Magnetic structure of pulsar winds

The expectation has long been that pulsar winds (with extension to stellar winds in some cases) are energetically dominated by their magnetic fields. Coroniti has recently argued that reconnection could transfer energy from magnetic reversals in the wind into particle energy. This process could have observational consequences in binary systems such as PSR 1957+20 because the wind properties could change before interacting with the companion. We show here that the reconnection process is simply inductive heating, which allows it to be calculated without appeal to existing phenomenology surrounding reconnection. Surprisingly, we find that the resultant wind will not necessarily shock, because it is not super-Alfvenic, but rather may simply decelerate smoothly to match nebular boundary conditions. Moreover, the two oppositely magnetized hemispheres are not casually disconnected, and the flow should then relax meridionally which should lead to formation of an equatorial neutral sheet. Such a neutral sheet will not dissipate the field inductively (unlike the sheets perpendicular to the flow), but may indeed be unstable to reconnection. A rather simple picture for magnetization of the wind emerges in which plasma is 'frozen' into the large-amplitude electromagnetic waves generated by the orthogonal magnetic dipole component of a rotating magnetic (neutron) star. As the plasma is convected away, it also pulls with it the aligned dipole magnetic field lines beyond the wind zone ('light cylinder'). At large distances from the star, the wave component is dissipated, leaving behind the wound-up field lines of the aligned component. It will be important to explore the torques exerted by these two components to eventually understand why pulsar magnetic fields do not seem to asympotically achieve alignment parallel or orthogonal to the spin axis.