Abstract.The classical vacuum gap model of Ruderman & Sutherland, in which spark-associated sub-beams of subpulse emission circulate around the magnetic axis due to the E × B drift of spark plasma filaments, provides a natural and plausible physical mechanism explaining the subpulse drift phenomenon. Moreover, this is the only model with quantitative predictions that can be compared with observations. Recent progress in the analysis of drifting subpulses in pulsars has provided a strong support for this model by revealing a number of sub-beams circulating around the magnetic axis in a manner compatible with theoretical predictions. However, a more detailed analysis revealed that the circulation speed in a pure vacuum gap is too high when compared with observations. Moreover, some pulsars demonstrate significant time variations in the drift rate, including a change of the apparent drift direction, which is obviously inconsistent with the E × B drift scenario in a pure vacuum gap. We attempted to resolve these discrepancies by considering a partial flow of iron ions from the positively charged polar cap, coexisting with the production of outflowing electron-positron plasmas. The model of such a charge-depleted acceleration region is highly sensitive to both the critical ion temperature T i ∼ 10 6 K (above which ions flow freely with the corotational charge density) and the actual surface temperature T s of the polar cap, heated by the bombardment of ultra-relativistic charged particles. By fitting the observationally deduced drift-rates to the theoretical values, we managed to estimate polar cap surface temperatures in a number of pulsars. The estimated surface temperatures T s correspond to a small charge depletion of the order of a few percent of the Goldreich-Julian corotational charge density. Nevertheless, the remaining acceleration potential drop is high enough to discharge through a system of sparks, cycling on and off on natural time-scales described by the Ruderman & Sutherland model. We also argue that if the thermionic electron outflow from the surface of a negatively charged polar cap is slightly below the Goldreich-Julian density, then the resulting small charge depletion will have similar consequences as in the case of the ions outflow. We thus believe that the sparking discharge of a partially shielded acceleration potential drop occurs in all pulsars, with both positively ("pulsars") and negatively ("anti-pulsars") charged polar caps.

We propose a new, self-consistent theory of coherent pulsar radio emission based on the non-stationary sparking model of Ruderman & Sutherland (1975), modified by Gil & Sendyk (2000) in the accompanying Paper I. According to these authors, the polar cap ( with a radius r p ≃ 10 4 P −0.5 cm ) is populated by about (r p /h) 2 sparks of a characteristic perpendicular dimension D approximately equal to the polar gap height scale h ∼ 5 × 10 3 P 3/7 cm, separated from each other also by about h. Each spark reappears in approximately the same place on the polar cap for a time scale much longer than its 10 µs life-time and delivers to the open magnetosphere a sequence of e − e + clouds which flow orderly along a flux tube of dipolar magnetic field lines. The overlapping of particles with different momenta from consecutive clouds leads to effective two-stream instability, which triggers electrostatic Langmuir waves at the altitudes of about 50 stellar radii. This is the only known instability which can develop at the low altitudes, where the observed pulsar radio emission originates. The electrostatic oscillations are modulationally unstable and their nonlinear evolution results in formation of "bunchlike" charged solitons. A characteristic soliton length along magnetic field lines is about 30 cm, so they are capable of emitting coherent curvature radiation at radio wavelengths. A perpendicular cross-section of each soliton at radiation altitudes follows from a dipolar spread of a plasma cloud with a characteristic dimension near the star surface of about D ≈ h ≈ 50 meters. The net soliton charge is about 10 21 fundamental charges, contained within a volume of about 10 14 cm 3 . For a typical pulsar, there are about 10 5 solitons associated with each of about 25 sparks operating on the polar cap at any instant. One soliton moving relativisticaly along dipolar field lines with a Lorentz factor of the order of 100 generates a power of about 10 21 erg/s by means of curvature radiation. Then the total power of a typical radio pulsar can be estimated as being about 10 27−28 erg/s. The energy of the soliton curvature radiation is supported by kinetic energy of secondary electron-positron plasma created by the primary beam produced by the accelerating potential drop within the polar gap. A significant fraction of kinetic energy generated by sparks is radiated away in form of the observed coherent radio emission.

A non-stationary polar gap model first proposed by Ruderman & Sutherland (1975) is modified and applied to spark-associated pulsar emission at radio wave-lengths. It is argued that under physical and geometrical conditions prevailing above pulsar polar cap, highly non-stationary spark discharges do not occur at random positions. Instead, sparks should tend to operate in well determined preferred regions. At any instant the polar cap is populated as densely as possible with a number of two-dimensional sparks with a characteristic dimension as well as a typical distance between adjacent sparks being about the polar gap height. Our model differs, however, markedly from its original 'hollow cone' version. The key feature is the quasi-central spark driven by pair production process and anchored to the local pole of a sunspot-like surface magnetic field. This fixed spark prevents the motion of other sparks towards the pole, restricting it to slow circumferential drift across the planes of field lines converging at the local pole. We argue that the polar spark constitutes the core pulsar emission, and that the annular rings of drifting sparks contribute to conal components of the pulsar beam. We found that the number of nested cones in the beam of typical pulsar should not excced three; a number also found by Mitra & Deshpande (1999) using a completely different analysis.Comment: 31 pages, 8 figures, accepted by Ap

We consider the curvature radiation of the point-like charge moving relativistically along curved magnetic field lines through a pulsar magnetospheric electron-positron plasma. We demonstrate that the radiation power is largely suppressed as compared with the vacuum case, but still at a considerable level, high enough to explain the observed pulsar luminosities. The emitted radiation is polarized perpendicularly to the plane of the curved magnetic filed lines coincides with that of extraordinary waves, which can freely escape from the magnetospheric plasma. Our results strongly support the coherent curvature radiation by the spark-associated solitons as a plausible mechanism of pulsar radio emission.