The effects of non-linearity on the growth rate constraint from velocity correlation functions

TL;DR

This paper investigates how non-linear effects and density-velocity correlations impact the measurement of the growth rate of cosmic structures from galaxy velocity data, revealing a significant potential bias of about 10%.

Contribution

It provides a perturbation theory-based model including non-linear and density-velocity effects up to fourth order, improving the accuracy of growth rate constraints from velocity correlations.

Findings

Non-linear effects cause a ~10% shift in fσ8 estimates.

Density-velocity correlation contributes significantly at small scales.

Next-generation surveys must account for these effects to avoid bias.

Abstract

The two-point statistics of the cosmic velocity field, measured from galaxy peculiar velocity (PV) surveys, can be used as a dynamical probe to constrain the growth rate of large-scale structures in the universe. Most works use the statistics on scales down to a few tens of Megaparsecs, while using a theoretical template based on the linear theory. In addition, while the cosmic velocity is volume-weighted, the observable line-of-sight velocity two-point correlation is density-weighted, as sampled by galaxies, and therefore the density-velocity correlation term also contributes, which has often been neglected. These effects are fourth order in powers of the linear density fluctuation $\delta_{\rm L}^4$, compared to $\delta_{\rm L}^2$ of the linear velocity correlation function, and have the opposite sign. We present these terms up to $\delta_{\rm L}^4$ in real space based on the standard perturbation theory, and investigate the effect of non-linearity and the density-velocity contribution on the inferred growth rate $f\sigma_8$, using $N$-body simulations. We find that for a next-generation PV survey of volume $\sim {\cal O}(500 \, h^{-1} \, {\rm Mpc})^3$, these effects amount to a shift of $f\sigma_8$ by $\sim 10$ per cent and is comparable to the forecasted statistical error when the minimum scale used for parameter estimation is $r_{\rm min} = 20 \, h^{-1} \, {\rm Mpc}$.