Testing primordial non-gaussianity with gravitational wave surveys

Next-generation Gravitational Wave (GW) observatories, such as the Einstein Telescope and Cosmic Explorer, are set to significantly enhance both the quantity and quality of signals detected from compact-object mergers. This dissertation examines the potential of analyzing binary black hole merger events independently, without relying on electromagnetic counterparts or redshift measurements, to constrain the local-type Primordial non-Gaussianity parameter, fₙₗ. Using a modified version of the Boltzmann code CLASS (Multi_CLASS) – adapted to account for fₙₗ signatures on the scale-dependent bias – the angular power spectra Cₗ are calculated for a source density resembling that of an Einstein Telescope-like detector, extending up to a maximum ℓ = 100. A custom Fisher matrix pipeline, which I developed, is then employed to forecast constraints within the ΛCDM+fₙₗ parameter space marginalizing over the clustering bias of the sources. Under a baseline detection rate of N = 4.8×10⁴ GW events per year, with Planck priors on all ΛCDM parameters and no prior on bias terms, the analysis predicts σ(fₙₗ) = 7.64. For a combined ET2CE network with N = 1.1×10⁵ GW events per year, the forecast improves to σ(fₙₗ) = 3.61, and in an optimistic scenario of 10⁶ detections per year by the Einstein Telescope alone it tightens further to σ(fₙₗ) = 0.9. As fₙₗ endures as a fundamental unresolved issue in cosmology, probing it with independent tracers can yield substantively complementary constraints.