Speaker
Description
Gravitational wave spectral sirens provide a powerful approach to measuring cosmological parameters — requiring neither electromagnetic counterparts nor galaxy catalogs — by leveraging population-level features in the distribution of compact binary mergers. With the Einstein Telescope (ET) set to deliver event catalogs three orders of magnitude larger than current ones, validating the scalability and consistency of existing inference frameworks is a pressing priority. In this talk, I present a blinded mock data challenge in which three independent public pipelines — ICARO, CHIMERA, and PYMCPOP-GW — are tested on simulated ET observations comprising the $\sim 12000$ highest SNR binary black hole mergers detectable in one year. I discuss their computational performance, cross-pipeline agreement, and resulting cosmological forecasts. Thanks to GPU acceleration, all pipelines process the expected ET event rates within feasible runtimes and recover mutually consistent cosmological and population parameters. Assuming a flat $\Lambda$CDM cosmology, we measure $H(z)$ at $z\sim1.5$ with $2.4\%$ precision, and achieve a mean precision of 2.8% on $H(z)$ across $0.7<z<1.8$. This translates into joint constraints of $\sim10\%$ on $H_0$ and $\sim26\%$ on $\Omega_{m,0}$. I further show that low-distance sources near population features dominate the constraining power on all cosmological parameters, while high-distance events primarily inform $\Omega_{m,0}$. These results establish a validated, performance-tested framework for spectral siren cosmology in the ET era.