A Observational Test Of Radio Beam Models Of Pulsar

The structure of radio emission beam of pulsar plays an important role not only in the study of pulsar radiation mechanism but also in the studies of pulsar population and the detection rate of pulsars. However, the structure of radio emission beam is an open question of long-term debate. At present, there are mainly three kinds of theoretical or empirical models, i.e. the conal beam, the fan beam and the patchy beam models. Previous tests for these models were limited by the small scale of available samples. This limitation is caused by the fact that the inclination and impact angles, the two angles used to determine the beam parameters, have only been constrained from observational data for a small number of pulsars. With the increasing high-quality data released over the recent years, it is now possible to perform a test using a large sample of pulsars.The inclination and impact angles can be obtained by fitting the data of polarization position angle(PPA) with the rotation vector model(RVM). To overcome the issues of the Levenberg-Marquardt non-linear fitting and the grid search methods, e.g. inefficiency and easily over-fitting, a Markov Chain Monte Carlo(MCMC) fitting algorithm, on the basis of the Bayesian Inference, is developed in this work. All the pulsars from the European Pulsar Network database with polarization profiles were reviewed. Finally, our MCMC algorithms successfully obtained the inclination and impact angles for 123 pulsars. Combining with a previous sample given by other authors, the largest sample of pulsars with known inclination and impact angles has hitherto been obtained.A test of radio emission beam model was performed with the above largest sample. The relationship between the observed impact angles and the pulse widths of integrated profiles were studied and compared with the predictions of the fan beam and the conal beam models. It was found that observational data favored the fan beam model while disfavored the conal beam model, which challenges the later one, the currently most popular emission beam model.Our MCMC algorithm has the advantages of being highly efficient, resistant to over-fitting and easy to estimate the uncertainty of parameters. In the future, it can be used to perform the RVM fitting effectively for larger samples obtained with FAST and SKA. The inclination and impact angles obtained in this work are useful for other studies, e.g. the radiation geometry, the emission mechanism, population studies and the evolution of pulsars.