X-ray Binaries

X-ray binaries, systems containing an accreting neutron star or stellar-mass black hole in orbit with a less evolved companion, are sites of some of the highest-energy phenomena in the local universe. As well as being copious emitters of X-rays, in some cases extending to the $\gamma$ -ray regime ( $\geq$ several hundred keV), it has been increasingly realised in recent years that they are often strong sources of bright and variable radio synchrotron emission. Reviews can be found in Hjellming & Han (1995) and Mirabel & Rodriguez (1999).

Broadly speaking, the radio emission from X-ray binaries can be grouped into systems with bright, highly variable radio emission which can be resolved into jet-like outflows, and a population of weaker, persistently-emitting sources.

Relativistic Ejections from X-ray Transients

The most obviously spectacular aspect of the radio emission from X-ray binaries is the large flaring events which have been resolved into relativistic outflows. The first example of such an outflow was in fact the rather unique and anomalous case of SS 433, from which precessing jets propagate with a velocity which is widely accepted to be more or less constant at 0.26c (Margon 1984; Vermeulen 1989). The jets from this source have distorted and elongated the surrounding W50 radio nebula (Dubner et al. 1998), which may have begun its existence as the supernova remnant associated with the formation of the compact object in SS 433.

$\begin{figure}% latex2html id marker 4098 \centering \centerline{\epsfxsize=6.0t... ...nto the black hole, as revealed by simultaneous X-ray observations.}\end{figure}$

While remaing a unique and fascinating source, the radio jets from SS 433 have been rather overshadowed in the past five years by the discovery of highly relativistic jets from some bright black-hole candidate (BHC) X-ray binaries. Mirabel & Rodriguez (1994) reported the first observations of apparent superluminal motions in our Galaxy, from the X-ray transient GRS 1915+105, and inferred a true bulk velocity for the ejections of $\geq 0.9c$ . Within a year the second superluminal jet source, GRO J1655-40 was discovered, and again true bulk velocities of $\geq 0.9c$ were inferred (Tingay et al. 1995; Hjellming & Rupen 1995). In the past year a third (probable) superluminal jet source, XTE J1748-288, has been discovered in the Galactic centre (Rupen, Hjellming & Mioduszewski 1998), and it appears that maybe all transient BHC systems may produce relativistic outflows when in outburst (Kuulkers et al. 1999).

GRS 1915+105 has been a particularly exciting source for attempts to understand the connection between synchrotron-emitting ejections and variability in the accretion disc. As well as producing multiple episodes of relativistic ejections (Mirabel & Rodriguez 1994; Rodriguez & Mirabel 1999; Fender et al. 1999 - see Fig 2.23), the source displays radio oscillations with quasi-periods observed between 10 - 120 min (Pooley & Fender 1997). These oscillations, remarkably, appear to have a flat spectrum extending to the near infrared (2 $\mu$ m - Fender et al. 1997; Eikenberry et al. 1998; Mirabel et al. 1998; Fender & Pooley 1998). These oscillations have been observed to be directly coupled with X-ray dips observed with the XTE satellite and modelled as the disappearance of the inner few 100 km of the accretion disc around the black hole (Belloni et al. 1997; Pooley & Fender 1997; Fender et al. 1997; Eikenberry et al. 1998; Mirabel et al. 1998). Hence for the first time coupling of high time-resolution radio, infrared and X-ray observations have revealed the repeated ejection and refill of the inner accretion disc around a black hole.

It now seems probable that the majority, maybe all, transient BHCs produce jets (of eight newly-discovered BHC transients in 1998-99, seven had reported radio counterparts). These sources often reach flux densities of $\geq 50$ mJy; in exceptional cases $\geq 1$ Jy. In general it seems that neutron star transients, while occasionally radio-emitters (e.g. Aql X-1, Cen X-4; see Hjellming & Han 1995) are fainter (typically $\leq 50$ mJy) than the BHCs. Whether this is because they are less efficient at particle ejection and/or matter ejection remains to be seen.

The Faint Persistent Population

Of the $\sim 200$ X-ray binary systems known, around 20% have been detected as radio sources; often transient in nature (Hjellming & Han 1995). However, there exist a dozen or so systems which are persistently detected at low ( $\leq 20$ mJy) levels. This handful of sources comprise the four persistent BHCs in our Galaxy, the six `Z-sources', containing low magnetic field neutron stars accreting at or near the Eddington limit, and the unusual `Atoll-source' GX 13+1. These systems are the brightest persistent X-ray sources in our Galaxy, and the best examples of compact objects steadily accreting at a significant fraction of the Eddington rate.

**Figure 2.24:** The radio luminosity of the persistently accreting black holes and neutron stars in our Galaxy. With the sensitivity of SKA, this plot could be extended to the entire local group of galaxies.
$\begin{figure}\centering \centerline{\epsfxsize=14.0truecm \epsfbox{stars/xray/newrd.ps} \message{Figure newrd.ps}} \end{figure}$

Fender & Hendry (1999) have found that these systems share a common mean radio luminosity at cm wavelengths, to within an order of magnitude. This luminosity is significantly higher than that inferred from the other classes of persistent X-ray sources, the `Atoll-sources' (which appear to be accreting at too low a rate in general), and the X-ray pulsars (in which we believe truncation of the accretion disc by the strong magnetic field of the neutron star inhibits formation of a jet). The luminosity function for the BHCs and Z sources is $S_{\nu {\rm , cm}} \sim 65/{d^2}$ mJy, where d is the distance to the source in kpc. Fig. 2.24 illustrates this relation and the detections and upper limits established for all persistent X-ray binaries for which there are good distance estimates. These sources, while less spectacular than the jet-producing transients, are our best chance to understand the process of steady accretion, particle acceleration and outflow around a neutron star or stellar-mass black hole.

The potential of the SKA

The SKA will revolutionise the study of jets, both transient and persistent, from X-ray binaries. With sensitivity better than the VLA and angular resolution at least as good as MERLIN, it will be possible to rapidly establish whether collimated, relativistic outflows from X-ray transients are the norm or a spectacular but very rare fireworks display. With typically 5 - 10 bright radio sources associated with X-ray binaries `on' at any one time across half of the sky, it would be possible to make high S/N multi-frequency monitoring observations daily. For sources which will be resolved on the longest baselines (including SS 433, GRS 1915+105 and possibly a large fraction of new transients) the snapshots available from such a monitoring program would allow a previously unimaginable continuous sequence of observations of jet structure which evolves daily. Within a few years of the completion of the SKA we could imagine more than doubling the sample of measured jet proper motions.

The SKA will also allow us to probe the populations of X-ray binaries in the entire local group of galaxies. Within a few hr we could detect the `faint' source population in the Magellanic Clouds, and within 8 hr in M31. One critical question to be addressed by such observations would be the ratio of accretion to ejection power, and its dependence upon environment and metallicity of the accreted matter (for example the brightest X-ray sources in the LMC appear to be systematically more luminous than their analogs in our Galaxy, possibly due to the lower metallicity).

The key features of the SKA that will realise the goals discussed above, and to truly advance our knowledge of the physics and populations of stellar-mass relativistic jet sources are: