4. Discussion and Conclusions
Using the value, deduced above, we can deduce the opening angle of the emission cone if the geometry of the pulsar is known, i.e. the relative orientation of the magnetic and spin axes to the observer's line-of-sight (e.g. Gil et al. 1984). Since the geometry of B0355+54 can be deduced from polarization data (Lyne & Manchester 1988, Rankin 1993), we find using the geometry published by Rankin(1993) 1. Assuming dipolar field lines (for a justification of this assumption see Kramer et al. 1997a), this can in turn now be translated into an emission height of only km. We believe that this height may be underestimated, since a detection of the trailing part of the profile, visible at lower frequencies, could result in a larger emission height. It becomes nevertheless clear that the emission is created close to the pulsar surface. We note that the good alignment of the 87 GHz profile with lower frequency data excludes the existence of significant magnetic multipole field components at this altitude, which in any case would have to be extremely strong in order to have a noticeable influence at a distance of larger than three stellar radii above the surface.
Although the involved uncertainties are large, the measured 87 GHz flux density for PSR B0355+54 may already have some very interesting implications. Using the measured flux density,the observed pulse width, and the pulsar geometry deduced from polarization data (Lyne & Manchester 1988,Rankin 1993) we can estimate the luminosity emitted by the pulsar at 87 GHz. We obtain (cf. Kramer 1995) a luminosity of about erg/s in a bandwidth of 500 MHz. This luminosity can be compared to the energy output of proposed emission models such as curvature radiation which was an early favoured model to explain pulsar radio emission (e.g. Ruderman & Sutherland 1975). We note that based on this calculated luminosity, Lesch et al. (1997) have demonstrated that the emission observed at mm-wavelengths (i.e. at frequencies higher than 32 GHz) can indeed be generated by an incoherent superposition of the radiation from coherently emitting volumes.
In order to obtain reliable information about the actual shape of the pulse spectra, higher sensitivity measurements are necessary to reduce the current uncertainties. We may envisage the recording of two polarizations and an increase in bandwidth.
© European Southern Observatory (ESO) 1997
Online publication: June 5, 1998