Tolerance to Astrophysical Model Uncertainty in Dark Siren Hubble Measurement with Third-generation Gravitational-wave Detectors

Abstract

Gravitational-wave (GW) events can serve as standard sirens for cosmology, as the luminosity distance to source can be directly measured from the waveform amplitude. Specifically, the ``dark'' siren method involves inferring cosmological parameters, e.g. the Hubble constant, by comparing the luminosity distance distribution and that of the redshift, typically obtained through a combination of galaxy survey catalog and theoretical models. Especially with the prospect of third-generation GW detectors, the statistical uncertainty of the Hubble measurement can be suppressed to a percent level. However, incorrect assumption in galaxy population models can lead to systematic bias, which becomes increasingly relevant as third-generation GW detectors can detect large-redshift sources beyond the reach of currently available galaxy catalogs. In this work, we adopt a Fisher information formalism and study the maximum model error tolerance given specific total error budget. We find that, to achieve a total error budget of 1% in the Hubble constant, the galaxy mass function redshift evolution should be known to O(1%). We find that galaxy redshift uncertainty, survey magnitude limit and GW angular localization error are important factors. We find galaxy clustering also improves model error tolerance, thus our results provide a conservative benchmark to future analysis using real galaxy catalogs. We also investigate the effective bright siren scenario and highlight that the dark siren selection strategy should be catered to measurement uncertainty and the target total error budget. This work thus highlights the challenge in the dark siren method and quantifies requirements on both the galaxy catalog and GW measurement to contribute to constraining cosmology.