Speaker: Nina Roth
The deviation of the initial density field from a Gaussian random field (“Primordial non-Gaussianity”, PNG) is generally expressed by one or more real-valued parameters, e.g. the skewness and kurtosis, and depends on the exact model of Inflation. Constraints on PNG from the Cosmic Microwave Background are predicted to be surpassed by future large-scale structure surveys, based on an additional scale-dependent bias between galaxies and dark matter caused by PNG of the local type.
Previous studies using this effect have mostly focused on constraining the skewness parameter, f_nl. However, it was shown that the additional galaxy bias induced by f_nl and that induced by the kurtosis parameter (g_nl) have the exact same scale-dependence. I will first discuss how to use N-body simulations to test this degeneracy and motivate Bayesian model selection when constraining these parameters.
On the observational side, estimates of PNG from galaxy surveys are severely hampered by systematic effects which mimic the scale-dependence on large scales. I will discuss how “mode-projection”, a novel blind systematics mitigation technique, can be used successfully to clean up a highly contaminated quasar dataset. These results incorporate uncertainties in the cosmological parameters, redshift distributions, shot noise, and the bias prescription, and are therefore very robust. In addition to the standard skewness and kurtosis parameters, we obtain constraints on a generalised model where f_nl is allowed to vary as a function of k. These are the strongest PNG constraints obtained to date using a single population of large-scale structure tracers, and are already at the level of pre-Planck constraints from the CMB.