Efficient and Robust Inference of Neutron Star Parameters From Pulse Profile Modeling

Abstract

Modeling of X-ray pulse profiles from millisecond pulsars offers a promising method of inferring the mass-to-radius ratios of neutron stars. Recent observations with NICER resulted in measurements of radii for two neutron stars using this technique. In this dissertation, I explore correlations between model parameters and the degree to which individual parameters can be inferred from pulse profiles, using an analytic model that allows for an efficient and interpretable exploration. I introduce a new set of model parameters that reduce the most prominent correlations and allow for an efficient sampling of posteriors. I use this framework to show that the uncertainties in the model parameters for neutron stars for which the polar cap temperature falls outside of the NICER energy range are significantly degraded. I also demonstrate that the degree of beaming of radiation emerging from the neutron star surface has a large impact on the uncertainties in the inferred model parameters. In particular, when a model with an incorrect beaming function is used to fit data, the inferred neutron star mass-to-radius ratio can be significantly larger than the true value. This has important implications for interpreting and assessing the radii of neutron stars inferred through pulse profile modeling.