The discovery, monitoring and environment of SGR J1935+2154

0.53 MB·

by  G. L. Israel

Mon.Not.R.Astron.Soc.000,000–000(0000) Printed5January2016 (MNLaTEXstylefilev2.2)
The discovery, monitoring and environment of SGRJ1935+2154
G.L. Israel,1⋆ P. Esposito,2,3 N. Rea,4,5 F. Coti Zelati,6,4,7 A. Tiengo,8,9,10
6 S. Campana,7 S. Mereghetti,8 G.A. Rodriguez Castillo,1 D. Go¨tz,11 M. Burgay,12
1
0 A. Possenti,12 S. Zane,13 R. Turolla,14,13 R. Perna,15 G. Cannizzaro,16 and J. Pons 17
2
1OsservatorioAstronomicodiRoma,INAF,viaFrascati33,I-00040MonteporzioCatone,Italy
n 2IstitutodiAstrofisicaSpazialeeFisicaCosmica-Milano,INAF,viaE.Bassini15,I-20133Milano,Italy
a 3Harvard-SmithsonianCenterforAstrophysics,60GardenStreet,Cambridge,MA02138,USA
J 4AntonPannekoekInstituteforAstronomy,UniversityofAmsterdam,Postbus94249,NL-1090-GEAmsterdam,TheNetherlands

3 5InstitutodeCienciasdel’Espacio(ICE,CSIC–IEEC),CarrerdeCanMagrans,S/N,08193,Barcelona,Spain
6Universita`dell’Insubria,viaValleggio11,I-22100Como,Italy

] 7INAF–OsservatorioAstronomicodiBrera,viaBianchi46,I-23807Merate(LC),Italy
E 8INAF–IstitutodiAstrofisicaSpazialeeFisicaCosmica,viaE.Bassini15,I-20133Milano,Italy
H 9IUSS–IstitutoUniversitariodiStudiSuperiori,piazzadellaVittoria15,I-27100Pavia,Italy
. 10INFN–IstitutoNazionalediFisicaNucleare,SezionediPavia,viaA.Bassi6,I-27100Pavia,Italy
h 11AIMCEA/Irfu/Serviced’Astrophysique,OrmedesMerisiers,F-91191Gif-sur-Yvette,France
p 12OsservatorioAstronomicodiCagliari,viadellaScienza5,09047,Cagliari,Italy
- 13MullardSpaceScienceLaboratory,UniversityCollegeLondon,HolmburySt.Mary,Dorking,SurreyRH56NT,UK
o
14DipartimentodiFisicaeAstronomia,Universita`diPadova,viaF.Marzolo8,I-35131Padova,Italy
r
t 15DepartmentofPhysicsandAstronomy,StonyBrookUniversity,StonyBrook,NY11794,USA
s 16Universita` diRoma”LaSapienza”,P.leA.Moro5,00185,Roma,Italy
a
17DepartamentdeF`ısicaAplicada,Universitatd’Alacant,Ap.Correus99,03080Alacant,Spain
[

1
v
7 5January2016
4
3
0
ABSTRACT
0
Wereportonthediscoveryofanewmemberofthemagnetarclass,SGRJ1935+2154,andon
.
1 its timing and spectral propertiesmeasured by an extensive observationalcampaign carried
0 outbetweenJuly2014andMarch2015withChandraandXMM-Newton(11pointings).
6
We discovered the spin period of SGRJ1935+2154 through the detection of coher-
1
ent pulsations at a period of about 3.24s. The magnetar is slowing-down at a rate of
:
v P˙=1.43(1)×10−11ss−1andwithadecreasingtrendduetoanegativeP¨ of−3.5(7)×10−19
i ss−2.Thisimpliesasurfacedipolarmagneticfieldstrengthof∼2.2×1014G,acharacteristic
X
age of about3.6kyrand,a spin-downluminosityLsd ∼1.7×1034ergs−1. The sourcespec-
r
a trum is well modelled by a blackbody with temperature of about 500eV plus a power-law
componentwithphotonindexofabout2.Thesourceshowedamoderatelong-termvariabil-
ity, with a flux decay of about 25% during the first four months since its discovery, and a
re-brighteningofthesameamountduringthesecondfourmonths.
The X-ray data were also used to study the source environment. In particular, we dis-
coveredadiffuseemissionextendingonspatialscalesfromabout1′′ uptoatleast1′ around
SGRJ1935+2154 both in Chandra and XMM-Newton data. This component is constant in
flux(atleastwithinuncertainties)anditsspectrumiswellmodelledbyapower-lawspectrum
steeperthanthatofthepulsar.Thoughascatteringhalooriginseemstobemoreprobablewe
cannotexcludethatpart,orall,ofthediffuseemissionisduetoapulsarwindnebula.
Key words: stars: neutron – stars: magnetars – X-rays: bursts – X-rays: individual:
SGRJ1935+2154
1 INTRODUCTION
Large observational and theoretical efforts have been devoted in
the past years to unveil the nature of a sample of peculiar high-
⋆ E-mail:[email protected] energypulsars,namelytheAnomalousX-rayPulsars(AXPs)and
2 G.L. Israelet al.
theSoftGamma-rayRepeaters(SGRs).Theseobjectsarebelieved
Table1.SummaryoftheSwift,ChandraandXMM-Newtonobservations
tobeisolatedneutronstarsandpoweredbytheirownmagneticen- usedinthisworkandcarriedoutbetweenJuly2014andMarch2015.
ergy,storedinasuper-strongfield,andarecollectivelyreferredto
asmagnetars(Duncan&Thompson1992;Paczynski1992).They Mission/Obs.ID Instrument Date Exposure
sharesimilartimingproperties(spinperiodP inthe2–12srange (ks)
and period derivative P˙ in the 10−13–10−11 ss−1, range). Their
X-rayluminosity,typicallyLX∼1033–1035ergs−1,generallyex- SSwwiifftt//660033448888000002 XXRRTT JJuull56 34..43
ceedstherotationalenergylossrate,whilethetemperaturesofthe
Swift/603488004 XRT Jul7 9.3
thermalcomponentobservedintheirspectraareoftenhigherthan
Swift/603488006 XRT Jul8 3.7
thosepredictedbymodelsofnon-magneticcoolingneutronstars.
Swift/603488008 XRT Jul13 5.3
Their (surface dipolar) magnetic fields inferred from the dipolar- Swift/603488009 XRT Jul13 3.0
loss formula are generally of the order of B∼1014 – 1015 G. Chandra/15874 ACIS-S Jul15 10.1
However, recently low dipole field magnetars have been discov- Swift/603488010 XRT Jul16 7.1
ered,whichbehaveastypicalmagnetarsbutwithdipolarmagnetic Chandra/15875 ACIS-Sa Jul28 75.4
fieldaslow as6×1012G, i.e.in therange of normal radio pul- Chandra/17314 ACIS-Sa Aug31 29.2
sars(Reaetal.2010):thesesourcespossiblystorelargemagnetic XMM/0722412501 EPIC Sep26 19.0
XMM/0722412601 EPIC Sep28 20.0
energy in other components of their magnetic field (Turollaetal.
XMM/0722412701 EPIC Oct04 18.0
2011;Reaetal.2013).
XMM/0722412801 EPIC Oct16 9.7
Sporadically, magnetars emit high energy (up to the MeV
XMM/0722412901 EPIC Oct24 7.3
range)burstsandflareswhichcanlastfromafractionofaseconds XMM/0722413001 EPIC Oct27 12.6
tominutes,releasing∼1038÷1047 ergs−1,oftenaccompaniedby XMM/0748390801 EPIC Nov15 10.8
long-lived(uptoyears)increasesofthepersistentX-rayluminosity XMM/0764820101 EPIC Mar25 28.4
(outbursts).Theseeventsmaybeaccompaniedortriggeredbyde-
a Datacollectedincontinuousclockingmode(CC).
formationsorfracturesoftheneutronstarcrustand/orlocal/global
rearrangements of the star magnetic field. The detection of these
energeticeventsprovidesthemainchanneltoidentifynewobjects (Stamatikosetal. 2014). Follow-up observations carried out by
ofthisclass. Chandra on 2014 July 15 and 29 allowed us to precisely lo-
Afundamentalquestionaboutmagnetarconcernstheirevolu- cate the source and detect its spin period (P=3.25s; Israeletal.
tionary link to their less magnetic siblings, the rotation-powered 2014)confirmingthatSGRJ1935+2154isindeedamagnetar.The
pulsars. A number of unexpected results, both from known and SGRJ1935+2154 position is coincident with the center of the
newly discovered magnetars, drasticallychanged our understand- Galacticsupernovaremnant(SNR)G57.2+0.8ofundeterminedage
ing of these objects. In 2004, while studying the emission prop- andatapossible,butuncertain,distanceof9kpc(Sunetal.2011;
ertiesofthebrightX-raytransientmagnetarXTEJ1810−197,the Pavlovic´etal.2013).
sourcewasdiscoveredtobeabrighttransientradiopulsar,thefirst InthispaperwereportontheresultsofanXMM-Newton and
oftheclass(Camiloetal.2006).Todayweknowthat4outofthe Chandra observational campaign covering the first 8 months of
about25knownmagnetars,areoccasionallyshiningasradiopul- SGRJ1935+2154’soutburst.Ourobservationalcampaignisongo-
sarsintheoutburstphase.Alltheradio“active”magnetarsarechar- ingwithXMM-Newton,anditslong-termresultswillbereported
acterizedbyaquiescentX-rayoverspin-downluminosityratioof elsewhere. We also report upper limitson the radio emission de-
LX/Lsd <1(Reaetal.2012). rived from Parkes observations (Burgayetal. 2014). We first re-
Energeticpulsarsareknowntoproduceparticleoutflows,of- portonthedataanalysis,thensummarizetheresultsweobtained
tenresultinginspectacularpulsarwindnebulae(PWNe)ofwhich fortheparameters,propertiesandenvironmentofthisnewmagne-
the Crab is the most famous example (Weisskopfetal. 2000). tar.Finallywediscussourfindingsinthecontestofthemagnetar
Magnetars are expected to produce particle outflows as well, ei- scenario.
therinquiescenceorduringoutburstsaccompanyingbrightbursts.
Giventhestrongmagneticfieldsassociatedwiththisclassofneu-
tron stars, the idea of a wind nebula around a magnetar is thus
promising. There has not been yet a confirmed detection of such 2 X–RAYOBSERVATIONS
anebula, butsomecases of“magnetically powered” X-raynebu-
2.1 Chandra
laearound pulsars withrelativelyhigh magnetic fieldshavebeen
suggested.Apeculiarextendedemissionhasbeenreportedaround ChandraobservationsofSGRJ1935+2154 werecarriedoutthree
the rotating radio transient RRAT J1819-1458 (Reaetal. 2009; timesduringJulyandAugust2014(seeTable1)inresponsetothe
Camero-Arranzetal.2013),withanominalX-rayefficiencyηX ∼ detectionofshortSGR-likeburstsfromthesource.Thefirstdataset
0.2, too high to be only rotationally powered. The authors sug- wasacquiredwiththeACIS-SinstrumentinFaintimaging(Timed
gestedthattheoccurrenceofthenebulamightbeconnected with Exposure)and1/8subarraymode(timeresolution:∼0.44s),while
the high magnetic field (B = 5×1013 G) of the pulsar. Similarly, the subsequent two pointings were obtained with the ACIS-S in
Younesetal.(2012)reportedthediscoveryofapossiblewindneb- Faint timing (Continuous Clocking) mode (time resolution 2.85
ulaaroundSwiftJ1834−0846,withanX-rayefficiencyηX ∼0.7 ms).
(butseeEspositoetal.2013foradifferentinterpretationinterms ThedatawerereprocessedwiththeChandraInteractiveAnal-
ofdustscatter). ysisofObservationssoftware(CIAO,version4.6)usingthecalibra-
SGRJ1935+2154 is a newly discovered member of the tionfilesavailableintheChandraCALDB4.6.3database.Thesci-
magnetar family, and was discovered thanks to the detection entificproductswereextractedfollowingstandardprocedures,but
of low-Galactic latitude short bursts by Swift on 2014 July 5 adoptingextractionregionswithdifferentsizeinordertoproperly
The monitoringof SGRJ1935+2154 3
subtracttheunderlyingdiffusecomponent(seeSection3.2andFig- 3 ANALYSISANDRESULTS
ure 1). Correspondingly, for the first observation (Faint imaging)
weused circular regions of 1.5′′ (and 3.0′′) radius for thesource 3.1 Position
(and diffuseemission) associated toabackground annular region WeusedtheChandraACIS-Sobservationcarriedouton2014July
with1.6′′ and 3.0′′ (10′′, 15′′) for the inner and outer radius, re- the15th, the only one inimaging mode, inorder toprecisely lo-
spectively.Furthermoreweusedrectangularboxesof3′′×2′′(and cateSGRJ1935+2154.Onlyonebrightsourcewasdetectedinthe
4′′×2′′)sidesalignedtotheCCDreadoutdirectionfortheremain- S7 CCD operating at 1/8 of the nominal field of view. The re-
ingtwoobservationsinCCmode.Forthebackgroundweusedtwo finedpositionofthesource,calculatedwithWAVDETECT,isR.A.=
rectangularboxesof1.5′′ ×1.5′′ (and2′′ ×2′′)atthesidesofthe 19h34m55s.5978s,Dec.=+21o53′47′.′7864(J2000.0;statisticalun-
sourceextractionregion.Forthespectra,theredistributionmatrices certaintyof0′.′02)witha90%confidenceleveluncertaintyradius
andtheancillaryresponsefileswerecreatedusingSPECEXTRACT. of 0′.′7. This position is consistent with that of SGRJ1935+2154
For the timing analysis, we applied the Solar system barycentre obtained by Swift: R.A. = 19h34m55s.68s, Dec. = +21o53′48′.′2,
correctiontothephotonarrivaltimeswithAXBARY. J2000.0, radius of 2′.′3 at 90% confidence level (Cummingsetal.
2014). Correspondingly, we are confident that the source we de-
tectedintheChandraimageisindeedthesourcefirstdetectedby
SwiftBATandlaterbyXRTandresponsiblefortheobservedSGR-
2.2 XMM-Newton likebursts.
XMM-Newton observations of SGRJ1935+2154 were carried out
betweenSeptemberandMarch2015(seeTable1)tomonitorthe
3.2 Spatialanalysis
sourcedecayandstudythesourceproperties.Weusedthedatacol-
lected with the European Photon Imaging Camera (EPIC), which Upon visual inspection of the X-ray images, it is apparent that
consistsoftwoMOS(Turneretal.2001)andonepn(Stru¨deretal. SGRJ1935+2154 isembedded inapatch of diffuse emission. To
2001) CCD detectors. The raw data were reprocessed using the assessthisindetail,webuiltforeachpnobservationaradialpro-
XMM-NewtonScienceAnalysisSoftware(SAS,version14.0)and fileinthe0.4–10keVbandandfitapointspreadfunction(approxi-
the calibration files in the CCF release of 2015 March. The pn matedbyaKingmodel;Readetal.2011)toit.Ineachinstance,the
operated in Full Window (time resolution of about 73 ms) while innerpartoftheprofilecanbefitbyaKingmodelwithusualcore
the MOSs were set in Small Window (time resolution of 300 radiusandslopevalues,whereasatradii≈30–40′′thedatastartto
ms), therefore optimized for the timing analysis. The intervals exceedsignificantlythemodelprediction.Sinceweobtainedcon-
of flaring background were located by intensity filters (see e.g. sistentresultsfromallthe2014observations,werepeatedthesame
DeLuca&Molendi2004)andexcludedfromtheanalysis.Source analysis on thestacked images inorder toimprove the signal-to-
photons were extracted from circles with radius of 40′′. The pn noiseratioofthedata.Wealsoselectedthephotonsinthe1–6keV
background was extracted froman annular region withinner and energy range, since the spectral analysis (see Section3.4) shows
outerradiiof45′′and90′′,respectively(alsointhiscasethechoice thatthediffuseemissionismoreprominentinthisband.Thecom-
wasdictated by thediffuse emissioncomponent; Section3.2 and bined2014XMM-NewtonprofileisshowninblackinFig.1.The
Figure1).PhotonarrivaltimeswereconvertedtotheSolarsystem diffuseemissionemergesat&30′′ fromSGRJ1935+2154andex-
barycenter usingthe SAStask BARYCENusingthesourcecoordi- tendstoatleast70′′.Itishowevernotpossibletodeterminewhere
nateasinferredfromtheChandrapointings(seeSection3.1).The the feature ends, because of both the low-signal to noise at large
ancillaryresponsefilesandthespectralredistributionmatricesfor distance from the point source and the gaps between the CCDs.
thespectral analysis weregenerated with ARFGEN and RMFGEN, The profile of the latest XMM-Newton dataset has been obtained
respectively.Inordertomaximizethesignaltonoiseratiowecom- separately fromtheremaining datasetsin order tolook for shape
bined,whenneeded,thespectrafromtheavailableEPICcameras variabilitiesofthediffusecomponentonlongtimescales.Thetwo
andaveragedtheresponsefilesusingEPICSPECCOMBINE.Inpar- pnprofilesareinagreementwithintheuncertainties(determinedby
ticular,thelattercommand wasroutinelyappliedforthestudyof usingaKolmogorov–Smirnovtestthatthereisasubstantialprob-
thedimdiffuseemission. ability(¿50%) that the twoprofiles have been extracted from the
same distribution), though a possible shift of the diffuse compo-
nent,towardslargerradii,mightbepresentinthe30′′-40′′ radius
interval.
2.3 Swift AsimilaranalysiswascarriedoutbyusingthelongestChan-
dradataset.ThoughthelatterisinCCmode,thefieldisnotpartic-
The Swift X-Ray Telescope (XRT) uses a front-illuminated
ularlycrowdedandonlyfaintpoint-likeobjectsaredetectedinthe
CCD detector sensitive to photons between 0.2 and 10 keV
fieldofview.Correspondingly,itisstillpossibletogatherinforma-
(Burrowsetal. 2005) . Two readout modes can be used: pho-
tionoversmallerscalesthanintheXMM-Newtondata.theACIS-
ton counting (PC) and windowed timing (WT). The PC mode
SPSFwassimulatedusingtheChandraRayTracer(ChaRT)and
provides images and a 2.5s time resolution; in WT mode only
Model of AXAF Response to X-rays (MARX v5.0.0-0) software
one-dimensional imaging is preserved with a time resolution of packages1.TheresultofthisanalysisisshowninblueinFigure1.
1.766ms. Data were processed with XRTPIPELINE (version 12), DiffuseemissionisclearlypresentintheChandradataandstarts
and ltered and screened with standard criteria, correcting for ef- becomingdetectableatadistanceof>1′′ fromthesource.Dueto
fective area, dead columns, etc. Events wereextracted from a 20
poor statistics we have no meaningful information at radii larger
pixel radius region around the source position. For spectroscopy
weusedthespectralredistributionmatricesinCALDB(20130101,
v014forthePC),whiletheancillaryresponselesweregenerated 1 Formoredetailsonthetasksseehttp://cxc.harvard.edu/chart/index.html
withXRTMKARF. andhttp://space.mit.edu/cxc/marx/index.html
4 G.L. Israelet al.
22:00:03.5 Table2.TiminEgporecshulTts0.(MJD) 56926.0
Validityrange(MJD) 56853.6–56976.4
57:07.1 PP˙((TT00))(s) 13..24435(016)5×0(11)0−11
P¨(T0)(s−1) −3.5(7)×10−19
54:10.7 νν˙((TT00))((HHzz)s−1) 0−.310.3861060(32)3(×1)10−12
ν¨(T0)(Hzs−2) 3.3(7)×10−20
21:51:14.3 rχm2νs(rde.soi.df.u)al(ms) 505.57(6)
Bp(Gauss) 2.2×1014
48:17.9 Lτcsd(y(re)rgs−1) 13.6700×1034
45:21.5 36.0 24.0 12.0 19:35:00.0 48.0 36.0 34:24.0 XMM-Newtonwas0.11±0.02ctss−1 and0.21±0.01ctss−1,re-
spectively.Coherentpulsationsataperiodofabout3.24swerefirst
6 discoveredinthe2014July29ChandradatasetcarriedoutinCC
0
1 mode(Israeletal.2014).Thepulseshapeisnearlysinusoidaland
pn 2014
doesnotshowvariationsasafunctionoftime.Alsothepulsedfrac-
5 pn 2015
0 tion,definedasthesemi-amplitudeofthesinusoiddividedbythe
1 ACIS 2014
−2 source average count rate, is timeindependent (withinuncertain-
n
mi 4 ties)andinthe17%÷21%range(1σuncertaintyofabout1.5%).
0
rc 1 Additionally,thepulseshapedoesnotdependontheenergyrange,
a
s 0 thoughashiftinphaseofabout0.16cyclesisclearlydetectedbe-
unt 00 tweenthesoft(0.5-1.5keV)andhard(3.0-12.0keV)energybands,
o 1 withhardphotonsanticipatingthesoftones(seeFigure2).
C
0 A refined value of P=3.244978(6) s (1σ confidence level;
0
1 epoch56866.0MJD)wasinferredbasedonaphase-coherentanal-
ysis. Due to thelong timeelapsed between the epoch of the first
perioddeterminationandthoseoftheotherChandraobservations
we were not able to further extend the timing solution based on
the Chandra data. Therefore, we inferred a new phase-coherent
o 2
ati solution by means of the seven XMM-Newton pointings carried
R
out between the end of September and mid November 2014 (red
1
filledcirclesinleft panel of Figure2). The new solution alsoin-
cludedafirstperiodderivativecomponent:P=3.2450656(2)sand
0 50 100 P˙=1.37(3)×10−11 s s−1 (1σ confidence level; epoch 56926.0
Radius (arcsec) MJD;χ2of3.1for4degreeoffreedom).
Figure 1. Top: 98ks-long XMM-Newton PN image of the region around Thelattertimingsolutionwasaccurateenoughtoincludethe
SGRJ1935+2154;the1.4GHzradiomapofSNRG57.2+0.8isalsoshown previousChandrapointings(blackfilledcirclesinleftpanelofFig-
(bluecontoursfromtheVLAGalacticPlaneSurvey,Stiletal.2006;upper ure2).Thefinaltimingsolution,encompassingthewholedataset,
image).TheXMM-NewtonimagehasbeensmoothedwithaGaussianfunc- isreportedinTable2andincludesasecondperiodderivativeacting
tionwitharadiusof4′′andmagentacontoursaredisplayedinordertoem- inthedirectionofdeceleratingtherateofperiodchangeP˙.Thein-
phasisetheextendedemissionaroundSGRJ1935+2154.Theblackdashed clusionofthenewP¨componenthasaF-testprobabilityof8×10−4
circle marks adistance of90′′ from theSGRJ1935+2154 position. Bot- and10−7 ofnotbeingneeded (whenconsideringonlytheXMM-
tom:2014and2015XMM-NewtonandChandrasurfacebrightness(black
Newtondatasetsorthewholetenpointingsinthefit,respectively).
crosses,purplesquaresandbluecrosses,respectively)asafunctionofthe
Moreover, the new timing solution implies a r.m.s. variability of
distance fromSGRJ1935+2154 compared with their Point Spread Func-
tions(PSF;redlines;lowerplot).TheratiosbetweenthedataandthePSF only55ms,correspondingtoatimingnoiseleveloflessthan2%,
areplottedinthelowestpanel. wellwithinthevaluerangeobservedinisolatedneutronstars.
Wenotethatthesecondperiodderivativewefoundisunlikely
toresultfromachange,asafunctionoftime,ofthepulseprofiles,
than∼15′′.Therefore,wearenotabletoassessifthediffusestruc- which are almost sinusoidal and show no evidence for variation
turesdetectedbyXMM-NewtonandChandraareunrelatedtoeach (seerightpanelofFigure2).
otherorlinkedsomehow. Wenoticethatthisanalysisisvalidundertheassumptionthat
thelocationandgeometryoftheemittingregionremainsconstant
throughouttheobservations,assuggestedbystudiesofothertran-
3.3 Timinganalysis
sientmagnetars(seePerna&Gotthelf2008;Albanoetal.2010).
The 0.5-10keV events were used to study the timing properties The accuracy of the timing solution reported in Table2 is
ofthepulsar. Theaveragecount rateobtainedfromChandra and not good enough to coherently include the March 2015 XMM-
The monitoringof SGRJ1935+2154 5
55 0
−− .5
3 (a) −
1
.5
Phase (rad) −20−15−10−20−15−10 nd shifted Intensity 2 ((bc)) 1.5−2.0 2.0−3.0 3.0−1 Eenergy (keV)
uals (s) 0.20.40.20.4 Normalized a ((de)) 2.0 0.5−12
d 00
Resi 0.20.2 1 (f)
−−
−50 0 50 0 0.5 1 1.5 2
Time (d − 16926.0 TJD) Phase
Figure2.Left:SGRJ1935+2154’sphaseevolutionasafunctionoftimefittedwithaalinearplusaquadraticplusacubiccomponents(upperpanel).The
residualswithrespecttoourbestphase-coherentsolutionarereportedinthelowerpanel,inunitsofseconds.BlackandreddpointsmarktheChandraand
XMM-Newtonobservations,respectively.Right:ChandraplusXMM-Newtonbackground-subtractedpulseprofiles(arbitraryshiftedonthey-axis).Fromtop
tobottomtheyreferto:(a)0.5-1.5keV,(b)1.5-2.0keV,(c)2.0-3.0keV,(d)3.0-12.0keVand(e)0.5-12.0keV.Thedashedorangecurvemarksthebestfit(by
assumingamodelwithtwosinusoids)ofprofile(a):asystematicshifttowardssmallerphases(advanceintime)asafunctionofenergyisevident.Profile(f)
hasbeenobtainedbyaligningprofilesfrom(a)to(d).
Newtondata.Correspondingly, weinferredtheperiodforthislat- freedom,hereafterd.o.f.,fortheChandraandXMM-Newtonspec-
estpointingsimilarlytowhat reportedabovefindingabestvalue tra,respectively).Acanonicaltwo-componentmodeloftenusedto
ofP=3.24528(6) s(95%confidencelevel;epoch57106.0MJD). modelmagnetars,i.e.anabsorbedBBplusPL,resultedinagood
This is less than 2σ away from the expected period extrapolated fit with reduced χ2 of 0.99 (280 d.o.f.) and 1.03 (405 d.o.f.) for
from the timing solution in Table2. The pulse profile parameters theChandraandXMM-Newtonspectra,respectively.Theinclusion
changed significantlywithrespect totheprevious datasetswitha of a further spectral component (the BB in the above procedure)
pulsedfractionofonly5±1%(1σ)andamoreasymmetricshape. wasevaluatedtohaveaformalF-testprobabilityequalto4.5σand
7.0σ (forChandra andXMM-Newton,respectively) ofbeingsig-
nificant.
3.4 Spectralanalysis A fluxvariation, of the order of about 25%, was clearly de-
tectedbetweentheChandraandXMM-Newton2014pointings.On
For the phase-averaged spectral analysis (performed with XSPEC
the other hand no significant flux variation was detected among
12.8.2 fitting package; Arnaud 1996) we started by considering
spectra of XMM-Newton observations. Correspondingly, in order
all the datasets together. Then, we concentrated on the 29 July
toincreasethestatisticsweproceededtocombinethesevenXMM-
2014data,beingthelongestandhigheststatisticsChandrapointing
Newton 2014 spectra together (we used the SAS task EPICSPEC-
(about 75ks effective exposure for 8200 photons) and the XMM-
COMBINE).ByusingthelatterspectrumweobtainaF-testproba-
Newton pn spectra (effective exposure time of about 105ks and
bilityof7.8σthattheBBcomponentinclusionissignificant.Inthe
22000events).AsummaryofthespectralfitsisgiveninTable3.To
upper panel ofFigure3theXMM-Newtoncombined sourcespec-
accountfortheabovereporteddiffusecomponent(seeSection3.2)
trum(inblack)isreportedtogetherwiththeChandraspectrumof
weused,asbackgroundspectraofthepoint-likecentralsource,the
thelongestpointing(inred;thetwofurtherChandraspectraarenot
regions we described in Section2.1 and 2.2 and from which we
showninFigureforclaritypurposes). Wenote that,withinabout
extractedlaterthediffusecomponentspectra.
1σuncertainties,theChandraandXMM-Newtonspectralparame-
Westartedbyfittingallthe10datasetscarriedoutduring2014
terareconsistentwitheachotherwiththeexceptionoftheflux.
separatelyleaving freetovary alltheparameters. Theabsorption
was forced to be free but the same among observations. Photons ThelatestXMM-Newtonpointing,carriedoutinMarch2015,
having energies below about 0.8keV and above 10keV were ig- wasnotcombinedwiththepreviousonesinordertolookforspec-
nored, owing to the very few counts from SGRJ1935+2154 (en- tralvariabilityonlongtimescales.WhilethePLplusBBspectral
ergychannelswererebinnedinawayofhavingatleast30events). decompositionholdsalsoforthisdataset,thefluxsignificantlyin-
Furthermore,alltheenergychannelsconsistentwithzeroafterthe creasedbyabout25%reachingalevelsimilartothatofthelongest
background subtractionwereignored. Theabundances usedwere ChandrapointinginJuly2014.ItisevidentfromTable3thatthe
those of Wilmsetal. (2000). The spectra were not fitted well by onlysignificantlychangedparameteristhefluxofthePLcompo-
any single component model such asa power-law (PL)or black- nent.
body (BB) which gave a reduced χ2 in the 1.2 – 1.8 range de- Due to the poor statistics of the Swift XRT spectra we only
pending on the used single component (282 and 407 degrees of inferred the 1-10keV fluxes by assuming the PL plus BB model
6 G.L. Israelet al.
1 thatthereisnosuggestionofvariabilityforthediffuseemission.A
SGR XMM PN combined (from the seven XMM-Newton pointings) spectrum for
SGR Chandra
thediffuseemissionwasobtained,inawaysimilartothatalready
Diffuse XMM
described for the source spectrum. The XMM-Newton combined
V−1 0.1 spectrumofthediffuseemissionandtheresultsofthespectralfit-
e
k ting for the PL model are shown in Figure3 and in Table3. Two
−1 factscanbeimmediatelyevinced:asimplemodelisnotagoodap-
s
s 0.01 proximationforthediffuseemissionandtheabsorbingcolumnis
nt
u significantlydifferentfromtheoneweinferredforthemagnetar.At
o
C presentstagewecannotexcludethatthetwothingsarerelatedto
10−3 eachother.Inparticular,wenotethatthelargestvaluesoftheresid-
uals originated from few “random” datapoints rather than by an
up-and-down trend(oftensuggestingawrongadoptedcontinuum
4
σ) component; see blue points in the lower panel of Figure3). Also
∆als ( 02 foobrsethrveadtiioffnusineeomrdiesrsitoonlowoekkfeoprtsspeepcatrraatlevdatrhiaeti2o0n1s5.UXnMfoMrt-uNneawtetolyn,
du thelowstatisticspreventedusincheckingifchangesinthespec-
esi −2 tral parameters are present. The inferred 1-10keV observed flux
R −4 is(1.67±00..0035)×10−13ergcm−2s−1,in agreement withthe2014
1 2 5 10 value.
Energy (keV)
16
3.5 Pre-outburstobservations
) 14
−13 m−2 Swift XRT observed SGRJ1935+2154 twice before its activation
0
1c 12 duringtheSwiftGalacticplanesurvey(seeCampanaetal.2014).
ux x g s −1 10 T00h0e4fi5r2s7t8o0b0s1e)r.vSatGioRnJt1o9o3k5p+l2a1c5e4oinsDfaercof3f0-a,x2i0s1(0∼f1o0r′)5a1n4dsw(oebdsied-
Fl(er riveda3σupperlimitof3.2×10−2ctss−1.
8
The second observation took place on Aug 28, 2011 for
1 10 100 617s(obsid00045271001). SGRJ1935+2154isdetectedatarate
Time (d since TJD 16843.0) (1.55±0.63)×10−2ctss−1.Assumingthesamespectralmodel
oftheXMM-Newtonobservations(seeSection3.4andTable3,we
Figure3.SpectraofSGRJ1935+2154andofthediffuseemissionaround
thepulsar.Fromtoptobottom:SGRJ1935+2154cumulativeXMM-Newton derive a1-10keV luminosity of (9.3±3.6)×1033 erg s−1 (in-
PN spectrum, the SGRJ1935+2154 Chandra ACIS spectrum of obser- cludinguncertaintiesinthecountrateandassumingadistanceof
vation 15875 and the cumulative XMM-Newton PN spectrum of the dif- 9kpc).
fuseemission(upper plot).Residuals (isσ units)areshownandreferto Thefieldwasalsoimaged duringtheROSATall-skysurvey
theabsorbed PL+BBmodelforSGRJ1935+2154andtoaPLmodelfor twice,butthehighcolumndensitypreventsanyfirmupperlimiton
thediffusecomponent.Timeevolutionfortheabsorbed1-10keVfluxof theobservedflux.
SGRJ1935+2154obtainedbyusingdatasetsfromSwift(blacktriangles),
Chandra(redsquares)andXMM-Newton(blackcircles).Thezeroonthe
x-axismarkstheSwiftBATtrigger.
4 RADIOOBSERVATIONS
The first radio follow-up observations of SGRJ1935+2154 were
obtainedbythecombinedXMM-Newtonspectrumandincludinga carriedouton9and14July2014fromtheOotyRadioTelescope
scalefactorwhichwasfreetovaryinordertotrackthefluxvaria- (ORT) and the Giant Meterwave Radio Telescope (GMRT), at
tionthroughtheoutburst.ThelowerpanelofFigure3includesall 326.5and610.0MHz,respectively(Surnisetal.2014).Nopulsed
the1-10keVobservedfluxesinferredfromtheSwift,Chandraand radioemissionwasfounddowntoafluxof0.4mJyand0.2mJyat
XMM-Newtonspectra.Itisevidentthatthesourceisstillvariable 326.5and610.0MHz(assuminga10%dutycycle),respectively.
aboveageneraldecaytrend. The source was observed with the Parkes radio telescope at
The same background regions used to correct the EPIC pn 10-cmand20-cminfourepochsbetween1and3August,shortly
sourcespectrawerethenassumedasareliableestimateofthedif- afterthedetectionofX-raypulsations(Israeletal.2014),andagain
fuseemission.Forthebackgroundofthediffuseemissionwecon- at10-cmon28September,almostsimultaneouslywithoneofour
sidered two regions laying far away (at a distance >4′) from the XMM-Newtonobservations.Observationsat10-cmwereobtained
pulsarandintwodifferentCCDsobtainingsimilarresultsinboth usingtheATNFDigitalFilterbanksDFB3(usedinsearchmodewith
cases. We first fit all the seven spectra together. The use of one a sampling time of 1 ms) and DFB4 (in folding mode) at a cen-
spectral component gave a relatively good fit with a reduced χ2 tralfrequencyof3100MHz,over1024MHzofbandwidth.20-cm
of 1.22 and 1.33 (107 d.o.f.) for an absorbed PL and BB model, observations wereacquired usingthereconfigurable digitalback-
respectively.Thenweleftfreetovaryalltheparametersresulting endHIPSR(HI-Pulsarsignalprocessor)withacentralfrequencyof
in a reduced χ2 of 1.15 and 1.18 (95 d.o.f.) for the PL and BB 1357MHz,a350MHzbandwidthandasamplingtimeof64µs.
model, respectively. Whileno improvement was achieved for the FurtherdetailsoftheobservationsaresummarizedinTable4.
BBmodelthePLmodelappearstovaryamongXMM-Newtonob- Data were folded in 120-s long sub-integrations using the
servationsatabout2.0σconfidencelevel.Therefore,weconclude ephemeris in Table2 and then searched over a range of periods,
The monitoringof SGRJ1935+2154 7
Table3.ChandraandXMM-Newtonspectralresults.Errorsareata1σconfidencelevelforasingleparameterofinterest.
Mission(Model) NHb Γ kT RBBc Fluxd Luminosityd χ2ν(dof)
(1022cm−2) (keV) (km) (10−12ergcm−2s−1) (1034ergs−1)
SOURCEEMISSION
CHANDRA(BB+PL) 2.0±0.4 2.8±0.8 0.45±0.03 1.9±0.2 1.24±0.06 3.1±0.5 0.97(165)
XMM(BB+PL) 1.6±0.2 1.8±0.5 0.47±0.02 1.6±0.1 0.89±0.05 1.7±0.4 1.02(74)
XMMe ” 1.6±0.2 2.1±0.4 0.48±0.02 1.6±0.2 1.19±0.06 2.4±0.5 0.93(109)
DIFFUSEEMISSION
XMM(PL) 3.8±0.4 3.8±0.3 −− −− 0.14±0.02 0.6±0.1 1.94(23)
a XSPECmodels;BB=BBODYRAD,PL=POWERLAW.
b WeusedtheabundancesofWilms,Allen&McCray(2000)
c Theblackbodyradiusiscalculatedatinfinityandforanarbitrarydistanceof9kpc.
d Inthe1–10keVenergyband;fluxesareobservedvalues,luminositiesarede-absorbedquantities.
e March2015XMM-Newtonobservation.
2014ChandraandXMM-Newtondatasetsandweinferredbotha
Table4.Thetablelistsforeachradioobservation:thedateandtime(UT)
ofthestartoftheacquisition(intheformyy-mm-dd-hh:mm);thereceiver first and second period derivative. These findings further confirm
used,eitherthe10-cmfeedofthecoaxial10-50cm(Granetetal.2005)or thatSGRJ1935+2154isindeedamagnetarwhichisslowing-down
the central beam of the 20-cm multibeam receiver (Staveley-Smithetal. atarateofabout halfamillisecondperyear.However,thistrend
1996);theintegrationtime;thefluxdensityupperlimitforapulsedsignal is slowing-down due to a negative P¨ (see Table2). The accurate
witha3.2speriod;thefluxdensityupperlimitforasinglepulseof32ms timingsolutionallowedusalsotoinferthedipolarmagneticfield
duration.FluxareexpressedinmJyunits. strength,anupperlimitonthetruepulsarageandthecorrespond-
ingspin-downluminosity(underusualassumptions).
UTStart Rec Tobs(h) Smin Ssmpin SGRJ1935+2154isaseeminglyyoungobject,6 3kyr,with
14-08-01-11:34 10-50cm 3.0 0.04 68 a Bp value (∼2.2×1014Gauss) well within the typical range of
14-08-02-11:22 10-50cm 3.0 0.04 68 magnetars. The X-ray emission ispulsed. The pulse shape is en-
14-08-03-12:29 20cm-MB 1.5 0.05 61 ergyindependent(withinuncertainties)anditisalmostsinusoidal
14-08-03-13:32 10-50cm 1.0 0.07 68 witha∼20%pulsedfraction(measuredasthesemi-amplitudeof
14-09-28-08:34 10-50cm 2.0 0.05 68 thesinusoiddividedbytheaveragecountrate)during2014.Itbe-
comeslesssinusoidalwithapulsedfractionofonly5%duringthe
latestXMM-Newtonobservation.Wedetectedanenergy-dependent
spanning ±1.5ms with respect to the X-ray value of any given phaseshift(∼0.16cyclesatmaxiumum),withthehardphotonsan-
observing epoch, and over dispersion measures (DM) up to1000
ticipatingthesoftones.Thisbehaviourisnotverycommonamong
pccm−3. known magnetars, 1RXSJ1708−4009 being a notable exception
Thedataacquiredinsearchmodewerealsoblindlysearched
(though with a different trend in energy; see Israeletal. 2001;
over DMs up to 1000 both for periodic signals and single dedis- Reaetal.2005).In1RXSJ1708−4009theshiftislikelyassociated
persed pulses. The 20-cm data were searched in real time using
tothepresenceofa(spinphase)variablehardX-raycomponentex-
HEIMDALL2,whilethe10-cmdatawereanalysedwiththepackage tendinguptoatleast100keV(Kuiperetal.2006;Go¨tzetal.2007).
SIGPROC(http://sigproc.sourceforge.net/).Nopulsedsignalwitha Similarly,thepulseprofilephaseshiftofSGRJ1935+2154 might
periodsimilartothatdetectedinX-rays,norsingledispersedpulses
beduetothepresenceofatleasttwodistinctcomponents(peaks)
werefounddowntoasignal-to-noiseratioof8.Table4liststheup-
with different weight at different energies. The non detection of
perlimitsobtainedateachepochandfrequency.
emissionfromSGRJ1935+2154atenergiesabove10keVdoesnot
allowustofirmlyassessthecauseoftheshift.
The source spectrum can be well described by the canoni-
5 DISCUSSION cal two-component model often applied to magnetars, i.e. an ab-
sorbed black body plus a power-law (kT∼0.5keV and Γ ∼2).
Thanks to an intensive Chandra and XMM-Newton observational
The SGRJ1935+2154 1-10keV observed flux of 1.5×10−12 erg
campaign of SGRJ1935+2154 covering the first 8 months since
cm−2 s−1 is among the lowest observed so far from magne-
the first bursts detected by SwiftBAT, we were able to infer the
tars at the beginning of their outbursts. Although it is possible
maintimingandspectralpropertiesofthisnewlyidentifiedmem-
thatwemissedtheoutburstonset(whichperhapsoccurredbefore
ber of themagnetar class.Inparticular, wediscovered strongco-
the first burst epoch), a backward search of burst activity in the
herent pulsations at a period of about 3.24s in a Chandra long
BATdataatthepositionofSGRJ1935+2154gavenegativeresults
pointingcarriedoutinJuly2014.Subsequently,byusingtheXMM-
(Cummings&Campana 2014). Emission from SGRJ1935+2154
Newtonobservations(spacedsotokeepthepulsephasecoherence
isdetectedinanarchivalSwiftXRTpointingin2011atafluxonlya
amongpointings) westartedbuildingatimingsolutionbymeans
factoroffewlowerthantheonedetectedsoonaftertheburstemis-
ofaphasefittingtechnique.Wewereabletophase-connectallthe
sion. At current stage we cannot exclude that the source has not
reachedthequiescentlevelorthatithasarelativelybrightquies-
2 seehttp://sourceforge.net/projects/heimdall-astro/forfurtherdetails centluminosity.Thislatterpossibilityispartiallysupportedbythe
8 G.L. Israelet al.
unusualpropertiesofSGRJ1935+2154whichdisplaysbothinter- ityofabout 2×1034 ergs−1.Atthepresentstageitisalsorather
valsoffluxweakeningabrighteningsuperimposedtoaslowdecay. difficulttoobtainareliablevalueofthequiescentluminositydue
WenotethatthelatestXMM-Newtonpointingoccurredlessthan20 to the uncertainties on the distance and the flux of the Swift pre-
daysfromtheKonus-Winddetectionofthefirstintermediateflare burstdetection.Ifadistanceof9kpcisassumed,theSwiftfaintest
fromthissource(Golenetskiietal.2015). fluxconverttoaluminosityofabout5×1033ergs−1whichresults
A significant diffuseemission, extending fromspatial scales inLX/Lsd∼0.25,closeto0.3limitingvalue.However,ifthedis-
of >1′′ up to more than 1′ around the magnetar, was clearly de- tanceislargerand/orthequiescentfluxisafactoroffewlargerthan
tectedbothbyChandraandXMM-Newton.Duetotheuseofdif- estimatedfromSwift,thesourcewouldmovetowardhighervalues
ferent instruments/modes at different epochs we werenot ableto ofLX,qui/Lsdinthe“radio-quiet”regionofthefundamentalplane
testifthediffusecomponentvariedintime(asexpectedinthecase (seeleftpanelofFigure2inReaetal.2012).Correspondingly,the
ofscatteringbydustcloudsontheline-of-sight)betweentheChan- nondetectionofradiopulsationsmightbenotthatsurprising.
draandXMM-Newtonpointings.AmongtheXMM-Newtonpoint- The uncertainty in the quiescent level of this new magnetar
ings the component does not change significantly. The Chandra makesanyattempttoinferitsevolutionaryhistoryratheruncertain.
data allowed us to sample the spatial distribution of the compo- Giventheshortcharacteristicage(afewkyrs,whichismostprob-
nentonlyuptoabout20′′ (atlargerradiiwearehamperedbythe ably representative of the true age given that no substantial field
statistics),whilethelower spatialresolution oftheXMM-Newton decayisexpected over suchatimespan), thepresent valueof the
pnallowedustodetectthediffuseemissiononlybeyondabout20′′. magneticfieldislikelynotthatdifferentfromthatatthemoment
Wedonotdetectanyfluxvariationforthediffuseemissionamong ofbirth.Theabovereviewedtimingcharacteristicswouldthenbe
theeight XMM-Newtonpointings despitethepulsar enhancement consistent witha quiescent bolometric luminosity of the order of
of about 20% between October 2014 and March 2015, a result ∼5×1033−34ergs−1(seeFig.11and12inVigano`etal.2013),
which would favour a magnetar wind nebula (MWN) interpreta- depending on theassumed magneticfieldgeometryand envelope
tion. ThePLmodel used tofitthe pnspectra impliesarelatively composition.
steepphotonindexofabout3.8whichissimilartowhatobserved Constraints on its outburst luminosity evolution can be put
for the candidate MWN around SwiftJ1834−0846 (Younesetal. from general considerations (see Pons&Rea 2012; Vigano` etal.
2012),butatthesametimeissteeperthanthePLphotonindexof 2013).Ifweassumethatthefluxderivedbythepre-outburstSwift
SGRJ1935+2154suggestingthatthedustscatteringscenariomight observations provides a correct estimate of the magnetar quies-
bemorelikely. cence,andwerelyonadistanceof9kpc,thenthesourceluminos-
InSwiftJ1834−0846twodiffusecomponentshavebeeniden- ityincreasesfromaquiescentlevelofLX,qui ∼7×1033ergs−1
tified:asymmetriccomponent around themagnetar extending up toa’detected’outburstpeakofLX,out ∼ 4×1034ergs−1.Such
to about 50′′ interpreted as a dust scattering halo (Younesetal. luminosity variation within the outburst (about a factor of 5) is
2012;Espositoetal.2013),andanasymmetriccomponentextend- rathersmallforamagnetarwithamedium-lowquiescentlevel(see
ing up to 150′′ proposed as a wind nebula (Younesetal. 2012). Fig.2ofPons&Rea2012).Inparticular,theoutburstpeaklumi-
The spectrum of the former component has a PL photon index nosityusually reaches about LX,out ∼ 5×1035ergs−1,due to
steeperthanthatofthemagnetar(whichhowever,atvariancewith thetypicalenergiesreleasedinmagnetars’crustalfractures(about
SGRJ1935+2154, isfittedwellbyasinglePLalonelikelydueto 1044−45erg;Pons&Rea2012;Perna&Pons2011),coupledwith
a very high absorption which hampers the detection of any soft estimatesof theneutrino cooling efficiencies(Pons&Rea2012).
BB), while the latter has a flatter spectrum. In order to compare Iftherearenointrinsicphysicaldifferencesbetweenthisoutburst
the properties of the diffuse emission around SwiftJ1834−0846 andothermagnetaroutbursts(seeRea&Esposito2011),thenwe
and SGRJ1935+2154, we fitted the Chandra and XMM-Newton canforeseetwopossibilitiestoexplaintherelativelylowmaximum
spectra of SGRJ1935+2154 with a PL alone obtaining a pho- luminositydetected.
ton index of 4.4±0.1 and 4.3±0.1 (we used only photons in the The first possibility is that we have missed the real outburst
1.5-8.0keV band similar to the case of SwiftJ1834−0846) im- peak of SGRJ1935+2154, which was then caught already during
plying that the diffuse component might have a spectrum flat- itsoutburstdecay.Inthiscasethequiescentluminosityclaimedby
ter than that of the magnetar and favouring the wind nebula sce- thearchival Swift observation mightbecorrect,and themagnetar
nario.Inthelattercasetheefficiencyatwhichtherotationalenergy hadafluxincreaseduringtheoutburst,butwecouldcatchitonly
loss of a pulsar, E˙rot, is radiated by the PWN is given by ηX = thanks to an SGR-like burst detected when the magnetar had al-
LX,pwn/E˙rot=(0.6×1034/1.7×1034)≃0.35,notthatdifferentfrom readycooleddownsubstantially.Giventhetypicaloutburstcooling
whatinferredfromsimilarcomponentsaroundSwiftJ1834−0846 curves,wecanroughlyestimatethatinthisscenarioweobserved
andRRATJ1819−1458(Younesetal.2012;Reaetal.2009).Fur- thesourceabout10-40daysafteritsrealoutburstonset.
therXMM-Newtonand/orChandraobservationstakenatfluxlevels The second possibility is that the source distance is farther
significantlydifferentfromthosewerecordedsofarshouldhelpin thantheassumedSNRdistanceof9kpc(notethatthemethodused
settlingthenatureofthediffuseemission. by Pavlovic´etal. 2013 to infer this distance implies a relatively
A search for radio pulsed emission from SGRJ1935+2154 largedegreeofuncertainty,evenafactoroftwoinbothdirections).
gavenegativeresultdowntoafluxdensityofabout0.5mJy(and Tohaveanoutburstpeakluminosityinlinewithothermagnetars,
70mJy for asinglepulse). Ithas beensuggested that whether or SGRJ1935+2154 should have a distance of ∼20-30kpc. At this
not a magnetar can also shine as a transient radio pulsar might distancetheassumedSwift quiescence levelwouldalsobelarger
depend on the ratio between its quiescent X-ray luminosity and (∼7×1034ergs−1),henceafactorof∼5inincreaseinluminos-
spin-downluminosity,giventhatallmagnetarswithdetectedradio ityintheoutburstwouldthenbeinlinewithwhatobserved(and
pulsedemissionhavethisratiosmallerthan∼0.3(Reaetal.2012), predicted) in other cases (see again Fig.2 of Pons&Rea 2012).
atvariancewithtypicalradio-quietmagnetarsthathavequiescent However,inthedirectionofSGRJ1935+2154 theGalaxyextends
X-rayluminositynormallyexceedingtheirrotationalpower.Based until∼14kpc(Houetal.2009)makingsuchalargedistancerather
onthecoherenttimingsolutionweinferredaspin-downluminos- unlikely.
The monitoringof SGRJ1935+2154 9
We then suggest that the very low peak flux of the detected Camero-Arranz A.,Rea N., Bucciantini N., McLaughlin M. A.,
outburst of SGRJ1935+2154 has no different physics involved SlaneP.,GaenslerB.M.,TorresD.F.,StellaL.,deOn˜aE.,Israel
with respect to other magnetar outbursts, but we have simply G.L.,CamiloF.,PossentiA.,2013,MNRAS,429,2493
missedtheonsetoftheoutburst.IfthefluxdetectedbySwiftbefore Camilo F., Ransom S. M., Halpern J. P., Reynolds J., Helfand
theoutburstwasitsquiescent level,weenvisagethattheoutburst D.J.,ZimmermanN.,SarkissianJ.,2006,Nature,442,892
onsetoccurredaboutamonthbeforethefirstX-rayburstdetection. Campana S.,de UgartePostigoA., Thoene C.C., Gorosabel J.,
Iffutureobservationswillsetthesourceatalowerquiescentlevel, ReaN.,CotiZelatiF.,2014,GRBCoordinatesNetwork,16535,
theoutburstpeakshouldhaveoccurredevenlongerbeforewefirst 1
detecteditsactivity. Cummings J. R., Barthelmy S. D., Chester M. M., Page K. L.,
2014,TheAstronomer’sTelegram,6294,1
CummingsJ.R.,CampanaS.,2014,TheAstronomer’sTelegram,
6299,1
DeLucaA.,MolendiS.,2004,A&A,419,837
ACKNOWLEDGEMENTS
DuncanR.C.,ThompsonC.,1992,ApJ,392,L9
The scientific results reported in this article are based on obser- EspositoP.,TiengoA.,ReaN.,TurollaR.,FenziA.,GiulianiA.,
vationsobtainedwiththeChandraX-rayObservatoryandXMM- Israel G. L., Zane S., Mereghetti S., Possenti A., Burgay M.,
Newton, an ESA science mission with instruments and contribu- StellaL.,Go¨tz D., Perna R., Mignani R. P., Romano P.,2013,
tions directly funded by ESA Member States and NASA. This MNRAS,429,3123
research has made use of software provided by the Chandra X- GolenetskiiS.,AptekarR.,Pal’ShinV.,FrederiksD.,SvinkinD.,
ray Center (CXC) in the application package CIAO, and of soft- ClineT.,HurleyK.,MitrofanovI.G.,GolovinD.,LitvakM.L.,
wares and tools provided by the High Energy Astrophysics Sci- SaninA.B.,vonKienlinA.,ZhangX.,RauA.,Savchenko V.,
ence Archive Research Center (HEASARC), which is a service BozzoE.,FerrignoC.,Boynton W.,FellowsC.,Harshman K.,
of the Astrophysics Science Division at NASA/GSFC and the EnosH.,StarrR.,2015,GRBCoordinatesNetwork,17699
HighEnergyAstrophysicsDivisionoftheSmithsonianAstrophys- Go¨tzD.,ReaN.,IsraelG.L.,ZaneS.,EspositoP.,GotthelfE.V.,
ical Observatory. Thisresearch isbased on observations withthe MereghettiS.,TiengoA.,TurollaR.,2007,A&A,475,317
NASA/UK/ASISwiftmission.WethankN.Schartelforapproving GranetC.,ZhangH.Z.,ForsythA.R.,GravesG.R.,DohertyP.,
theXMM-NewtonNovember2014observationthroughtheDirec- GreeneK.J.,JamesG.L.,SykesP.,BirdT.S.,SinclairM.W.,
tor’sDiscretionaryTimeprogramandthestaffoftheXMM-Newton MooreyG.,ManchesterR.N.,2005,IEEEAntennasandPropa-
Science Operation Center for performing the Target of Opportu- gationMagazine,47,13
nityobservations.Similarly,wethankB.Wilkesforapprovingthe HouL.G.,HanJ.L.,ShiW.B.,2009,A&A,499,473
Chandra August 2014 observation through theDirector’sDiscre- Israel G.,Oosterbroek T., StellaL., Campana S.,Mereghetti S.,
tionary Time program and the Chandra staff for performing the ParmarA.N.,2001,ApJ,560,L65
TargetofOpportunityobservations.WethanktheSwiftdutyscien- Israel G. L., Rea N., Zelati F. C., Esposito P., Burgay M.,
tistsandscienceplannersformakingtheseobservationspossible. MereghettiS.,PossentiA.,TiengoA.,2014,TheAstronomer’s
TheParkesradiotelescopeispartoftheAustraliaTelescopewhich Telegram,6370,1
isfundedbytheCommonwealthofAustraliaforoperationasaNa- KuiperL.,HermsenW.,denHartogP.R.,CollmarW.,2006,ApJ,
tionalFacilitymanagedbyCSIRO.TheauthorswarmlythankPhil 645,556
Edwards for the prompt scheduling of the observations and John PaczynskiB.,1992,ActaAstronomica,42,145
Reynoldsforreleasingpartofthetelescopetimeofhisproject.We Pavlovic´ M. Z., Urosˇevic´ D., Vukotic´ B., Arbutina B., Go¨ker
thanksJulesHalpernforusefulcomments.NRissupportedbyan U¨.D.,2013,ApJS,204,4
NWOVidi Grant,and by grantsAYA2012-39303 and SGR2014- PernaR.,GotthelfE.V.,2008,ApJ,681,522
1073.ThisworkispartiallysupportedbytheEuropeanCOSTAc- PernaR.,PonsJ.A.,2011,ApJ,727,L51
tionMP1304(NewCOMPSTAR). PonsJ.A.,ReaN.,2012,ApJ,750,L6
Rea N., Esposito P., 2011, in Torres D. F., Rea N., eds, Astro-
physicsandSpaceScienceProceedings,High-EnergyEmission
fromPulsarsand theirSystems.Springer, Heidelberg, pp247–
273
REFERENCES
Rea N., Esposito P., TurollaR., Israel G. L., Zane S., StellaL.,
AlbanoA.,TurollaR.,IsraelG.L.,ZaneS.,NobiliL.,StellaL., Mereghetti S., Tiengo A., Go¨tz D., Go¨g˘u¨s¸ E., Kouveliotou C.,
2010,ApJ,722,788 2010,Science,330,944
ArnaudK.A.,1996,inJacobyG.H.,BarnesJ.,eds,Astronomical ReaN.,IsraelG.L.,PonsJ.A.,TurollaR.,Vigano` D.,ZaneS.,
DataAnalysisSoftwareandSystemsVVol.101ofAstronomical EspositoP.,PernaR.,PapittoA.,TerreranG.,TiengoA.,Salvetti
SocietyofthePacificConferenceSeries,XSPEC:TheFirstTen D.,2013,ApJ,770,65
Years.ASP,SanFrancisco,pp17–20 ReaN.,IsraelG.L.,TurollaR.,EspositoP.,MereghettiS.,Go¨tz
BurgayM.,IsraelG.L.,ReaN.,PossentiA.,ZelatiF.C.,Esposito D.,ZaneS.,TiengoA.,HurleyK.,FerociM.,StillM.,Yershov
P.,MereghettiS.,TiengoA.,2014,TheAstronomer’sTelegram, V.,2009,MNRAS,396,2419
6371,1 ReaN.,McLaughlinM.A.,GaenslerB.M.,SlaneP.O.,StellaL.,
BurrowsD.N.,HillJ.E.,NousekJ.A.,KenneaJ.A.,WellsA., ReynoldsS.P.,BurgayM.,IsraelG.L.,PossentiA.,Chatterjee
OsborneJ.P.,AbbeyA.F.,BeardmoreA.,MukerjeeK.,Short S.,2009,ApJ,703,L41
A. D. T., Chincarini G., Campana S., Citterio O., Moretti A., Rea N., Oosterbroek T., Zane S., TurollaR., Me´ndez M., Israel
PaganiC.,TagliaferriG.,GiommiP.,CapalbiM.,TamburelliF., G.L.,StellaL.,HaberlF.,2005,MNRAS,361,710
AngeliniL.,2005,SpaceSci.Rev.,120,165 ReaN.,PonsJ.A.,TorresD.F.,TurollaR.,2012,ApJ,748,L12
10 G.L. Israelet al.
ReadA.M.,RosenS.R.,SaxtonR.D.,RamirezJ.,2011,A&A,
534,A34
StamatikosM.,MalesaniD.,PageK.L.,SakamotoT.,2014,GRB
CoordinatesNetwork,16520,1
Staveley-Smith L., Wilson W. E., Bird T. S., Disney M. J., Ek-
ers R. D., Freeman K. C., Haynes R. F., Sinclair M. W., Vaile
R. A., Webster R. L., Wright A. E., 1996, Publications of the
AstronomicalSocietyofAustralia,13,243
Stil J. M., Taylor A. R., Dickey J. M., Kavars D. W., Martin
P.G.,RothwellT.A.,BoothroydA.I.,LockmanF.J.,McClure-
GriffithsN.M.,2006,AJ,132,1158
Stru¨der L., Briel U., Dennerl K., Hartmann R., Kendziorra E.,
MeidingerN.,PfeffermannE.,ReppinC.,AschenbachB.,2001,
A&A,365,L18
Sun X. H., Reich P., Reich W., Xiao L., Gao X. Y., Han J. L.,
2011,A&A,536,A83
SurnisM.P.,KrishnakumarM.A.,MaanY.,JoshiB.C.,Manoha-
ranP.K.,2014,TheAstronomer’sTelegram,6376,1
TurnerM.J.L.,AbbeyA.,ArnaudM.,BalasiniM.,BarberaM.,
BelsoleE.,BennieP.J.,BernardJ.P.,BignamiG.F.,BoerM.,
2001,A&A,365,L27
TurollaR.,ZaneS.,PonsJ.A.,EspositoP.,ReaN.,2011, ApJ,
740,105
Vigano` D.,ReaN.,PonsJ.A.,PernaR.,AguileraD.N.,Miralles
J.A.,2013,MNRAS,434,123
WeisskopfM.C.,TananbaumH.D.,VanSpeybroeckL.P.,O’Dell
S. L., 2000, in X-Ray Optics, Instruments, and Missions III.
Edited by Truemper, J. E. and Aschenbach, B. Proceedings of
theSPIE.SPIE,BellinghamWAVol.4012ofSocietyofPhoto-
Optical Instrumentation Engineers (SPIE) Conference Series,
ChandraX-rayObservatory(CXO):overview.pp2–16
WilmsJ.,AllenA.,McCrayR.,2000,ApJ,542,914
YounesG.,KouveliotouC.,KargaltsevO.,PavlovG.G.,Go¨g˘u¨s¸
E.,WachterS.,2012,ApJ,757,39
ThispaperhasbeentypesetfromaTEX/LaTEXfilepreparedbythe
author.