Lagrangian bias of generic large-scale structure tracers

TL;DR

This paper introduces a multivariate expansion scheme for modeling the clustering bias of large-scale structure tracers, accounting for multiple stochastic variables and symmetries, with applications to peaks and non-Gaussian bias.

Contribution

It develops a general, symmetry-based multivariate expansion method for biased tracers in large-scale structure, extending previous local bias models.

Findings

Explicit bias parameters for density peaks derived

Derived the BBKS formula using the new formalism

Computed non-Gaussian bias with primordial trispectrum

Abstract

The dark matter halos that host galaxies and clusters form out of initial high-density patches, providing a biased tracer of the linear matter density field. In the simplest local bias approximation, the halo field is treated as a perturbative series in the average overdensity of the Lagrangian patch. In more realistic models, however, additional quantities will affect the clustering of halo-patches, and this expansion becomes a function of several stochastic variables. In this paper, we present a general multivariate expansion scheme that can parametrize the clustering of any biased Lagrangian tracer, given only the variables involved and their symmetry (in our case rotational invariance). This approach is based on an expansion in the orthonormal polynomials associated with the relevant variables, so that no renormalization of the coefficients ever occurs. We provide explicit expression for the series coefficients, or Lagrangian bias parameters, in the case of peaks of the linear density field. As an application of our formalism, we present a simple derivation of the original BBKS formula, and compute the non-Gaussian bias in the presence of a primordial trispectrum of the local shape.