3.3. Distribution of B_reg, with galactic radius

The use of synchrotron radiation is often employed to get the total magnetic field strength. The average 'equipartition'-derived magnetic field strength follows from the assumption of energy-density equipartition between cosmic-ray particles and magnetic fields, the assumption or knowledge of the ratio between cosmic-ray electrons and protons (and its variation with particle energy), the synchrotron spectral index , the extent of the radio emission along the line of sight, and the volume filling factor. Still, large uncertainties in these factors lead to only moderate errors in the total magnetic field strengths (e.g., Beck 1996; Heiles 1996b; Fitt & Alexander 1993).

The use of the Faraday rotation is often employed to get the uniform component of the magnetic field strength. The average 'Faraday'- derived magnetic field component follows from the assumption or knowledge of the density distribution of the thermal (non-synchrotron) electrons in the Galaxy. Since pulsars are often near the galactic plane, their emitted light often travels through the spiral arms. A distribution model for thermal electrons is thus used to analyse pulsar dispersion measures DM and pulsar rotation measures RM. The recent model of Taylor and Cordes (1993) for the distribution of thermal electrons in the galactic disk, employing 4 spiral arms, has been often employed, and its accuracy has been confirmed by Weisberg et al. (1995), and very slightly modified by Heiles et al. (1996) in the innermost areas. The use of this model allows small distance errors to pulsars, with a relative random error r / r = 0.20 (e.g., Fig. 7 in Taylor and Cordes 1993). In this case, after looking through 4 spiral arms, one has an error r = 0.2 × 4 arm widths = 0.8 arm, enabling a proper study to be done all the way to the region of the galactic center (e.g., Vallée 1996b).