ebook img

MAXI J1659-152: The shortest orbital period black-hole transient in outburst PDF

2.7 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview MAXI J1659-152: The shortest orbital period black-hole transient in outburst

Astronomy&Astrophysicsmanuscriptno.kuulkers˙maxi (cid:13)c ESO2013 January15,2013 − MAXIJ1659 152: The shortest orbital period black-hole transient in outburst E.Kuulkers1,C.Kouveliotou2,T.Belloni3,M.CadolleBel1,J.Chenevez4,M.D´ıazTrigo5,J.Homan6,A.Ibarra1,J.A. Kennea7,T.Mun˜oz-Darias3,8,J.-U.Ness1,A.N.Parmar1,A.M.T.Pollock1,E.P.J.vandenHeuvel9,andA.J.vander Horst9,10 1 EuropeanSpaceAstronomyCentre(ESA/ESAC),ScienceOperationsDepartment,28691VillanuevadelaCan˜ada(Madrid),Spain e-mail:[email protected] 3 2 AstrophysicsOffice,ZP12,NASA/MarshallSpaceFlightCenter,Huntsville,AL35812,USA 1 3 INAF-OsservatorioAstronomicodiBrera,ViaE.Bianchi46,I-23807Merate(LC),Italy 0 4 NationalSpaceInstitute,TechnicalUniversityofDenmark,JulianeMariesVej30,2100Copenhagen,Denmark 2 5 ESO,Karl-Schwarzschild-Strasse2,85748GarchingbeiMu¨nchen,Germany 6 MITKavliInstituteforAstrophysicsandSpaceResearch,70VassarStreet,Cambridge,MA02139,USA n 7 DepartmentofAstronomy&Astrophysics,ThePennsylvaniaStateUniversity,525DaveyLab,UniversityPark,PA16802,USA a 8 SchoolofPhysicsandAstronomy,UniversityofSouthampton,SO171BJ,UnitedKingdom J 9 AstronomicalInstitute‘AntonPannekoek’,UniversityofAmsterdam,SciencePark904,1098XHAmsterdam,TheNetherlands 1 10 UniversitiesSpaceResearchAssociation,NSSTC,320SparkmanDrive,Huntsville,AL35805,USA 1 Received;accepted ] E ABSTRACT H MAXIJ1659−152isabrightX-raytransientblack-holecandidatebinary systemdiscoveredinSeptember2010.Wereporthereon . h MAXI,RXTE,Swift,andXMM-Newtonobservationsduringits2010/2011outburst.Wefindthatduringthefirstoneandahalfweek p oftheoutbursttheX-raylightcurvesdisplaydropsinintensityatregularintervals,whichweinterpretasabsorptiondips.Aboutthree - weeks intothe outbursts, again drops inintensity areseen. These dips have, however, a spectral behaviour opposite tothat of the o absorptiondips,andarerelatedtofastspectralstatechanges(hencereferredtoastransitiondips).Theabsorptiondipsrecurwitha r t periodof2.414±0.005hrs,whichweinterpretastheorbitalperiodofthesystem.ThisimpliesthatMAXIJ1659−152istheshortest s periodblack-holecandidatebinaryknowntodate.Theinclinationoftheaccretiondiskwithrespecttothelineofsightisestimated a tobe65–80◦.Weproposethecompaniontotheblack-holecandidatetobeclosetoanM5dwarfstar,withamassandradiusofabout [ 0.15–0.25M⊙and0.2–0.25R⊙,respectively.Wederivethatthecompanionhadaninitialmassofabout1.5M⊙,whichevolvedtoits 2 currentmassinabout5–6billionyears.Thesystemisrather compact(orbitalseparationof &1.33R⊙),andislocatedatadistance v of8.6±3.7kpc,withaheightabovetheGalacticplaneof2.4±1.0kpc.ThecharacteristicsofshortorbitalperiodandhighGalactic 0 scaleheightaresharedwithtwoothertransientblack-hole candidateX-raybinaries,i.e.,XTEJ1118+480andSwiftJ1735.5−0127. 4 WesuggestthatallthreearekickedoutoftheGalacticplaneintothehalo,ratherthanbeingformedinaglobularcluster. 8 Keywords.Accretion,accretiondisks–binaries:close–Stars:individual:MAXIJ1659−152–X-rays:binaries 5 . 4 0 1. Introduction naries(LMXBs),i.e.,binariescontainingalow-mass(typically 2 .1M )companionorbitinganeutronstarorablackhole. 1 ⊙ TransientX-raysourceshavebeenextensivelystudiedsincethe : BHXBs are primarilydiscoveredwhentheyenteroutbursts v adventofX-rayastronomy.Mostareinbinarysystemscompris- characterisedbyincreasedX-rayluminosities,bya factor ofat i ingatleastonecompactobject,whichiseitheraneutronstaror X least 100 in a few days, and a variety of spectral and tempo- ablackhole.Determiningthenatureofthecompactobjectisnot ralvariabilitystates.Themostcontrastedstatesaretheso-called r trivial,asanumberoftheirX-raycharacteristicsarecommonto a ‘hard’andthe‘soft’states. Theformer,originallywas alsode- both types of objects. Black hole X-ray binaries (BHXBs) are finedasa‘low’state,owingtotherepresentationofthespectrum uniquelyidentifiedbyamassdeterminationofthecompactob- by a power law with spectral index 1.4–2.0 up to several hun- ject in excess of 3M . The first two confirmed BHXBs were ⊙ dredsofkeV.Conversely,thelatter‘soft’,or‘high’,stateischar- CygX-1 (Webster & Murdin 1972, Bolton 1972) and LMCX- acterised by a tenfold increase of the ∼2–10keV flux. Several 3(Cowleyetal.1983);thesesystemswere,however,persistent other states have been defined in the literature, depending also X-raysources. The veryfirst transientsource with a confirmed ontheevolutionofthecombinedspectralandtemporalvariabil- blackhole,A0620−00(McClintock&Remillard1986),wasdis- ityproperties(forextensivereviews,seeHoman&Belloni2005, coveredin1975,whenitreachedintensitiesof∼50Crab(Elvis Belloni2010,andMcClintock&Remillard2006,Remillard& et al. 1975). To date there are 20 known BHXBs, of which 17 McClintock2006). areknowntobetransient(e.g.,Remillard&McClintock2006). TheBHXBtransientsbelongtotheclassoflow-massX-raybi- On 2010 September 25 08:05 UTC, the Swift/Burst Alert Telescope (BAT) triggered on a source located roughly 17◦ above the Galactic plane. The source was initially designated Sendoffprintrequeststo:E.Kuulkers as Gamma-Ray Burst (GRB) 100925A(Manganoet al. 2010), 2 E.Kuulkersetal.:MAXIJ1659−152 and was monitored with the Swift/X-Ray Telescope (XRT). Our team was involved in various Target of Opportunity Interestingly, the source flux did not decline during the next (ToO)observationstakenwithRXTE,SwiftandXMM-Newton. several hours, as is usually the trend with GRBs. This unusual Here we present a detailed account of the X-ray and UV light behaviour and the source location near the Galactic bulge in- curvesofMAXIJ1659−152,taken overthe courseof the2010 dicated that it might not be a GRB but a new Galactic source outburst with these satellites and with MAXI. A preliminary (Kahn2010).Laterthatday,theMonitorofAll-skyX-rayImage account of the results can be found in Kuulkers et al. (2010b, (MAXI) team reportedthe detection of a new hard X-ray tran- 2010d,2012a)andBellonietal.(2010).TheMAXI,RXTEand sient, MAXIJ1659−152, whose position was consistent with Swiftdatahavealreadybeen(partly)describedelsewhere,how- GRB100925Aandwhichhadbrightenedsince2010September ever, the focus was more on the combined spectral and timing 25 02:30 UTC (Negoro et al. 2010). The discovery initiated behaviourof MAXIJ1659−152(Kalamkaretal. 2011,Kennea severalmulti-wavelengthobservations(groundandspace-based; et al. 2011, Mun˜oz-Darias et al. 2011b, Shaposhnikov et al. e.g.,vanderHorstetal.2010,Kuulkersetal.2010b,2010c;see 2012, Yamaoka et al. 2012). We find that MAXIJ1659−152is alsoKuulkersetal.2012a). indeed viewed at a high inclination. It still has the shortest or- TheGalacticoriginofMAXIJ1659−152wasalsosuggested bital period known to date, and is, therefore, a rather compact the next day from a combined UV-X-ray spectral-energy dis- BHXBtransient.WepresentinSect.2adetaileddescriptionof tribution analysis using data from the Swift/XRT and Swift/UV thedatasetsusedinouranalysis,andinSect.3theresultsofour Optical Telescope (UVOT) (Xu 2010). Finally, optical spec- timing analysis of the variations of the source light curve. We troscopydatataken with the ESO/VeryLargeTelescope(VLT) discuss the observedlightcurve dipsin Sect. 4 and providean X-shooterinstrumentshowedvariousbroademissionlinesfrom interpretationonthecharacteristicsofthebinarysystem andits theBalmerandPaschenseriesofHandofHeII,aswellasCaII distance. andNaIabsorptionlinesfromtheinterstellarmedium,allatred- shiftzero,clinchingtheGalacticnatureofthesource.Moreover, the double-peaked profiles of the emission lines indicated that 2. Observations the source was an X-ray binary(de UgartePostigo et al. 2010, Kaur et al. 2012). The binary nature of the source has by now In this paper we only concentrate on the light curves of the been established in several studies of its spectral and tempo- various instruments onboard MAXI (Matsuoka et al. 2009), ral properties. Kennea et al. (2010) reported frequent intensity RXTE(Bradtetal.1993),Swift(Gehrelsetal.2004)andXMM- drops in the X-ray light curve, attributed possibly to eclipses Newton1 (Jansenetal. 2001).We referto Table1 foran obser- by a companion star. Kuulkers et al. (2010d, 2012a) estab- vation log of the latter 3 satellites. The XMM-Newton spectral lishedaperiodof∼2.42hrusingtheX-rayMultiMirrormission analysisisdeferredto a futurepaper.Fora spectralanalysisof (XMM-Newton)data,whichwaslaterconfirmedbyBellonietal. theRXTEandSwiftdatawerefertoMun˜oz-Dariasetal.(2011b) (2010,Rossi X-ray Timing Explorer(RXTE);also Kuulkerset andKenneaetal.(2011),respectively,aswellasYamaokaetal. al.2012a)andKenneaetal.(2011;Swift).Kurodaetal.(2010) (2012).Allourlightcurveshavebeensubjectedtoabarycentric have reported optical variations up to 0.1mag, consistent with correctionusingthestandardtoolsavailable. a double-peakedmodulation at a period of 2.4158±0.0003hrs. WeconcludedthatthisistheshortestBHXBorbitalperiodmea- 2.1.XMM-Newton suredasyet(Kuulkersetal. 2010d,2012a),andsuggestedthat thesourcewasmostlikelyviewedatahighinclination(Kuulkers WeusedScienceAnalysisSystem(SAS)version11.0.0together etal.2012a). withthelatestcalibrationfilestoanalysetheXMM-Newtondata. ThesourcewasidentifiedasaBHXBthroughthefasttiming TheEuropeanPhotonImagingCamera(EPIC)-MOS(Turneret behaviourobservedwiththeRXTE/ProportionalCounterArray al. 2001) cameras were not used during the observation in or- (PCA),whichwassimilartothatseeninstellar-massblack-hole der to allocate their telemetry to the EPIC-pn camera (Stru¨der transients (Kalamkar et al. 2011, Kennea et al. 2011, Mun˜oz- et al. 2001) and to avoid full scientific buffer in the latter. The Darias et al. 2011b, Shaposhnikov et al. 2012, Yamaoka et al. EPIC-pnwasusedintimingmode,whilsttheReflectionGrating 2012).Estimatesofthemassofthe compactobjectrangefrom Spectrometer instruments (RGS1 and RGS2; den Herder et al. 2.2–3.1M (Kenneaetal.2011),to3.6–8.0M (Yamaokaetal. 2001)wereoperatinginthestandardspectroscopymode.Dueto ⊙ ⊙ 2012),to 20±3M (Shaposhnikovet al. 2012).Partof the dis- the brightnessof the sourcethe EPIC-pndata were affectedby ⊙ crepancycanberesolvedbytakingintoaccountthespinofthe pile-up(seebelow).Similarly,thecountratewasalsoabovethe blackhole(Kenneaetal.2011,Yamaokaetal.2012).Thesource limitsofpile-upfortheRGS2,forwhichtheread-outisslower distance estimates range between 1.6−4.2kpc (Miller-Jones et byafactoroftwocomparedtothatofRGS1sinceAugust2007 al.2011)and8.6kpc(Yamaokaetal.2011). (seesection3.4.4.8oftheXMM-NewtonUsersHandbook). After the main outburst, MAXIJ1659−152 continued to Standarddatareductionprocedures(SAStasks epprocand show low-level activity (e.g., Kennea et al. 2011). The flux rgsproc) were used to obtain EPIC-pn and RGS calibrated of MAXIJ1659−152 was observed to exhibit sudden intensity eventfiles.Weusedthetaskepfastontheeventfilestocorrect changes, e.g., during 2011 May, when Swift and Chandra de- forchargetransferinefficiency(CTI)effectsseenintheEPIC-pn tectedareflare.Thesourcethendecayedagainoveraperiodof timingmodewhenhighcountratesarepresent. ∼2 months (Yang & Wijnands 2011a, 2011b, 2011c, Jonker et The count rate in the EPIC-pn was close to, or above, the al.2012).Itseemedtoenteritsquiescentstateduringthesecond 800ctss−1 level,atwhichX-rayloadingandpile-upeffectsbe- halfof2011(Yang&Wijnands2011d,Russelletal.2011,Kong comesignificant.Pile-upoccurswhenmorethanonephotonis etal.2011).However,foraBHXBitwasstilltoobrighttobein readinapixelduringaread-outcycle.Thiscausesphotonloss, truequiescence(Jonkeretal. 2012).A possiblequiescentopti- calcounterparthasbeenreportedbasedonobservationsdoneon 1 WenotethattheXMM-Newtonobservationsweretakensimultane- 2010 June 19 (about3 monthsbefore the start of the outburst) ouslywithoneoftheINTEGRALToOobservationsofthesource,see and2012March23(Kongetal.2010,Kong2012). Kuulkersetal.(2012a). E.Kuulkersetal.:MAXIJ1659−152 3 Table1.LogofX-rayobservationswithRXTE,SwiftandXMM-NewtonduringthemainoutburstofMAXIJ1659−152presented inthispaper,orderedalongthestarttimeoftheobservation. Day1 Starttime(UT) Exp.2 ObsID Satellite Day1 Starttime(UT) Exp.2 ObsID Satellite 0.3 2010-09-2507:49 19667 00434928000 Swift 16.6 2010-10-1115:30 2028 95108-01-25-00 RXTE 1.0 2010-09-2600:07 16214 00434928001 Swift 17.0 2010-10-1200:36 3348 95108-01-26-00 RXTE 1.6 2010-09-2613:31 9909 00434928002 Swift 17.7 2010-10-1216:27 2453 95108-01-27-00 RXTE 2.0 2010-09-2700:13 19108 00434928003 Swift 18.4 2010-10-1309:43 2697 95108-01-28-00 RXTE 2.7 2010-09-2716:15 51916 0656780601 XMM-Newton 18.8 2010-10-1319:24 1284 00434928023 Swift 3.0 2010-09-2800:53 16910 95358-01-02-00 RXTE 19.5 2010-10-1411:37 1705 00434928025 Swift 3.3 2010-09-2807:06 10009 00434928005 Swift 19.9 2010-10-1421:32 3516 95108-01-30-00 RXTE 4.1 2010-09-2901:58 1706 95358-01-02-01 RXTE 20.0 2010-10-1500:28 1240 00434928026 Swift 4.2 2010-09-2905:14 2318 00434928007 Swift 20.2 2010-10-1505:27 3444 95118-01-01-00 RXTE 5.1 2010-09-3002:00 1410 95358-01-02-02 RXTE 20.6 2010-10-1513:27 1390 00434928027 Swift 5.2 2010-09-3005:37 2719 00434928008 Swift 20.7 2010-10-1516:24 3554 95118-01-01-01 RXTE 6.2 2010-10-0105:45 3369 95358-01-03-00 RXTE 21.0 2010-10-1600:34 1305 00434928028 Swift 6.2 2010-10-0105:45 2594 00434928009 Swift 21.1 2010-10-1603:25 3374 95118-01-02-00 RXTE 6.4 2010-10-0110:44 2130 95108-01-02-00 RXTE 21.5 2010-10-1611:45 1325 00434928029 Swift 7.1 2010-10-0202:31 1763 95358-01-03-01 RXTE 22.0 2010-10-1700:38 1390 00434928030 Swift 7.1 2010-10-0202:37 2369 00434928010 Swift 22.2 2010-10-1704:25 829 95118-01-03-01 RXTE 7.5 2010-10-0211:50 1905 95108-01-03-00 RXTE 22.8 2010-10-1718:55 2464 95118-01-03-00 RXTE 7.8 2010-10-0218:22 1175 95108-01-04-00 RXTE 23.0 2010-10-1800:46 1289 00434928032 Swift 8.1 2010-10-0301:41 994 95108-01-05-00 RXTE 23.7 2010-10-1816:36 3556 95118-01-04-00 RXTE 8.1 2010-10-0302:36 1720 00434928011 Swift 24.0 2010-10-1900:24 3027 95118-01-05-00 RXTE 8.2 2010-10-0304:50 3352 95358-01-03-02 RXTE 24.0 2010-10-1900:50 1290 00031843001 Swift 8.5 2010-10-0311:20 1737 95108-01-06-00 RXTE 24.6 2010-10-1913:31 1245 00031843002 Swift 8.9 2010-10-0321:03 1378 95108-01-07-00 RXTE 24.9 2010-10-1920:51 2358 95118-01-05-01 RXTE 9.1 2010-10-0402:40 3314 00434928012 Swift 25.0 2010-10-2000:56 1290 00031843003 Swift 9.1 2010-10-0402:43 3352 95108-01-08-00 RXTE 25.3 2010-10-2006:17 3495 95118-01-06-00 RXTE 9.5 2010-10-0410:52 2216 95108-01-09-00 RXTE 25.5 2010-10-2012:21 1089 00031843004 Swift 9.7 2010-10-0417:19 1309 95108-01-10-00 RXTE 25.7 2010-10-2017:16 3468 95118-01-06-01 RXTE 10.1 2010-10-0502:47 3414 00434928013 Swift 26.0 2010-10-2100:53 985 00031843005 Swift 10.6 2010-10-0513:43 1847 95108-01-11-00 RXTE 26.1 2010-10-2102:40 2883 95118-01-07-01 RXTE 10.8 2010-10-0520:02 1676 95108-01-12-00 RXTE 26.6 2010-10-2113:46 985 00031843006 Swift 11.1 2010-10-0602:53 3284 00434928014 Swift 26.7 2010-10-2116:48 3214 95118-01-07-00 RXTE 11.4 2010-10-0609:48 2397 95108-01-13-00 RXTE 27.0 2010-10-2200:37 893 95118-01-08-00 RXTE 11.7 2010-10-0617:55 1443 95108-01-14-00 RXTE 27.0 2010-10-2200:57 1264 00031843007 Swift 12.1 2010-10-0701:17 3409 95108-01-15-00 RXTE 27.2 2010-10-2205:51 1040 00031843008 Swift 12.1 2010-10-0702:47 1575 00434928015 Swift 27.8 2010-10-2219:22 1940 95118-01-09-00 RXTE 12.4 2010-10-0709:28 2337 95108-01-16-00 RXTE 29.2 2010-10-2405:54 1162 95118-01-10-00 RXTE 12.7 2010-10-0715:38 1595 00434928016 Swift 30.2 2010-10-2505:24 888 95118-01-11-00 RXTE 12.7 2010-10-0715:47 1602 95108-01-17-00 RXTE 31.0 2010-10-2600:39 868 95118-01-12-00 RXTE 13.0 2010-10-0723:37 858 95108-01-18-00 RXTE 32.5 2010-10-2712:25 2197 95118-01-13-00 RXTE 13.0 2010-10-0800:00 1015 95108-01-18-01 RXTE 33.5 2010-10-2811:54 774 95118-01-14-00 RXTE 13.1 2010-10-0802:52 1630 00434928017 Swift 34.5 2010-10-2911:31 1152 95118-01-15-00 RXTE 13.7 2010-10-0816:55 1662 95108-01-19-00 RXTE 35.3 2010-10-3007:54 1412 95118-01-15-01 RXTE 14.0 2010-10-0900:49 1900 95108-01-20-00 RXTE 36.3 2010-10-3107:24 1458 95118-01-16-00 RXTE 14.1 2010-10-0902:58 1584 00434928019 Swift 37.2 2010-11-0105:20 1467 95118-01-16-01 RXTE 14.5 2010-10-0911:39 2309 95108-01-21-00 RXTE 38.0 2010-11-0200:14 1715 95118-01-17-00 RXTE 14.7 2010-10-0915:49 1595 00434928020 Swift 39.0 2010-11-0301:11 1457 95118-01-17-01 RXTE 15.1 2010-10-1001:27 1009 00434928021 Swift 40.1 2010-11-0401:15 786 95118-01-18-00 RXTE 15.1 2010-10-1003:05 3507 95108-01-22-00 RXTE 41.0 2010-11-0500:16 2348 95118-01-19-00 RXTE 15.5 2010-10-1012:42 1629 00434928022 Swift 42.2 2010-11-0604:32 1139 95118-01-20-00 RXTE 15.7 2010-10-1015:54 1914 95108-01-23-00 RXTE 44.1 2010-11-0801:58 1608 95118-01-21-00 RXTE 16.2 2010-10-1104:03 3513 95108-01-24-00 RXTE Note1. Day=MJD-55464,whereMJD55464correspondstoUT2010,September25,0:00. Note2. Totalon-sourceexposuretimeinsec. patternmigrationfromlowertohigherpatterntypesandharden- ofsingle-anddouble-pixeleventswhichdeviatefromstandard ingofthespectrum,becausethechargesdepositedbymorethan values in case of significantpile-up, as a diagnostic tool in the one photon are added up before being read out(see the XMM- pn camera timing mode data and found that the spectrum was NewtonUsersHandbookformoreinformationonpile-up). affectedbypile-up.Next,weextractedseveralspectraselecting single anddoubletimingmodeevents(patterns0 to 4) butdif- Sincepile-upcausessignificantspectraldistortionandade- ferentspatialregionsforthesource.Sourceeventswerefirstex- cline in the count rate measured by XMM-Newton, we investi- tractedfroma62′′(15columns)wideboxcentredonthesource gatedindetailitspresencebeforeextractingthetimeseries.We position(Region1). Nextwe excluded2,4,6,8and10 columns used the SAS task epatplot,which utilises the relative ratios 4 E.Kuulkersetal.:MAXIJ1659−152 Table2.LogofobservationswiththeXMM-Newton/OM. the standard selection criteria for bright sources in our analy- sis. We included data when the elevation angle of the source Time3 Filter Flux Magnitude abovetheEarthhorizonwas morethan10◦, andusedonlysta- (ksec) (10−15ergcm−2s−1Å−1) ble pointings,i.e., those with offsetanglesless than 0.02◦. The 0.00 UVW1 1.82±0.02 15.75±0.02 light curves were corrected for background,as estimated from 4.52 UVW1 1.81±0.02 15.76±0.02 thebackgroundmodelforbrightsources. 9.05 UVW1 1.79±0.02 15.77±0.02 13.57 UVW1 1.79±0.02 15.77±0.02 18.09 UVW1 1.92±0.03 15.69±0.02 2.3.Swift 22.61 UVM2 1.18±0.05 16.45±0.05 27.13 UVM2 1.32±0.05 16.33±0.04 We utilised the methods described by Evans et al. (2009) to 31.65 UVM2 1.22±0.05 16.41±0.05 extract XRT (Hill et al. 2004, Burrows et al. 2005) X-ray 37.97 UVM2 1.35±0.05 16.31±0.04 light curves in the energy range 0.3–10keV, with corrections 42.50 UVM2 0.95±0.03 16.69±0.03 for the effects of pile-up, hot-columns and hot-pixels applied. These light curves, at 100s time resolution,were extracted us- Note3. StartoftheexposureinksecrelativetoUTC2010September ing the most accurate available localisation in the XRT co- 2716:24:36(=MJD55466.68375). ordinate system, derived from late-time PC mode data taken on 2011 February 6, 134 days after the initial detection of MAXIJ1659−154,whenthesourcewasnotaffectedbypile-up from the centre of Region 1 (Regions 2–6) and extracted one (Kenneaetal.2011).WeusedtheXRT/WTlightcurvesbetween spectrumforeachofthedefinedregions.Wefoundthatthespec- 2010September25andOctober22(seeTable1),wheneverthe trumwasfreeofpile-upafterremovingthecentral10columns. pointing offset was smaller than about 3.5′. This led to a total Then,we used this free of pile-upeventlist to extractthe time good-timeexposureofabout123ksec. seriesshowninthispaper. We used the BAT (Barthelmy et al. 2005) 15–50keV light InthecaseoftheRGS,weusedtable11intheXMM-Newton curves(seeKrimmetal.2006)2generatedon2011October25. Users Handbook to determine which CCDs were affected by Wedidnotusetimebinswithlessthan500sexposuretime(see pile-up.WefoundthatCCDs6,7and8fromRGS2wereabove Kenneaetal.2011). thepile-uplimitsand,therefore,wedidnotusethemforanaly- sis. TheEPIC-pnandRGStimeseriesat1sand100stimeres- 2.4.MAXI olution were producedusing the epiclccorrand rgslccorr tasks,respectively. The Gas Slit Camera (GSC; Mihara et al. 2011) is part of the IntheEPIC-pntimingmode,therearenosource-freeback- payloadoftheMAXImission,onboardtheInternationalSpace groundregions,sincethepoint-spreadfunctionofthetelescope Station(ISS).Onlythosedata(version0.3)whentheinstrument extendsfurtherthanthecentralCCDboundaries.Inthecase of was operating at a high voltage of 1650V were taken into ac- RGS, since MAXIJ1659−152 was very bright, its spectrum is count.Inthispaperwe onlyfocusonthe lightcurvesaveraged notsignificantlymodifiedbythe ‘real’ backgroundwhich con- over an ISS orbit and on a daily basis, in the total 2–20keV tributeslessthan1%tothetotalcountrateinmostoftheband- band.3 width.Therefore,wechosenottosubtractthe‘background’ex- tractedfromtheouterregionsofthecentralCCD(seealsoDone &DiazTrigo2010,Ngetal.2010). 3. Results The final RGS light curve was calculated by combining ThemainoutburstlightcurveofMAXIJ1659−152hasalready RGS1andRGS2,addingbothorders1and2. been shown in various papers, as well as a multitude of en- Theopticalmonitor(OM,Talavera2009,Masonetal.2001) ergyrangeandtimeresolutioncombinations.Authorsalsoused wasoperatedinEPICimagingmodewiththetwofiltersUVW1 different combinations of instruments. We refer to Kalamkar (λ∼2500–3500Å) and UVM2 (λ∼2000–2600Å). Ten consecu- et al. (2011), Mun˜oz-Darias et al. (2011b) and Shaposhnikov tive exposures of 4200s duration each were taken, first five in et al. (2012) for the RXTE/PCA light curves of the averages the UVW1, andthenfivemorein the UVM2filter. We derived perobservation,invariousX-raybands.Yamaokaetal. (2012) thesourcemagnitudesfromtheoutputofstandardextractionus- showedtheRXTE/PCAdataatatimeresolutionof16sinvar- ingtheSAStask omichain.InTable2,the averagefluxesand ious X-ray bands. Kalamkar et al. (2011) and Yamaoka et al. magnitudesineachfilterarelistedforeachoftheexposures. (2012)alsoshowtheone-dayaveragedMAXI/GSClightcurves. Additionally, the latter authors provided the daily-averaged Swift/BATdata.Kenneaetal.(2011)showedtheSwift/BATdata 2.2.RXTE onasatelliteorbittimescale,togetherwiththeSwift/XRTrates RXTEmonitoredthesourcemorethanoncedailythroughoutthe at a 100s time resolution. Both Kalamkar et al. (2011) and outburst(seeKalamkaretal.2011,Mun˜oz-Dariasetal.2011b, Shaposhnikov et al. (2012) provided daily hardness averages Shaposhnikov et al. 2012, Yamaoka et al. 2012). We used 64 basedontheRXTE/PCAdata,whilstKenneaetal.(2011)pro- PCA(Jahodaetal.2006)observationsfrom2010September28 vided hardnesscurvesduring the outburstusing the Swift/XRT toNovember8(seeTable1);thetotalgood-timeexposurewas data. about138ks. Weherepresentanoverviewoftheoutburstbehaviour,com- ForthePCAdataweusedtheFTOOLSanalysissuite(ver- bining the information at soft and hard energies, as well as sion 6.9).We producedlightcurvesfromPCU2 with 16-sbins over the full PCA energy range (using Standard 1 data), i.e., 2 http://swift.gsfc.nasa.gov/docs/swift/results/transients/ 2–60keV, and over three energy ranges, i.e., 2.1–4.9keV, 4.9– index.html 9.8keV and 9.8–19.8keV (using Standard 2 data). We applied 3 http://maxi.riken.jp E.Kuulkersetal.:MAXIJ1659−152 5 the main outburst and up to before the soft X-ray peak of the 25−27/9/2010 main outburst (i.e., up to about day 10, which we denote out- 0 5 burstepochA)ismainlyduetoperiodicdipsinthelightcurves, 1 which will be described in more detail in the next subsection, Sect.3.2.1.Frombeforethepeakoftheoutburstuptoaboutday V) 22 (outburst epoch B), variability within an observationis still e 3−10 k 100 stheeeno,baslebreviattiloesnsss(streoen,ge..gT.,hFisigv.a5rcia)b.DiliutyrinisgdouuetbtuorsfltaerpinogchdBuritnhge 0. s s (−1 vavaetiroange(sfleuexavlsaorieKsablaemtwkeaerne1t5a–l.202%011fr)o.mOnobdsaeyrvsa2ti2o–n3t1o.5ob(oseurt-- ct T burstepochC),apartfromageneraldecreasingtrend,theinten- R X sityappearstobevaryingbetweenalowandahighvalue.Flux 0 5 variationson a similar scale are seen within three observations of that period (days 23.7, 24.0 and 26.1, see Sect. 3.2.2). The lastpartofthemainoutburst(days30.5–44.1;outburstepochD) showsarathersmooth,butnotlinearorpower-law/exponential 0 like (see below), decay. We note that the observations during 0.5 1 1.5 2 2.5 Time (MJD − 55464) outburstepochDhaveonlyarelativelyshortduration(seeTable 1);theyexhibitnostrongvariabilitylikethatseenintheearlier Fig.2. Swift/XRT 0.3–10keV light curve during the first few outburstepochs. daysoftheoutburstofMAXIJ1659−152,whenSwiftobserved the source during every satellite orbit (days 0.3–2.6). Dipping activityisclearlyapparentonaregularbasis. In Figs. 1d and e the MAXI/GSC and Swift/BAT light curves of the outburst are presented, integrated over a satel- at various time resolutions. We first describe the overall out- lite orbit. We note that the Swift/XRT (and Swift/UVOT, see burst light curve and then focus on two independent intensity Kennea et al. 2011) coverage was up to day 27.3 (see Table 1 variationsseen, i.e., ‘absorptiondips’ and ‘transitiondips’. We and Kennea et al. 2011). The difference in rise time to maxi- subsequently present a timing study of the data taken during mum betweenthe softand hard energyband,noted above,can the time period when absorption dips were seen. We use days be clearly seen. This is also borne out by the hardness curve since MJD55464 to describe the epoch of time. This date is shown in Fig. 1f: just after discovery,the outburst is hard, and closeto theMAXI/GSCandSwift/BAT triggersof theoutburst then softens during the following week (see also Kennea et al. (MJD55464.10and55464.34,respectively,seeSect.1). 2011, using Swift/XRT). Near the end of the main outburst, i.e.,aroundday35,MAXIJ1659−152’sradiationhardensagain (seealsoKalamkaretal.2011,Shaposhnikovetal.2012,using 3.1.Overalloutburstlightcurve RXTE/PCA). The end of the main outburst is well covered by theMAXI/GSC(seealsoabove),and,asnotedearlier,itshowsa In the top panel of Fig. 1 we show the overall outburst light smooth,power-law/exponentiallikedecayinthe2–20keVband. curve of MAXIJ1659−152, using daily averages, at relatively The hardening around day 35 and the fact that RXTE/PCA is softenergies(2–20keV;Fig.1a)andhardenergies(15–50keV; sensitivetohard(&20keV)X-rayscausestheRXTE/PCAlight Fig.1b).Afterafastriseofacoupleofdays,MAXIJ1659−152’s curvetodeviatefromthepower-law/exponentiallikedecayde- soft intensity fluctuates by 20–30% on a daily basis (see also scribedatsofterX-rays(.20keV). Kalamkaretal.2011),ontopofageneralslowdeclineininten- sity (∼0.1ctscm−2s−1 per10days),uptoaboutday30.Itthen shows an exponential-like decay (with a decay constant of of ∼7days)uptoaboutday100.Inthehardenergyband(Fig.1b), Between about days 18 and 28, the MAXI/GSC and MAXIJ1659−152reachesapeakinintensitywithin3daysafter Swift/BAT intensities seem to modulate on a several day time thestart.Itsubsequentlydecreasesinasomewhatirregularfash- scale (Figs. 1d and e). Before and after this time period this is ion (but smoother than with respect to the soft energy band), notevident.AcloserlookattheMAXI/GSCandSwift/BATlight until about day 65 (see also Kennea et al. 2011). After days curves(Fig.1gandh)showsapossibleperiodicvariationonan 100 and 65, MAXIJ1659−152 falls below the detection limits about 3-day time scale, for about 3 cycles. Indeed, a pure lin- of MAXI/GSCand Swift/BAT,respectively.In Fig. 1a we have ear trend does not describe the data well in this time period: also indicatedthe timesof the postmain-outburstX-ray obser- χ2 = 3.3 for 81 degrees of freedom (dof) and χ2 = 2.1 for red red vationswithSwiftandChandra,aswellasradioobservations,as 123dof,fortheMAXI/GSCandSwift/BATlightcurves,respec- reportedintheliterature(Kenneaetal.2011,Yangetal.2011a, tively. Adding a quadratic term does not significantly improve 2011b,Yang& Wijnands2011a,2011b,2011c,2011d,Miller- the situation (χ2 = 3.4 for 80 dof and 1.8 for 122 dof, re- red Jonesetal.2011,Jonkeretal.2012). spectively).Asinusoidalplusconstant,linearandquadraticterm In the middle panelof Fig. 1 we zoom in on the main part does significantly improve the fit, although it is still not ideal of the outburst, from just before the start of the outburst up to (χ2 = 2.2for77dofandχ2 = 1.3for119dof,respectively). red red thepartwhenMAXIJ1659−152wastooclosetotheSuntobe The modulation may thus be not purely sinusoidal. Assuming observedduringdedicatedpointedobservationswithRXTEand thesignalisreal,wederiveaperiodof3.15±0.05days(uncer- Swift. tainty is determined by using ∆χ2 = 1) in the Swift/XRT light Fig.1cshowstheoutburstasseenbytheRXTE/PCAinits curveand3.04±0.09daysintheSwift/BAT,i.e.,consistentwith whole sensitive energy band (2–60keV), at a 16s time resolu- eachother.Thephaseofthesinusoidisabout0.25daysearlier tion. The variability during an observation in the first part of fortheSwift/BATwithrespecttothatoftheMAXI/GSC. 6 E.Kuulkersetal.:MAXIJ1659−152 3.2.Recurrentintensityvariations 3.2.1. Absorptiondips 0 0 2 Periodic drops of intensity, or dips, are observed shortly after the start of the outburst, at day 0.3, up to day 8.2, first in the Swift/XRT light curves(e.g., Fig. 2, days0.3–2.6,Fig. 3b, day 3.3), then in the XMM-Newton/RGS and EPIC-pn light curves s−1 (Fig. 3c and d, days 2.7–3.3), and finally in the RXTE/PCA s ct 0 curves(e.g.,Fig.3a,days3.0–3.4andFig.5a,day7.1).Aclear 0 1 recurrence time of ≃0.1 days is best observed in the XMM- Newton/RGS and EPIC-pn light curves, thanks to the continu- ous coverage. In addition to the periodic dips, a linear rise in theout-of-dipintensityisobservedintheSwift/XRTlightcurve a (Fig.2;seealsoKenneaetal.2011),whichextendsthroughout 27/9/2010 the XMM-Newton observations (Fig. 3c and d) and is consis- 0 2.88 2.89 2.9 tentwiththeMAXI/GSClightcurve(Fig.1d),wherethesource Time (MJD − 55464) reached its first plateau at soft X-rays, .20 keV, around day 4 (i.e.,justaftertheXMM-Newtonobservations). Thedipsshowirregularstructurewhichlastsbetweenabout 0 5 and 40min. Occasionally,the XMM-Newton/RGS and EPIC- 0 2 pnlightcurvesshowshallowerdipactivityathalftherecurrence time (see, e.g.,Figs. 3c and d nearday 3.15).Thedepth of the dips varies between ≃90% and 50% of the average out-of-dip- interval intensity in the light curves of Swift/XRT and XMM- s−1 Newton/RGSandEPIC-pn(Figs.2and3),extractedwithatime cts 0 resolutionof100s. 10 InFig.4weshowtwoepochsofthedipactivityatahigher timeresolutionof1sasseenbytheXMM-Newton/EPIC-pn.The two epochsare one dip-cycleapart, i.e., about0.1day. Clearly, the dip morphologychanges from cycle to cycle. Fast dipping b activityisobserved,whichcanlastupto30min.Thesefastdips 27−28/9/2010 0 haveadurationof.30s,andoftenassmallas1s.Occasionally, 2.97 2.98 2.99 3 before and/or after the fast dips the intensity reaches the per- Time (MJD − 55464) sistent level seen outside the dipping intervals. During the fast Fig.4. Zoom in from Fig. 3d on the XMM-Newton/EPIC-pn dipstheintensitycandropdowntoabout15%ofthepersistent (0.2–15keV)lightcurvesaroundthetimeofdipactivityaround intensity, indicating that the shallower dips observed in Fig. 3 day2.9(a)and3.0(b),respectively.The timeresolutionis1s. areaconsequenceofaveragingasmallernumberoffastdipsin Alldatawithafractionalexposureofmorethan0.5perbinare coarsertimebins. shown. Next, we examine the hardness values of the XMM- Newton/EPIC-pn (Fig. 3e) and RXTE/PCA (see Fig. 5) light curves. The hardness ratio is defined as the ratio of the count ray binaries (Sect. 4.1.1), we refer to these dips as ‘absorption rates in the 2–10keV to the 0.6–2keV bands for the EPIC-pn dips’. data,and4.9–9.8keVtothe2.1–4.9keVbandsforthePCAdata. We note that UV data simultaneous to the XMM- Duringthedipsthesourcehardens.Thiscolourbehaviourdur- Newton/EPIC-pndataareavailablefromtheXMM-Newton/OM ingdipsisalsoconfirmedbytheSwift/XRTdata(seeKenneaet (seeFig.3f).AlthoughtheUV isvariable,theredoesnotseem al. 2011). We observe in particulara strongerhardeningas the tobeacorrelationwithX-rayintensity,inparticularwith theX- dip becomes deeper. As the out-of-dip intensity increases, the ray dipping. Unfortunately,however,the time resolution is too dippingbecomesshallowerandthechangesinthehardnessra- coarse to investigate in detail the UV light curve on the X-ray tiolesspronounced(seeFigs.5aandb).Thisisshowninmore dip structure time scale. Therefore, we do not discuss the UV detail in Fig. 5g, where we plot the changes of hardness ratio dataanyfurther. asafunctionofcountrate.Theshapeofthecurvetracedbythe PCAdatafromday7.1to9.1to12.7isremarkablysimilartothe 3.2.2. Transitiondips shapeofthecurveshowninfigure4ofBoirinetal.(2005)forthe classical LMXB dipper4U1323−62as the source movesfrom On three occasions (days 23.7, 24.0 and 26.1; see Fig. 6), the deepdippingtoshallowdippingandfinallytoapersistentstate.4 X-ray light curves show further pronounced variations. The Using the absorptiondip activityephemerisfrom Sect. 3.3, we RXTE/PCAlightcurveonday23.7(Fig.6a)resemblesthatseen findthatthedipsrecuratphase0.4–0.6(see,e.g.,Fig.5a–f). duringobservationswithabsorptiondips.However,thehardness Sincethedipmorphologyandhardnessbehaviourresembles (Fig.6d,g)correlateswiththeX-rayintensity,incontrasttothe thedipphenomenonencounteredinvarioushigh-inclinationX- absorptiondips.TheRXTE/PCAlightcurvesondays24.0and 26.1 (Fig. 6b and c) show the presence of two intensity levels 4 Note that the differences in the numbers are due to the different which differ by about 30%, between which the source fluctu- instruments used for the curves, RXTE/PCA in Fig. 5g and XMM- ates. The transitions are fast, i.e., they occur within tenths of Newton/EPIC-pninfigure4ofBoirinetal.(2005). seconds.The time spentin the upperlevelis about1–2min on E.Kuulkersetal.:MAXIJ1659−152 7 day24.0andatleast6–14minonday26.1.Thetimespentinthe tionwasdonebykeepingthetimetagsandrandomlydistribut- lowerlevelisabout12toatleast20minonday24.0andabout ingtheintensitiesofthedatasetswhichwereinputtotheperi- 26minonday26.1.Theselightcurveshavebeenreferredtoas odogramprograms.Thiswasrepeated1000timesandtheresult- ’flip-flop’lightcurves(seeKalamkaretal.2011).However,the ingaveragedperiodogramwasusedtoevaluatethesignificance intensity fluctuationscorrespondto fast source state transitions ofthepeaks/minima.We findthatthevaluesofthepower(LS) (seeSect.4.1.2),sowerefertothemas‘transitiondips’.Thein- oramplitudeΘ(PDM)arenarrowlydistributedaround1forall tensityvariationsseenduringthetransitiondipsaremoreorless periodsinvestigated:thevariancesare1fortheLSvaluesofall ofthesameorderasthevariationsintheaverageintensityfrom theinstruments,and0.001,0.001,0.002and0.007forthePDM observation to observation in outburst epoch C (see Fig. 1 and values of the RGS, EPIC-pn, XRT and PCA, respectively. We Sect.3.1;seealsoKalamkaretal.2011).ContrarytoKalamkar refertothislevelasthenoiselevel. et al. (2011),we find clear hardnesschangesduringthe transi- Another way to characterise the uncertainty in the period, tiondipslightcurves(Fig.6eandf).Thehardnessbehaviouris which is more conservative, is to use the width of the peak or similartothatseenonday23.7,againoppositetothatseendur- minimumoftheperiodograms,e.g.,thehalf-widthathalfmaxi- ing the absorptiondip episodes, i.e., the sourceis harderat the mumorminimum(HWHM). upperintensitylevel,andsoftenswhenthesourcetransitstothe Theresultsforthedifferentinstrumentsarediscussedinthe lowerlevel(Figs.6handi). nextsubsections,andasummaryofthebest-foundperiodsnear Although the phasing5 of the transition dips on day 26.1 0.1dayswiththeirassociatederrorsisgiveninTable3.Wefind, is consistent with the phasing of the absorption dips (see thattheHWHMvaluesareafactorof25–200timeslargerthan Sect.3.2.1),thephasingofthetransitiondipsondays23.7and those derived by the measured standard deviations of the po- 24.0isclearlynot.Observationsinbetweendays24.0and26.1 sitions of the peaks or minima, as described above. Since the donotshowanydippingbehaviour(evenbetweenphase0.4and spread in the best-foundperiodsis of the order of the HWHM 0.6), and indicate that the transition dips do not recur with the values,weusethesevaluesasafinalindicatoroftheuncertainty 0.1daytimescale. inthederivedperiods. 3.3.Timinganalysisofabsorptiondipactivity 3.3.1. XMM-Newton/EPIC-pnandRGS The Swift/XRT, XMM-Newton/EPIC-pn and RGS, and RXTE/PCA data indicate that the absorption dips occur WeusedtheEPIC-pnandtheRGSdatawith100stimeresolu- regularly,i.e.,aboutevery∼0.1day(Sect.3.2.1),fromthestart tion.ThehighestpeakintheLSperiodogramsisataperiodnear ofthemainoutburstupto aboutday10(outburstepochA,see 0.1day(toptwoleftpanelsofFig.7),i.e.,0.1004±0.0090days Sect.3.1).InLMXBdippers(Sect.4.1.1)absorptiondipsrecur and0.1010±0.0071days,respectively,fortheEPIC-pnandRGS withtheorbitalperiod.Usingourratherlargetimebaselinewe datasets.ThreedeepminimaarevisibleintheRGSandEPIC- can establish an accurateperiodof the recurringdip activity in pnPDMperiodograms,near0.1,0.2and0.4days(toptworight MAXIJ1659−152. panelsofFig.7).Thedeepestminimumisnotatthebestperiod We performed a Lomb-Scargle (LS; Lomb 1976, Scargle foundwiththeLSsearch,however. 1982)periodsearch,aswellasaPhaseDispersionMinimisation Inspection of the folded light curves on the three periods (PDM;Stellingwerf1978)periodsearch,onourdatasetstaken foundwiththePDM,revealsthatonlywhenfoldingthedataon duringoutburstepochA. The datasetsofeach instrumentwere the ≃0.1dayperiodis the absorptiondipactivity(i.e., atcount treated separately. The LS and PDM searches were done over rates.140cs−1and.75cs−1fortheEPIC-pnandRGS,respec- the period range 0.01–0.5days, with a frequency interval of tively) clustered within 0.2 in phase space. For the other two 0.001day−1. For the PDM search we used 20 phase bins with periods,thesameabsorptiondipactivityisdistributedalongall a phase bin width of 0.056. In the LS periodogram a peak in- phases.Thisstrengthensourconclusionthatthefundamentalpe- dicates a dominating period in the data set, whilst in a PDM riodisnear0.1days,sinceabsorptiondipactivityisexpectedto periodogramaminimumindicatesadominatingperiod. occuratrestrictedorbitalphases(seeSect.4.1.1). The error on a period found was computed by construct- ing 1000 synthesised data sets. These data sets were obtained bydistributingeachdatapointarounditsobservedvalue,byan 3.3.2. Swift/XRT amountgivenbyitserrorbarmultipliedbyanumberoutputby a Gaussian random-numbergeneratorwith zero mean and unit A LS and PDM search on the XRT data duringoutburstepoch variance. The measured standard deviation of the positions of A did not reveal any significant period, except for the satellite the deepest minima in Θ or highest peak in power in the re- orbital period around the Earth (see below). This we attribute sulting periodogramswas taken as the error. These latter peri- to the variation in intensity of the overall main outburst light odograms were done in a narrow range around ≃0.1 day, i.e., curve,whichisofthesameorderasthedropsinintensityduring between 0.0909 and 0.1111 days, with a frequency interval of theabsorptiondips(seeFig.2andKenneaetal.2011).Thein- 0.00009days. creaseintheout-of-dipintensityisrathergradualduringthefirst 6days.Thereafter,theout-of-diplightcurvevariesirregularlyon To check the significance of the peaks and minima found atimescaleofdays.We,therefore,firstdetrendedthedatausing in the LS and PDM diagrams, respectively,we randomisedthe amulti-orderpolynomial(seealsoKenneaetal.2011);athird- data and calculated the resulting periodogram.The randomisa- orderpolynomialdescribestheoverallout-of-diplightcurveup 5 By folding the data on the recurrence time of the absorption dip toaboutoutburstday6sufficientlywell.Includingdataafterday activity,≃0.1day(seeSects.3.2.1and3.3). 6 contaminatedour periodsearch significantly, and had the ef- 6 Thechoiceofnumberofphasebinsandphasebinwidthisrather fect of diminishing the peak and minima in our LS and PDM arbitrary;testswithvariousnumbersofphasebinsanddifferentphase searches,respectively,near0.1day.We,therefore,usethesedata binwidthyieldedconsistentresults. onlyuptoday6,intheremainderofthissubsection. 8 E.Kuulkersetal.:MAXIJ1659−152 The highest peak in the LS periodogram is at 0.1003±0.0010 days (see Fig. 7, left panel).7 Several min- ima are found in the PDM periodogram (Fig. 7, right panel), 4 XRT window including those seen at the same periods as in the PDM peri- 0. odogramsoftheXMM-Newtondata(Sect.3.3.1).Theminimum er w at0.1005±0.0010daysisclearlynotthedominantperiodinthis o PDMsearch.However,inspectionofthefoldedlightcurveson p 2 0. the various periods with peaks or minima in the LS and PDM searches,respectively,showthat,again,onlyfortheperiodnear 0.1 day, the dip activity is clustered within 0.2 in phase space, 0 asexpectedforabsorptiondips(seeSect.4.1.1). We notethatperformingtheperiodsearchafterrenormalis- ing the XRT data in a similar manneras we did forthe RXTE 0.3 PCA window data(seeSect.3.3.3)didnotrevealsignificantpowerattheabove reportedperiods.ThisisbecausetheXRTobservationswere in er 2 generalshorterthantheRXTE/PCA,resultinginaverageswhich w 0. o arehigherwhenthereisadippingperiod,andthereforethevari- p ationsduetodippingarediminished. 0.1 To inspectwhetherthe fundamentalperiodis related to the datasampling(suchasduetothesatelliteorbit),weconstructed 0 aFouriertransform(FT)ofthewindowfunction.Thiswindow 0.01 0.02 0.05 0.1 0.2 0.5 function was determined by setting the intensities of the time Period (days) series to zero. Forthe FT we used the same frequencysettings as the LS and PDM searches. The largest peak in the FT pe- Fig.8. The Fourier transforms of the window functions using riodogram is evidently at the satellite orbit period around the theRXTE/PCA(top)andSwift/XRT(bottom)data.Thepeaksin Earthofabout0.067days(Fig.8,top).Thusourbestperiodis powerdensityspectraareconsistentwiththesatelliteorbitalpe- notrelatedtoanypeakintheFTperiodogram,andcanthusbe riodsaroundtheEarth,i.e.,0.065days(94min)and0.067days considered to be intrinsic to the source (see also Kennea et al. (96min),respectivelyfortheRXTE/PCAandSwift/XRT. 2011). XMM-Newton and XRT period searches (Fig. 7, bottom left).8 3.3.3. RXTE/PCA Weattributethistothehighlynon-sinusoidalnatureofthemod- TheoverallvariabilityofthePCAdata(seeSect3.1andFig.1c), ulation in the renormalisedlightcurvesat 16s time resolution, preventedtheLSandPDMperiodogramstoshowanysignificant andthepossibleimperfectionofourmethodofrenormalisation peaksorminima,respectively,exceptatthesatellite orbitalpe- fortheLSsearch. riodaroundthe Earth(see below).DetrendingthePCA dataas donefortheXRTdidnotimproveourperiodsearches,however. 3.3.4. Recurrenceperiodofdippingactivity ThisisduetothefactthatthePCAdatasampling(aswellasthe energyrange)isdifferentfromthatoftheXRTdata.Moreover, The XMM-Newton and XRT LS period searches and the therearesignificantday-to-dayvariationsintheaverageout-of- PCA PDM period search reveal one common best period, dipPCAintensity,whichcannotbedescribedbyasimplepoly- i.e., ∼0.1 day, with peaks/minima in the periodograms well nomial.We,therefore,renormalisedthePCAlightcurvesduring above/below the noise level. Inspection of the individual light the whole outburstepochA in oursearch forperiodicities.For curves of the various instruments, and the folding of the data eachobservationinterval(generallycorrespondingtoanRXTE onthe≃0.1dayperiodleadingtorestrictedphaserangeofdip- satellite orbit) we determined the mean count rate. This value ping (see below), supports the main period to be at that value. was subtracted from the light curves corresponding to each of We attribute the peaks/minimaat half this period in the XMM- theseobservationintervals. Newtonperiodogramstotheintermediatedippingepisodesseen. The lowest minimum in the PDM periodogram is at a pe- Thepeaks/minimaatmultipletimestheabovequotedperiodare riod of 0.10058±0.00022days (Fig. 7, bottom right). A PDM duetothefactthatthemorphologyofthedippingchangesfrom searchinthreeenergybands(2–5keV,5–10keVand10–20keV, cycletocycle,aswellasthefactthatnotalldippingperiodsare seeSect.2.2)showsthedeepestminimaatthesameperiod(see sampledwellenough. Table3).Again,theperiodquotedabovedoesnotcoincidewith The RXTE/PCA provides the strongest constraint on the thestrongpeakatabout0.065daysintheFTofthePCAwindow period. We, therefore, conclude that the fundamental period function(Fig.8,bottompanel). of dipping activity is at 0.10058±0.00022days. We arbitrarily A LS search on the PCA data did not reveal a peak near set the zero point of the absorption dip activity ephemeris to ≃0.1day period,nornearany of the otherperiodsfoundin the the first data point of the first RXTE/PCA observation at day 3.0. Therefore, the absorption dip activity ephemeris is T = 0 MJD55467.039561+0.10058(22)×E,whereEisthecyclenum- 7 We note that the period reported by Kennea et al. (2011), beroftheperiod. 0.1008±0.0037 days, was determined using the LS search on the de- trended first 12.8 days of the XRT/WT data. Their error in the pe- riodisderivedbyfittingaGaussiantotheLomb-Scargleperiodogram 8 ThehighestpeakintheLSperiodogramisat0.0335days.However, aroundthepeak,andtakingavalueof2.7σ,whereσisthewidthofthe folding the data at this period reveals the absorption dips to be dis- Gaussian. tributedalongallphases. E.Kuulkersetal.:MAXIJ1659−152 9 Table3.ResultsofthePDMandLSperiodsearchesinthelightcurvesforthedifferentinstruments. PDM LS Instrument period error HWHM period error HWHM (energy) (days) (days) (days) (days) (days) (days) RGS(0.3–2keV) 0.10070 0.00013 0.0042 0.100985 0.000098 0.0071 EPIC-pn(0.2–15keV) 0.09931 0.00006 0.0049 0.100394 0.000190 0.0090 XRT(0.3–10keV) 0.10054 0.00004 0.0010 0.100282 0.000011 0.0010 PCA(2–60keV) 0.100582 0.000001 0.00022 — — — PCA(2–5keV) 0.100590 0.000018 0.00021 — — — PCA(5–10keV) 0.100590 0.000019 0.00023 — — — PCA(10–20keV) 0.100580 0.000024 0.00022 — — — 0 We constraintheinclinationofMAXIJ1659−152to bebe- 0 1 tween about65◦ and 80◦ fromthe presenceof the periodicab- y sit sorptiondips,duetomaterialinthelineofsightthatobscuresup n nte 0 to about90% of the total emission at given cycles, and the ab- d i senceofeclipses.Webasethelowerlimitonthesizeofthebulge e aliz (andnotonthediskopeninganglewhichhasbeengenerallyesti- enorm −100 mWahtietdet&obHeo≃lt1(21◦9,8e2.g).,edsteimJoantegdetthael.s1iz9e96o,fBtahyelebsuslgeetarle.s2p0o1n0si)-. r ble for absorption dips as 19◦±6◦ for the LMXB 4U1822−37. TakingthisastypicalforLMXBs,wecanthussetalowerlimit 0 0.2 0.4 0.6 0.8 1 phase on the inclination of 65◦. We note that, if the accretion disk is tilted orwarped, the lower limit forthe inclinationcould be as Fig.9. ThedetrendedRXTE/PCA (2–60keV)16s datafolded lowas55◦,takingintoaccountthatforgenericLMXBs,adisk ontheephemerisgiveninSect.3.3.4. tilt of about 10◦ is expected (Foulkes et al. 2010). An upper limit for the inclination of 80◦ is derived from the absence of ThefoldeddetrendedPCAcurveduringthedipepochusing eclipses(e.g.,Horne1985),the spectraltypeof thecompanion thisephemerisisshowninFig.9.Thedippingactivityhasaduty star (M5V, see Sect. 4.2.1), and the fact that the companion is cycle of about 0.2 in phase, around phase 0.5. The mid-point fillingitsRochelobe(see,e.g.,Motchetal.1987).Hereweare betweenstartandendofthe dutycyclecorrespondsto T = assuming that the source of emission is point-like (for an ex- 0,dip MJD55467.0904±0.0005. tendedsourceonewouldnotbeabletoseefulleclipsesforany oftheLMXBs). TheabsorptiondipsinMAXIJ1659−152sharemanyofthe 4. Discussion propertiesofclassicalabsorptiondippingsystems.Theychange from period to period, they are fast, and the obscuration can 4.1.Originofintensityvariations be large, i.e., down to about 90% of the total intensity (see, 4.1.1. Absorptiondips e.g., White & Mason 1985, Parmar & White 1988). Boirin et al. (2005) and D´ıaz Trigo et al. (2006) were able to model the WefindregularabsorptiondipsintheX-raylightcurvesduring changes in both the narrow X-ray absorption features and the theoutburstofMAXIJ1659−152betweenoutburstdays0.3and continuumduring the dips from all the brightdippingLMXBs 8.2.Absorptiondipsrecurattheorbitalperiodofthesystemand observedbyXMM-Newtonbyanincreaseinthecolumndensity arethoughttobecausedbyobscurationbymateriallocatedina and a decrease in the amount of ionisation of a photo-ionised thickenedouterregion(‘bulge’)ofthe accretiondiskdueto its absorbing plasma. The changes in the hardness ratio observed interaction with the inflowing gas stream from the companion in the dips in MAXIJ1659−152are consistent with absorption (e.g., White & Swank 1982, Walter et al. 1982; see discussion by neutraland photo-ionisedplasma, in the sense that they are below). The presence of absorption dips allows a precise mea- energydependent.A furthersupportto the existenceof neutral surementoftheorbitalperiodandisasignatureofhighinclina- and photo-ionised plasma is the presence of various stages of tion(e.g.,White&Swank1982,White&Mason1985;seealso dipping: persistent, shallow and deep dipping states. The fact D´ıazTrigoetal.2009). We determine the recurrence period of the dips to be 2.414±0.005hrs (see also Kuulkerset al. 2012a,Kennea et al. note,however,thattheoutburstamplitudeintheoptical(∼6mag,Rauet 2011).Byanalogywithotherclassicaldippers,weidentifythis al.2011)isincompatiblewithsuchashortperiod,basedontheempir- icalrelationbetweentheoutburstamplitudeandorbitalperiod(Shabaz periodwiththeorbitalperiodofasystem.Thefastestrevolving &Kuulkers1998),seealsoSect.4.6.Anothershort-periodbinarycan- binary was SwiftJ1753.5−0127(3.2443hrs; Zurita et al. 2008, didate, MAXIJ1305−704, was put forward recently by Kennea et al. Durant et al. 2009). As suggested by Kuulkers et al. (2010d, (2012b).Thisnewtransient(Satoetal.2012)showedabsorptiondips. 2012a), if the compactobject in MAXIJ1659−152is indeed a Aperiodicityof∼1.5hrswasreported(Kenneaetal.2012b), butthis black hole (Kalamkar et al. 2011, Kennea et al. 2011, Mun˜oz- wasputintodoubtbyKuulkersetal.(2012b).Ithasbeensuggestedthat Darias et al. 2011b, Shaposhnikov et al. 2012, Yamaoka et al. MAXIJ1305−704isaBHXBbasedonoutburstopticalamplitude,blue 2012), its ≃2.4hrs period is the shortest among the currently opticalspectralenergydistributionandhardX-rayspectrum(Greineret knownBHXBsample(see,e.g.,Ritter&Kolb2003).9 al.2012;seealsoKenneaetal.2012a),aswellastheoccurrenceofa statetransition(Suwaet al.2012). However, thesearefeatureswhich 9 ApossibleexceptionmaybeSwiftJ1357.2−0933,basedonanin- areseeninneutronstarLMXBsaswell(seeSuwaetal.2012,Kennea directestimateoftheorbitalperiodof2hrbyCasaresetal.(2011).We etal.2012b). 10 E.Kuulkersetal.:MAXIJ1659−152 that dips become shallower and less energy dependent as the changes could be discerned as well, with behaviour similar to countrateincreasescouldbeaconsequenceofthephoto-ionised thatseenforMAXIJ1659−152.Thecleardifferenceinhardness plasmabecomingmoreandmoreionisedandtransparentasitis behaviour of the absorption and transition dips, as well as the illuminated by the X-rays of the central region. However, we factthattheyoccurredinwellseparatedphasesoftheoutburst, note that a definite confirmation of an increase of neutral and suggeststhatthetwophenomenahaveadifferentorigin. ionisedplasmaduringdipsforMAXIJ1659−152isonlypossi- bleafterspectralanalysis. ThetransitiondipsinMAXIJ1659−152occurredduringthe Absorption dippingin otherLMXBs occurs mainly around first soft excursion of the source (see Fig. 10). During this pe- orbitalphase0.7–0.9,whereeclipsesareexpectedatphasezero riodthecountratedifferencesbetweenconsecutiveobservations if we view the accretion disk edge-on, i.e., when the compan- wereofthesameorderasthecountratechangesseenduringthe ion star is closest to us and in front of the neutron star (e.g., transitiondips(withthesofterobservationshavinglower count Parmar&White1988).Occasionally,absorptiondipswitha0.5 rates),suggestingthatadditionaltransitionstookplacebetween phasedifferencewithrespecttothephaseatwhich‘regulardips’ observations. occurare observed.These ‘anomalousdips’,or‘secondaryab- Intheothersourcesinwhichtransitionsdipshavebeenseen, sorptiondips’,werealsoseeninotherdippingsystems(suchas thetransitionswereoftenaccompaniedbypronouncedchanges XB1916−053, e.g., White & Swank 1982, Walter et al. 1982, in the power density spectra (Miyamoto et al. 1991, Takizawa Smale et al. 1988, Boirin et al. 2004). They are explained as etal.1997,Homanetal.2001,Casellaetal.2004).Thepower being due to material migrating to the other side after impact densityspectrafromthetwoobservationsthatshowedthetran- with the disk, or the accretion stream partly freely overflow- sitions dips in MAXIJ1659−152 (days 24.0 and 26.1) were ingthediskorbouncingofthediskrimandthenoverflow(see, not of high enough quality to detect significant changes in the e.g.,Frank et al. 1987,Armitage& Livio 1998,andreferences power-densityspectralproperties.However,byanalysingtheav- therein). In the latter case the flow may impact the disk near eraged powerdensity spectra from the combined high and low the circularisation radius, either causing a second bulge (see, countratelevelsduringthefirst‘softexcursion’,Kalamkaretal. e.g.,Frank et al. 1987,Armitage& Livio 1998,andreferences (2011)wereabletoseeindicationsforanadditionalbroadbump therein)orbouncingofthediskagain(Kunzeetal.2001). around7–8Hzinthepowerdensityspectraofthelowcountrate The absorption dips appear for only part of the outburst selection.Wenotethatthetransitionsinothersourcesoftenin- in MAXIJ1659−152. This has been observed already for volve so-called ‘type B’ quasi-periodic oscillations, QPOs, ei- other BHXBs undergoing an outburst, like 4U1630−47 and ther in the low or high count rate level power density spectra. GROJ1655−40 (Kuulkers et al. 1998, 2000, Tomsick et al. ThebroadexcessseenbyKalamkaretal.(2011)istoobroadto 1998). It is plausible that for transient BHXBs, changes in the beidentifiedasatypeBQPO,andismorelikelytobeapeaked accretionmode cause the appearanceordisappearanceof dips. noisecomponent. Kuulkersetal.(2000)interpretedthe(deep)absorptiondipsdur- ingtheriseandplateauphaseoftheoutburstinGROJ1655−40 Transition dips are most likely the result from instabilities as due to an absorbingmediumwhich is filamentaryin nature. intheinneraccretionflow(seeMiyamotoetal.1991foranex- These filaments couldbe due to the stream of materialcoming ample interpretation), but their exact origin remains unknown. from the companion star splashing into the accretion disk and The observations of MAXIJ1659−152 do not provide signifi- overflowing above and below the impact region (e.g., Livio et cant new insights into the nature of these instabilities, but they al.1986;seeKuulkersetal.2000,andreferencestherein).Ifthe doshowthattransitiondipscanalso befoundinstatesthatare inclination is high enough, the impact region itself comes also slightly softer than those in which they have been observed in intothelineofsight(e.g.,Franketal.1987).However,thepres- othersources(i.e., states in which typeB QPOs are observed). enceofabsorptionfeaturesallaroundtheorbitforneutronstars ThistrendisveryclearlyseeninFig.10,wherethetypeBQPOs (e.g., Parmar et al. 2002) shows that at least part of the photo- alloccurwithinahardnessrangeof0.36–0.40,whilstalltransi- ionisedplasmaisdistributedequatoriallyalongthewholeplane tiondipsoccurathardness.0.36. ofthedisk,indicatingthatabsorptionisduetoastructureinthe diskratherthanbyfilaments.Inthatscenario,thecausefor the Aroundthetimeoftheoccurrenceoftransitiondipswefind disappearance of dips in BHXBs could be, e.g., a strong ioni- somemarginalevidenceforthesoftandhardX-raylightcurves sation of the plasma in bright (but hard) states of the outburst, to modulate on a ≃3 day period. We speculate that this period whichrendersthe plasma transparentandthereforeinvisibleto may be related to a disk precession period. Systems with ex- us.Alternatively,achangeofthestructureofthediskcoulddi- treme mass ratio’s (i.e., q . 0.33), like MAXIJ1659−152,are minishthethicknessofthebulgeandcausetheabsorptiondips vulnerableto a 3:1 orbital resonancewithin the accretion disk. todisappear. This causes the disk to be eccentric and to slowly precess on time scales of days to weeks, which may be discernable in the lightcurves(Whitehurst1988,Whitehurst&King1991,Lubow 4.1.2. Transitiondips 1991a,b; see also Haswell et al. 2001). Periodic variations are As noted by Kalamkar et al. (2011), during the second epoch alsoforeseeninthismodelwithaperiodslightlylongerthanthe of the outburst rapid and sharp flux variations — transition orbitalperiod.However,inLMXBsthisphenomenonisinclina- dips — were seen, resembling the ‘flip-flop’ and ‘dip’ light tion dependent:in systems with a high orbital inclination only curvesin,e.g.,GX339−4(Miyamotoetal.1991),GS1124−68 orbitalmodulationsduetotheheatedfaceofthecompanionstar (Takizawaetal.1997)andXTEJ1859+226(Casellaetal.2004). are expected (Haswell et al. 2001). This is consistent with the In all these cases 10–20% changes in intensity were seen. fact that the X-ray and optical light curves show the same (or- We find that MAXIJ1659−152 softens when the intensity de- bital)period(seeSect.1).Thetransitiondipsoccurredduringthe creases, in contrast to the hardening seen during the absorp- timeofthe≃3daymodulation.Possibly,thenon-axisymmetric tion dips. In the case of GX339−4 (Miyamoto et al. 1991) accretiondiskmodulatestheinneraccretionflow,givingriseto and XTEJ1859+226 (Casella et al. 2004) significant hardness thesporadictransitiondipsintheX-raylightcurves.

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.