Linear and non-linear bias: predictions vs. measurements

TL;DR

This paper compares various methods of measuring galaxy bias parameters, revealing consistent relations and small variations, which can enhance cosmological model constraints using clustering statistics.

Contribution

It introduces direct bias measurement techniques and compares them with traditional correlation-based methods, highlighting their consistency and potential for improved cosmological constraints.

Findings

Bias measurements vary by about 5% across methods and halo masses.

A tight relation exists between linear and quadratic bias parameters.

Results are consistent across different cosmologies and redshifts.

Abstract

We study the linear and non-linear bias parameters which determine the mapping between the distributions of galaxies and the full matter density fields, comparing different measurements and predictions. Associating galaxies with dark matter haloes in the MICE Grand Challenge N-body simulation we directly measure the bias parameters by comparing the smoothed density fluctuations of haloes and matter in the same region at different positions as a function of smoothing scale. Alternatively we measure the bias parameters by matching the probability distributions of halo and matter density fluctuations, which can be applied to observations. These direct bias measurements are compared to corresponding measurements from two-point and different third-order correlations, as well as predictions from the peak-background model, which we presented in previous articles using the same data. We find an overall variation of the linear bias measurements and predictions of $\sim 5 \%$ with respect to results from two-point correlations for different halo samples with masses between $\sim 10^{12} - 10^{15}$ $h^{-1}M_\odot$ at the redshifts $z=0.0$ and $0.5$. Variations between the second- and third-order bias parameters from the different methods show larger variations, but with consistent trends in mass and redshift. The various bias measurements reveal a tight relation between the linear and the quadratic bias parameters, which is consistent with results from the literature based on simulations with different cosmologies. Such a universal relation might improve constraints on cosmological models, derived from second-order clustering statistics at small scales or higher-order clustering statistics.