The 2010 nova outburst of the symbiotic Mira V407 Cyg
U. Munari, V.H. Joshi, N.M. Ashok, D.P.K. Banerjee, P. Valisa, A. Milani, A. Siviero, S. Dallaporta, F. Castellani
aa r X i v : . [ a s t r o - ph . S R ] N ov Mon. Not. R. Astron. Soc. , 1–7 (2010) Printed 20 September 2018 (MN L A TEX style file v2.2)
The 2010 nova outburst of the symbiotic Mira V407 Cyg ⋆ U. Munari , V.H. Joshi , N.M. Ashok , D.P.K. Banerjee , P. Valisa , A. Milani ,A. Siviero , S. Dallaporta , F. Castellani INAF Astronomical Observatory of Padova, 36012 Asiago (VI), Italy Astronomy and Astrophysics Division, Physical Research Laboratory, Navrangapura, Ahmedabad - 380009, Gujarat, India ANS Collaboration, c/o Osservatorio Astronomico, via dell’Osservatorio 8, 36012 Asiago (VI), Italy
Accepted .... Received ....; in original form ....
ABSTRACT
The nova outburst experienced in 2010 by the symbiotic binary Mira V407 Cyg hasbeen extensively studied at optical and infrared wavelengths with both photometricand spectroscopic observations. This outburst, reminiscent of similar events displayedby RS Oph, can be described as a very fast He/N nova erupting while being deeplyembedded in the dense wind of its cool giant companion. The hard radiation fromthe initial thermonuclear flash ionizes and excites the wind of the Mira over greatdistances (recombination is observed on a time scale of 4 days). The novae ejecta isfound to progressively decelerate with time as it expands into the Mira wind. This isdeduced from line widths which change from a FWHM of 2760 km s − on day +2.3to 200 km s − on day +196. The wind of the Mira is massive and extended enoughfor an outer neutral and unperturbed region to survive at all outburst phases. Key words:
Stars: novae – Stars: symbiotic stars – Miras
The symbiotic binary V407 Cyg consists of an accretingwhite dwarf and an O-rich Mira companion pulsating witha 745 day period. Miras with such a long pulsation periodare generally OH/IR sources with a very thick dust envelopewhich prevents direct observation of the central star at op-tical wavelengths. The much thinner dust envelope in V407Cyg is probably due to the presence of the WD companionwhose orbital motion, hard radiation field in quiescence andviolent mass ejection during outbursts inhibits dust forma-tion in a large fraction of the Mira wind (Munari et al. 1990,hereafter M90).V407 Cyg was discovered by Hoffmeister (1949) asNova Cyg 1936, just at the time when its Mira was pass-ing through maximum brightness. No spectroscopic obser-vations confirming it as a genuine nova outburst were how-ever available. What actually occured is unclear because (i)the object was discovered and remained at B ≈ B >
19 minimum (cf Figure 1 in M90), but at the sametime (ii) the peak brightness was much smaller than B ∼ ⋆ Tables 1-4 available in electronic form only active phases during the 1990’s were reported and discussedby Munari et al. (1994), Kolotilov et al. (1998, 2003, here-after K98 and K03), and some earlier ones can be spottedin the historical light-curves of V407 Cyg by Munari andJurdana-ˇSepi´c (2002) and M90.The 2010 outburst of V407 Cyg was discovered onMarch 10.813 UT by Nishiyama and Kabashima (2010) at V =7.6 mag. This was at an unsurpassed brightness level inthe star’s recorded photometric history thereby underscor-ing the peculiarity and importance of the event. The firstspectroscopic confirmation and analysis of the outburst wasgiven by Munari et al. (2010a) who described the event asa He/N nova expanding within the wind of the Mira com-panion. The similarity with RS Oph was also pointed out.In the following weeks and months the outburst was inten-sively monitored over several wavelength regimes viz. in γ -rays (Abdo et al. 2010, Cheung et al. 2010), radio (Krausset al. 2010, Giroletti et al. 2010, Bower et al. 2010, Nestoraset al. 2010, Gawronski et al. 2010, Pooley 2010), SiO maser(Deguchi et al. 2010), and infrared (Joshi et al. 2010).So far, apart from brief circulars, no comprehensive re-port on the photometric and spectroscopic evolution of V407Cyg at optical and IR wavelengths is available. The aim ofthis Letter is thus to provide a first report; follow-up paperswill present a more detailed analysis and modeling of thehuge amount of data we have and are still collecting. c (cid:13) U. Munari et al.
Figure 1.
Photometric evolution of V407 Cyg over the last 16 years, or 7 Mira’s pulsation cycles. The dashed line is hand-drawn toprovide a guide through the long-lasting active phase that peaked in 1998. The right panels provide a zoom over the 2010 outburst.
Optical photometry was recorded with several small tele-scopes operated by the ANS Collaboration in northern Italy,all equipped with CCDs and photometric
UBVR C I C filters.Corrections for bias, dark, and flat fields were applied in theusual manner. Photometric calibration and correction forcolor equations was performed for all instruments againstthe same UBVR C I C sequence calibrated by Henden and Mu-nari (2000) around V407 Cyg. Our photometry of V407 Cygcovering the 2010 outburst is presented in Table 1 and plot-ted in Figure 1 and 2.Optical spectroscopy was obtained with different tele-scopes: Asiago 1.82m + Echelle spectrograph (20000 resolv-ing power), Asiago 1.22m + B&C spectrograph (low reso-lution mode), Varese 0.6m + multi mode spectrograph. Ajournal of the observations is given in Table 2. With theVarese 0.6m telescope we obtained both low resolution andEchelle spectra. The latter were recorded both in unbinned(resolving power 17000, marked ech in Table 2) and binnedmode (resolving power 10000, marked echB in Table 2). Allspectra (including Echelle ones) were calibrated in absolutefluxes by observations of several spectrophotometric stan-dards during the night. Their zero-points were then checkedagainst simultaneous BVRI photometry by integrating theband transmission profiles on the fluxed spectra.Near-IR observations were carried out in the J, H, K bands at the Mt. Abu 1.2m telescope during the early out-burst phase. The spectra were obtained at a resolution of ∼ The observations presented in this paper show that the vio-lent outburst experienced by V407 Cyg in March 2010 wasa thermonuclear runaway (TNR), the same event that pow-ers a normal nova eruption. In normal novae, the ejectedmaterial essentially expands freely into a void circumstellarmedium. However, in V407 Cyg, the fast ejecta have to ex-pand into the dense and slow wind of the Mira companion,and are thus progressively slowed down as the pre-existingcircumstellar material is swept up in an expanding shell.Noteworthy, the pre-existing circumstellar material offers anideal ionization target for the hard radiation from the initialTNR flash.The similarity with the outburst displayed by the cele-brated RS Oph is evident (Bode 1987, Evans et al. 2008, andreferences therein). The latter is a symbiotic binary, with anorbital period of 460 days and an M giant filling its Rochelobe (Schaefer 2009) which transfers material to a massiveWD (Hachisu et al. 2007). Similar nova eruptions have beenseen also in the symbiotic binaries and recurrent novae TCrB, V745 Oph and V3890 Sgr (Schaefer 2010).What occured in V407 Cyg is well illustrated by theevolution of the H α profile (Figure 5) and its width and in-tegrated flux (top panels of Figure 2). At the earliest stages,the H α profile is dominated by a sharp component super-posed on a much broader one, as first noted by Munari etal. (2010a).The sharp component, identical to that in quiescencebut enormously brighter (cf profiles for 2008 and 2009 inFigure 5), is due to the sudden ionization of a large fraction c (cid:13) , 1–7 he 2010 outburst of V407 Cyg Figure 2.
Lower panels : early optical and infrared photometricevolution of V407 Cyg during the 2010 outburst. The dashed lineis hand-drawn for guidance.
Upper panels : evolution in width andintegrated flux of the H α emission line. ‘Nova ejecta’ refer to thebroad component (cf. sect. 3), ‘Mira wind’ to the superimposednarrow one (see Figure 5). of the Mira’s wind by the flash of energetic radiation pro-duced by the TNR event. The wind of the Mira does notas yet get perturbed kinematically, as proven by the pre-served sharpness of the H α profile that increased its emissiv-ity by two orders of magnitude compared to quiescence. Theflux of hard photons, however, is not large enough to ion-ize the whole Mira wind, as indicated by the persistence ofthe sharp absorption component which maintains the sameheliocentric radial velocity as in quiescence ( −
50 km/sec). The intensity of the H α sharp component rapidly declinessubsequently (cf Figure 2), with a recombination time scaleof 4 days, which can be written as t rec = 0 . (cid:16) T e K (cid:17) . (cid:16) n e cm − (cid:17) − ≈
100 hours (1)following Ferland (2003). It corresponds to a density ofabout 5 × cm − for the fraction of the Mira wind ion-ized by the TNR initial flash. The point at day +12.3 inFigure 2, e.g. the last epoch at which a narrow componentcould still be resolved in the H α profiles of Figure 5, deviatesfrom the t rec =4 days of earlier points. By this time, the novaejecta has begun to turn optically thin and the hard radia-tion field of the central star (presumably still burning hydro-gen at its surface during the constant luminosity phase) ishot and intense enough to produce coronal emission lines, asreported by Munari et al. (2010b). The same radiation field,leaking through the optically thin ejecta, is also responsiblefor sustaining the ionization of the circumstellar gas not yetreached by the expanding shell.The broad component of the V407 Cyg H α profiles inFigure 5, originates instead in the material ejected at highvelocity, as in any normal nova. The broad spectrum nicelymatches that of a normal ”He/N” nova (Williams 1992) asillustrated by the low resolution optical and infrared spec-tra for days +2.3 and +4.2 in Figure 3 and 4 respectively.A He/N spectrum is typical of fast novae and of RS Ophtoo. The nova ejecta is rapidly decelerated while trying toexpand through the surrounding Mira wind and the distinc-tion between a sharp and a broad component to the emis-sion lines is then progressively attenuated, disappearing twoweeks past optical maximum. As more material is swept bythe expanding shell, the velocity continues to decrease. Fig-ure 2 illustrates the temporal evolution of the FWHM (in kmsec − ) of the broad component of H α , which is accuratelyfitted by the expressionFWHM = 4320 − t + 2635(log t ) − t ) (2)including later phases characterized by 400, 280 and 200km s − on days +48.2, +105 and +196 respectively. Thesame trend is shared also by the hydrogen lines dominatingthe infrared spectra of Figure 4. For comparison the FWHMof H α in quiescence was stable at ∼
120 km s − (cf profilesfor 2008 and 2009 in Figure 5). Figure 5 shows the emer-gence of [NII] 6548, 6584 ˚A doublet two months past opticalmaximum. It did not originate in the expanding material,but instead in the outer wind of the Mira, external to theexpanding shell. This is proved since its profile FWHM of ∼
110 km s − is much sharper than that of the adjacent H α and identical to the width in quiescence. The existence of anouter region of the Mira wind not yet reached on day +196by the already greatly slowed down ejecta (cf the sharp ab-sorption component at −
50 km s − in Figure 5), leads us tospeculate that some part of the ejecta could remain boundto the binary system and could be re-accreted at later timesby the WD.The light-curve of V407 Cyg over the last 15 years ispresented in Figure 1. It is characterized by three main com-ponents: (1) the 745 day pulsation of the Mira (sinusoiddrawn as a solid line), which dominates the light-curve atreddest wavelengths; (2) the presence of a limited ampli-tude, slow evolution active phase (dashed line in the V band c (cid:13) , 1–7 U. Munari et al.
Figure 3.
A sample of the absolutely fluxed, optical spectra of V407 Cyg that we collected during the 2010 outburst. The 10 December2007 spectrum shows the quiescence spectrum at the time of Mira brightness maximum. Only some of the emission lines are identified.
Figure 4.
A sample of our absolutely fluxed, infrared spectra of V407 Cyg for the 2010 outburst. Major emission lines are identified. panel) that peaked in intensity in 1998/99 (described in de-tail by K98 and K03) when it rivalled in V the brightnessof the Mira but went unnoticeable in I C . This correspondsto the typical, non-TNR outbursts that essentially all sym-biotic stars have experienced several times in their recordedphotometric history; and (3) the violent, rapid and brightTNR outburst of 2010. The latter overwhelmed the bright-ness of the Mira at optical wavelengths, but only equalledit in the K band (cf. data in Table 4 with the long term JHK light-curve of the Mira presented by K98 and K03).Figure 2 presents a zoomed view on the earliest evolution ofthe 2010 outburst in the
BV R C I C JHK bands. The opticalmaximum was reached at V =7.1 on March 10.8 UT, andthe subsequent decline was very fast and characterized by t V =5.9 and t V =24 days. The decline was similarly fast inT CrB, V745 Sco, RS Oph and V3890 Sgr that showed t =4, 5, 7 and 9 days, respectively. The V -band light-curve inFigure 1 shows a distinct knee at day +47. By analogy with RS Oph (cf. Hachisu et al. 2006), it could mark the end ofthe stable H-burning on the WD.The outburst evolution seen in the I C panel in Figure 1could appear in conflict with the expected underlying pulsa-tion cycle of the Mira. Indeed, the pulsation of the latter isknown to be highly variable from cycle to cycle (K98, Munariand Jurdana-Sepic 2002), with puzzling sharp minima oc-curring at various pulsation phases (Kiziloglu and Kiziloglu2010; some of them are visible also in the light-curve of Fig-ure 1 in 2007 and 2009), and that could be related to theunusual nature of the Mira in V407 Cyg. In fact, Miras ofsuch a long pulsation period are usually the central stars ofOH/IR sources and their thick dust cocoon prevent themfrom being visible in the optical. As remarked by M90, thepresence of the hot and outbursting WD companion, coulddisturb the formation of the dust cocoon and thus make theMira in V407 Cyg visible at optical wavelengths. c (cid:13) , 1–7 he 2010 outburst of V407 Cyg Figure 5.
Evolution of the H α profile of V407 Cyg during the 2010 outburst. Older profiles are given for reference purposes, and pertainto the brightness peak during the 1997-2006 active phase (1999), and to quiescence at the time of a Mira minimum (2008) and maximum(2009). The three spectra marked with ∗ are courtesy of Christian Buil and fluxed by us against our low-resolution spectra. REFERENCES
Abdo A. A., et al., 2010, Sci, 329, 817Bode M. F., 1987, RS Ophiuchi (1985) and the recurrentnova phenomenon, Utrecht: VNU Science PressBower G.C., Forster R., Cheung C.C., 2010, ATel, 2529, 1Cheung C. C., Donato D., Wallace E., Corbet R., DubusG., Sokolovsky K., Takahashi H., 2010, ATel, 2487, 1Deguchi S., Matsunaga N., Takahashi S., Kuno N.,Nakashima J., 2010, ATel, 2519, 1Evans A., Bode M. F., O’Brien T. J., Darnley M. J., 2008,RS Oph (2006) and the Recurrent Nova Phenomenon, ASPConf. Ser. vol 401, ASP San FranciscoFerland G. J., 2003, ARA&A, 41, 517Gawronski M. P., et al., 2010, ATel, 2511, 1Giroletti, K.E., Corbel S., Sokolovsky K., Fuhrmann L.,Schinzel F., Cheung C.C., 2010, ATel, 2536, 1Hachisu I., et al., 2006, ApJ, 651, L141Hachisu I., Kato M., Luna G. J. M., 2007, ApJ, 659, L153Henden A., Munari U., 2000, A&AS, 143, 343Henden A., Munari U., 2008, BaltA, 17, 293Hoffmeister C., 1949, QB835,Joshi, V., Ashok, N.M., Banerjee, D.P.K. 2010, CBET,2210, 1Jurdana-ˇSepi´c R., Munari U., 2010, PASP, 122, 35 Kiziloglu U., Kiziloglu N., 2010, IBVS, 5947, 1Kolotilov E. A., Munari U., Popova A. A., TatarnikovA. M., Shenavrin V. I., Yudin B. F., 1998, AstL, 24, 451Kolotilov E. A., Shenavrin V. I., Shugarov S. Y., YudinB. F., 2003, ARep, 47, 777Krauss M., et al., 2010, ATel, 2741, 1Meduza 2010, http://var2.astro.cz/EN/Munari U., Margoni R., Stagni R.,1990, MNRAS, 242, 653Munari U., Bragaglia A., Guarnieri M. D., Sostero G., Lep-ardo A., Yudin B. F., 1994, IAUC, 6049, 2Munari U., Jurdana-ˇSepi´c R., 2002, A&A, 386, 237Munari U., Siviero A., Valisa P., Dallaporta S., 2010a,CBET, 2204, 3Munari U., Siviero A., Valisa P., 2010b, ATel, 2546, 1Naik S., Banerjee D. P. K., Ashok N. M., 2009, MNRAS,394, 1551Nestoras I., et al., 2010, ATel, 2506, 1Nishiyama, K., Kabashima, F., 2010, CBET 2199Pooley G., 2010, ATel, 2514, 1Schaefer B. E., 2009, ApJ, 697, 721Schaefer B. E., 2010, ApJS, 187, 275Williams R. E., 1992, AJ, 104, 725This paper has been typeset from a TEX/ L A TEX file preparedby the author. c (cid:13) , 1–7 U. Munari et al.
Table 1.
Our optical photometry of V407 Cyg during the 2010outburst. JD ⊙ V ± ǫ B − V ± ǫ V − R C ± ǫ V − I C ± ǫ Table 2.
Journal of our optical spectroscopic observations ofV407 Cyg during the 2010 outburst and a few earlier epochs. date UT ∆ t expt disp λ start λ end tel.(day) (sec) (˚A/pix) (˚A) (˚A)1999-06-09 02:19 600 ech ech ech echB ech ech ech ech ech ech ech ech ech ech echB echB ech Table 3.
Journal of our infrared spectroscopic observations ofV407 Cyg during the 2010 outburst. The integration times inbrackets are for the J, H and K bands respectively date UT ∆t expt disp λ start λ end tel.(day) (sec) ( µ m/pix) ( µ m) ( µ m)2010 03 14 0:06 3.2 (15,5,10) 0.00096 1.080 2.400 1.2m2010 03 15 0:29 4.2 (10,5,7) 0.00096 1.080 2.400 1.2m2010 03 16 0:24 5.2 (10,5,20) 0.00096 1.080 2.400 1.2m2010 03 16 23:55 6.2 (10,5,6) 0.00096 1.080 2.400 1.2m2010 03 22 23:36 12.2 (15,10,20) 0.00096 1.080 2.400 1.2m2010 03 23 23:35 13.2 (15,20,15) 0.00096 1.080 2.400 1.2m2010 03 28 23:24 18.2 (10,10,15) 0.00096 1.080 2.400 1.2m2010 03 29 23:35 19.2 (10,7,20) 0.00096 1.080 2.400 1.2m2010 03 30 23:26 20.2 (40,30,20) 0.00096 1.080 2.400 1.2m2010 03 31 23:35 21.2 (20,20,15) 0.00096 1.080 2.400 1.2m2010 04 08 22:29 29.1 (30,10,40) 0.00096 1.080 2.400 1.2m2010 04 10 0:00 30.2 (15,10,10) 0.00096 1.080 2.400 1.2m2010 04 30 23:26 51.2 (25,10,30) 0.00096 1.080 2.400 1.2m2010 05 03 22:08 54.1 (90,20,40) 0.00096 1.080 2.400 1.2m c (cid:13) , 1–7 he 2010 outburst of V407 Cyg Table 4.
Our
JHK photometry of V407 Cyg during the 2010outburst.
JD date
J ǫ J H ǫ H K ǫ K c (cid:13)000