Astron. Astrophys. 323, 395-398 (1997)

4. Conclusion and discussion

We have shown that if the different emission heights of the core and conal components are considered, there should be a phase shift between their beam centers. That will cause depolarization and position-angle jump(s) in some longitude range of the integrated profile. If the relative phase shifts of three beam centers are investigated (core, inner cone and outer cone), the same conclusion could be obtained, but the simulations and observed situations will be more complicated as the superposition of three, instead of two, beam components are involved.

As subpulses are thought to be emitted from a small emission region of a given beam component (core or cone), then if one observes the polarization of a subpulse around the so-called "singular points", the seperated polarization modes should be obtained. On the one side, the core beam emission dominates, on the other side, the conal emission dominates. The observed position angle of subpulses should more or less concentrate around the dashed position angle curves of the two beam components in Fig. 3b. Though perhaps this is not the orthogonal modes observed over a much larger longitude range (Stinebring et al. 1984a, b); however, due to the relative shift of beam centers, possible non-orthogonal polarization modes should appear over a remarkable longitude range. At a given longitude the position angle of a subpulse can be one of two perpendicular or nearly perpendicular states. The existence of both modes at the same time or rapid transitions between them will also lead to a reduction in the percentage polarization all the way down to nearly zero, as seen from the simulations above.

If only the first condition in Eq. (6) is approximately satisfied at some longitude, that region of the profile can be largely depolarized, but the PA jumps will not be as the linear polarization of the integrated pulse cannot be reduced exactly to zero.

What we have considered in this paper is the longitudinal shift of pulsar beam centers caused by retardational delay, however, it seems quite possible that there are other geometrical factors which also produce latitudinal shifts of the beem centers. The different maximum sweep rates of the position-angle curves of the core and conal components seperated by Rankin (1988) imply different impact angles for these beam components, which further suggest latitude shifts of these beam centers.

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de