On the Depolarization of Discrete Radio Sources by Faraday Dispersion

A study is made of the implications of the recent polarization measurements for the structures of discrete radio sources and the source–observer media. Simple models of wavelength-dependent depolarizing mechanisms are investigated and it is found that most are incompatible with the observations of Gardner & Whiteoak. The models of internal Faraday dispersion predict a lower polarization at 30 cm than is observed. It is suggested that the depolarization of the Crab nebula is produced by Faraday rotation in the filamentary shell that surrounds the nebula. Such filaments could also exist in the outer regions of extragalactic sources. A complex number representation is used for the state of linear polarization and a Faraday dispersion function is defined to describe the distribution of polarized radiation with respect to Faraday depth. The persistence of polarization at 30 cm, after partial depolarization between 10 cm and 20 cm, implies that the radiation is spread over a large range of Faraday depths. The observed linearity of the plot of the angle of polarization against wavelength squared for most sources implies that it is justifiable to make an assumption which enables one to calculate the Faraday dispersion function of a source from the dependence of its polarization on wavelength. Estimates are given for upper limits to the densities of internal ionized gases in the sources for which we have polarization measurements.