Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 321, 513-518 (1997)

Previous Section Next Section Title Page Table of Contents

5. Conclusions

The results of our monitoring of four pulsars at frequencies near 30 GHz indicate that strong modulations of the flux-density exist at those frequencies. The modulations ranging from 20-100% of the average flux density come as a surprise in a region where the ISM is thought to be inactive and the pulsars are fading in intensity.

While we can rule out instrumental effects for the observed flux modulations, we cannot completely exclude a propagation effect due to the ionized interstellar medium. However, the observed characteristics are inconsistent with the expected weak interstellar scintillations and, though an unexpected interstellar propagation effect cannot be ruled out, we consider propagation effects as unlikely. Therefore, we are left with the conclusion that the observed large flux variations on time scales of ten to twenty minutes are most likely of intrinsic rather than extrinsic origin. We suggest that they are inherited from the emission process and might be due to a loss of coherence also indicated by other radiation properties at mm-wavelengths (see X96, Kramer & Xilouris 1996). Actually, single pulse measurements revealed an increasing pulse-to-pulse fluctuation towards high radio frequencies (1 to 8 GHz) for some pulsars (Bartel et al. 1980, hereafter B80). Assuming that single pulse observations were possible at 32 GHz, an interesting comparison follows. We have extrapolated the strength of the fluctuations measured by B80 to 32 GHz. The values obtained are as large as [FORMULA] (B0329+54), [FORMULA] (B0355+54), [FORMULA] (B1929+10), and [FORMULA] (B2021+51). These values are even larger than the indices presented in this work. However, the latter refer to 5 min averages instead of single pulses. Smoothing simulated sequences of single pulses exhibiting the modulation index values extrapolated from B80, to our five minutes sub-integrations, we find that the observed modulation indices are still by factors of 4 to 8 higher than expected. The tendency reported by B80 of increasing modulation after a critical frequency usually located above 1 GHz, persists at mm-wavelengths and even becomes more severe than suggested by the power law determined by B80. B80 has attributed the observed erratic nature of pulsar emission to a progressive loss of coherency with frequency, which has also been suggested by K96 as the reason for the observed apparent spectral turn-up at mm-wavelengths.

Even though the location of the emission region in pulsar magnetospheres is somewhat uncertain, most current studies (e.g. Cordes 1978, Blaskiewicz et al. 1991, Phillips 1992, X96, Kramer et al. in press) seem to indicate that the radiation is created at a distance of a few percent of the light cylinder radius in a stratified mode with higher frequencies closer to the pulsar surface (radius-to-frequency mapping, RFM). Progressively increasing plasma densities and higher magnetic fields encountered close to the stellar surface might change the physical conditions in the regions that are responsible for the high radio frequency radiation. While at low frequencies the RFM is prominent, it has been shown that towards high frequencies RFM reaches a saturation and that the emission is essentially radiated from the same magnetospheric region (X96). Changes in the environment of the emission could be reflected in the modulations that we observe.

The same changes in the environment of the emission could also be responsible for the earlier reported spectral turn-up (K96). An association between these two phenomena is naturally tempting. Perhaps, the unusual spectral behaviour of B1929+10 and B2021+51 could be explained by assuming that these pulsars were observed at maximum pulsar activity. However all our observing sessions at different epochs lasted longer than an hour. Therefore, the resulting mean value was averaged over several independent time scales of the variation. Also, within the measurement uncertainties, the average flux densities taken at different epochs appear consistent. Therefore, effects of the short time variations reported here are largely suppressed from the average spectra presented by K96. Concluding, the observed flux variations cannot account for the unusual spectral behaviour earlier reported by K96 although both phenomena may well have a common physical origin.

Summarizing, we have observed unexpected flux density variations at mm-wavelengths which seem to be intrinsic to the pulsar radio emission mechanism. Further observations from 10 to 30 GHz will be needed to answer the yet open questions about their origin. A detailed study of the frequency dependence of this new variability appears to be most promising.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 30, 1998