Fast and accurate mock catalogue generation for low-mass galaxies

TL;DR

This paper introduces a fast, accurate method using COLA simulations to generate large numbers of low-mass galaxy mock catalogues efficiently, suitable for large-scale structure analysis.

Contribution

The authors develop and validate a new framework combining COLA simulations with HOD to produce numerous accurate low-mass galaxy mocks efficiently.

Findings

COLA simulations accurately reproduce velocities for low-mass halos

The method produces 600 mock catalogues with 3% velocity power spectrum error

The code is publicly available for community use

Abstract

We present an accurate and fast framework for generating mock catalogues including low-mass halos, based on an implementation of the COmoving Lagrangian Acceleration (COLA) technique. Multiple realisations of mock catalogues are crucial for analyses of large-scale structure, but conventional N-body simulations are too computationally expensive for the production of thousands of realisations. We show that COLA simulations can produce accurate mock catalogues with a moderate computation resource for low- to intermediate- mass galaxies in $10^{12} M_\odot$ haloes, both in real and redshift space. COLA simulations have accurate peculiar velocities, without systematic errors in the velocity power spectra for k < 0.15 h/Mpc, and with only 3-per-cent error for k < 0.2 h/Mpc. We use COLA with 10 time steps and a Halo Occupation Distribution to produce 600 mock galaxy catalogues of the WiggleZ Dark Energy Survey. Our parallelized code for efficient generation of accurate halo catalogues is publicly available at github.com/junkoda/cola_halo.