V1006 Cygni: Dwarf Nova Showing Three Types of Outbursts and Simulating Some Features of the WZ Sge-Type Behavior
Taichi Kato, Elena P. Pavlenko, Alisa V. Shchurova, Aleksei A. Sosnovskij, Julia V. Babina, Aleksei V. Baklanov, Sergey Yu. Shugarov, Colin Littlefield, Pavol A. Dubovsky, Igor Kudzej, Roger D. Pickard, Keisuke Isogai, Mariko Kimura, Enrique de Miguel, Tamas Tordai, Drahomir Chochol, Yutaka Maeda, Lewis M. Cook, Ian Miller, Hiroshi Itoh
aa r X i v : . [ a s t r o - ph . S R ] D ec Publ. Astron. Soc. Japan (2014) 00(0), 1–6doi: 10.1093/pasj/xxx000 V1006 Cygni: Dwarf Nova Showing Three Typesof Outbursts and Simulating Some Features ofthe WZ Sge-Type Behavior
Taichi K
ATO , Elena P. P
AVLENKO , Alisa V. S
HCHUROVA , Aleksei A. S
OSNOVSKIJ , Julia V. B
ABINA , Aleksei V. B
AKLANOV , Sergey Yu. S
HUGAROV , , Colin L
ITTLEFIELD , Pavol A. D
UBOVSKY , Igor K
UDZEJ , Roger D. P
ICKARD , , Keisuke I
SOGAI , Mariko K
IMURA , Enrique de M
IGUEL , , Tam ´as T
ORDAI , Drahomir C
HOCHOL , Yutaka M
AEDA , Lewis M. C
OOK , Ian M
ILLER , Hiroshi I
TOH , Department of Astronomy, Kyoto University, Kyoto 606-8502, Japan Crimean Astrophysical Observatory, p/o Nauchny, 298409, Republic of Crimea Taras Shevchenko National University of Kyiv, Glushkova ave., 4, 03127, Kyiv, Ukraine Sternberg Astronomical Institute, Lomonosov Moscow State University, Universitetsky Ave.,13, Moscow 119992, Russia Astronomical Institute of the Slovak Academy of Sciences, 05960, Tatranska Lomnica, theSlovak Republic Department of Physics, University of Notre Dame, 225 Nieuwland Science Hall, NotreDame, Indiana 46556, USA Vihorlat Observatory, Mierova 4, 06601 Humenne, Slovakia The British Astronomical Association, Variable Star Section (BAA VSS), Burlington House,Piccadilly, London, W1J 0DU, UK Departamento de F´ısica Aplicada, Facultad de Ciencias Experimentales, Universidad deHuelva, 21071 Huelva, Spain Center for Backyard Astrophysics, Observatorio del CIECEM, Parque Dunar,Matalasca ˜nas, 21760 Almonte, Huelva, Spain Polaris Observatory, Hungarian Astronomical Association, Laborc utca 2/c, 1037 Budapest,Hungary Kaminishiyamamachi 12-14, Nagasaki, Nagasaki 850-0006, Japan Center for Backyard Astrophysics Concord, 1730 Helix Ct. Concord, California 94518, USA Furzehill House, Ilston, Swansea, SA2 7LE, UK Variable Star Observers League in Japan (VSOLJ), 1001-105 Nishiterakata, Hachioji,Tokyo 192-0153, Japan ∗ E-mail: ∗ [email protected] Received 201 0; Accepted 201 0
Abstract
We observed the 2015 July–August long outburst of V1006 Cyg and established this objectto be an SU UMa-type dwarf nova in the period gap. Our observations have confirmed thatV1006 Cyg is the second established object showing three types of outbursts (normal, longnormal and superoutbursts) after TU Men. We have succeeded in recording the growing stage c (cid:13) Publications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0 of superhumps (stage A superhumps) and obtained a mass ratio of 0.26–0.33, which is closeto the stability limit of tidal instability. This identification of stage A superhumps demonstratedthat superhumps indeed slowly grow in systems near the stability limit, the idea first introducedby Kato et al. (2014). The superoutburst showed a temporary dip followed by a rebrightening.The moment of the dip coincided with the stage transition of superhumps, and we suggestthat stage C superhumps is related to the start of the cooling wave in the accretion disk. Weinterpret that the tidal instability was not strong enough to maintain the disk in the hot statewhen the cooling wave started. We propose that the properties commonly seen in the extremeends of mass ratios (WZ Sge-type objects and long-period systems) can be understood as aresult of weak tidal effect.
Key words: accretion, accretion disks — stars: novae, cataclysmic variables — stars: dwarf novae —stars: individual (V1006 Cygni)
Cataclysmic variables (CVs) are composed of a white dwarf anda red (or brown) dwarf supplying matter to the white dwarf,forming an accretion disk. Dwarf novae are a class of CVscharacterized by outbursts. SU UMa-type dwarf novae are asubclass of dwarf novae which show superoutbursts in additionto normal outbursts. During superoutbursts, superhumps havingperiods a few percent longer than the orbital periods ( P orb ) areobserved and are considered to be the defining characteristicsof SU UMa-type dwarf novae [For general information of CVs,SU UMa-type dwarf novae and superhumps, see e.g. Warner(1995a)]. The origin of superhumps and superoutbursts is cur-rently understood as the consequence of the 3:1 resonance in theaccretion disk resulting tidal instability combined with thermalinstability (thermal-tidal instability model; Osaki 1989; Osaki,Kato 2013a). Only systems having mass ratios ( q = M /M )smaller than ∼ P orb for systems with short P orb (cf. Kato et al.2009; Kato et al. 2015). In long- P orb systems, however, there have been a number of objects showing a strong decrease ofthe superhump periods [the best-known examples are MN Draand UV Gem, see subsection 4.10 in Kato et al. (2009)]. Theorigin of strongly negative period derivatives had remained amystery. Kato et al. (2014) proposed a working hypothesis thatwhat looked like strongly negative period derivatives for stageB superhumps in such systems are actually caused by stage A-B transition based on the photometrically detected P orb in MNDra. Kato et al. (2014) suggested that the 3:1 resonance growsslower in systems having large q close to the stability borderof the resonance and that this is observed as long-lasting stageA. This interpretation violated the received wisdom that long-lasting stage A reflects the small q , as is typically seen in WZSge-type dwarf novae (Kato 2015), which have completely op-posite properties to long-orbital systems. Since the discussionin Kato et al. (2014) was based on the yet unconfirmed P orb ofMN Dra, further confirmation is clearly needed. We present thedetection of long-lasting growing superhumps in a long- P orb system V1006 Cyg having a spectroscopically established P orb .V1006 Cyg was discovered as a dwarf nova (S 7844)(Hoffmeister 1963a; Hoffmeister 1963b) with a photographicrange of 16–18 mag. Gessner (1966) derived an outburst cy-cle length of 13.5 d. Bruch et al. (1987) recorded another out-burst. Bruch, Schimpke (1992) obtained a spectrum and es-tablished this object to be a dwarf nova. Due to the initiallyreported faintness, this object had not been well studied sincethen. Since 2005, AAVSO observers started regular monitor-ing using CCDs and recorded an outburst reaching V =13.6 on2006 June 24. This outburst lasted at least for 4 d. Sheets et al.(2007) performed a radial-velocity study and obtained P orb of0.09903(9) d. Since this period places the object in the periodgap, the slowly fading 2006 outburst was suspected to be a su-peroutburst. Upon a bright (unfiltered CCD magnitude 13.6)outburst on 2007 August 14, a search for superhumps was con-ducted (vsnet-alert 9471). Although short-term modulationswere detected, these variations were later found to be orbital ublications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0 variations (vsnet-alert 9489). This observation and the obser-vation of a similar bright outburst in 2009 September (vsnet-alert 11490, 11508) were examined in detail by Pavlenko et al.(2014) and the absence of superhumps was confirmed.On 2015 July 12, ASAS-SN (Shappee et al. 2014; Danilet etal. in preparation) detected an outburst at V =14.1 (vsnet-alert18846). This outburst was detected sufficiently early and theinitial phase of the outburst was observed. The object reached V =13.6 and low-amplitude superhumps were detected on July15–16 (vsnet-alert 18851). The superhumps were observed toglow until July 18 (BJD 2457222) (see figure 1). The data were obtained under campaigns led by the VSNETCollaboration (Kato et al. 2004). We also used the public datafrom the AAVSO International Database . Time-resolved ob-servations were performed in 13 different locations by using30cm-class telescopes (table 2). We deal with observations un-til August 7. The data analysis was performed just in the sameway described in Kato et al. (2009) and Kato et al. (2014) andwe mainly used R software for data analysis. In de-trendingthe data, we divided the data into four segments in relation tothe outburst phase and used locally-weighted polynomial re-gression (LOWESS: Cleveland 1979) except the rising segmentof a rebrightening. During the rising phase of the rebrighten-ing a third order polynomial fitting was instead used. The timesof superhumps maxima were determined by the template fittingmethod as described in Kato et al. (2009). The times of all ob-servations are expressed in barycentric Julian Days (BJD). The amplitudes of superhumps before BJD 2457220 weresmall, indicating that we recorded the growing stage (stage A)of superhumps. Although the observations before BJD 2457220were short and the initial night suffered from poor conditions,observations between BJD 2457221 and 2457223 were of suf-ficient quality to determine the period in the early phase (table3). The O − C analysis (17 ≤ E ≤
28) and PDM analysis yieldedperiods of 0.1073(2) d and 0.1076(1) d, respectively. The periodof 0.1075 d (average of the two methods) is 8.5% longer than P orb , giving an exceptionally large fractional superhump ex-cess. Since the amplitudes of superhumps on BJD 2457221 al-ready approached the maximum, these periods are likely shorterthan the true period of stage A superhumps, because the pres-sure effect starts to dominate when superhumps fully grow and < > . The R Foundation for Statistical Computing: < http://cran.r-project.org/ > . Fig. 1. O − C diagram of superhumps in V1006 Cyg (2015). (Upper:) O − C diagram. We used a period of 0.10541 d for calculating the O − C residuals.The superhump maxima up to E = 28 are stage A superhumps, maximabetween E =36 and E = 94 have a positive period derivative and are identi-fied as stage B superhumps. After this, the period decreased to a constantone (stage C superhumps). (Middle:) Amplitudes of superhumps. The am-plitudes were small around E = 0 . The O − C diagram suggests that stageA-B transition occurred somewhere between E = 28 and E = 36 . Thesuperhump amplitudes monotonously decreased during the superoutburst.After E = 100 , the amplitudes became large (0.3–0.4 mag) when the objectfaded. (Lower:) Light curve. The data were binned to 0.035 d. The initialoutburst detection was on BJD 2457215.9, 3 d before the start of our ob-servation. It took 6 d for this object to fully develop stage B superhumps.The maximum on BJD 2457232 is a rebrightening. The maximum on BJD2457238 is the first normal outburst of regular series of outbursts following asuperoutburst. reduces the precession rate (Kato, Osaki 2013). By using thisperiod as an approximate period of stage A superhumps andwith the strong expectation that the true period of stage A super-humps is longer than this period, we have been able to resolvethe ambiguity in the cycle counts between BJD 2457220 and2457221. The cycle counts in table E2 are based on this iden-tification. The resultant mean period of stage A superhumpsbetween BJD 2457219 and 2457223 by the O − C analysis is0.1093(3) d, which we consider the best period from the presentobservations. This period gives ǫ ∗ of 0.094(3), which corre-spond to q =0.34(2).The duration of stage A was at least 32 cycles. Although thetrue duration of the growing phase of superhumps is unknownin this object due to the observational gap, the close similarity ofthe O − C diagram and variation of superhumps amplitudes be-tween V1006 Cyg and MN Dra (figure 2) suggests it took longtime to develop superhumps in this system. It also took 6 d ( ∼ Publications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0 cycles) since the outburst detection to fully develop stage B su-perhumps based on the O − C diagram. Since this object has rel-atively frequent outbursts, it it likely that the Case A outburst ofOsaki, Meyer (2003) classification should occur in this object.However, the observed delay in superhumps likely reflects thelong growth time of superhumps (just like the Case B outburstfor low- q system; in most ordinary SU UMa-type dwarf novae,the growing stage of superhumps is rarely recorded a few daysafter the outburst detection, e.g. Kato et al. 2009). This delayof appearance of stage B superhumps is unusually long for anordinary SU UMa-type dwarf nova and is even comparable toextreme WZ Sge-type objects (cf. Kato 2015).After reaching the maximum amplitude, the superhump pe-riod became short as in stage B superhumps in ordinary SUUMa-type dwarf novae (Kato et al. 2009; Kato, Osaki 2013).We identified 40 ≤ E ≤
98 as stage B and obtained a mean pe-riod of 0.10541(4) d and a period derivative P dot = ˙ P /P of +20 . . × − . The O − C analysis indicates that the timesof superhumps for E ≥
106 can be very well expressed by a pe-riod of 0.10444(5) d, which we consider the period of stage Csuperhumps.
As described in subsection 3.1, the modern method using stageA superhumps gives a very large mass ratio of q =0.34(2). Sincethe early part of the observations was not ideally obtained, wegive a firmly determined lower limit of the period of stage Asuperhumps (0.1075 d), which corresponds to q ≥ q = q ∼ As described in Kato, Osaki (2013), the precession rate of stageC superhumps can be used to estimate the disk radius if the massratio is known, since the pressure effect can be neglected in coolpost-superoutburst disks. If the disk radius can be estimated, wecan constrain the mass ratio (e.g. Kato et al. 2013b). The mea-sured ǫ ∗ for stage C superhumps of V1006 Cyg is 0.0518(10).This value corresponds to a disk radius of 0.37 A , where A isthe binary separation, for q =0.26 and 0.34 A for q =0.34. The stage B-C transition described in subsection 3.1 occurredduring the rapid decline from the superoutburst plateau. Thisfeature is different from the behavior from the one in “textbook”SU UMa-type dwarf novae, in which stage B-C transition usu-ally occurs during the later part of the superoutburst plateau andis usually associated with a small brightening trend (Kato et al.2009). In V1006 Cyg, the object instead faded and a rebright-ening was recorded after the transition. Six days after this re-brightening, the object underwent another outburst. As judgedfrom the subsequent behavior (E. Pavlenko et al. in prepara-tion), this outburst was the first normal outburst of the regularcycle of normal outbursts. The origin of stage B-C transition is still poorly understood.In the present case, it appears that the cooling front started be-fore the termination of the plateau phase, since the first rebright-ening occurred only three days after the rapid fading. Althoughsuch early occurrence of a rebrightening is rarely met in ordi-nary SU UMa-type dwarf novae, similar one was observed inthe long- P orb system MN Dra (see figure 3 in Antipin, Pavlenko2002). We propose that the mass ratio close to the stability bor-der of the 3:1 resonance in V1006 Cyg made it difficult to main-tain the tidal instability, and the thermal and tidal instabilitiesdecoupled as proposed for ER UMa-type objects and WZ Sge-type rebrightenings as suggested by Hellier (2001). AlthoughHellier (2001) considered that small q is responsible for thisphenomenon, we can extend the same discussion to objects withlarge q close to the stability border of the 3:1 resonance.It would be worth mentioning that stage B-C transition isnot usually observed in WZ Sge-type objects (Kato 2015). Itis possible that rapid fading from the superoutburst plateau (of-ten seen as a temporary dip) in WZ Sge-type dwarf novae hasthe same properties as in V1006 Cyg. The common behavior(termination of the plateau phase before appearance of stage Csuperhumps, dip-like fading and rebrightening) in WZ Sge-typeobjects and objects having mass ratios close to the stability bor-der may be understood in a unified way: the small effect of thesmall tidal torque is unable to maintain the hot state when thecooling front starts. Up to this work, TU Men was the only established dwarf novathat shows three types of outbursts (normal, long normal andsuperoutbursts; the names here are given in modern sense)(Warner 1995b; Bateson et al. 2000). The only other possibleexample is NY Ser (Pavlenko et al. 2014) which showed out-bursts with intermediate durations without superhumps. Since The identification as a rebrightening is also based on the similarity of thebehavior with the SU UMa-type dwarf nova QZ Ser in the period gap (T.Ohshima et al. in prep.), which shows only very infrequent outbursts. ublications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0 there were outbursts in V1006 Cyg lasting more than six days(2007) and more than five days (2009) without superhumps(Pavlenko et al. 2014), the present detection of a genuine su-peroutburst makes V1006 Cyg the second case showing threetypes of outbursts. This indicates that P orb above the periodgap is not an essential condition for displaying such behavior. The interpretation that superhumps slowly grow in systems withmass ratios close to the stability limit was first presented inKato et al. (2014) for MN Dra. Although there still remainsuncertainty about P orb of MN Dra, and the identification of su-perhumps stages remained somewhat unclear, the present de-tection of growing superhumps in V1006 Cyg has establishedthis interpretation. In table 1, we list the objects having longorbital (or superhump) periods and long-lasting stage A super-humps. The suspected orbital periods for MN Dra and CRTSJ214738.4 + + Acknowledgments
This work was supported by the Grant-in-Aid “Initiative forHigh-Dimensional Data-Driven Science through Deepening ofSparse Modeling” (25120007) from the Ministry of Education,Culture, Sports, Science and Technology (MEXT) of Japan.This work also was partially supported by grants of RFBR 15-32-50920 and 15-02-06178, 14-02-00825 and by the VEGAgrant No. 2/0002/13.
Supporting information
Additional supporting information can be found in the onlineversion of this article: Figure 2, Tables 1, 2, 3.Supplementary data is available at PASJ Journal online ( in-cluded at the end in this astro-ph version ). References
Antipin, S. V., & Pavlenko, E. P. 2002, A&A, 391, 565Bateson, F., McIntosh, R., & Stubbings, R. 2000, Publ. Variable StarsSect. R. Astron. Soc. New Zealand, 24, 48Bruch, A., Fischer, F.-J., & Wilmsen, U. 1987, A&AS, 70, 481Bruch, A., & Schimpke, T 1992, A&AS, 93, 419Cleveland, W. S. 1979, J. Amer. Statist. Assoc., 74, 829Gessner, H. 1966, Ver¨off. Sternw. Sonneberg, 7, 61Hellier, C. 2001, PASP, 113, 469 Hoffmeister, C. 1963a, Mitteil. Ver¨anderl. Sterne, 751Hoffmeister, C. 1963b, Astron. Nachr., 287, 169Kato, T. 2015, PASJ, 67, 108Kato, T., et al. 2014, PASJ, 66, 90Kato, T., et al. 2015, PASJ, 67, 105Kato, T., et al. 2013a, PASJ, 65, 23Kato, T., et al. 2009, PASJ, 61, S395Kato, T., Monard, B., Hambsch, F.-J., Kiyota, S., & Maehara, H. 2013b,PASJ, 65, L11Kato, T., & Osaki, Y. 2013, PASJ, 65, 115Kato, T., Uemura, M., Ishioka, R., Nogami, D., Kunjaya, C., Baba, H., &Yamaoka, H. 2004, PASJ, 56, S1Murray, J., Warner, B., & Wickramasinghe, D. 2000, New Astron. Rev.,44, 51Osaki, Y. 1989, PASJ, 41, 1005Osaki, Y., & Kato, T. 2013a, PASJ, 65, 50Osaki, Y., & Kato, T. 2013b, PASJ, 65, 95Osaki, Y., & Meyer, F. 2003, A&A, 401, 325Pavlenko, E. P., et al. 2014, PASJ, 66, 111Pavlenko, E. P., et al. 2010, Astron. Rep., 54, 6Shappee, B. J., et al. 2014, ApJ, 788, 48Sheets, H. A., Thorstensen, J. R., Peters, C. J., Kapusta, A. B., & Taylor,C. J. 2007, PASP, 119, 494Smith, A. J., Haswell, C. A., Murray, J. R., Truss, M. R., & Foulkes, S. B.2007, MNRAS, 378, 785Warner, B. 1995a, Cataclysmic Variable Stars (Cambridge: CambridgeUniversity Press)Warner, B. 1995b, Ap&SS, 226, 187Whitehurst, R. 1988, MNRAS, 232, 35
Publications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0
V1006 Cyg
MN Dra
Fig. 2.
Comparison of O − C diagrams and variation of superhump amplitudes in V1006 Cyg and MN Dra. The data for MN Dra are taken from Kato et al.(2014). The behavior of O − C diagrams are almost the same in both systems. In MN Dra with denser observations in the early part, the slow growth of thesuperhump amplitudes for two nights were recorded. Based on the similarlity of the O − C diagrams, we consider that the superhump amplitudes also grewslowly in V1006 Cyg. Table 1.
Comparison of SU UMa-type objects with long phase of stage A superhumps
Object P orb ∗ P A † P B ‡ P C § dur k q ReferencesV1006 Cyg (2015) 0.09903(9) 0.1093(3) 0.10541(4) 0.10444(5) ≥
32 0.34(2) This workMN Dra (2012) 0.0998(2) 0.10993(9) 0.10530(6) – ≥
39 0.327(5) Kato et al. (2014)MN Dra (2013) 0.0998(2) 0.1082(1) 0.10504(7) – ≥
18 0.258(5) Kato et al. (2014)CRTS J214738.4 + ≥
21 0.204(11) Kato et al. (2015)OT J064833.4 + ≥
38 – Kato et al. (2015) ∗ Orbital period (d). † Period of stage A superhumps (d). ‡ Period of stage B superhumps (d). § Period of stage C superhumps (d). k Duration of stage A (cycles).
Determined from stage A superhumps. ublications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0 Table 2.
Superhump maxima of V1006 Cyg (2015) E max ∗ error O − C † N ‡ − − − − − − − − − − − ∗ BJD − † Against max = 2457219 . . . ‡ Number of points used to determine the maximum.
Publications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0
Table 3.
Log of observations of V1006 Cyg (2015)
Start ∗ End ∗ N † Code ‡ filter § Start ∗ End ∗ N † Code ‡ filter § ∗ JD − † Number of observations. ‡ Key to observers: COO (L. Cook), CRI (Crimean Astrophys. Obs.), DPV (P. Dubovsky), deM (E. deMiguel), IMi (I. Miller), Ioh (H. Itoh), KU (Kyoto U., campus obs.), Kaz (Kazan’ Univ. Obs.), LCO (C.Littlefield), Mdy (Y. Maeda), RPc (R. Pickard), Shu (S. Shugarov team), Trt (T. Tordai). §§