E. T. Harlaftis

ULTRACAM: an ultrafast, triple‐beam CCD camera for high‐speed astrophysics

Charge-Coupled Devices (CCD) have revolutionised observational astronomy since they were introduced in the 1970s. Their only limitation is their inability to operate at high frame rates. This has meant that faint objects that vary on timescales of less than of order seconds cannot be temporally resolved, ruling out, for example, the study of variability in compact objects, such as white dwarfs, neutron stars and black holes. ULTRACAM is a triple-beam camera designed to overcome this limitation and enable the study of astrophysics on fast timescales using CCDs. The project was awarded £292,034 in July 1999 by the Particle Physics and Astronomy Research Council and the instrument saw first light on 16 May 2002 on the William Herschel Telescope, on budget and three months ahead of schedule. This thesis describes my contribution to the ULTRACAM project. It begins with a description of the motivation for building ULTRACAM and lists its functional and performance requirements. An instrument design which meets these requirements is then presented, followed by a description of the manufacture and integration phase of the project. A whole chapter is then devoted to an in-depth analysis of the commissioning data obtained on the WHT, which verifies that ULTRACAM performs to specification. The thesis concludes with some suggestions for enhancements and future work.

HS 2331+3905: The cataclysmic variable that has it all

We report detailed follow-up observations of the cataclysmic variable HS 2331+3905, identified as an emissionline object in the Hamburg Quasar Survey. An orbital period of 81.08 min is unambiguously determined from the detection of eclipses in the light curves of HS 2331+3905. A second photometric period is consistently detected at P 83.38 min,∼2.8% longer than P orb , which we tentatively relate to the presence of permanent superhumps. High time resolution photometry exhibits short-timescale variability on time scales of 5−6 min which we interpret as non-radial white dwarf pulsations, as well as a coherent signal at 1.12 min, which is likely to be the white dwarf spin period. A large-amplitude quasi-sinusoidal radial velocity modulation of the Balmer and Helium lines with a period ∼3.5 h is persistently detected throughout three seasons of time-resolved spectroscopy. However, this spectroscopic period, which is in no way related to the orbital period, is not strictly coherent but drifts in period and/or phase on time scales of a few days. Modeling the far-ultraviolet to infrared spectral energy distribution of HS 2331+3905, we determine a white dwarf temperature of T eff 10 500 K (assuming M wd = 0.6 M ), close to the ZZ Ceti instability strip of single white dwarfs. The spectral model implies a distance of d = 90 ± 15 pc, and a low value for the distance is supported by the large proper motion of the system, µ = 0.14 yr −1 . The non-detection of molecular bands and the low J, H, and K fluxes of HS 2331+3905 make this object a very likely candidate for a brown-dwarf donor.

The rotational broadening and the mass of the donor star of GRS 1915+105

The binary parameters of the microquasar GRS 1915+105 have been determined by the detection of Doppler-shifted 12 CO and 13 CO lines in its K-band spectrum (Greiner et al. 2001b). Here, we present further analysis of the same K-band VLT spectra and we derive a rotational broadening of the donor star of v sin i = 26±3 km s -1 from the 12 CO/ 13 CO lines. Assuming that the K-type star is tidally locked to the black hole and is filling its Roche-lobe surface, then the implied mass ratio is q = M d /M g = 0.058 ± 0.033. This result, combined with (P, K, i) = (33,5 d, 140 km s -1 , 66°) gives a more refined mass estimate for the black hole, M x = 14.0 ± 4.4 M ○. , than previously estimated, using an inclination of i = 66° ± 2° (Fender et al. 1999) as derived from the orientation of the radio jets and a more accurate distance. The mass for the early K-type giant star is M d = 0.81 ± 0.53 M ○. , consistent with a more evolved stripped-giant donor star in GRS 1915+105 than, for example, the donor star of the prototype black-hole X-ray transient. V404 Cyg which has the longest binary period after GRS 1915+105.

The 1996 outburst of GRO J1655-40: the challenge of interpreting the multiwavelength spectra

We report on the results of a multiwavelength campaign to observe the soft X-ray transient (SXT) and superluminal jet source GRO J1655-40 in outburst using HSTRXTECGRO together with ground-based facilities. This outburst was qualitatively quite different from other SXT outbursts and from previous outbursts of this source. The onset of hard X-ray activity occurred very slowly, over several months, and was delayed relative to the soft X-ray rise. During this period, the optical fluxes declined steadily. This apparent anticorrelation is not consistent with the standard disc instability model of SXT outbursts, nor is it expected if the optical output is dominated by reprocessed X-rays, as in persistent low-mass X-ray binaries. Based on the strength of the 2175-Angstroms interstellar absorption feature we constrain the reddening to be E(B-V)=1.2+/-0.1, a result which is consistent with the known properties of the source and with the strength of interstellar absorption lines. Using this result we find that our dereddened spectra are dominated by a component peaking in the optical, with the expected nu^1/3 disc spectrum seen only in the ultraviolet. We consider possible interpretations of this spectrum in terms of thermal emission from the outer accretion disc and/or secondary star, both with and without X-ray irradiation, and also as non-thermal optical synchrotron emission from a compact self-absorbed central source. In addition to the prominent Heii 4686-Angstroms line, we see Bowen fluorescence lines of Niii and Oiii, and possible P Cygni profiles in the ultraviolet resonance lines, which can be interpreted in terms of an accretion disc wind. The X-ray spectra broadly resemble the high-soft state commonly seen in black hole candidates, but evolve through two substates. Taken as a whole, the outburst data set cannot readily be interpreted by any standard model for SXT outbursts. We suggest that many of the characteristics could be interpreted in the context of a model combining X-ray irradiation with the limit-cycle disc instability, but with the added ingredient of a very large disc in this long-period system.

SW Sextantis stars: the dominant population of cataclysmic variables with orbital periods between 3 and 4 h

We present time-series optical photometry of five new cataclysmic variables (CVs) identified by the Hamburg Quasar Survey (HQS). The deep eclipses observed in HS 0129+2933 (= TT Tri), HS 0220+0603 and HS 0455+8315 provided very accurate orbital periods of 3.35129827(65), 3.58098501(34) and 3.56937674(26) h, respectively. HS 0805+3822 shows grazing eclipses and has a likely orbital period of 3.2169(2) h. Time-resolved optical spectroscopy of the new CVs (with the exception of HS 0805+3822) is also presented. Radial velocity studies of the Balmer emission lines provided an orbital period of 3.55 h for HS 1813+6122, which allowed us to identify the observed photometric signal at 3.39 h as a negative superhump wave. The spectroscopic behaviour exhibited by all the systems clearly identifies them as new SW Sextantis (SW Sex) stars. HS 0220+0603 shows unusual N ii and Si ii emission lines suggesting that the donor star may have experienced nuclear evolution via the CNO cycle. These five new additions to the class increase the number of known SW Sex stars to 35. Almost 40 per cent of the total SW Sex population do not show eclipses, invalidating the requirement of eclipses as a defining characteristic of the class and the models based on a high orbital inclination geometry alone. On the other hand, as more SW Sex stars are identified, the predominance of orbital periods in the narrow 3–4.5 h range is becoming more pronounced. In fact, almost half the CVs which populate the 3–4.5 h period interval are definite members of the class. The dominance of SW Sex stars is even stronger in the 2–3 h period gap, where they make up 55 per cent of all known gap CVs. These statistics are confirmed by our results from the HQS CVs. Remarkably, 54 per cent of the Hamburg nova-like variables have been identified as SW Sex stars with orbital periods in the 3–4.5 h range. The observation of this pile-up of systems close to the upper boundary of the period gap is difficult to reconcile with the standard theory of CV evolution, as the SW Sex stars are believed to have the highest mass-transfer rates among CVs. Finally, we review the full range of common properties that the SW Sex stars exhibit. Only a comprehensive study of this rich phenomenology will prompt to a full understanding of the phenomenon and its impact on the evolution of CVs and the accretion processes in compact binaries in general.

THE MASS RATIO AND THE DISK IMAGE OF THE X-RAY NOVA GS 2000+25

Keck observations of the black hole candidate GS 2000+25 in quiescence have recently been used by Filippenko, Matheson, and Barth to determine a mass function f(Mx) = 4.97 ± 0.10 solar mass for the compact object. Our reanalysis of the data confirms this result (5.01 ± 0.12 solar mass). We estimate a mass ratio of q = Mc/Mx = 0.042 ± 0.012 from the rotational broadening of the companion star, v sin i = 86 ± 8 km s-1. From q and the companion stars radial velocity Kc, we derive the mass of the compact object Mx = (5.44 ± 0.15) sin-3 i solar mass and the mass of the companion star Mc = (0.23 ± 0.02) sin-3 i solar mass. Constraints on the inclination (75 degrees > i > 47 degrees) lead to 6.04 < Mx < 13.9 and 0.26 < Mc < 0.59 (1-sigma) in solar mass units. We determine a spectral type of K3-K6 for the companion star, which contributes more than 72% of the light at red wavelengths (94% ± 5% for K5 V). Our analysis shows that the companion is an undermassive star, slightly evolved but not a subgiant. Consistent with the work of Filippenko et al. we detect Li I lambda-6708 absorption (equivalent width 150 ± 85 mA) in the spectrum of the companion. A Doppler image of the system shows evidence for a bright spot which contributes 10% of the emission-line flux density, and arises in the range 0.3-0.6 RL1 of the accretion disk, where L1 is the inner Lagrangian point. Along the trajectory of the gas stream, the velocities are initially ballistic and gradually become Keplerian.

HS 2237+8154: On the onset of mass transfer or entering the period gap?

We report follow-up observations of a new white dwarf/red dwarf binary HS 2237+8154, identified as a blue variable star from the Hamburg Quasar Survey. Ellipsoidal modulation observed in the R band as well as the radial velocity variations measured from time-resolved spectroscopy determine the orbital period to be Porb = 178.10 +- 0.08 min. The optical spectrum of HS 2237+8154 is well described by a combination of a Teff = 11500 +- 1500 K white dwarf (assuming log g = 8) and a dM 3.5 +- 0.5 secondary star. The distance implied from the flux scaling factors of both stellar components is d = 105 +- 25 pc. Combining the constraints obtained from the radial velocity of the secondary and from the ellipsoidal modulation, we derive a binary inclination of i = 50-70 and stellar masses of and Mwd = 0.47-0.67 M and Msec = 0.2-0.4 M. All observations imply that the secondary star must be nearly Roche-lobe filling. Consequently, HS 2237+8154 may be either a pre-cataclysmic variable close to the start of mass transfer, or - considering its orbital period - a cataclysmic variable that terminated mass transfer and entered the period gap, or a hibernating nova.

Cataclysmic variables from a ROSAT/2MASS selection - I. Four new intermediate polars

We report the first results from a new search for cataclysmic variables (CVs) using a combined X-ray (ROSAT)/infrared (2MASS) target selection that discriminates against background active galactic nuclei. Identification spectra were obtained at the Isaac Newton Telescope for a total of 174 targets, leading to the discovery of 12 new CVs. Initially devised to find short-period low-mass-transfer CVs, this selection scheme has been very successful in identifying new intermediate polars. Photometric and spectroscopic follow-up observations identify four of the new CVs as intermediate polars: 1RXS J063631.9+353537 P(orb)similar or equal to 201 min, P-spin= 1008.3408 s or 930.5829 s), 1RXS J070407.9+262501 (P(orb)similar or equal to 250 min, P-spin= 480.708 s) 1RXS J173021.5-055933 (P-orb= 925.27 min, P-spin= 128.0 s), and 1RXS J180340.0+401214 (P-orb= 160.21 min, P-spin= 1520.51 s). RX J1730, also a moderately bright hard X-ray source in the INTEGRAL/IBIS Galactic plane survey, resembles the enigmatic AE Aqr. It is likely that its white dwarf is not rotating at the spin equilibrium period, and the system may represent a short-lived phase in CV evolution.

A doppler map and mass-ration constraint for the black-hole x-ray nova ophiuchi 1977.

We have reanalyzed Keck observations of Nova Oph 1977 to extend the work done by Filippenko et al. (1997), who recently determined a mass function f(M_x) = 4.86 +/- 0.13 M_o for the compact object. We constrain the rotational broadening, v sin i < 79 km/s, at the 90% confidence level, which gives a mass ratio q < 0.053. The K-type companion star of Nova Oph 1977 contributes 28-37% of the light at red wavelengths. The abnormal LiI 6708 absorption line from the companion star is not detected (EW < 0.12 A), in contrast to four other X-ray binaries. An Halpha Doppler image of the system shows emission from the companion star in addition to the accretion disk.

Spiral shocks in the accretion disc of IP Peg during outburst maximum

In response to our recent discovery of spiral arms in the accretion disc of IP Peg during rise to outburst, we have obtained time-resolved spectrophotometry of IP Peg during outburst maximum. In particular, indirect imaging of He II 4686, using Doppler tomography, shows a two-arm spiral pattern on the disc image, which confirms repeatability over different outbursts. The jump in He II intensity (a factor of more than 2) and in velocity (∼ 200–300 km s-1) clarifies the shock nature of the spiral structure. The He II shocks show an azimuthal extent of ∼ 90 °, a shallow power-law emissivity ∼ V-1, an upper limit of 30 ° in opening angle, and a flux contribution of 15 per cent of the total disc emission. We discuss the results in view of recent simulations of accretion discs which show that spiral shocks can be raised in the accretion disc by the secondary star.