We introduce a technique for gravitational-wave analysis, where Gaussian process regression is used to emulate the strain spectrum of a stochastic background using population-synthesis simulations. This leads to direct Bayesian inference on astrophysical parameters. For PTAs specifically, we interpolate over the parameter space of supermassive black-hole binary environments, including 3-body stellar scattering, and evolving orbital eccentricity. We illustrate our approach on mock data, and assess the prospects for inference with data similar to the NANOGrav 9-yr data release.
@article{arxiv.1612.02817,
title = {Constraints On The Dynamical Environments Of Supermassive Black-hole Binaries Using Pulsar-timing Arrays},
author = {Stephen R. Taylor and Joseph Simon and Laura Sampson},
journal= {arXiv preprint arXiv:1612.02817},
year = {2017}
}
Comments
Published in Physical Review Letters. 7 pages, 5 figures. Jupyter notebook and synthesized supermassive black-hole binary populations are available at https://github.com/stevertaylor/gw_pta_emulator