The topology and polarization of sub-beams associated with the `drifting' sub-pulse emission of pulsar B0943+10 - I. Analysis of Arecibo 430- and 111-MHz observations

Abstract

The `drifting' sub-pulses exhibited by some radio pulsars have fascinated both observers and theorists for 30 years, and have been widely regarded as one of the most critical and potentially insightful aspects of their emission. Moreover, Ruderman & Sutherland, in their classic model, suggested that such regular modulation was produced by a system of sub-beams, rotating around the magnetic axis under the action of E×B drift. Such `drift' sequences have thus been thoroughly studied in a number of pulsars, but it has proven difficult to verify the rotating sub-beam hypothesis, and thus to establish an illuminating connection between the phenomenon and the actual physics of the emission. Here, we report on detailed studies of pulsar B0943+10, whose nearly coherent sequences of `drifting' sub-pulses have permitted us to identify their origin as a system of sub-beams that appear to circulate around the magnetic axis of the star. We introduce several new techniques of analysis, and we find that both the primary and secondary features in the fluctuation spectra of the star are aliases of their actual values. We have also developed a method of tracing the underlying pattern responsible for the observed sequences, using a `cartographic' transform and its inverse, permitting us to study the characteristics of the polar emission `map' and to confirm that such a `map' in turn represents the observed sequence. We apply these techniques to the study of three different Arecibo observations: a 1992 430-MHz sequence which includes a transition from the highly organized `B' profile mode of the star to its disorganized `Q' mode; a 1972 430-MHz `B'-mode sequence; and a 1990 111-MHz `B'-mode sequence. The `B'-mode sequences are consistent in revealing that the emission pattern consists of 20 sub-beams, which rotate around the magnetic axis in about 37 periods or 41s. Even in the `Q'-mode sequence, we find evidence of a compatible circulation time. The similarity of the sub-beam patterns at different radio frequencies strongly suggests that the radiation is produced within a set of columns, which extend from close to the stellar surface up through the emission region and reflect some manner of `seeding' phenomenon at their base. The sub-beam emission is then tied neither to the stellar surface nor to the field. While the origin of the `memory' responsible for the stability of the pattern over several circulation times is unknown, the hollow conical form of the average pattern is almost certainly the origin of the conal beam forms observed in most pulsars.