An empirical model for the radio emission from pulsars

A pulsar model is proposed which involves the entire magnetosphere in the production of the observed coherent radio emission. The observationally inferred regularity of peaks in the pulsar profiles of ‘slow’ pulsars is shown to suggest that inner and outer cones of emission near the polar cap interact with and ‘mirror’ two rings in the outer magnetosphere: one where the null line intersects the light cylinder, and another where it intersects the boundary of the corotating dead zone. The observed dependency of conal type on period is shown to follow naturally from the assumption that cones only form when the mirror intersection points lie between two fixed heights from the surface, suggesting that a feedback system exists between the surface and the mirror points, accomplished by a flow of charges of opposite sign in either direction. In their flow to and from the mirror points, the particles execute an azimuthal drift around the magnetic pole, thereby creating a ring of discrete ‘emission nodes’ close to the surface. Motion of the nodes is observed as subpulse ‘drift’, which is interpreted here as a small residual component of the real particle drift. The nodes can move in either direction or even remain stationary, and can differ in the inner and outer cones. A precise fit is found for the drifting subpulses of PSR 0943+10. Azimuthal interactions between different regions of the magnetosphere depend on the angle between the magnetic and rotation axes and influence the conal type, as observed. The model sees ‘slow’ pulsars as being at the end of an evolutionary development where the outer gap region no longer produces pair cascades, but is still the intermittent source of low-energy pairs in a magnetosphere-wide feedback system.