2009ApJ...701.1243D

Accurate measurement of pulsar distances via astrometry using very long baseline interferometry enables the improvement of Galactic electron density distribution models, improving distance estimates for the vast majority of pulsars for which parallax measurements are unavailable. However, pulsars at southern declinations have been under-represented in previous interferometric astrometry campaigns, due to the dominance of northern hemisphere instrumentation for astrometry. In order to redress this imbalance, we have conducted a two-year astrometric campaign targeting eight southern pulsars with the Australian Long Baseline Array. The program summarized in this paper has resulted in the measurement of seven new pulsar parallaxes, with success on objects down to a mean flux density of 800 µJy at 1600 MHz. Our results highlight the substantial uncertainties that remain when utilizing free electron density models for individual pulsar distances. Until this study, PSR J0630-2834 was believed to convert 16% of its spin-down energy into X-rays, but our measured parallax distance of 332⁺⁵²_–40 pc has revised this value to <1%. In contrast, PSR J0108-1431 was found to be almost a factor of 2 more distant than previously thought, making its conversion of spin-down energy to X-rays the most efficient known (>1%). The 8.5 s radio pulsar J2144-3933 was found to be closer than previously predicted, making its apparent 1400 MHz radio luminosity the lowest of any known pulsar (20 µJy kpc²). We have examined the growing population of neutron stars with accurate parallaxes to determine the effect of distance errors on the underlying neutron star velocity distribution, and find that typical distance errors may be biasing the estimated mean pulsar velocity upwards by 5%, and are likely to exaggerate the distribution's high-velocity tail.