Non-perturbative halo clustering from cosmological density peaks

TL;DR

This paper develops non-perturbative predictions for halo clustering in cosmology, accurately matching simulations at various stages and highlighting the need for advanced models at non-linear scales.

Contribution

It introduces a non-perturbative approach to predict halo correlation functions at all scales, improving upon perturbative bias models and incorporating displacement effects.

Findings

Good agreement with N-body simulations for Lagrangian halo clustering.

Accurate predictions for halo infall velocities and velocity dispersions.

Limitations identified in modeling halo displacements at non-linear scales.

Abstract

Associating the formation sites of haloes with the maxima of the smoothed linear density field, we present non-perturbative predictions for the Lagrangian and evolved halo correlation functions that are valid at all separations. In Lagrangian space, we find significant deviations from the perturbative bias calculation at small scales, in particular, a pronounced exclusion region where $\xi=-1$ for maxima of unequal height. Our predictions are in good agreement with the Lagrangian clustering of dark matter proto-haloes reconstructed from N-body simulations. Our predictions for the mean infall and velocity dispersion of haloes, which differ from the local bias expansion, show a similar level of agreement with simulations. Finally, we displace the initial density peaks according to the Zeldovich approximation in order to predict the late-time clustering of dark matter haloes. While we are able to reproduce the early evolution of this conserved set of tracers, our approximation fails at the collapse epoch (z=0) on non-linear scales r<10Mpc/h, emphasizing the need for a non-perturbative treatment of the halo displacement field.