Modest set of simulations of local-type primordial non-Gaussianity in the DESI era

Boryana Hadzhiyska, Lehman H. Garrison, Daniel J. Eisenstein, Simone FerraroPhysical Review. D/Physical Review. D 2024 · DOI

TL;DR

This paper introduces ABACUSPNG, a new suite of N-body simulations designed to study primordial non-Gaussianity (PNG) in the context of upcoming galaxy surveys like DESI. By combining these simulations with advanced halo models, the authors investigate how galaxy assembly biases affect PNG measurements and provide initial forecasts for detecting PNG with DESI data.

What Problem Does This Paper Address?

Detecting primordial non-Gaussianity (PNG) is a crucial goal in cosmology, as it can reveal fundamental physics of the early Universe. While large-scale structure surveys offer a powerful way to probe PNG, interpreting these signals is challenging due to complex astrophysical effects and the non-linear evolution of cosmic structures. Specifically, the scale-dependent bias of galaxies, which is sensitive to PNG, is also influenced by how halos and galaxies form and evolve (assembly bias). This paper aims to disentangle these effects by using high-fidelity simulations and advanced halo models to understand the relationship between PNG, galaxy bias, and halo assembly bias, ultimately improving our ability to constrain PNG from future surveys.

Methods

The authors introduce ABACUSPNG, a set of N-body simulations that incorporate local-type PNG by modifying the initial conditions with a non-Gaussian primordial gravitational potential. These simulations are validated against theoretical predictions from perturbation theory for the matter bispectrum and power spectrum. To study the impact of PNG on galaxy surveys, they employ halo occupation distribution (HOD) models, specifically those developed for DESI galaxy samples (LRGs and QSOs). They also utilize the "Separate Universe" approach, a theoretical tool that allows them to estimate the PNG-induced bias ($b_\phi$) by simulating different cosmological parameters. Finally, they perform Fisher forecasts to estimate the precision with which $f_{NL}$ (the amplitude of local PNG) can be constrained by DESI data, considering the uncertainties in both $f_{NL}$ and $b_\phi$.

Datasets

The primary "dataset" introduced in this paper is the ABACUSPNG simulation suite. This suite consists of 10 simulations, each with a box size of 2 Gpc/h and particle mass of $1.01 \times 10^{10} M_\odot/h$. These simulations are designed to cover a range of local-type PNG amplitudes ($f_{NL} = \{-100, -30, 0, 30, 100\}$) with two realizations each. The authors also leverage existing HOD model fits to early DESI 1% survey data for LRGs and QSOs, as described in Yuan et al. (2023).

Key Insights and Ideas

  • Simulations as a Bridge: ABACUSPNG provides a crucial bridge between theoretical predictions of PNG and the complex reality of galaxy formation, allowing for more robust interpretation of observational data.
  • Assembly Bias is Key: The study highlights that halo assembly bias (how halo properties like concentration, shear, and accretion rate influence their clustering) significantly impacts the measurement of $b_\phi$, the bias parameter related to PNG. This means that simply accounting for halo mass isn't enough; understanding assembly bias is vital for accurate PNG constraints.
  • $b_\phi$ is Sensitive to Halo Properties: The authors find a strong response of $b_\phi$ to halo concentration and accretion rate, suggesting that these properties, beyond just halo mass, must be considered when modeling galaxy bias and its relation to PNG.
  • DESI Data Offers Promise: The paper demonstrates that by combining realistic HOD models for DESI galaxies with the ABACUSPNG simulations, it's possible to make initial forecasts for constraining $f_{NL}$, even with current uncertainties in astrophysical parameters.

Key Results

  • The ABACUSPNG simulations show good agreement with theoretical predictions for the matter bispectrum and power spectrum, validating the simulation methodology.
  • The study finds a strong correlation between $b_\phi$ and halo assembly bias properties (concentration, shear, accretion rate), indicating that these factors are important for accurate PNG measurements.
  • Initial forecasts for DESI data suggest that the error on $f_{NL}$ will be around 15, 8, and 7 for LRGs at $z=0.5$, LRGs at $z=0.8$, and QSOs at $z=1.4$, respectively. These errors are larger than they would be if $b_\phi$ were perfectly known, emphasizing the importance of understanding galaxy assembly bias.

Limitations and Future Work

  • Simulation Resolution: While ABACUSPNG offers a good balance of volume and resolution, the authors note that higher resolution might be needed for certain analyses, particularly for lower-mass halos.
  • HOD Model Simplifications: The HOD models used are based on fits to early DESI data and may not capture all the complexities of galaxy formation. Future work could explore more sophisticated HOD models or even hydrodynamical simulations.
  • Cosmic Variance: Despite using multiple simulations and the Separate Universe approach, cosmic variance remains a limiting factor, especially for the $b_\phi$ measurements.
  • Future Work: The authors suggest further investigation into the link between galaxy formation physics (e.g., through hydrodynamical simulations) and halo assembly bias to improve PNG constraints. They also plan to explore the impact of other PNG shapes and more complex HOD extensions. The ABACUSPNG simulations are publicly available, encouraging further research in this area.

Summary updated

Abstract

A measurement of a primordial non-Gaussianity (PNG) signal through late- or early-Universe probes has the potential to transform our understanding of the physics of the primordial Universe. While large-scale structure observables in principle contain vital information, interpreting these measurements is challenging due to poorly understood astrophysical effects. Luckily, $N$-body simulations, such as the abacuspng set presented in this study, consisting of 9 boxes, each of size ${L}_{\mathrm{box}}=2\text{ }\text{ }\mathrm{Gpc}/h$ and particle mass of $1.01\ifmmode\times\else\texttimes\fi{}{10}^{10}{M}_{\ensuremath{\bigodot}}/h$, provide a viable path forward. As validation, we find good agreement between the simulations and our expectations from one-loop perturbation theory (PT) and the ``separate universe'' method for the matter bispectrum, matter power spectrum and the halo bias parameter associated with PNG, ${b}_{\ensuremath{\phi}}$. As a science application, we investigate the link between halo assembly bias and ${b}_{\ensuremath{\phi}}$ for halo properties known to play a vital role in accurately predicting galaxy clustering: concentration, shear (environment), and accretion rate. We find a strong response for all three parameters, suggesting that the connection between ${b}_{\ensuremath{\phi}}$ and the assembly history of halos needs to be taken into account by future PNG analyses. We further perform the first study of the ${b}_{\ensuremath{\phi}}$ parameter from fits to early DESI data of the luminous red galaxy (LRG) and quasistellar object (QSO) samples and comment on the effect on ${f}_{\mathrm{NL}}$ constraints for the allowed galaxy-halo models (note that $\ensuremath{\sigma}[{f}_{\mathrm{NL}}]\ensuremath{\propto}\frac{\ensuremath{\sigma}[{b}_{\ensuremath{\phi}}]}{{b}_{\ensuremath{\phi}}}$). We find that the error on ${f}_{\mathrm{NL}}$ is 15, 8, 7 for the LRGs at $z=0.5$ and $z=0.8$ and QSOs at $z=1.4$, respectively, suggesting that a thorough understanding of galaxy assembly bias is warranted so as to perform robust high-precision analysis of local-type PNG with future surveys.
BibTeX
@article{hadzhiyska2024modest,
  author = {Boryana Hadzhiyska and Lehman H. Garrison and Daniel J. Eisenstein and Simone Ferraro},
  title = {{Modest set of simulations of local-type primordial non-Gaussianity in the DESI era}},
  journal = {Physical review. D/Physical review. D.},
  year = {2024},
  doi = {10.1103/physrevd.109.103530},
}
Figure