Interactions in the Double Pulsar System

The two pulsars, separated by less than 3 light-seconds, show strong evidence that the emission of each influences the observed radio waves from the other, permitting the investigation of many plasma physics phenomena. Here we summarize the various effects and their interpretations, and refer the reader to the original papers for full details.

2.1 Eclipses of A

One of the most astonishing features of the double-pulsar system is that its orbit is inclined so nearly parallel to our line of sight that the emission from A is actually extinguished in an eclipse for about 30 seconds when A passes behind B (Lyne et al., 2004). The eclipse is asymmetric in time, with a slow ingress and a steep egress, and shows only a weak dependence on observing frequency, being slightly longer in duration at low frequencies (Kaspi et al.,

2004). Detailed investigations of the flux density of A as a function of time show that the eclipse is in fact not smooth nor always total, but is strongly modulated at half the spin period of the B pulsar for most of its overall duration (McLaughlin et al., 2004c, see Figure 1). An elegantly simple geometric model has been put forth to explain these variations (Lyutikov & Thompson,

2005). In brief, B's plasma-filled magnetosphere is represented as dominated by a dipolar component, misaligned with B's spin axis. As B spins, the dipole field lines move in and out of a configuration in which they can intercept A's radio beam and extinguish it via a form of synchrotron absorption. This model leads to estimates of the angles B's spin axis makes to the orbital angular momentum and to its own dipole axis (Lyutikov & Thompson, 2005).

Orbital Phase (degrees)

Fig. 1. The pulsed flux density of A in arbitrary units, as a function of orbital phase (true anomaly re-referenced to the ascending node), and averaged over three orbits. The vertical dashed line indicate the arrival times of pulses from B. This plot clearly shows the overall dip in observed A flux due to the eclipse, as well as the modulation of the flux at twice the spin frequency of the B pulsar. Figure adapted from McLaughlin et al., (2004c).

Orbital Phase (degrees)

Fig. 1. The pulsed flux density of A in arbitrary units, as a function of orbital phase (true anomaly re-referenced to the ascending node), and averaged over three orbits. The vertical dashed line indicate the arrival times of pulses from B. This plot clearly shows the overall dip in observed A flux due to the eclipse, as well as the modulation of the flux at twice the spin frequency of the B pulsar. Figure adapted from McLaughlin et al., (2004c).

2.2 Orbital-phase and Long-term Modulation of B's Profile

The vast majority of pulsars show very stable "integrated profiles" when several hundred to several thousand pulses are summed. From the start, however, it was clear that B's average profile displayed very different shapes at different orbital phases; moreover the pulsar was only bright at two orbital phases as observed with the 64-m Parkes telescope (Lyne et al., 2004). More sensitive observations with the 100-m Green Bank Telescope (GBT) showed that B was in fact visible at all orbital phases, although with much reduced flux (Ransom et al., 2005). There are also severe timing systematics for B (Ransom et al., 2005), particularly during the two bright phases; in fact, in the most recent published timing solution, the brightest orbital phases are left out of the timing analysis altogether (Kramer et al., 2006). The orbital phase modulation is likely due to the effects of A's strong particle wind interacting with the magnetosphere of B. The simple dipole model used to explain the A eclipse (Lyutikov & Thompson, 2005), and distorted based on the geometry of the Earth's magnetosphere, has been able to replicate approximately the orbital phases at which B is bright and dim (Lyutikov, 2005).

B also shows strong profile shape evolution over secular (long-term) timescales (Burgay et al., 2005), with the two bright phases each shrinking in duration, and the average profile shapes also changing. This is likely interpretable as being due to geodetic precession (e.g., Barker & O'Connell, 1975; Weisberg et al., 1989). After the second supernova explosion, it is likely that B's spin axis was misaligned with the total angular momentum vector (this is further supported by the dipole model of the modulation of A's eclipse (Lyutikov & Thompson, 2005)), and its precession about the total angular momentum vector would naturally lead to different observed pulse shapes with time (Burgay et al., 2005). It is however difficult to completely rule out a contribution to the long-term profile evolution from A's influence (Burgay et al., 2005). The combined orbital-phase-dependent and long-term profile evolutions necessitate a detailed matrix of reference profiles in timing B (Kramer et al., 2006).

Interestingly, no long-term evolution has yet been seen in A's pulse profile (Manchester et al., 2005; Kramer et al., 2006), although this pulsar might also reasonably be expected to have its spin axis misaligned with the total angular momentum and hence to undergo geodetic precession as well. Unless we happen to have caught the pulsar at a special phase of the precession cycle (Manchester et al., 2005), it is likely that the spin-orbit misalignment is actually small for this pulsar. This would be consistent with recent modeling of the properties of the supernova explosion that created B (e.g. Willems et al., 2006; Piran & Shaviv, 2005; Stairs et al., 2006).

2.3 Short-term Modulation of B's Profile

The profile variation of B on orbital timescales contains a further short-term direct contribution from the interaction with A: over a small range of orbital phase, individual pulses of B show a "drifting subpulse" phenomenon in which a strong pulse feature migrates earlier in time with each successive B pulse (McLaughlin et al., 2004b). The drift timescale corresponds exactly to the beat frequency of the A and B pulses. These modulations are seen only near one of the orbital phases at which the line between A and B is in the plane of the sky. The physics behind the modulation is presently unclear, but the interaction must be induced by the radiation at the 44 Hz spin frequency of A (McLaughlin et al., 2004b).

2.4 Unpulsed Emission

A comparison of the initial radio observations of J0737—3039A/B with the Parkes single-dish telescope and the Australia Telescope Compact Array interferometer suggested that the combined flux densities of the pulsed emission from A and B could not account for the full flux density observed with the interferometer (Burgay et al., 2003), and that there might therefore be significant unpulsed radio emission, perhaps arising in the region where the winds of the two pulsars collide (Lyne et al., 2004). This unpulsed emission has however not been seen in follow-up observations with the Very Large Array (Chatterjee et al., 2005) and could therefore have been spurious or due to a problem in the data analysis.

Meanwhile, there is excellent evidence for X-ray emission from the system (McLaughlin et al., 2004a; Pellizzoni et al., 2004; Campana et al., 2004). While several models have proposed to account for the existence of X-ray emission (e.g., Zhang & Loeb, 2004; Granot & Meszaros, 2004), the origin of the observed flux is unclear: in particular, there is no evidence yet for modulation at either the orbital period or either of the two pulsar spin periods. More sensitive observations should soon yield a solution to this problem; meanwhile we note that X-ray emissions from the DNS PSR B1534+12 show modulation with orbital phase and likely arise from the interaction of the (recycled) pulsar's wind with that of its companion (Kargaltsev et al., 2006).

Was this article helpful?

0 0

Post a comment