Phenomenological power spectrum models for H alpha emission line galaxies from the Nancy Grace Roman Space Telescope

The High Latitude Spectroscopic Survey (HLSS) is the reference baseline spectroscopic survey for NASA's Nancy Grace Roman Space Telescope, measuring redshifts of similar to 10M H alpha emission line galaxies over a 2000 deg(2) footprint at z = 1-2. In this work, we use a realistic Roman galaxy mock catalogue to explore optimal phenomenological modelling of the measured power spectrum. We consider two methods for modelling the redshift-space distortions (Kaiser squashing and another with a window function on beta that selects out the coherent radial infall pairwise velocities, M-A and M-B, respectively), two models for the non-linear impact of baryons that smear the baryon acoustic oscillation signal (a fixed ratio between the smearing scales in the perpendicular and parallel dimensions and another where these smearing scales are kept as free parameters, P-dw(k|k(*)) and P-dw(k|Sigma(perpendicular to), Sigma(parallel to)), respectively), and two analytical emulations of non-linear growth (one employing the halo model and another formulated from simulated galaxy clustering of a semi-analytical model, F-HM and F-SAM, respectively). We find that the best model combination employing F-HM is P-dw(k|k(*)) * F-HM * M-B, while the best combination employing F-SAM is P-dw(k|k(*)) * F-SAM * M-B, which leads to unbiased measurements of cosmological parameters. We compare these to the Effective Field Theory of Large-Scale Structure perturbation theory model P-EFT(k|Theta), and find that our simple phenomenological models are comparable across the entire redshift range for k(max) = 0.25 and 0.3 h Mpc(-1). We expect the tools that we hav e dev eloped to be useful in probing dark energy and testing gravity using Roman in an accurate and robust manner.

Automated summary

In this study, the optimal phenomenological models for measuring the power spectrum were explored using a realistic Roman galaxy mock catalogue. The best model combinations were found to be P-dw(k|k(*)) * F-HM * M-B and P-dw(k|k(*)) * F-SAM * M-B. These models were comparable to the Effective Field Theory of Large-Scale Structure perturbation theory model across the entire redshift range. The developed tools are expected to be useful in accurately and robustly probing dark energy and testing gravity using the Roman telescope.