Photometric Studies of New Southern SU UMa-type dwarf novae, FL Triangulum Australe and CTCV J0549-4921
Akira Imada, Taichi Kato, L.A.G. Monard, Rod Stubbings, Makoto Uemura, Ryoko Ishioka, Daisaku Nogami
aa r X i v : . [ a s t r o - ph ] A p r PASJ:
Publ. Astron. Soc. Japan , 1– ?? , c (cid:13) Photometric Studies of New Southern SU UMa-type dwarf novae, FLTriangulum Australe and CTCV J0549-4921
Akira Imada , Taichi Kato , L.A.G. Monard , Rod Stubbings ,Makoto Uemura , Ryoko Ishioka , and Daisaku Nogami Department of Astronomy,Faculty of Science, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japana [email protected] Bronberg Observatory, CBA Pretoria, PO Box 11426, Tiegerpoort 0056, South Africa Tetoora Observatory, Tetoora Road, Victoria, Australia Hiroshima Astrophysical Science Center, Hiroshima University, Hiroshima 739-8526, Japan Subaru Telescope, National Astronomical Observatory of Japan 650 North A’ohoku Place, Hilo,HI 96720, U.S.A. Hida Observatory, Kyoto University, Kamitakara, Gifu 506-1314, Japan (Received ; accepted )
Abstract
We report time-resolved optical CCD photometry on newly discovered SU UMa-type dwarf novae, FLTrA and CTCV J0549-4921. During the 2006 August outburst, we detected superhumps with a periodof 0.59897(11) days for FL TrA, clarifying the SU UMa nature of the system. On the first night of ourobservations on FL TrA, the object showed no superhumps. This implies that it takes a few days forfull development of superhumps. The superhump period variation diagram of FL TrA was similar to thatobserved in some WZ Sge stars and short period SU UMa-type stars. This indicates that the system isclosely related to WZ Sge stars and SU UMa stars having short orbital periods. For CTCV J0549-4921,the candidates of the mean superhump period are 0.083249(10) days and 0.084257(8) days, respectively.Due to a lack of the observations, we cannot determine the true superhump period, but the latter periodis favorable. Using the ASAS-3 archive, it turned out that the system shows only four outbursts over thepast 6 years. The outburst amplitude of CTCV J0549-4921 was relatively small, with about 4.5 mag. Onepossibility is that mass evaporation may play a role during quiescence.
Key words: accretion: accretion discs — stars: cataclysmic — stars: dwarf novae — stars: individual(FL Triangulum Australe, CTCV J0549-4921) — stars: novae, cataclysmic variables — stars: oscillations
1. Introduction
Dwarf novae are a subclass of cataclysmic variables thatconsist of a white dwarf (primary) and a late-type star(secondary). The secondary star fills its Roche lobe, trans-ferring the matter into the primary via inner Lagragianpoint (L1). Then the accretion disc is formed aroundthe white dwarf. The accretion disc shows various mod-ulations both in outburst and quiescence (for a review,see Warner 1995; Hellier 2001; Lasota 2001; ConnonSmith 2007).SU UMa-type stars are a subclass of dwarf no-vae (Osaki 1989; Osaki 1996; Patterson et al. 2005;Osaki (2005)). The systems basically exhibit two typesof eruptions: normal outburst which lasts a few days andsuperoutburst which lasts about two weeks. During thesuperoutburst, modulations having an amplitude of ∼ Table 1.
Observation log of FL TrA during the 2005 Augustsuperoutburst. a End a N b Jul. 27 579.2491 579.3572 225Jul. 28 580.2378 580.5585 451Jul. 29 581.2032 581.5047 426Jul. 30 582.3141 582.5241 297Jul. 31 583.2259 583.4713 347Aug. 1. 584.1918 584.4763 360Aug. 2. 585.2243 585.3945 241 a HJD - 2453000. b Number of exposure.
2. FL TrA
FL TrA was first cataloged in Meinunger (1970) inwhich the system was numbered S 5770 TrA with thevariable type of UG. Downes, Shara (1993) tabulated cat-aclysmic variable stars, including FL TrA, in which thevariable was categorised as UG with the magnitude rangeof 15.5p − h m s , Dec: − ◦ ′ ′′ . No outburstof FL TrA was reported to the VSNET (Kato et al. 2004)until 2005. We have suspected the WZ Sge subclass ofthe object. Spectroscopic observations were carried outby Mason, Howell (2003) in which the object showed aspectrum of a common G-type star. Mason, Howell (2003)pointed out the misidentification of FL TrA.On 2005 July 27, Rod Stubbings reported to theVSNET that FL TrA appeared to be in outburst witha visual magnitude of 15.0 ([vsnet-alert 8574]). He fur-ther noticed that the position of the system looked slightlynorth from the above mentioned coordinate. It turned outthat FL TrA was misidentified as USNOB1 0281-0691553(RA:16 h m s .
4, Dec: − ◦ ′ ′′ . The precise coordinate of the system is RA:16 h m s . − ◦ ′ ′′ .
0, where no optical counterpart exists inthe USNO B1 catalog, which indicates the magnitude inquiescence may be fainter than 21 mag.
Time resolved CCD photometric observations were car-ried out from 2005 July 27 to 2005 August 2 at BronbergObservatory in South Africa using a 32 cm Schmidt-Cassegrain telescope equipped with a SBIG ST-7XMECCD camera. We tabulate journal of observations in ta-ble 1. All of the observations were performed with 30sec exposure time. The total data points amounted to2347. No filter was used during the observations. Theunfiltered data are close to the R c system. After debi-asing and flat-fielding, we performed aperture photom-etry using AIP4WIN software. As a comparison star,we used USNO A2.0 0225 − h m s . − ◦ ′ ′′ . B =14.2, R = 13.2), whose constancy h http://archive.stsci.edu/prepds/cvcat/index.html i M agn i t ude HJD-2400000
Fig. 1.
Light curves of FL TrA during the 2005 July/Augustsuperoutburst. The vertical and horizontal axis indicate dif-ferential magnitude and the fractional HJD, respectively. Themagnitude of the comparison star is 13.2 in R . The starshowed almost the constant decline from HJD 2453581 (2005July 28) at a rate of 0.13 mag d − . R e l a t i v e m ag . HJD 2453579+
Fig. 2.
Enlarged light curve on HJD 2453579 (2005 July 27),the first night of our run. The light curves provide no evidenceof superhumps during this phase. was checked by some stars located in the same image. The1-sigma error for each differential magnitude is of an or-der of 0.03 mag, which is small enough to perform thefollowing analysis, including exploring superhump periodand profile variations. Heliocentric corrections to our runwere applied before the following analysis.
Figure 1 shows light curves of FL TrA during the 2005July/August superoutburst. At the onset of our obser-vations, FL TrA was at the magnitude of 15.0 on 2005July 27, after which the system almost constantly de-clined at the rate of 0.13(1) mag d − . This decline rateis a typical value among SU UMa-type dwarf novae. Ascan be seen in figure 2, the light curve showed almostno feature on 2005 July 27 (HJD 2453579), indicatingthat superhumps did not yet develop. After subtracting alinear decline trend of daily light curves, we performeda period analysis of the phase dispersion minimization(PDM) method (Stellingwerf 1978) applied between HJD2453580 and 2453285. Figure 3 displays the results ofthe PDM analysis, by which we determined 0.059897(11)days as the best estimated period during this stage. Ao. ] dwarf nova FL TrA and CTCV J0549-4921 3 t he t a period Fig. 3.
Theta diagram of the superoutburst of FL TrA fromHJD 2453580 to HJD 2453585. The arrow shows the best es-timated superhump period, 0.059897(11) days with 99% sig-nificance level. statistical F-test provided the confidence level of 99 %.The 1-sigma error was calculated using the Lafler-Kinmanmethod (Fernie 1989).We present daily averaged light curves in figure 4. Theselight curves are folded with the above obtained period. OnHJD 2453580 (2005 July 28), the profile is characteristic ofsuperhumps, with the mean amplitude of about 0.2 mag,from which we first confirmed the SU UMa nature of FLTrA. On 2005 July 29, the amplitude of the superhumpswas at the maximum value of about 0.3 mag. No eclipsefeature was detected during the observations, indicating alow-to-mid inclination of FL TrA.In order to investigate the variations of the superhumpperiod during the plateau phase, we measured the timingsof the superhump maxima listed in table 2. The typicalerror is an order of 0.002 days. A linear fitting yielded asthe following equation,
HJD ( max ) = 2453580 . . × E, (1)where the parentheses denote 1-sigma error for eachvalue. By using the above ephemeris, we draw an O − C diagram, which is displayed in figure 5. The best fittedquadratic for 16 < E <
84 can be represented as follows: O − C =5 . . × − − . . × − E + 2 . . × − E . (2)The above obtained value implies that the superhumpperiod may increase since HJD 2453581 with P dot = ˙ P / P = +8.4(5.0) × − . The present photometric studies and the previousarchival survey reasonably qualified FL TrA as a newmember of SU UMa-type dwarf novae with short pe-riods. The outburst amplitude of FL TrA exceeded 6mag, suggestive of a large amplitude SU UMa-type dwarfnovae (TOAD, Howell et al. 1995). Unfortunately, the -0.2 0 0.2 0.4 0.6 0.8 1 1.2 -0.5 0 0.5 1 1.5 2 2.5 R e l a t i v e m ag . Phase July 28July 29July 30July 31Aug. 1Aug. 2
Fig. 4.
Phase averaged light curve during the superoutburstfolded by 0.059897 days. The abscissa and ordinate denotephase and differential magnitude, respectively. A rapid riseand slow decline, characteristic of superhumps, are visible. -0.01-0.005 0 0.005 0.01 0 30 60 90 580 582 584 586 O - C ( d ) Cycle countHJD-2453000
Fig. 5. O − C diagram of FL TrA. Each datapoint of themaximum timing of superhumps is listed in table 2. The solidcurve means the best fitting quadratic described in equation(2) for 16 < E <
84. It should be noted that the cycle countbetween 0 < E < A.Imada et al. [Vol. ,
Table 2. superhump timing maxima E a Time b error c a Cycle count. b HJD - 2453000. c error in a unit of days.lack of baseline during the early stage of the superout-burst prevented us from further investigating whetherdouble-peaked humps existed, which is exclusively ob-served among WZ Sge stars in early phase of super-outburst (Osaki, Meyer 2002; Kato 2002a; Pattersonet al. 2002).As for superhump period changes, the estimated pos-itive P dot derivative indicates that the superhump pe-riod increases during the plateau stage. Such systemsinclude all of confirmed WZ Sge-type dwarf novae (Katoet al. 2001), as well as SU UMa-type dwarf novae withshort superhump periods (Oizumi et al. 2007). Recently,Uemura et al. (2005) found that a short period SU UMastar TV Crv shows two types of P dot . The 2001 super-outburst of TV Crv showed positive P dot , while the 2004superoutburst showed almost constant P sh . The big dif-ference between the two superoutbursts is not only thedifferent P dot , but also the light curves themselves: a pre-cursor was present for the 2004 superoutburst while it wasabsent for the 2001 superoutburst. One interpretation isthat an appearance of the positive or constant/negative P dot depends on the maximum radius of the accretion discduring the superoutburst (Uemura et al. 2005). Uemuraet al. (2005) have further stated that systems which showboth types of P dot will be restricted to short period SUUMa stars, because the tidal truncation radius shouldbe significantly larger than the 3:1 resonance radius (seeOsaki, Meyer 2003). With this respect, the 2005 superout-burst of FL TrA had a large disc radius, which is consistentwith the large amplitude of the outburst.Another important finding is that unfittable cycle counts by the equation (2) exists at the earliest stage ofour run. These correspond to 0 < E < < E <
4, thesuperhump period keeps constant, while an abrupt changeof the superhump period occurred after
E >
4. The ori-gin of the abrupt period change remained unknown, whichshould be elucidated in the future observations.
3. CTCV J0549-4921
CTCV J0549-4921 (hereafter CTCV J0549) was firstidentified as a candidate of cataclysmic variables afterspectroscopic observations in the Calan-Tololo Survey(Maza et al. 1989). The optical spectrum shows H α andHeI 5876 emission (Tappert et al. 2004), indicating thedwarf nova nature of the system. Tappert et al. (2004)pointed out there is no evidence of the secondary star inthe optical spectrum because of the absence of TiO bands.Optical observations during quiescence and outburst werealso performed by Tappert et al. (2004). During the qui-escence, Tappert et al. (2004) found photometric orbitalmodulations with the period of 0.080218(70) days, whichthey interpreted as the orbital period of CTCV J0549.Tappert et al. (2004) pointed out the shape of the mod-ulation is reminiscent of quiescent light curve of WZ Sge.During the outburst, CTCV J0549 brightened up to V = 13.75 on 1996 October 5. However, superhumps werenot detected. It is likely Tappert et al. (2004) observeda normal outburst of SU UMa-type dwarf novae. In con-junction with the above observations, CTCV J0549 hasbeen a promising candidate for SU UMa-type dwarf no-vae.On 2006 April 2, a brightening of the star was discov-ered by L.A.G. Monard ([vsnet-alert 8896]) at the mag-nitude of 13.8, who detected the rising phase of out-burst. On 2006 April 4, we first detected superhumps ofCTCV J0549, and confirmed the SU UMa nature of theobject. Long-term monitoring by the ASAS-3 have de-tected 3 outbursts, of which one was possibly a superout-burst. This occurred in 2004 January. The precise coor-dinate of the system is RA:05 h m s .
4, Dec: − ◦ ′ ′′ ,where the 2MASS counterpart of CTCV J0549 yields J = 15.619(50), H = 15.210(81), and K = 14.869(112), re-spectively (Imada et al. 2006b). Time resolved CCD photometric observations were car-ried out from 2006 April 2 to 2006 April 12. The ob-serving site and instrument are the same as described insection 2.2. The journal of observations is summarized intable 3. All of the observations were performed with 30-sec exposure time with no filter. The total data pointsof our run amounted to 1979. For obtained data, weperformed the same manner as mentioned in section 2.2.We used USNO A2.0 0375 − h m s . Table 3.
Observation log of CTCV J0549 during the 2006April superoutburst. a End a N b Apr. 2 828.2200 828.3839 452Apr. 4 830.2275 830.3658 286Apr. 5 831.1955 831.3463 428Apr. 6 832.1991 832.3524 428Apr. 12 838.1970 838.3339 385 a HJD - 2453000. b Number of exposure. M agn i t ude HJD-2400000
Fig. 6.
Light curves of CTCV J0549 during the 2006 Aprilsuperoutburst. The ordinate means the ASAS-3 V and R magnitude. The magnitude of the comparison star is 12.7 in R . The filled circles show time resolved CCD observations.The crosses indicate the ASAS-3 light curves, which contains0.2 mag error originated from the modulation of superhumps.The negative observation was performed by the ASAS-3 onHJD 2453849, when the object was fainter than 14.4 in V . Dec: − ◦ ′ ′′ . B = 13.3, R = 12.8) as a comparisonstar, whose constancy was checked by some stars in thesame image. The 1-sigma error for each differential mag-nitude is of an order of 0.01 mag. Heliocentric correctionsto our run were applied before the following analyses. The overall light curves during our run are presented infigure 6, in which we also demonstrate the ASAS-3 posi-tive and negative observations. The discrepancy betweenour CCD observations and the ASAS-3 archive is large,presumably due to different filters between the site andthe “snapshot” in the ASAS-3 photometry. Nevertheless,it is well determined that the bright maximum of CTCVJ0549 was on HJD 2453831 with the magnitude of ∼ − − as the ris-ing rate. Although our observations were absent betweenHJD 2453833-2453837, during which the magnitude wasassumed to decline linearly, the decline rate could be esti-mated to be 0.12(1) mag − . The value is typical for usualSU UMa-type dwarf novae during the plateau phase.Enlarged light curves for each night are depicted in fig-ure 7, after subtracting linear rising or declining trend. -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0 0.05 0.1 0.15 0.2 D i ff. m ag . fractional HJD April 2April 4April 5April 6April 12 Fig. 7.
Daily light curves after removing the linear trend.There were almost no modulation on April 2. The growthtime of superhumps is as short as 2 days. The amplitude ofsuperhumps is the largest on April 5 with 0.4 mag. Thereis no evidence for an eclipse, which indicates of low to midinclination system.
Table 4. superhump timing maxima E a Time b error c a Cycle count. b HJD - 2453000. c error in unit of days.As can be seen in this figure, there are no features onHJD 2453829 (2006 April 2), corresponding to the ris-ing phase, while prominent superhumps are shown fromHJD 2453831 (2006 April 4). Therefore, we first confirmedCTCV J0549 as an SU UMa star.In order to determine the mean superhump period dur-ing the plateau phase, we performed the PDM method(Stellingwerf 1978). The strongest periodicity can befound at 0.084257(8) days. However, due to the lack ofour observations, we cannot rule out the second strongest A.Imada et al. [Vol. , t he t a period (d) Fig. 8.
Theta diagram of CTCV J0549 applied to the plateauphase. Two possible periods of superhumps were found,0.083249(10) days and 0.084257(8) days. Due to the lack ofthe observations, we cannot specify the exact period of super-humps. period, 0.083249(10) days. Hence, we carried out anotherapproach to determine the mean superhump period bymeasuring the maximum timing of the superhumps. Wetabulate the result on table 4. The best fitting linear re-gression is yielded in the following equation:
HJD ( max ) = 2453830 . . × E. (3)The above equation favors the former period of the su-perhump, 0.084257(8) days. If the quiescent modulationsreflect the orbital period of the system, and the mean su-perhump period is P sh =0.084257(8) days, then the frac-tional superhump period excess is ∼ P orb and P sh should be measured in the future observations. We first confirmed the SU UMa nature of CTCV J0549by the detection of superhumps. Although the mean su-perhump period cannot be determined, the period exceeds0.08 days, which we safely qualify CTCV J0549 as a longperiod SU UMa star. This is also supported by quiescentphotometric observations (Tappert et al. 2004).The most remarkable fact for CTCV J0549 is thatthe object has shown only 4 outbursts over the past 6years. According to the ASAS-3 archive, the recordedoutbursts were 2001 Febuary 12, 2001 September 20, 2004January 21, and the present superoutburst. We summa-rize recorded outbursts monitored by the ASAS-3 in table5. Judging from table 5, only two are superoutbursts, oneis a normal outburst, and we cannot distinguish the typefor one outburst. If we do not miss any superoutburstsince 2001, a supercycle of CTCV J0549 is estimated as ∼
800 days. This is one of the longset values among SUUMa-type dwarf novae (Kato et al. 2001). Inactive sys-tems having a similar superhump period include QY Per( P sh = 0.07681 days, Kato et al. 2000), EF Peg ( P sh =0.08705 days, Kato 2002b) and V725 Aql ( P sh = 0.09909 Table 5.
Recorded outbursts by the ASAS-3.
Time a Mag. b error c type d < < < < < < < < a HJD - 2400000. b Mean magnitude in V . c V . d N: Normal outburst. S: Superoutburst.days, (Uemura et al. 2001)). Although the exact mecha-nism of the long supercycle still remains unknown, massevaporation during quiescence might be a possible expla-nation for the origin of the outburst and quiescent prop-erties (Meyer, Meyer-Hofmeister 1994; Lasota et al. 1995;Mineshige et al. 1998). This may be consistent with rel-atively small amplitude of 4.5 mag of CTCV J0549. Asfor evporation, Mineshige et al. (1998) predicted that qui-escent superhumps could be observed even during quies-cence if evaporation works in the acretion disc. Oizumiet al. (2007) also argued that a peak separation variationof an optical spectrum during quiescence is a powerfultool to check whether or not the evaporation works inthe accretion disc. Future spectroscopic observations arerequired to elucidate the nature of CTCV J0549.
4. Summary