# Debiasing Cosmic Gravitational Wave Sirens

## TL;DR

This paper presents a model-independent statistical approach using Gaussian process regression to accurately estimate cosmological parameters from gravitational wave data, reducing bias and enabling rigorous tests of cosmological models.

## Contribution

It introduces a Gaussian process-based method for bias-free reconstruction of the Hubble parameter, enhancing the analysis of gravitational wave sirens independently of specific cosmological models.

## Key findings

- Gaussian process regression effectively removes bias in $H(z)$ reconstruction.
- Model-independent combination with supernova data improves parameter estimation.
- Redshift systematic control must reach spectroscopic precision to prevent bias.

## Abstract

Accurate estimation of the Hubble constant, and other cosmological parameters, from distances measured by cosmic gravitational wave sirens requires sufficient allowance for the dark energy evolution. We demonstrate how model independent statistical methods, specifically Gaussian process regression, can remove bias in the reconstruction of $H(z)$, and can be combined model independently with supernova distances. This allows stringent tests of both $H_0$ and $\Lambda$CDM, and can detect unrecognized systematics. We also quantify the redshift systematic control necessary for the use of dark sirens, showing that it must approach spectroscopic precision to avoid significant bias.