We investigate the birth and evolution of isolated radio pulsars using a population synthesis method , modeling the birth properties of the pulsars , their time evolution , and their detection in the Parkes and Swinburne Multibeam ( MB ) surveys .
Together , the Parkes and Swinburne MB surveys \citep [ ] [ ] 2001MNRAS.328…17M , 2001MNRAS.326..358E have detected nearly 2/3 of the known pulsars and provide a remarkably homogeneous sample to compare with simulations .
New proper motion measurements \citep [ ] [ ] 2002ApJ…571..906B , 2003AJ….126.3090B and an improved model of the distribution of free electrons in the interstellar medium , NE2001 \citep [ ] [ ] NE2001 , also make revisiting these issues particularly worthwhile .
We present a simple population model that reproduces the actual observations well , and consider others that fail .
We conclude that : pulsars are born in the spiral arms , with the birthrate of 2.8 \pm 0.5 pulsars/century peaking at a distance \sim 3 kpc from the Galactic centre , and with mean initial speed of 380 ^ { +40 } _ { -60 } km s ^ { -1 } ; the birth spin period distribution extends to several hundred milliseconds , with no evidence of multimodality , implying that characteristic ages overestimate the true ages of the pulsars by a median factor ¿2 for true ages ¡30,000 yr ; models in which the radio luminosities of the pulsars are random generically fail to reproduce the observed P - \dot { P } diagram , suggesting a relation between intrinsic radio luminosity and ( P, \dot { P } ) ; radio luminosities L \propto \sqrt { \dot { E } } provides a good match to the observed P - \dot { P } diagram ; for this favored radio luminosity model , we find no evidence for significant magnetic field decay over the lifetime of the pulsars as radio sources ( \sim 100 Myr ) .