Orbital periods of cataclysmic variables identified by the SDSS – IV. SDSS J220553.98+115553.7 has stopped pulsating

We present time-series Very Large Telescope (VLT) spectroscopy and New Technology Telescope (NTT) photometry of the cataclysmic variable SDSS J220553.98+115553.7, which contains a pulsating white dwarf. We determine a spectroscopic orbital period of   P _orb= 82.825 ± 0.089 min  from velocity measurements of the Hα emission line. A period analysis of the light curves reveals a dominant periodicity at   P _phot= 44.779 ± 0.038 min  which is not related to the spectroscopic period. However, the light curves do not exhibit a variation at any frequency which is attributable to GW Lib-type pulsations, to a detection limit of 5 mmag. This non-detection is in contrast to previous studies which have found three pulsation frequencies with amplitudes of 9–11 mmag at optical wavelengths. Destructive interference and changes to the thermal properties of the driving layer in direct response to accretion can be ruled out as causes of the non-detection. Alternatively, it is feasible that the object has cooled out of the instability strip since a previous (unobserved) dwarf nova superoutburst. This would be the first time this behaviour has been seen in a cataclysmic variable pulsator. Another possibility is that changes in the surface characteristics, possibly induced by accretion phenomena, have modified the surface visibility of the pulsation modes. Further observations, particularly improved constraints on the time-scale for changes in the mode spectrum, are needed to distinguish among possible explanations.     

SDSS unveils a population of intrinsically faint cataclysmic variables at the minimum orbital period

We discuss the properties of 137 cataclysmic variables (CVs) which are included in the Sloan Digital Sky Survey (SDSS) spectroscopic data base, and for which accurate orbital periods have been measured. 92 of these systems are new discoveries from SDSS and were followed-up in more detail over the past few years. 45 systems were previously identified as CVs because of the detection of optical outbursts and/or X-ray emission, and subsequently re-identified from the SDSS spectroscopy. The period distribution of the SDSS CVs differs dramatically from that of all the previously known CVs, in particular it contains a significant accumulation of systems in the orbital period range 80–86 min. We identify this feature as the elusive 'period minimum spike' predicted by CV population models, which resolves a long-standing discrepancy between compact binary evolution theory and observations. We show that this spike is almost entirely due to the large number of CVs with very low accretion activity identified by SDSS. The optical spectra of these systems are dominated by emission from the white dwarf photosphere, and display little or no spectroscopic signature from the donor stars, suggesting very low mass companion stars. We determine the average absolute magnitude of these low-luminosity CVs at the period minimum to be  〈 M_g 〉= 11.6 ± 0.7  . Comparison of the SDSS CV sample to the CVs found in the Hamburg Quasar Survey and the Palomar Green Survey suggests that the depth of SDSS is the key ingredient resulting in the discovery of a large number of intrinsically faint short-period systems.     

The space density of cataclysmic variables: constraints from the ROSAT North Ecliptic Pole survey

We use the ROSAT North Ecliptic Pole (NEP) survey to construct a small, but purely X-ray flux-limited sample of cataclysmic variable stars (CVs). The sample includes only four systems, two of which (RX J1715.6+6856 and RX J1831.7+6511) are new discoveries. We present time-resolved spectroscopy of the new CVs and measure orbital periods of 1.64 ± 0.02 and 4.01 ± 0.03 h for RX J1715.6+6856 and RX J1831.7+6511, respectively. We also estimate distances for all the CVs in our sample, based mainly on their apparent brightness in the infrared. The space density of the CV population represented by our small sample is  1.1^+2.3_−0.7× 10⁻⁵ pc⁻³  . We can also place upper limits on the space density of any subpopulation of CVs too faint to be included in the NEP survey. In particular, we show that if the overall space density of CVs is as high as  2 × 10⁻⁴ pc⁻³  (as has been predicted theoretically), the vast majority of CVs must be fainter than   L _X≃ 2 × 10²⁹ erg s⁻¹  .     

Time-resolved optical observations of five cataclysmic variables detected by INTEGRAL

The European Space Agency γ-ray telescope, INTEGRAL , is detecting relatively more intrinsically rare cataclysmic variables (CVs) than were found by surveys at lower energies. Specifically, a large fraction of the CVs that are INTEGRAL sources consists of asynchronous polars and intermediate polars (IPs). IP classifications have been proposed for the majority of CVs discovered by INTEGRAL , but, in many cases, there is very little known about these systems. In order to address this, I present time-resolved optical data of five CVs discovered through INTEGRAL observations. The white dwarf spin modulation is detected in high-speed photometry of three of the new CVs (IGR J15094−6649, IGR J16500−3307 and IGR J17195−4100), but two others (XSS J12270−4859 and IGR J16167−4957) show no evidence of magnetism, and should be considered unclassified systems. Spectroscopic orbital period ( P _orb) measurements are also given for IGR J15094−6649, IGR J16167−4957, IGR J16500−3307 and IGR J17195−4100.     

Roche tomography of cataclysmic variables – I. Artefacts and techniques

Roche tomography is a technique used for imaging the Roche-lobe-filling secondary stars in cataclysmic variables (CVs). In order to interpret Roche tomograms correctly, one must determine whether features in the reconstruction are real, or the result of statistical or systematic errors. We explore the effects of systematic errors using reconstructions of simulated data sets, and show that systematic errors result in characteristic distortions of the final reconstructions that can be identified and corrected. In addition, we present a new method of estimating statistical errors on tomographic reconstructions using a Monte Carlo bootstrapping algorithm, and show this method to be much more reliable than Monte Carlo methods which 'jiggle' the data points in accordance with the size of their error bars.     

Infrared spectroscopy of cataclysmic variables – III. Dwarf novae below the period gap and nova-like variables

We present K -band spectra of the short-period dwarf novae YZ Cnc, LY Hya, BK Lyn, T Leo, SW UMa and WZ Sge, the nova-like variables DW UMa, V1315 Aql, RW Tri, VY Scl, UU Aqr and GP Com, and a series of field dwarf stars with spectral types ranging from K2 to M6.
The spectra of the dwarf novae are dominated by emission lines of H  i and He  i . The large velocity and equivalent widths of these lines, in conjunction with the fact that the lines are double-peaked in the highest inclination systems, indicate an accretion disc origin. In the case of YZ Cnc and T Leo, for which we obtained time-resolved data covering a complete orbital cycle, the emission lines show modulations in their equivalent widths that are most probably associated with the bright spot (the region where the gas stream collides with the accretion disc). There are no clear detections of the secondary star in any of the dwarf novae below the period gap, yielding upper limits of 10–30 per cent for the contribution of the secondary star to the observed K -band flux. In conjunction with the K -band magnitudes of the dwarf novae, we use the derived secondary star contributions to calculate lower limits to the distances to these systems.
The spectra of the nova-like variables are dominated by broad, single-peaked emission lines of H  i and He  i – even the eclipsing systems we observed do not show the double-peaked profiles predicted by standard accretion disc theory. With the exception of RW Tri, which exhibits Na  i , Ca  i and ¹²CO absorption features consistent with a M0V secondary contributing 65 per cent of the observed K -band flux, we find no evidence for the secondary star in any of the nova-like variables. The implications of this result are discussed.     

The influence of selection effects on the observed cataclysmic variable population: modelling and application to the Palomar–Green sample

Large differences between the properties of the known sample of cataclysmic variable stars (CVs) and the predictions of the theory of binary star evolution have long been recognized. However, because all existing CV samples suffer from strong selection effects, observational biases must be taken into account before it is possible to tell whether there is an inconsistency. In order to address this problem, we have modelled the impact of selection effects on observed CV samples using a Monte Carlo approach. By simulating the selection criteria of the Palomar–Green (PG) Survey, we show that selection effects cannot reconcile the predictions of standard CV evolution theory with the observed sample. More generally, we illustrate the effect of the biases that are introduced by magnitude limits, selection cuts in U − B and restrictions in Galactic latitude.