Simultaneous measurements of giant pulses from the Crab pulsar were taken at two widely spaced frequencies, using the real-time detection of a giant pulse at 1.4 GHz at the Very Large Array to trigger the observation of that same pulse at 0.6 GHz at a 25 m telescope in Green Bank, WV. Interstellar dispersion of the signals provided the necessary time to communicate the trigger across the country via the Internet. About 70% of the pulses are seen at both 1.4 and 0.6 GHz, implying an emission mechanism bandwidth of at least 0.8 GHz at 1 GHz for pulse structure on timescales of 1 to 10 ks. The giant pulse spectral indices fall between [2.2 and [4.9, which may be compared to the average main pulse value for this pulsar of [3.0. The arrival times at both frequencies display a jitter of 100 ks within the window deÐned by the average main pulse proÐle and are tightly correlated. This tight correlation places limits on both the emission mechanism and the frequency-dependent propagation within the magnetosphere. At 1.4 GHz, the giant pulses are resolved into several closely spaced components. Simultaneous observations at 1.4 and 4.9 GHz show that the component splitting is frequency independent. We conclude that the multiplicity of components is intrinsic to the emission from the pulsar, and reject the hypothesis that this is the result of multiple imaging as the signal propagates through the perturbed thermal plasma in the surrounding nebula. At both 1.4 and 0.6 GHz, the pulses are characterized by a fast rise time and an exponential decay time that are correlated. At 0.6 GHz, the rise time is not resolved. The rise and fall times of the 1.4 GHz components vary from component to component and from pulse to pulse. The pulse broadening, with its exponential decay form, is most likely the result of multipath propagation in intervening ionized gas. These decay times, and that seen in contemporaneous 0.3 GHz average pulse data, are large compared to normal conditions for the Crab pulsar. The most likely location for the perturbed plasma is the interface region between the pulsar-driven synchrotron nebula and the overlying supernova ejecta.

We present evidence that the integrated proÐles of some millisecond pulsars exhibit severe changes that are inconsistent with the moding phenomenon as known from slowly rotating pulsars. We study these proÐle instabilities in particular for PSR J1022]1001 and show that they occur smoothly, exhibiting longer time constants than those associated with moding. In addition, the proÐle changes of this pulsar seem to be associated with a relatively narrowband variation of the pulse shape. Only parts of the integrated proÐle participate in this process, which suggests that the origin of this phenomenon is intrinsic to the pulsar magnetosphere and unrelated to the interstellar medium. A polarization study rules out proÐle changes due to geometrical e †ects produced by any sort of precession. However, changes are observed in the circularly polarized radiation component. In total we identify four recycled pulsars that also exhibit instabilities in the total power or polarization proÐles due to an unknown phenomenon (PSRs J1022]1001, J1730 [2304, B1821[24, and J2145[0750). The consequences for highprecision pulsar timing are discussed in view of the standard assumption that the integrated proÐles of millisecond pulsars are stable. As a result we present a new method to determine pulse times of arrival that involves an adjustment of relative component amplitudes of the template proÐle. Applying this method to PSR J1022]1001, we obtain an improved timing solution with a proper-motion measurement of [17^2 mas yr~1 in ecliptic longitude. Assuming a distance to the pulsar as inferred from the dispersion measure, this corresponds to a one-dimensional space velocity of 50 km s~1.

No abstract

No abstract