Galaxy Bias and its Effects on the Baryon Acoustic Oscillations Measurements

TL;DR

This study quantifies the impact of galaxy bias on Baryon Acoustic Oscillation measurements using 44 high-resolution ABACUS N-body simulations with 12 Halo Occupation Distributions to model biased tracers. It applies a one-step reconstruction based on the Zel'dovich approximation to mitigate large-scale bulk flows and redshift-space distortions, analyzing the propagator and the acoustic-scale shift through a flexible power-spectrum fitting framework. The authors find that low-bias tracers produce negligible shifts in the acoustic scale, while high-bias tracers exhibit modest shifts that are largely eliminated after reconstruction, reducing systematic distance errors to the ~0.07–0.15% level. Across real and redshift space, reconstruction also tightens the correlations between biased-tracer and mass-case BAO measurements and brings the observed scatter in line with Fisher-matrix expectations. The results support the robustness of BAO as a standard ruler for current surveys and provide guidance for interpreting future, higher-precision BAO measurements in the presence of galaxy bias.

Abstract

The baryon acoustic oscillation (BAO) feature in the clustering of matter in the universe serves as a robust standard ruler and hence can be used to map the expansion history of the universe. We use high force resolution simulations to analyze the effects of galaxy bias on the measurements of the BAO signal. We apply a variety of Halo Occupation Distributions (HODs) and produce biased mass tracers to mimic different galaxy populations. We investigate whether galaxy bias changes the non-linear shifts on the acoustic scale relative to the underlying dark matter distribution presented by Seo et al (2009). For the less biased HOD models (b < 3), we do not detect any shift in the acoustic scale relative to the no-bias case, typically 0.10% \pm 0.10%. However, the most biased HOD models (b > 3) show a shift at moderate significance (0.79% \pm 0.31% for the most extreme case). We test the one-step reconstruction technique introduced by Eisenstein et al. (2007) in the case of realistic galaxy bias and shot noise. The reconstruction scheme increases the correlation between the initial and final (z = 1) density fields achieving an equivalent level of correlation at nearly twice the wavenumber after reconstruction. Reconstruction reduces the shifts and errors on the shifts. We find that after reconstruction the shifts from the galaxy cases and the dark matter case are consistent with each other and with no shift. The 1-sigma systematic errors on the distance measurements inferred from our BAO measurements with various HODs after reconstruction are about 0.07% - 0.15%.

Figures (10)

• Figure 1: The average power spectrum from 44 simulations (HOD 1a, real space) divided by the no wiggle power spectrum, $P(k)/Pnwl(k)$. The no-wiggle power spectrum represents the power spectrum without the BAO peaks EH98. Thus, we clearly see the BAO peaks appear in the power spectrum. We also start to see the increasing power on small scales (large $k$) as non-linear growth starts to dominate. Plotted in the dashed line is the average power spectrum after reconstruction. We see that our reconstruction scheme reduces the effect of non-linear structure formation on small scales and restores information into the acoustic peaks. The dot-dashed line representes the linear power spectrum divided by the no-wiggle power spectrum.
• Figure 2: The correlation between the initial and final density field as described by Eq. \ref{['eq:Ck']} for HOD 1a. The solid line represents real space while the dashed line represents redshift space. We see that redshift distortions reduce the correlation between the initial and final ($z = 1$) density fields. We are most interested in the correlation between the density fields at BAO wavenumbers ($k \approx 0.09 - 0.2 h{\rm\;Mpc^{-1}}$).
• Figure 3: The propagator, $G(k)$ for three HOD models and the mass case. The top panel shows real space and the bottom panel shows redshift space. We see that while redshift distortions reduce the correlation between the density fields, they do not spread the individual HOD models from the mass case.
• Figure 4: The propagator $G(k)$ for three HOD models and includes the effects of reconstruction. The top panel shows real space and the bottom panel shows redshift space. In both panels, the lines maintaining higher correlation at smaller scales are for the cases after reconstruction. We see that in both real and redshift space, reconstruction restores information on smaller scales.
• Figure 5: Scatter in the $\alpha$ values as a function of bias. bias = 1 refers to the mass case. The plot also shows the effective number density ($\rm n_{\rm eff}$) in units of $h^{3}{\rm\;Mpc^{-3}}$ for each HOD set. From the plot, we see that reconstruction dramatically reduces the error in $\alpha$ for all HOD models in both real and redshift space. We note that, the redshift space with reconstruction errors are very close to the real space with reconstruction errors even for the highest biased case. This shows that reconstruction works well for all biased tracers.