Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 342, L57-L61 (1999)

Previous Section Next Section Title Page Table of Contents

3. Discussion

Based on a simple propagational scenario for natural polarization modes in pulsar magnetospheres we were able to show that for the available data set of three pulsars, theory and observations are in very good agreement. In these objects the polarization changes smoothly from linear to circular with increasing frequency. The principal theoretical [FORMULA]-functional dependence of of this change, expressed in Eq. (2) is independent of all pulsar parameters (as long as a mono-energetic background plasma is concerned). The individual pulsar parameters enter Eq. (2) only as scaling factors, hence yielding a parallel displacement of the function. Thus the comparison of the calculated and observed functional dependence is a strong test for the theory. As shown in Sect. 2 the rate of change of R with frequency is consistent with Eq. (2).

We note that for all three pulsars the same combinations of [FORMULA] and PLR result into quantitative agreements between the measured points and the theoretical curve, although these objects are very different in their parameters P and [FORMULA]. Obviously not only the [FORMULA] dependence is in agreement with the observation, but also the scaling of the function as it is defined through the prefactor in Eq. (2) for these three pulsars. This scaling coincidence is remarkable but does not necessarily mean, that all pulsars have to scale in the same manner.

Of course the number of pulsars which clearly show this effect is too small to make a statistically rigorous statement. This is especially true as it cannot be predicted at which frequency this effect should occur and hence for which other pulsars we should have observed it. We emphasize that it is rather difficult to observe such a frequency behavior of polarization since it requires a significant increase of the circular polarization to result into some decrease of the linear because they are added in quadrature.

Pulsars exhibit steep radio spectra, a fact which reduces the total number of available profiles in full polarization to about 100 at 4.9 GHz with many of them at a fairly low signal-to-noise ratio. As pulsars are also known to depolarize towards high frequencies, the signal-to-noise-ratio of the polarized emission is often very low, hardly allowing a credible determination of the ratio between linear and circular polarization. Nevertheless, it has been shown at low frequencies that the linear polarization reduces with frequency while such a trend is not evident for circular polarization (e.g. Han et al. 1998). This result qualitatively fits into the same trend of changing the ratio between the two polarizations. Taking all points mentioned above into account we are not surprised that only three pulsars have been found yet which clearly show this effect and allow an opportunity to test the theory.

In this paper we have used the frequency dependence of integrated profiles. These integrated profiles of course differ strongly from the individual pulses. It has been shown e.g. by Stinebring et al. (1984), that these individual pulses are often highly circularly polarized also at low frequencies. This is not necessarily in contradiction to the ideas described in this paper as the conditions within the magnetosphere are known to be highly non-stationary and unstable. The location at which the radiation decouples from the plasma (PLR) might therefore vary strongly from pulse to pulse. However, the integrated profile represents an average over these individual pulses, corresponding to an average PLR. A crucial test whether the described effect is responsible for the circular polarization would be the simultaneous observation of single pulses in full polarization over large frequency intervals.

Summarizing, we may say that the observations of the three objects support the theoretical approach that the polarization characteristics - and especially the high frequency occurrence of strong circular polarization - of at least these three pulsars are explainable in terms of propagating natural wave modes in the magnetosphere. Thus, it may be worth to follow that line of thought from the theoretical point of view and to try to get some more high quality data of the general polarization properties at low and high radio frequencies.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: February 23, 1999