A unification theory is proposed in which almost all of outburst behavior in dwarf novae may be understood in a unified way within the basic framework of the disk instability model. In particular, a rich variety of outburst behavior of cataclysmic variables below the period gap may be understood by the thermal-tidal instability model in which the coupling of the two intrinsic instabilities in the accretion disk plays a unique role.

The AM CVn stars are mass transferring, interacting binary, white dwarf systems with orbital periods of 15…45 minutes. Hydrogen is completely lost from these systems, and we observe small helium disks which may show thermal and tidal instabilities if the mass transfer rate is large enough (Osaki 1995). A tidal instability brings the disk into a superoutburst state, and in the light curve we may observe superhumps. Based on the observed periods and mass transfer rates we can divide the AM CVn stars into three groups:

1. A: In permanent superoutburst: AM CVn and EC 15330-1403

2. B: Regular superoutbursts: CR Boo, V803 Cen and CP Eri

3. C: Not yet observed superoutburst: GP Com

Group A consists of systems with disks which are too hot to decline into a low state. These disks are in a constant superoutburst state, analogous to nova-like CVs which show no outbursts, but still exhibit permanent superhumps (Skillman & Patterson 1993). Group B shows normal outbursts which can trigger superoutbursts, analogous to the VY Scl dwarf novae. The group C object may also be a superhumper, but with very infrequent outbursts analogous to the SU UMa stars (Warner 1995). In the following we will discuss evidence for superoutbursts in these systems, and the likelihood for these systems to develop elliptical disks.

Accretion through Roche lobe overflow in the helium cataclysmic binary AM CVn is considered and its emitted spectrum is computed using model stellar atmospheres. The accretion disc in AM CVn is helium-rich and such discs have hitherto not been constructed. We have computed the models by numerically solving the radiative transfer equation for helium opacities. The method adopted is that of treating the accretion disc as a set of optically thick and geometrically thin concentric annuli. Each annulus is treated individually as a blackbody or as a stellar atmosphere. We find that our description of AM CVn is consistent with both photometric and spectroscopic measurements of the system which have been made using the Nordic Optical Telescope (NOT) and the International Ultraviolet Explorer (IUE). We find that the stellar model synthetic spectrum provides a better fit to the observed spectrum than the Planckian disc spectrum. Also, for a given accretion rate, the synthetic spectrum composed of the sum of a series of stellar atmospheres is flatter than the corresponding blackbody sum.

I will briefly review recent progress in our understanding of the global propagation of accretion disk instabilities operating in interacting binary systems in which the accreting star is a ~ 10 M⊙ black hole. A detailed discussion of this work can be found in Cannizzo, Chen & Livio (1995).

The X-ray light curves of the brightest and best studied X-ray novae – systems such as A0620–00, GS2000+25, GS1124–68, and GRO J0422+32 – are notable in two respects: the rise times are fast (a few days) and the decays have an exponential character, where the e–folding time constant is about 30 to 40 d (Mineshige, Yamasaki, & Ishizaka 1993, hereafter MYI). It is noteworthy that outbursts in dwarf novae also decay exponentially, with the time constant being about a factor of 10 shorter. In this contribution I will discuss the constraints on the physics of accretion disks which the (observed) exponential decay enables us to make.

We present time-dependent, irradiated, accretion disk models for the black hole X-ray novae in the first hundred days of the dwarf nova-like outbursts, including the rise, precursor, maximum and the secondary re-flare. This work is based on the disk instability model (Kim, Mineshige & Wheeler 1996, Kim, Wheeler & Mineshige 1996). The model is reasonably consistent with the observed optical light curves. The irradiators are the central hot region around the black hole, and the corona or chromosphere above the accretion disk. In addition, we include the time-dependent shadowing effect and consequent blocking of the outer portions of the disk from the central irradiator. We find the stagnation phenomenon whereby the disk stays in the intermediate temperature stage between the hot and cool state. This can explain the recently discovered optical precursor rise prior to the maximum light in Nova Sco 1994 (Bailyn et al. 1995: see Fig. 1). We suggest the secondary re-flare after the maximum is due to the coupled effects of the irradiation and stagnation. In the model, the stagnation phenomenon during the rise results from the partial ionization and molecular opacity. In addition, we find irradiation-induced stagnation during the decay phase, which is consistent with the observed secondary re-flare in X-ray novae (see Fig. 1). In the overall evolution of model outbursts in the first hundred days, the outer disk is blocked from the irradiation and, in turn, the companion star may not be strongly irradiated. This suggests that there is no appreciable increase of mass transfer rate during the decay prior to the secondary re-flare, unlike the behaviour in the mass transfer burst models.

Recently discovered intermediate polars extend the limits of known behaviour. They now fill the range between, and overlap with, nonmagnetic CVs and AM Hers. They have a wide range of asynchronism, a full range of field strengths, and several different accretion modes and pulsation mechanisms. I highlight (i) a puzzle over the symmetry between magnetic poles, (ii) the X-ray spectra, (iii) the similarity with polars, and (iv) the mass of EX Hya.

I review recent progress in understanding the accretion geometry and angular momentum flows in magnetic CVs. The recent discovery that AE Aqr is an ejector demonstrates that the criterion for disc formation depends on the spin rate of the white dwarf, in addition to the strength of its magnetic field. It seems likely that AE Aqr represents one phase of a cyclic evolution followed by magnetic CVs. An important consequence is that an observed white dwarf spin rate is not uniquely determined by current properties of the system (e.g. magnetic field and instantaneous accretion rate), but also by its history.

The large-scale accretion flow in the intermediate polars (IPs) is still a matter of vigorous debate. It is known that the magnetic field of the white dwarf (WD) controls the accretion flow close to the surface, channeling the plasma onto the polecaps and giving rise to X-ray emission modulated at the WD spin period (Pspin). After their discovery it was assumed that IPs were the WD analogues of the pulsing X-ray binaries, where a magnetic neutron star accretes from a disrupted accretion disc. However, a number of authors have pointed out that the criteria for disc formation in IPs are less certain than those for the X-ray binaries.

The simplest possible criterion for disc formation in a binary is that the accretion flow should be able to orbit freely about the primary star (see Frank, King & Raine 1991 for a review). In non-magnetic systems this is merely the condition that the minimum approach distance of the free stream (Rmin) should exceed the radius of the primary. The situation in magnetic systems is more complex, as the magnetic field of the primary presents an obstacle to the infalling accretion stream. In many treatments of IPs it is assumed that the plasma stream is able to orbit freely about the WD until the ram pressure of the stream is of the same order as the magnetic pressure ρv2 ~ B2/8π, where ρ is the stream density, v the stream velocity and B the local magnetic field strength. This condition fixes the magnetospheric radius, Rmag, inside which the magnetic field is assumed to thread the stream material and direct the accretion flow along the fieldlines.

The enigmatic flaring activity of AE Aqr, which extends from the radio to the X-ray bands, may be the exotic outcome of a relatively feeble accretion flow onto the rapidly spinning white dwarf. We summarize and interpret the results of recent observations by HST and ROSAT of the aperiodic flares. Optical and UV flares are tightly correlated and are clearly involved in powering the emission lines. The spectrum of a flare consists of a Balmer continuum, and a plethora of emission lines, which vary synchronously. The large optical/UV flares are not accompanied by large X-ray outbursts. The radial velocity curves of the UV lines, and of Hα, suggest that they originate in the accretion stream. The observational results dis-favor scenarios invoking coronal activity on the secondary star or instabilities at the disk/magnetosphere interface. Rather, they support a recently proposed scenario in which the accretion flow consists of blobs which are shocked upon encountering the white dwarf magnetosphere.

We propose a model for the acceleration of particles by magnetic pumping to drive the radio emission from AE Aqr.

The UV continuum of YY Dra exhibits a sinusoidal modulation with 15.5% semi-amplitude and period 264.7 s, commensurate with one half the white dwarf spin period. The Civ line profile varies with spin phase, implying accretion onto, and visibility of, both magnetic poles. The line width is greatest at the phases of continuum pulsation maximum, when the wings extend out to ±3000 km s−1. This implies a minimum white dwarf mass of 0.62 M⊙, and constrains the angle between the line of sight and the magnetic axis.

The 71-s modulation in the light curve of the old nova DQ Her has a mean semi-amplitude of 1.8% at 2180 Å, although the amplitude varies by at least a factor of two from orbit to orbit. The disturbance in the amplitude and phase of the modulation during eclipses is similar to that observed at visual wavelengths. We conclude that at 2180 Å, as at visual wavelengths, the modulated flux is reprocessed and comes from the accretion disk.

RXJ0558+5353 was discovered and classified as an intermediate polar during the ROSAT all-sky survey and subsequent optical follow up programme (Haberl et al. 1994). A further pointed ROSAT observation revealed a spin periodicity of 272.74 s and a soft X-ray component well represented by an absorbed 57 eV blackbody, the spin modulation being due principally to intensity variations of the component (rather than variable absorption). Optical spectroscopy showed the orbital period to be 4.15 h. In this paper we show the true spin period to be 545.4555(8) s (Allan, Horne & Hilditch 1995), twice the published X-ray period, and also discuss preliminary results of the first spin-resolved spectroscopy of RXJ0558+5353.

I report on the independent discovery of a feature in BG CMi which resembles a rotational disturbance (RD). A similar effect has already been reported in the He IIλ 4686 line, whereas here it is seen in the Balmer lines, 0.25 orbital cycles prior to that reported previously. These results suggest that BG CMi is at a relatively high inclination.

We present simultaneous multiwavelength observations of the intermediate polar PQ Gem (Mason et al. 1992) obtained on 1993 February 18 and 19 using the Hatfield Polarimeter on UKIRT.

The data are folded on the 13.9 m spin period in Fig. 1. The light curves are double peaked at long wavelengths, with dips at phase ~ 0.15 and phase ~0.65, but almost sinusoidal in the U and B bands where the phase ~0.65 dip is absent. The percentage of circular polarisation also varies with the spin cycle, most notably in the I band, with both positive and negative excursions. The peaks in the positive and negative polarisation occur at phase ~ 0.15 and phase ~ 0.65 respectively, approximately coincident with the two intensity dips.

Piirola, Hakala & Coyne (1993) modeled the optical/IR light curve of RE 0751+14 assuming a uniform shock structure and neglecting the hard X-ray emission. In this paper, we model the light curves at optical/IR and hard X-ray wavelengths and include the effects of the shock structure.

We base our model on accretion onto a white dwarf with a displaced magnetic dipole for a range of likely white dwarf masses. We find that the observed intensity variations of X-rays and in the I band over one spin period largely determine the position of the emission regions. Furthermore, the observed maximum X-ray flux constrains the specific accretion rate. We deduce that the magnetic field at the pole is likely to be in the range 9 .. .21 MG, which is consistent with the estimates of Piirola et al. (1993). It had been proposed previously that there must exist asynchronous rotators with sufficiently strong magnetic fields such that the binaries will evolve into AM Her binaries (Chanmugam & Ray 1984; King, Frank & Ritter 1985). With this deduced high magnetic field RE 0751+14 is the most likely example of such a system known to date.

TX Col (1H0547–407; Porb = 5.72 h) is an intermediate polar (IP) with spin and synodic (beat) periods of 1911s and 2106 s respectively (Buckley & Tuohy 1989). EXOSAT observations (Tuohy et al. 1986) showed the beat period dominated at higher energies (ME; 1… 9keV), while the spin period was stronger in softer X-rays (LE1; 0.1… IkeV). The high flux level, and pulse fraction (>70%), of the soft X-rays were unique in an IP at that time. Now three other IPs, recently discovered from the ROSAT survey, have been found to show a true separate soft X-ray component, similar to the polars (Motch & Haberl 1994).

The strength of the synodic X-ray period in TX Col was, until the recent discovery of the variable polarised IP, RX J1712.6–2414 (Buckley et al. 1995), unique. This was taken as evidence for direct accretion onto the magnetosphere (e.g. Mason, Rosen & Hellier 1988), through a component of the accretion stream overflowing the disc (Hellier 1992).

The flow of matter, the accretion geometry and the energy release in polars (AM Her binaries) are to a large extent dominated by the strong magnetic field of the accreting white dwarf. In this review I summarize the main tools used to investigate the magnetic fields and the accretion streams, using optical and X-ray spectroscopy and photometry.

We solved the stationary one-dimensional two-fluid radiation hydrodynamic equations including cyclotron radiation for a wide range of mass flow rates. Here, we discuss the implications for accretion phenomena on the white dwarfs in AM Her binaries.

There is growing evidence of complex white dwarf magnetic fields in CVs (e.g. Meggitt & Wickramasinghe 1989; Robinson & Cordova 1994). The properties of systems with non-dipolar white dwarf magnetic fields have been investigated, and it was found that a non-negligible quadrupole component is essential for the magnetic locking of AM Her binaries (AM Hers) (Wu & Wickramasinghe 1993). In a recent study the ratio of the quadrupole to dipole component of the surface field of white dwarfs was shown to grow by a factor of 10 during the white dwarf cooling lifetime (Muslimov, van Horn & Wood 1995), thus the quadrupole field component is long lived and significant.