Valery I. Altunin

Giant pulses from PSR B1937+21 with widths ≤15 nanoseconds and Tb ≥ 5 × 1039 K, the highest brightness temperature observed in the universe

Giant radio pulses of the millisecond pulsar B1937+21 were recorded with the S2 VLBI system at 1.65 GHz with NASA/JPLs 70 m radio telescope at Tidbinbilla, Australia. These pulses have been observed as strong as 65,000 Jy with widths ≤15 ns, corresponding to a brightness temperature of Tb ≥ 5 × 1039 K, the highest observed in the universe. The vast majority of these pulses occur in 5.8 and 8.2 μs windows at the very trailing edges of the regular main pulse and interpulse profiles, respectively. Giant pulses occur, in general, with a single spike. Only in one case of 309 was the structure clearly more complex. The cumulative distribution is fitted by a power law with index -1.40 ± 0.01 with a low-energy but no high-energy cutoff. We estimate that giant pulses occur frequently but are only rarely detected. When corrected for the directivity factor, 25 giant pulses are estimated to be generated in one neutron star revolution alone. The intensities of the giant pulses of the main pulses and interpulses are not correlated with each other nor with the intensities or energies of the main pulses and interpulses themselves. Their radiation energy density can exceed 300 times the plasma energy density at the surface of the neutron star and can even exceed the magnetic field energy density at that surface. We therefore do not think that the generation of giant pulses is linked to the plasma mechanisms in the magnetosphere. Instead we suggest that it is directly related to discharges in the polar cap region of the pulsar.

Pulsar microstructure and its quasi-periodicities with the S2 VLBI system at a resolution of 62.5 nanoseconds

We report a study of microstructure and its quasi-periodicities of three pulsars at 1.65 GHz with the S2 VLBI system at a resolution of 62.5 ns, by far the highest for any such statistical study yet. For PSR B1929+10 we found in the average cross-correlation function (CCF) broad microstructure with a characteristic timescale of 95 ± 10 µs and confirmed microstructure with characteristic timescales between 100 and 450 µs for PSRs B0950+08 and B1133+16. On a finer scale PSRs B0950+08, B1133+16 (component II) and B1929+10 show narrow microstructure with a characteristic timescale in the CCFs of ∼10 µs, the shortest found in the average CCF or autocorrelation function (ACF) for any pulsar, apart perhaps for the Crab pulsar. Histograms of microstructure widths are skewed heavily toward shorter timescales but display a sharp cutoff. The shortest micropulses have widths between 2 and 7 µs. There is some indication that the timescales of the broad, narrow, and shortest micropulses are, at least partly, related to the widths of the components of the integrated profiles and the subpulse widths. If the shortest micropulses observed are indeed due to beaming then the ratio, γ, of the relativistic energy of the emitting particles to the rest energy is about 20000, independent of the pulsar period. We predict an observable lower limit for the width of micropulses from these pulsars at 1.65 GHz of 0.5 µs. If the short micropulses are instead interpreted as a radial modulation of the radiation pattern, then the associated emitting sources have dimensions of about 3 km in the observers frame. For PSRs B0950+08 and B1133+16 (both components) the micropulses had a residual dispersion delay over a 16-MHz frequency difference of ∼2 µs when compared to that of average pulse profiles over a much larger relative and absolute frequency range. This residual delay is likely the result of propagation effects in the pulsar magnetosphere that contribute to limiting the width of micropulses. No nanopulses or unresolved pulse spikes were detected. Cross-power spectra of single pulses show a large range of complexity with single spectral features representing classic quasi-periodicities and broad and overlapping features with essentially no periodicities at all. Significant differences were found for the two components of PSR B1133+16 in every aspect of our statistical analysis of micropulses and their quasi-periodicities. Asymmetries in the magnetosphere and the hollow cone of emission above the polar cap of the neutron star may be responsible for these differences. the observed intensity fluctuations can be either caused by a longitudinal modulation of the radiation pattern over the cross section of the polar magnetic field lines or by a radial modula- tion of the radiation pattern along the opening polar magnetic field lines, or, again, by a combination of both. The longitudi- nal modulation is most likely related to the stationary geome- try of the emission beam fixed to each of the poles of the ro- tating neutron star. The radial modulation is likely related to plasma bunching and linked to the elementary emission mech- anism. In this model the spectrum of the radio emission is a function of the radial distance from the neutron star, and the beam width is frequency dependent. High frequency radiation is emitted closer to the neutron star and the beam is narrower, low frequency radiation is emitted further out and the beam is broader, reflecting the opening of the polar magnetic field

Microstructure of pulsar radio pulses measured with a time resolution of 62.5 ns at 1650 MHz

We present an analysis of pulsar observations carried out on two frequency channels at 1634 MHz and 1650 MHz with a time resolution of 62.5 ns on the 70-m radio telescope of the NASA Deep Space Network in Tidbinbilla. The data were recorded using the S2 system, intended primarily for VLBI observations. Microstructure with characteristic timescales of 270, 80, and 150 µs was detected in pulsars B0833-45, B1749-28, and B1933 + 16, respectively. The distribution of microstructure timescales for the Vela pulsar (B0833-45) is characterized by a gradual growth with decreasing timescale to 200 µs; the distribution has a maximum at 20–200 µs and falls off sharply for timescales below 20 µs. The statistical relation between the microstructure modulation index m and the corresponding timescale τµ can be approximated by the power law dependence R∝τ⊙0.5; i.e., the intensity is higher for micropulses with longer durations. This contradicts the predictions of nonlinear models for the formation of micropulses by supercompact soliton wave packets. In all the pulsars studied, the time delays of the micropulses between the two frequency channels deviate from the expected dispersion laws for the interstellar plasma. In particular, the micropulses in the low-frequency channel arrive earlier than predicted by the dispersion measures derived previously from the mean pulse profiles. The deviation from the dispersion delay is determined most accurately for B0833-45, and is 4.9±0.2 µs. Such anomalous delays are probably associated with the effects of propagation of the radio emission within the pulsar magnetosphere.

VLBI Observations of Supernova 1993J: The First 1000 Days

VLBI images of supernova SN 1993J in M81 from 50 to 1000 d after shock breakout (Figure 1) show an expanding shell with an increasingly complex brightness distribution (see also Bartel et al. 1995, Marcaide et al., these Proceedings, p. 353, and references therein). The shock is decelerating, with the radius 0 oc f°-837±o.025_ co mbining this result with model fits to the radio light curve (Van Dyk et al. 1994) gives a power-law index for the circumstellar density profile (p oc R~) of 6 = 1 -69^^23 which is consistent with equipartition and supported by X-ray observations. Phase-referencing with respect to the nucleus of M81 (cf. Bietenholz et al. 1994, 1996, these Proceedings, p. 201) suggests that the center of the shell is offset slightly from the origin of the explosion, although the shell itself remains highly circularly symmetric (cf. Bartel et al. 1994a, b, Rupen et al. 1994). In the 8.4 GHz image from 17 December 1995 we find enhanced emission near the explosion center; however the corresponding 5 GHz image shows no such enhancement, and any central source if present at other times must be considerably fainter. It is intriguing that spectral index maps made from 5 and 8.4 GHz data show a consistently flatter spectrum towards the center than around the rim of the shell. Finally, data taken at 5 and 8.4 GHz on 10 May 1995 set limits to the linear polarization of 10% at both frequencies on scales of 1 mas (5 X 10cm), suggesting that either the magnetic field is tangled on such scales or the material along the line-of-sight depolarizes any initially-aligned emission.