Reconstructing the Baryonic Acoustic Oscillations in the presence of photo-$z$ uncertainties

Kwan Chuen Chan, Guo‐Yuan Lu, Xin WangarXiv 2023 · DOI · arXiv

TL;DR

Cosmologists use Baryonic Acoustic Oscillations (BAO) as a "standard ruler" to measure the universe, but nonlinear gravity blurs this ruler over time. This paper introduces a new method to "sharpen" the BAO signal in photometric surveys—where distance measurements are notoriously fuzzy—by adapting 3D reconstruction techniques into a streamlined 2D transverse approach.

What Problem Does This Paper Address?

The Baryonic Acoustic Oscillations (BAO) are essentially "frozen" sound waves from the early universe, visible today as a slight preference for galaxies to be separated by a specific distance (the sound horizon). This makes them a premier tool for measuring the expansion history of the universe.

However, two major hurdles exist:

  1. Nonlinear Evolution: Over billions of years, gravity moves matter around, "smearing" the BAO signal and making the standard ruler harder to read.
  2. Photo-z Uncertainties: While spectroscopic surveys provide precise 3D positions, they are expensive. Photometric surveys (using colors to estimate distance) cover much larger volumes but have significant errors in the radial (line-of-sight) direction, known as photo-$z$ uncertainties.

While "reconstruction" algorithms—which mathematically move galaxies back toward their original positions—are standard for spectroscopic data, they haven't been widely applied to photometric data because the radial "smearing" from photo-$z$ errors is so severe.

Methods

The authors adapt the Zel’dovich reconstruction technique for the photometric realm. In a standard spectroscopic survey, you solve a 3D Poisson equation ($\nabla^2 \phi = \delta$) to find the displacement potential.

The key innovation here is the Transverse Potential Equation. Since photo-$z$ errors effectively destroy information along the line of sight, the authors show that the reconstruction can be treated as a 2D problem within "tomographic slabs."

They tested three variations:

  • 3D Inversion: Solving the full 3D equation in photo-$z$ space.
  • Full 2D Equation: Solving for the potential using both surface density and a radial potential term.
  • Density-only 2D: A simplified version where only the 2D surface density of the slab drives the reconstruction.

Surprisingly, they found that the Density-only 2D approach works remarkably well because the radial terms are naturally suppressed by the photo-$z$ smearing itself.

Datasets

The study utilized comoving N-body simulations to validate the method:

  • Simulation Specs: A flat $\Lambda$CDM model in a cubic box with a side length of 1000 Mpc/$h$ containing $512^3$ particles.
  • Tracers: They analyzed both the underlying dark matter field and "halos" (clusters of dark matter where galaxies live).
  • Synthetic Photo-z: They injected realistic Gaussian noise ($\sigma_z = 0.03$) into the radial coordinates to mimic real-world imaging surveys like DES or LSST.

Key Insights and Ideas

  • The 2D Shortcut: The most significant "aha moment" is that you don't need complex 3D math to fix photometric data. Because the radial information is already "washed out" by photo-$z$ errors, the transverse (side-to-side) reconstruction is almost entirely driven by the 2D map of the galaxy distribution on the sky.
  • Coherence Length Matters: The reason reconstruction is less effective for photo-$z$ data than for spec-$z$ data is that the "displacement field" (the map of how things moved) has a shorter coherence length in photo-$z$ space. Essentially, the errors make the movement look more chaotic/noisy.
  • RSD is Negligible: In spectroscopic surveys, Redshift Space Distortions (RSD) are a major headache. In photometric surveys, the photo-$z$ error is so much larger than the RSD effect that the latter can be largely ignored during the reconstruction process.

Key Results

  • Signal Enhancement: For the matter field at $z=0$, the reconstruction successfully sharpened the BAO peaks in the power spectrum, increasing the signal strength by about 0.5%.
  • Improved Correlation: The "propagator" (a measure of how well the current field correlates with the initial universe) showed significant gains. At $z=0$, the cross-correlation improved by a factor of 1.2 at $k_\perp = 0.2$ and 1.4 at $k_\perp = 0.3$.
  • Method Agreement: The simplified 2D density-only reconstruction yielded results nearly identical to the full 3D inversion, proving that the streamlined 2D approach is sufficient for future survey pipelines.

Limitations and Future Work

  • Shot Noise in Halos: While the method worked perfectly on the dense dark matter field, the results for the halo samples were noisier. The authors note that the halo density in their simulation was too low to fully showcase the gains, suggesting that denser samples (like those expected in the DES Y6 data) will be needed for a clear detection.
  • Curved Sky: This study used the "plane-parallel approximation" (assuming the sky is flat). For massive, wide-area surveys, the formalism will need to be updated to account for the curvature of the sky.
  • Advanced Algorithms: The authors focused on the standard Zel’dovich approximation. They suggest that more advanced, iterative reconstruction methods could potentially squeeze even more information out of photometric maps.

Reference: Chan, K. C., Lu, G., & Wang, X. (2024). Reconstructing the Baryonic Acoustic Oscillations in the presence of photo-z uncertainties. MNRAS.

Summary updated

Abstract

The reconstruction method has been widely employed to improve the Baryon Acoustic Oscillations (BAO) measurement in spectroscopic survey data analysis. In this study, we explore the reconstruction of the BAO signals in the realm of photometric data. By adapting the Zel'dovich reconstruction technique, we develop a formalism to reconstruct the transverse BAO in the presence of photo-$z$ uncertainties \change{under the plane-parallel approximation}. We access the performance of the BAO reconstruction through comoving $N$-body simulations. The transverse reconstruction potential can be derived by solving a 2D potential equation, with the surface density and the radial potential contribution acting as the source terms. The solution is predominantly determined by the surface density. As is evident in dense samples, such as the matter field, the transverse BAO reconstruction can enhance both the strength of the BAO signals and their cross correlation with the initial conditions. At $z=0$, the cross correlation is increased by a factor of 1.2 at $ k_\perp = 0.2 \, \mathrm{Mpc}^{-1}h $ and 1.4 at $ k_\perp = 0.3 \, \mathrm{Mpc}^{-1}h $, respectively. We contrast the 2D potential results with the 3D Poisson equation solution, wherein we directly solve the potential equation using the position in photo-$z$ space, and find good agreement. Additionally, we examine the impact of various conditions, such as the smoothing scales and the level of photo-$z$ uncertainties, on the reconstruction results. We envision the straightforward application of this method to survey data.
BibTeX
@misc{chan2023reconstructing,
  author = {Kwan Chuen Chan and Guo‐Yuan Lu and Xin Wang},
  title = {{Reconstructing the Baryonic Acoustic Oscillations in the presence of photo-$z$ uncertainties}},
  year = {2023},
  eprint = {2311.12611},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2311.12611},
}

Loading related content...

Figure