP.K.G.,BergerE.,2015,ApJ,808,189
Gu¨delM.,BenzA.O.,1993,ApJ,405,L63 WilliamsP.K.G.,CookB.A.,BergerE.,2014,ApJ,785,9
GuentherE.W.,ZapateroOsorioM.R.,MehnerA.,Mart´ınE.L., WilsonW.E.,etal.,2011,MNRAS,416,832
2009,A&A,498,281
HallinanG.,AntonovaA.,DoyleJ.G.,BourkeS.,BriskenW.F., ThispaperhasbeentypesetfromaTEX/LATEXfilepreparedby
GoldenA.,2006,ApJ,653,690 theauthor.
HallinanG.,etal.,2007,ApJ,663,L25
HallinanG.,etal.,2015,Nature,523,568
HancockP.J.,GaenslerB.M.,MurphyT.,2011,ApJ,735,L35
HawleyS.L.,GizisJ.E.,ReidI.N.,1996,AJ,112,2799
Kao M. M., Hallinan G., Pineda J. S., Escala I., Burgasser A.,
BourkeS.,StevensonD.,2015,preprint, (arXiv:1511.03661)
Leto G., Pagano I., Linsky J. L., Rodon`o M., Umana G., 2000,
A&A,359,1035
LombN.R.,1976,Ap&SS,39,447
LynchC.,MutelR.L.,Gu¨delM.,2015,ApJ,802,106
MachadoM.E.,AvrettE.H.,VernazzaJ.E.,NoyesR.W.,1980,
ApJ,242,336
McLean M., Berger E., Irwin J., Forbrich J., Reiners A., 2011,
ApJ,741,27M
McLeanM.,BergerE.,ReinersA.,2012,ApJ,746,23
McMullin J. P., Waters B., Schiebel D., Young W., Golap K.,
2007, in Shaw R. A., Hill F., Bell D. J., eds, Astronomical
SocietyofthePacificConferenceSeriesVol.376,Astronomical
DataAnalysisSoftwareandSystemsXVI.p.127
MohantyS.,BasriG.,ShuF.,AllardF.,ChabrierG.,2002,ApJ,
571,469
Mutel R. L., Molnar L. A., Waltman E. B., Ghigo F. D., 1998,
ApJ,507,371
Neuh¨auserR.,etal.,1999,A&A,343,883
NeupertW.M.,1968,ApJ,153,L59
OstenR.A.,JayawardhanaR.,2006,ApJ,644,L67
OstenR.A.,WolkS.J.,2009,ApJ,691,1128
OstenR.A.,HawleyS.L.,BastianT.S.,ReidI.N.,2006a,ApJ,
637,518
MNRAS000,1–10(2016)