COMMISSIONS 27 AND 42 OF THE IAU INFORMATION BULLETIN ON VARIABLE STARS Number 6xxx

Transcription

1 COMMISSIONS 27 AND 42 OF THE IAU INFORMATION BULLETIN ON VARIABLE STARS Number 6xxx Konkoly Observatory Budapest 2 December 2016 HU ISSN CORRELATION OF Hα EMISSION FLUX WITH B & V MAGNITUDES IN THE ECLIPSING BINARY VV Cep POLLMANN, E. 1 ; VOLLMANN, W. 2 ; BENNETT, P.D. 3 1 Emil-Nolde Straße 12, Leverkusen, Germany 2 Dammäckergasse 28/D1/20, A-1210 Wien, Austria 3 Department of Physics & Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, Canada Abstract VV Cephei (= HR 8383 = HD ) is the brightest eclipsing M supergiant binary (M2 Iab + B0-2? V) in the sky, and is a massive binary with one of the longest known orbital periods (7430 days = years) of any eclipsing system. As such, this binary provides a unique opportunity to study the chromosphere, outer atmosphere and stellar wind of a red supergiant star that is spectroscopically similar to the well-studied, bright star Betelgeuse (α Ori). Near eclipse, VV Cep offers a rare chance to observe lines of sight to the hot companion near eclipse that probe deep into the extended outer atmosphere of the M supergiant star. With the next eclipse beginning in August 2017, and lasting nearly two years ( 650 days) from 1st to 4th contact, an extensive observational campaign is planned. In this paper, we report the discovery of a correlation between the visual brightness and the Hα emission flux in VV Cep out of eclipse. Introduction The VV Cep binary is a red supergiant of mass M, with a hot, presumably main-sequence, early B-type companion of comparable mass. The two stars are sufficiently well separated that Roche lobe mass transfer does not occur at present, and given the high orbital eccentricity (e = 0.346, Wright 1977), probably has not occurred over the evolutionary history of the system. (Orbits of binaries undergoing Roche lobe mass transfer circularize on short timescales). In the ultraviolet (UV), the hot companion appears embedded in a region of circumstellar gas, as inferred from the persistent veiling that obscures the UV photospheric spectrum of the B-star companion. The gas around the companion is probably a result of wind accretion from the massive wind of the M supergiant. In VV Cep, Hα emission is especially prominent, with peak fluxes of times that of the (M star) continuum. This Hα emission exhibits radial velocity behaviour opposite to that of the M supergiant, implying a source near the hot companion (Wright 1977). Even though this emission declines sharply for higher Balmer lines, at times Balmer emission remains visible from levels up to n 16. Balmer continuum emission is often observed at wavelengths shortward of 3700 Å, and sometimes dominates this part of the UV spectrum (Bennett & Bauer 2015). In the UV, lines of Fe II also appear strongly in emission, and are probably pumped by Lyman-α and Lyman-β emission (Bennett & Bauer 2015). The great width of these Fe II emission lines (with wings out to 300 km s 1 ) suggests the lineforming region is in Keplerian rotation around the B star companion, as these velocities are far larger than any other observed in the circumstellar environment of VV Cep. Although the source of the companion s emission spectrum is usually attributed to accretion of circumstellar gas from the M star onto the hot star, it is likely that the emission luminosity comes not from the release of gravitational energy, but from recombination of circumstellar hydrogen photoionized by the B star s Lyman continuum. The Hα emission is variable on both short timescales of 150 days and longer timescales of several years. The slow variability in Hα flux appears to correlate with the orbital separation of the two stars, being larger when the companion is near periastron (Bennett, private communication). However, the cause of the fast, short-term variability was unknown until now.

2 Observations and Results VV Cep is a 5th magnitude system of variable visual brightness, with V magnitudes ranging from Due to its high declination (+64 ), VV Cep is circumpolar and well-suited for year-round observations at northern mid-latitude sites. In preparation for the VV Cep international campaign, contemporaneous observations of B and V band DSLR photometry and Hα emission equivalent width (EW) have been obtained over the past three years. A time series of V photometry from W. Vollmann is shown in Figure 1, and Hα EWs from the ARAS spectroscopy group ( are presented in Figure 2. Figure 1. DSLR V magnitude of VV Cep, observed by W. Vollmann Figure 2. Hα EW of VV Cep, observed by the ARAS Spectroscopy group 2

3 Figure 3 shows the correlation between the Hα emission EW and contemporaneous V photometry of VV Cep for the period January 2014 to October 2016 (JD { ) obtained by W. Vollmann (DSLR, AAVSO & BAV-Germany), B. Hassforther (DSLR, BAV-Germany) and G. Samolyk (CCD, AAVSO data base). We nd the Hα EW is correlated with V magnitude, with a correlation coefficient of R = 0.84 over this period. Although this relationship is reminiscent of the well-known relationship between stellar brightness and Be star emission fluxes (e.g., Harmanec 1983), the cause must be different for VV Cep because most of the visible continuum flux comes from the M supergiant (V 5.0), and not the 7th magnitude B-type companion responsible for the Hα emission. Note that the V band data also includes contributions from the bright Hα emission fluxes, which somewhat complicates interpretation of the results. One way of confrming the correlation of the Figure 3 result is to obtain simultaneous photometry in other bands, e.g., B band. B photometry has the additional advantage that contamination from the weaker Hγ and Hβ Balmer emission lines is much less than from Hα, which contributes significantly to the integrated V-band light of VV Cep. Figure 4 shows a preliminary period analysis of the V photometry of Figure 1, demonstrating that the period of the brightness change lies close to the 150-day period first proposed by Hayasaka et al. (1971). However, many other periods have been reported in the literature over the years: e.g., periods of 60, 110, 114, 116, and 280 days (Graczyk et al. 1999; Saito et al. 1980; McCook et al. 1978; Baldinelli et al. 1979; Pfeiffer et al. 1989). This behaviour suggests the short-term variability is somewhat irregular in nature, and probably has a substantial stochastic component. Indeed, the plot of the V variability, phased to a period days (Figure 4), shows a substantial fraction of this variability remains unexplained by this periodic oscillation. Figure 3. Correlation of VV Cep Hα EW with V magnitude 3

4 Figure 4. Period analysis of the data in Fig. 1; Period = ± 1.2 days W. Vollmann and G. Samolyk also derived B photometry, concurrent with the DSLR V observations (Figure 5). However because of the lower DSLR pixel sensitivity of Vollmann's B photometry, these data are not as precise, nor as accurate, as the V photometry. Vollmann used the Johnson B brightness of the reference stars, but did not transform to the Johnson B system, and therefore there is an offset in the derived B magnitudes compared to Samolyk's more accurate Johnson B magnitudes. Nevertheless, a similar inverse correlation is also seen between the B magnitudes and Hα EW. The addition of this B photometry confirms the inverse nature of the correlation between the Hα emission fluxes and M star brightness shown in Figure 3, derived from a large number of independent DSLR, CCD, and spectroscopic observations by a large number of individual observers over a period of several years. Figure 5. Hα EW vs. DSLR blue (B)-Band Photometry of VV Cep 4

5 Conclusions The observations presented here span the period from JD to JD , corresponding to orbital phases , measured from zero phase at mid-eclipse. The amplitude of the photometric variability is ΔV mag, whereas the amplitude of the total eclipse of the B star is only is ΔV 0.15 mag. The M supergiant is about 2 magnitudes brighter at V than the hot companion. Therefore, changes in disk size and brightness of the B-type companion similar to those observed in Be stars could only result in very limited photometric variability, and so cannot be the cause of the observed V-band photometric variability in VV Cep. Instead, the observed photometric variability must be intrinsic to the M supergiant. The period and photometric amplitude of the variability is quite similar to that resulting from irregular pulsation in other, single late-type supergiants. All of this strongly suggests that the photometric variability observed for VV Cep is intrinsic, and due to irregular pulsation of the M star. In that case, the variable Hα emission flux must be a consequence of the M supergiant's pulsation. One possibility is that the local wind density near the hot companion varies with the change in M star radius over the pulsational cycle, resulting in a higher ambient gas density near the hot star when the M star is at maximum radial extent. This higher circumstellar density results in a more efficient conversion of B-star Lyman continuum photons into Balmer line emission. At maximum radius, the effective temperature of the M star is near minimum, with redder B and V colours, and fainter B and V magnitudes, consistent with the observed Hα-V relation. The cause of the variable emission from the B star companion of VV Cep, and especially the very prominent Hα emission, has remained a mystery for many years. In this paper, we present the first quantitative connection between the variability of the companion's emission flux and the photometric variability of the M supergiant. Furthermore, since the only obvious physical link between these physically separated sources (the VV Cep B and M stars) is the M star wind, we propose that it is the indirect effect of M supergiant pulsation on the wind properties that is responsible for the variability of the Hα (and other Balmer line) emission flux. Therefore, if confirmed by further observation, the prominent Hα emission in VV Cep (and similar stars such as KQ Pup) may provide a useful diagnostic tool for monitoring the behaviour of the supergiant winds in these binaries. Acknowledgements We are grateful to Sara and Carl Sawicki (Alpine, Texas, USA) for their helpful improvements and suggestions in language. We are grateful to the other observers of the ARAS spectroscopy group: J. N. Terry, B. Koch, O. Thizy, E. Betrand, O. Garde, F. Teyssier, T. Lester, J. Foster, Ch. Buil, M. Schwarz, J. Montier, J. J. Boussat, Dong Li, J. Guarro, & D. Hyde, for their contribution of spectra of VV Cep used to determine for the H EW. We wish to thank Bela Hassforther (Deutsche Arbeitsgemeinschaft Veränderliche Sterne, BAV-Germany) for supplying additional V photometry. Finally, we acknowledge, with thanks, the use of variable star observations contributed to the AAVSO International Database by Gerard Samolyk. 5

6 References: Baldinelli, L., Ghedini, S., Marmi, S. 1979, IBVS, 1675 Bennett, P.D., Bauer, W.H. 2015, in Giants of Eclipse: The Zeta Aurigae Stars and Other Binary Systems, Astrophysics and Space Science Library, Vol. 408, ed. T.B. Ake, & E. Griffin, (Heidelberg: Springer), 85 Graczyk, D., Mikolajewski, M., Janowski, J. L. 1999, IBVS, 4679 Harmanec, P., 1983, Hvar Obs. Bull., 7, 55 Hayasaka, T., Saijo, K., Sato, H., Saito, M., Kitamura, M. 1977, Tokyo Astr. Bull. Second Series, 247, 2865 Hutchings, J. B., Wright, K. O. 1971, MNRAS, 155, 203 McCook, G., P., Guinan, E., F. 1978, IBVS, 1385 Pfeiffer, R. J., Maffei, J. C. 1989, BAAS, 21, 792 Saito, M., Sato, H., Saijo, K., Hayasaka, T. 1980, PASJ 32, Wright, K. O. 1977, JRASC, 71, 152 6