Performance study of Lagrangian methods: reconstruction of large scale peculiar velocities and baryonic acoustic oscillations

Abstract

NoAM for "No Action Method" is a framework for reconstructing the past orbits of observed tracers of the large scale mass density field. It seeks exact solutions of the equations of motion (EoM), satisfying initial homogeneity and the final observed particle (tracer) positions. The solutions are found iteratively reaching a specified tolerance defined as the RMS of the distance between reconstructed and observed positions. Starting from a guess for the initial conditions, NoAM advances particles using standard N-body techniques for solving the EoM. Alternatively, the EoM can be replaced by any approximation such as Zel'dovich and second order perturbation theory (2LPT). NoAM is suitable for billions of particles and can easily handle non-regular volumes, redshift space, and other constraints. We implement NoAM to systematically compare Zel'dovich, 2LPT, and N-body dynamics over diverse configurations ranging from idealized high-res periodic simulation box to realistic galaxy mocks. Our findings are (i) Non-linear reconstructions with Zel'dovich, 2LPT, and full dynamics perform better than linear theory only for idealized catalogs in real space. For realistic catalogs, linear theory is the optimal choice for reconstructing velocity fields smoothed on scales > ∼ 5 h −1 Mpc. (ii) all non-linear back-in-time reconstructions tested here, produce comparable enhancement of the baryonic oscillation signal in the correlation function.

Velocity–density correlations from the cosmicflows-3 distance catalogue and the 2MASS Redshift Survey

“…We use 30 mock catalogs of 2MRS extracted by Keselman & Nusser (2016) Table 3 in Planck Collaboration et al 2016). These are dark matter only simulations, but baryonic feedback is too weak to play any significant role over the relevant level of accuracy and scales considered (Hellwing et al 2016) The simulations are available with 5 different resolutions.…”

Constrained realizations of 2MRS density and peculiar velocity fields: growth rate and local flow

“…It has been shown in Keselman & Nusser (2017) that for scales 5 ℎ −1 Mpc such a linear reconstruction yields generally better results than nonlinear methods when applied to realistic mock redshift catalogues. The authors of Keselman & Nusser (2017) suspect this to be caused by the higher sensitivity of nonlinear reconstructions to the information on small scales, which are more strongly contaminated by non-linear RSDs (e. g. fingers-of-god) and selection effects. Altogether, this yields a numerically very efficient, conservative reconstruction algorithm.…”

Laguerre reconstruction of the BAO feature in halo-based mock galaxy catalogues

“…Gravitational evolution modifies the shape of the feature, potentially biasing and degrading the cosmological constraints which current and future datasets can provide. This has driven the development of many algorithms which seek to restore the feature to its original shape [5][6][7][8][9][10][11][12].…”