Of the (approximately) four hundred thousand million stars in our Galaxy, more than half are not single stars like our
Sun, but occur in binary or multiple systems. Binary star systems come in many flavours: red stars orbiting blue stars, huge
stars orbiting tiny stars, black holes orbiting blue giants, red stars orbiting neutron stars, and so on. In a CV binary,
one star is a white dwarf: a collapsed star with the mass of the Sun in the volume of the Earth (our Sun will become a white
dwarf in about four and a half billion years). The other star is a red dwarf rather like our Sun, but redder and less massive.
The red dwarf and the white dwarf orbit each other once every few hours: they are so close together that the average CV
system would fit comfortably into our Sun. When we observe CVs, we can't resolve the two stars: they appear on the sky as
a point source.
The red star in a CV is so close to the white dwarf that it becomes tidally distorted --- gas is stripped off the red star
and falls towards the white dwarf. Because of conservation of angular momentum, the infalling gas can't plunge directly onto
the surface of the white dwarf. In systems where the white dwarf doesn't have an appreciable magnetic field, the infalling
gas forms a disc -- an accretion disc -- with the white dwarf at its centre. The gas in the disc spirals down towards the
white dwarf, radiating its gravitational potential energy away as it goes. The accretion disc usually outshines both the red
star and the white dwarf in visible light. If the white dwarf is strongly magnetic, other interesting things happen: more
about that in the section on magnetic CVs.
The CVs with accretion discs come in several flavours. The first to be discovered were the novae,
originally novae stella: new stars. They draw attention to themselves by their stupendous amplitude of variation --- 6 to
19 magnitudes (that's factors of about 100 to several million in brightness) over a period of months or years. The novae with
the largest outburst amplitudes fade the fastest. Nova outbursts are due to thermonuclear runaways of the hydrogen-rich material
that has accreted onto the white dwarf. Most novae known have only been observed to undergo one nova outburst, but several
are recurrent novae. The recurrent nova T Pyx is particularly interesting: it was recently
observed to possess collimated jets emanating from the central regions of the accretion disc (Shahbaz et al. 1997, Astrophysical
Journal, 484, 59). It is the first CV system in which jets have been observed.
Another group of non-magnetic CVs is the dwarf novae. Their outbursts are not quite
as spectacular as those of the novae (in outburst, dwarf novae are a mere factor of 6 to 100 brighter than in quiescence),
but the outbursts occur more often. In general, the more frequent the outburst, the smaller the amplitude of outburst. Examples
of the extremes of the dwarf nova phenomenon are V1159 Ori, which has outbursts once every four days with an amplitude of
about two magnitudes; and WZ Sge, which shows outbursts once every thirty years, and the outburst ampitude rivals that of
a classical nova outburst. Both these extremes of behaviour occur in the dwarf novae with the shortest orbital periods. A
dwarf nova outburst is thought to be a large release of gravitational potential energy caused by a temporary enhancement of
the rate of mass transfer through the disc.
An interesting subset of the dwarf novae are the SU Ursa Majoris stars. They show two
distinct kinds of outburst: normal dwarf nova outbursts, and superoutbursts, which last 5-10 times longer and are slightly
brighter than the usual dwarf nova outbursts. Most superoutbursting dwarf novae have short orbital periods (less than 2 hours).
Some of the SU UMa stars with long outburst intervals show interesting rebrightenings on the way back to quiescence after
During a superoutburst, a SU UMa star shows an additional modulation of the light curve, a superhump, which is caused by
precession of the accretion disc. Superhumps show up in the light curve as a modulation with a period slightly longer (a few
percent) than the orbital period.
Another group of CVs with accretion discs are the nova-like variables. The difference
between the nova-like variables and dwarf novae is that nova-like variables don't undergo dwarf nova outbursts. This is because
the rate of transfer of matter in their discs is (by-and-large) stable, and the overall brightness varies only slightly about
its mean level. In addition, the rate of mass transfer in the discs of nova-like variables is much higher than that in quiescent
dwarf novae, and the accretion discs are thus very bright.
The nova-likes with the highest mass transfer rates are the SW Sextantis
stars. This intriguing group of stars has many unusual properties. For example, the emission lines in their spectra
behave in an unexpected way. If you observe a CV system edge-on, so that your line of sight is along the plane of the disc,
you expect to see double-peaked emission lines. This is because the gas in one half of the disc is moving towards you, and
its emission is thus blue-shifted, while the gas in the other half is moving away from you, and its emission is red-shifted.
Despite being edge-on systems, the SW Sextantis stars have single-peaked emission lines. For more details of the other weird
properties of SW Sextantis stars, and some recent models that explain them, see e.g. Hellier (1996, Astrophysical Journal,
471, 949) and Dhillon, Marsh & Jones (1997, Monthly Notices of the Royal Astronomical Society).
Some nova-like variables show superhumps, like those seen in the SU UMa stars in superoutburst. Unlike the superoutbursting
systems, however, these nova-likes have superhumps in their light curves all the time. They are the permanent
superhumpers. An interesting group of CV systems that show permanent superhumps is the AM
Canum Venaticorum stars. They are helium-rich CVs: no hydrogen has ever been detected in any of them. They have
much shorter orbital periods than the hydrogen-rich systems: the two stars of AM CVn itself whizz around each other every
quarter of an hour. It seems that the two crucial criteria for a system to show superhumps are the following: first, there
must be a high rate of mass transfer through the accretion disc; and secondly, the mass-donating red star must be much less
massive than the accreting white dwarf (its mass should be no more than a quarter of the white dwarf's mass). Interestingly,
the superhump phenomenon is not confined to CVs: they have also been observed in the weird neutron star binary SS433 and in black hole soft X-ray transients (see e.g. O'Donoghue & Charles,
Monthly Notices of the Royal Astronomical Society, 282, 191).
Magnetic CVs are systems where the white dwarf has an appreciable magnetic field (several
tens of millions of Gauss). Because the matter in the accretion stream is partially ionized, it can't form a disc (because
charged particles can't cross field lines, only spiral around them). Instead, the gas is threaded onto the field lines and
plunges straight down onto the magnetic poles of the white dwarf. Because the accretion occurs almost perpendicularly to the
surface of the white dwarf, copious X-ray and EUVB emission is liberated at the poles.
In systems where the magnetic field of the white dwarf is strong enough to synchronize the rotation of the white dwarf
with that of the binary, the red star and the white dwarf rotate essentially as rigid bodies. These are the polars,
or AM Herculis stars. The magnetic field in these stars is so strong (it's about fifty
million times the strength of the Earth's magnetic field) that no accretion disc can form.
One of the most distinctive properties of the light emitted by polars is that it is both linearly and circularly polarized
(this is one of the reasons why polars are called polars. The other reason is that the name was recommended by the Polish
astronomers Krzeminski & Serkowski). By observing changes in the intensity of polarized emission over time, we can discover
the magnetic field strength of the white dwarf and the magnetic field geometry.
In the intermediate polars (which are thought to have lower magnetic fields than the
polars), the spin period of the white dwarf is shorter than the orbital period. The accretion process in the intermediate
polars is through a disc with a disrupted inner radius (where the magnetic field is powerful enough to influence the flow
of the gas), or an accretion stream as in the polars, or either, or both (just to prevent you from falling asleep). The intermediate
polars with the shortest spin periods are the DQ Herculis stars.
New CV systems are being discovered all the time. Each new member of the CV zoo presents us with new properties that add
to our knowledge of the CV phenomenon. From radio to TeV gamma rays, cataclysmic variables continue to provide fascinating
new areas of research.
Adopted from an article written by Kate Harrop-Allin, 22 March 1998