Allen W. Shafter

On the Nova Rate in the Galaxy

The Galactic nova rate is poorly known, with estimates in the literature ranging from as few as 11 to as many as 260 yr-1. At the lower end of the spectrum (50 yr-1), the predictions are based on scalings from extragalactic nova surveys, while estimates based on extrapolations of Galactic nova observations suggest rates that are significantly higher, in the range 50-100 yr-1 or more. In an attempt to reconcile this difference, the nova rate, based on Galactic nova observations, is recomputed. If the stellar mass distribution is axisymmetric about the Galactic center, a new estimate of the Galactic nova rate of ~35 ± 11 yr-1 is deduced. Although this value is marginally consistent with the highest estimates based on extragalactic surveys, the agreement is not entirely satisfactory. It is pointed out that a departure from axial symmetry, such as that caused by the presence of a Galactic bar, can potentially lower the estimated nova rate (perhaps approaching 50%) if the bar is sufficiently large, and has its long axis pointed toward the Sun. Alternatively, or perhaps in addition, it is possible that previous extragalactic surveys may have missed a significant fraction of novae (up to a factor of ~2) because of extinction internal to the galaxies. If this latter possibility can be ruled out, for example, through infrared surveys of nearby galaxies, the Galactic nova observations may provide additional support for the existence of a Galactic bar.

On the outburst recurrence time for the accretion disk limit cycle mechanism in dwarf novae

An attempt is made here to elucidate the dwarf nova outburst time scale in terms of basic physics, and to place limits on the model by using the observed recurrence times. A simple analytical model is developed to describe the limit cycle process, and an expression is obtained for the interval between eruptions. It is shown that the smallness of the amount of mass accreted onto the central white dwarf during eruption can be explained in terms of the propagation of heating and cooling fronts. The dependency of the fraction of the disk mass accreted on the model parameters is derived, and the result is used to construct a semianalytic expression for the recurrence time.

On the Spatial Distribution, Stellar Population, and Rate of Novae in M31

As part of a program to study the stellar population of novae, multiepoch Hα surveys of M31 have been undertaken. In the first survey, conducted between 1990 and 1993, a total of 53 novae were discovered in 11 CCD fields oriented generally along the major axis of M31. The second survey was conducted between 1995 and 1997, yielding a total of 29 novae in five fields including the outer regions of M31s disk. In both surveys, the spatial distribution of novae is found to follow the background galactic light. However, after decomposing the background B light into its bulge and disk components, the nova distribution is found to follow the bulge light better than the overall light, with the disk light providing a significantly poorer fit. As a check on the possibility that extinction in the disk of M31 is biasing the results, an analysis of the spatial distribution of M31s planetary nebulae—the space density of which is expected to be only weakly affected by stellar population—was also considered. Unlike the nova distribution, the planetary nebula distribution is found to follow the overall background B light remarkably well, establishing that, regardless of extinction, the nova distribution is more centrally concentrated than the planetary nebula distribution. To quantify the results, a maximum likelihood analysis was conducted, and it suggests that ~70% of the novae arise from the bulge, although a bulge fraction as low as 50% cannot be ruled out by the data. A Monte Carlo simulation that makes use of the properties of observed Hα light curves was employed in order to translate the observed nova rate in the central region of M31 to a total bulge nova rate. The procedure was applied to the bulge field of our survey as well as to the Hα bulge survey of Ciardullo and coworkers and to a bulge subset of Arps extensive B-band survey. In all three cases, the results are remarkably consistent, suggesting a total bulge nova rate of ~25 ± 4 yr-1. Based on the bulge nova rate, an Hα mean lifetime versus Hα absolute magnitude relation is derived, which can be used to estimate the nova rate in galaxies of known distance that have been surveyed in Hα. A second-order relation given by log τc(days) = -(4.78 + 2.10Mc + 0.162M) provides the best fit to the data. Finally, the bulge rate is extrapolated to the entire galaxy assuming the disk-to-bulge nova ratio determined in the maximum likelihood analysis. A global nova rate of ~37 is found, which is somewhat higher than previous estimates of ~30 yr-1 for the overall nova rate in M31. If the luminosity-specific disk and bulge nova rates are comparable, the overall nova rate in M31 could approach 50 yr-1.

Exquisite Nova Light Curves from the Solar Mass Ejection Imager (SMEI)

We present light curves of three classical novae (KT Eridani, V598 Puppis, V1280 Scorpii) and one recurrent nova (RS Ophiuchi) derived from data obtained by the Solar Mass Ejection Imager (SMEI) on board the Coriolis satellite. SMEI provides near complete sky-map coverage with precision visible-light photometry at 102-minute cadence. The light curves derived from these sky maps offer unprecedented temporal resolution around, and especially before, maximum light, a phase of the nova eruption normally not covered by ground-based observations. They allow us to explore fundamental parameters of individual objects including the epoch of the initial explosion, the reality and duration of any pre-maximum halt (found in all three fast novae in our sample), the presence of secondary maxima, speed of decline of the initial light curve, plus precise timing of the onset of dust formation (in V1280 Sco) leading to estimation of the bolometric luminosity, white dwarf mass and object distance. For KT Eri, Liverpool Telescope SkyCamT data confirm important features of the SMEI light curve and overall our results add weight to the proposed similarities of this object to recurrent rather than to classical novae. In RS Oph, comparison with hard X-ray data from the 2006 outburst implies that the onset of the outburst coincides with extensive high velocity mass-loss. It is also noted that two of the four novae we have detected (V598 Pup and KT Eri) were only discovered by ground-based observers weeks or months after maximum light, yet these novae reached peak magnitudes of 3.46 and 5.42 respectively. This emphasizes the fact that many bright novae per year are still overlooked, particularly those of the very fast speed class. Coupled with its ability to observe novae in detail even when relatively close to the Sun in the sky, we estimate that as many as 5 novae per year may be detectable by SMEI.

The role of the dwarf nova period distribution in understanding the evolution of cataclysmic variables

Within the context of disk instability theory, cataclysmic variables possessing accretion disks (i.e., non-magnetic systems) are expected to exhibit disk instabilities that result in dwarf nova eruptions if the mass transfer rates fall below a critical level. It is argued that the eruptive characteristics of cataclysmic variables can therefore be used to infer relative mass transfer rates among nonmagnetic cataclysmic variables. Here the dwarf nova period distribution is used to constrain the variation of mass transfer with orbital period

A SPECTROSCOPIC AND PHOTOMETRIC SURVEY OF NOVAE IN M31

We report the results of a multi-year spectroscopic and photometric survey of novae in M31 that resulted in a total of 53 spectra of 48 individual nova candidates. Two of these, M31N 1995-11e and M31N 2007-11g, were revealed to be long-period Mira variables, not novae. These data double the number of spectra extant for novae in M31 through the end of 2009 and bring to 91 the number of M31 novae with known spectroscopic classifications. We find that 75 novae (82%) are confirmed or likely members of the Fe II spectroscopic class, with the remaining 16 novae (18%) belonging to the He/N (and related) classes. These numbers are consistent with those found for Galactic novae. We find no compelling evidence that spectroscopic class depends sensitively on spatial position or population within M31 (i.e., bulge versus disk), although the distribution for He/N systems appears slightly more extended than that for the Fe II class. We confirm the existence of a correlation between speed class and ejection velocity (based on line width), as in the case of Galactic novae. Follow-up photometry allowed us to determine light-curve parameters for a total of 47 of the 91 novae with known spectroscopic class. We confirm that more luminous novae generally fade the fastest and that He/N novae are typically faster and brighter than their Fe II counterparts. In addition, we find a weak dependence of nova speed class on position in M31, with the spatial distribution of the fastest novae being slightly more extended than that of slower novae.

TT Arietis - The low state

A comprehensive photometric and spectroscopic study of the low state of the novalike variable TT Ari has been made with observations spanning the UV to IR spectral regions. The roughly 5 mag drop in the systems luminosity from the high state indicates that the mass-transfer rate decreased by more than two orders of magnitude. The drastic reduction in the luminosity of the accretion disk between the high and low states enabled the white dwarf component to be directly observed for the first time. The broad absorption profiles at various lines and the UV energy distribution are best fitted by a hot white dwarf. A lower limit of 200 pc for the distance to TT Ari is derived, and the behavior of TT Ari is compared to that of another novalike variable, MV Lyr. The long-term photometric behavior of TT Ari makes it a typical member of the VY Scl subclass of novalike variables.

Novae in External Galaxies: M51, M87, and M101

As part of a program to determine the stellar population of novae, we have conducted a multiepoch Hα survey of the galaxies M51, M87, and M101. A total of nine and 12 novae were detected in the spiral galaxies M51 and M101, respectively, during four epochs of observation, and two epochs of observation yielded a total of nine novae in the giant elliptical galaxy M87. After correcting for the effective survey time and for the fraction of luminosity sampled, we find global nova rates of 18 ± 7, 91 ± 34, and 12 ± 4 novae per year for M51, M87, and M101, respectively. After normalizing to the total K-band luminosity of each galaxy, we estimate luminosity-specific nova rates for M51, M87, and M101 of 1.09 ± 0.47, 2.30 ± 0.99, and 0.97 ± 0.38 novae per year per 1010 solar luminosities in K. When we compare these data with measured values for the luminosity-specific nova rates of other galaxies, we find no compelling evidence for a significant variation with Hubble type. Possible ramifications of this result are discussed within the context of current theoretical models for nova production in galaxies.

M31N 2007-11d: A Slowly Rising, Luminous Nova in M31

We report a series of extensive photometric and spectroscopic observations of the luminous M31 nova M31N 2007-11d. Our photometric observations coupled with previous measurements show that the nova took at least 4 days to reach peak brightness at R ≃ 14.9 on 2007 November 20 UT. After reaching maximum, the time for the nova to decline 2 and 3 mag from maximum light (t 2 and t 3) was ~9.5 and ~13 days, respectively, establishing that M31N 2007-11d was a moderately fast declining nova. During the novas evolution, a total of three spectra were obtained. The first spectrum was obtained one day after maximum light (5 days post-discovery), followed by two additional spectra taken on the decline at two and three weeks post-maximum. The initial spectrum reveals narrow Balmer and Fe II emission with P Cygni profiles superimposed on a blue continuum. These data, along with the spectra obtained on the subsequent decline, clearly establish that M31N 2007-11d belongs to the Fe II spectroscopic class. The properties of M31N 2007-11d are discussed within the context of other luminous novae in M31, the Galaxy, and the LMC. Overall, M31N 2007-11d appears to be remarkably similar to Nova LMC 1991, which was another bright, slowly rising, Fe II nova. A comparison of the available data for luminous extragalactic novae suggests that the ≳ 4 day rise to maximum light seen in M31N 2007-11d may not be unusual, and that the rise times of luminous Galactic novae, usually assumed to be ≾ 2 days, may have been underestimated.

Tomographic Analysis of Hα Profiles in HDE 226868/Cygnus X-1

We present high-quality Hα spectra of the massive X-ray binary HDE 226868 = Cyg X-1 that were made in 1985 and 1986. We measured radial velocities using the He I λ6678 line, and we use the resulting orbital solution together with a Hipparcos light curve of the ellipsoidal variation to arrive at a revised period estimate of P = 5.59977 ± 0.00002 days. We analyze the orbital phase-related variations in Hα emission using a Doppler tomography algorithm, and we show that the profile variations are well matched by a linear combination of two components: a P Cygni profile that moves with the motion of the supergiant star and an emission component that follows an antiphase but phase-shifted velocity curve (semiamplitude Kem = 68 km s-1 and radial velocity maximum at 0 = 0.86; these parameters are similar to those for the He II λ4686 emission line). We use a vector decomposition of this motion to argue that the latter emission component forms between the stars in a focused wind flow from the supergiant to the unseen companion. We searched for but found no clear evidence of Hα emission from the accretion disk surrounding the companion (any such emission contributes less than 3% of the continuum intensity). The results demonstrate that the Hα line provides an important probe of the mass transfer process that eventually fuels the X-ray source.