TL;DR
This paper introduces AbacusPNG, a new suite of cosmological N-body simulations designed to help disentangle primordial non-Gaussianity (PNG) signals from astrophysical effects in galaxy surveys. By validating these simulations against theoretical predictions and applying them to realistic DESI galaxy samples, the authors demonstrate that halo assembly bias significantly impacts PNG measurements, highlighting the need for improved galaxy-halo connection models for future precision cosmology.
What Problem Does This Paper Address?
Understanding the very early Universe, specifically the origin of cosmic structure, is a major goal of modern cosmology. The standard model of inflation predicts that the initial density fluctuations were nearly Gaussian. However, deviations from this Gaussianity, known as primordial non-Gaussianity (PNG), could reveal new physics beyond simple single-field inflation. Measuring PNG is challenging because astrophysical processes and the non-linear evolution of structure can mimic or mask these subtle primordial signals. Specifically, the scale-dependent galaxy bias, a key observable for detecting local-type PNG, is degenerate with the intrinsic halo assembly bias. This degeneracy means that accurately measuring PNG requires a deep understanding of how galaxy formation depends on halo properties beyond just their mass.
Methods
The authors introduce the AbacusPNG simulation suite, an extension of the existing Abacus simulations. These simulations incorporate a primordial non-Gaussian gravitational potential, allowing them to generate cosmological volumes with varying levels of local-type PNG.
Key methodological aspects include:
- N-body Simulations: AbacusPNG consists of 9 large-volume (2 Gpc/h) simulations with high particle mass ($1.01 \times 10^{10} M_\odot/h$), designed to capture large-scale structure relevant for galaxy surveys.
- PNG Implementation: Local-type PNG is implemented by modifying the initial Gaussian primordial potential field according to a specific theoretical prescription.
- Validation: The simulations are rigorously validated against theoretical predictions from perturbation theory (PT) for the matter bispectrum and power spectrum.
- Separate Universe Technique: This technique is used to predict and measure the halo bias parameter ($b_\phi$) associated with PNG by simulating universes with slightly altered cosmological parameters.
- Halo Assembly Bias Study: The paper investigates the relationship between $b_\phi$ and halo properties like concentration, shear (environment), and accretion rate.
- HOD Modeling: Halo Occupation Distribution (HOD) models, fitted to early DESI data, are used to create synthetic galaxy catalogs (LRGs and QSOs) within the simulations.
- Fisher Forecasts: The authors use a Fisher formalism to estimate the impact of uncertainties in $b_\phi$ on the constraints for the PNG parameter $f_{NL}$.
Datasets
- AbacusPNG Simulations: A new suite of 9 N-body simulations with varying local-type PNG amplitudes ($f_{NL} = \{-100, -30, 0, 30, 100\}$) and two realizations each. These simulations are publicly available.
- AbacusSUMMIT Simulations: Used for comparison and validation, particularly for the "Separate Universe" technique.
- DESI 1% Survey Data: Used to constrain the HOD models for LRGs and QSOs, which are then applied to the AbacusPNG simulations.
Key Insights and Ideas
- PNG and Halo Assembly Bias are Intertwined: The study reveals a strong correlation between the PNG-induced bias ($b_\phi$) and halo assembly bias properties (concentration, shear, accretion rate). This means that simply measuring the scale-dependent galaxy bias isn't enough to isolate PNG; one must also account for how galaxy properties influence this bias.
- Concentration is a Key Driver: The response of $b_\phi$ to halo concentration is found to be stronger than to accretion rate, suggesting that the internal structure of halos plays a crucial role in how PNG imprints itself on galaxy clustering.
- Future Surveys Need Better Galaxy-Halo Models: The significant impact of assembly bias on $b_\phi$ implies that current galaxy-halo connection models, often focused solely on halo mass, are insufficient for precise PNG measurements with upcoming surveys like DESI.
- AbacusPNG is a Valuable Tool: The AbacusPNG simulations provide a robust platform for studying PNG effects and validating theoretical models in a regime relevant to current and future galaxy surveys.
Key Results
- Simulation Validation: AbacusPNG simulations show good agreement with theoretical predictions for matter bispectrum and power spectrum, and the $b_\phi$ parameter derived from simulations aligns with the "Separate Universe" technique.
- Strong Assembly Bias Dependence: The halo assembly bias properties (concentration, shear, accretion rate) show a significant response of $b_\phi$, particularly in low-mass halos.
- Impact on $f_{NL}$ Constraints: When accounting for the uncertainty in $b_\phi$ derived from DESI-like galaxy samples (LRGs and QSOs), the error on the PNG parameter $f_{NL}$ increases by a factor of 1.5-2 compared to scenarios where $b_\phi$ is precisely known.
- Specific $f_{NL}$ Errors: The paper reports projected errors on $f_{NL}$ for LRGs at $z=0.5$ and $z=0.8$, and QSOs at $z=1.4$ to be 15, 8, and 7, respectively, after accounting for $b_\phi$ uncertainties.
Limitations and Future Work
- Resolution Effects: While AbacusPNG has a larger volume, its mass resolution is coarser than AbacusSUMMIT, which might impact studies of the lowest and highest halo masses.
- Noise in Fitting: The fitting procedure for $b_\phi$ from simulation ratios can be noisy, especially for high-mass halos due to limited statistics.
- Model Degeneracies: The paper highlights the degeneracy between assembly bias and halo mass, which is difficult to break with clustering data alone.
- Future Work: The authors suggest that future work should focus on:
- Improving galaxy-halo connection models to better account for assembly bias.
- Utilizing weak lensing data in conjunction with clustering to break degeneracies and obtain tighter constraints on $b_\phi$.
- Exploring the role of hydrodynamical simulations to better understand the complex link between galaxy formation and PNG.
- Investigating the impact of other halo properties and more complex HOD extensions.
Summary updated