The full-sky Spherical Fourier-Bessel power spectrum in general relativity

Federico Semenzato, Daniele Bertacca, Alvise Raccanelli

TL;DR

This paper introduces a new theoretical framework for analyzing galaxy clustering on the cosmic lightcone, incorporating relativistic effects from General Relativity (GR) that are crucial for upcoming large-scale sky surveys. They demonstrate that neglecting these GR effects can lead to significant, systematic biases in cosmological parameter estimations, particularly for the non-Gaussianity parameter $f_\mathrm{NL}$.

What Problem Does This Paper Address?

Future galaxy surveys like SPHEREx, Euclid, and Roman will map vast regions of the sky with unprecedented precision. Analyzing this data requires highly accurate theoretical models to extract cosmological information. Traditional methods often rely on approximations like the "plane-parallel" approximation, which is valid only for small sky fractions and linear scales. However, these new surveys will probe large angular scales and volumes where the curvature of spacetime and relativistic effects become significant. These "GR projection effects" distort the observed positions and redshifts of galaxies, and their accurate modeling is essential to avoid systematic errors in cosmological parameter inference. Specifically, the paper addresses the challenge of accurately modeling these effects in a way that is computationally tractable for large sky surveys and quantifies their impact on key cosmological parameters, such as $f_\mathrm{NL}$, which probes deviations from a purely Gaussian early universe.

Methods

The authors extend the Spherical Fourier-Bessel (SFB) formalism, a natural choice for analyzing full-sky data in spherical coordinates, to include all relevant relativistic effects. Key methodological advancements include:

  • Full GR Inclusion: Incorporating all projection and GR effects, moving beyond approximations like the Limber approximation.
  • Efficient Numerical Implementation: Developing a code that handles multi-bin correlations and a full non-diagonal covariance matrix, which is computationally intensive but necessary for accuracy.
  • Novel Cross-Bin Correlation Prescription: Introducing a new method to analyze correlations between different radial bins, which is shown to be crucial for isolating specific GR signals.
  • Best-Fit Shift (BFS) Analysis: Quantifying the systematic bias introduced in parameter estimations when GR effects are neglected, by comparing the results of a full GR model with a simplified Newtonian model.

Datasets

The paper uses simulated data representative of the planned SPHEREx mission for its forecasts. They model the galaxy radial distribution and bias based on publicly available data for SPHEREx. The analysis is generalized to hold for any full-sky survey. No new observational datasets are introduced.

Key Insights and Ideas

  • GR Effects are Not Negligible: Neglecting relativistic projection effects can lead to significant, systematic biases in cosmological parameter estimates, particularly for $f_\mathrm{NL}$. This bias is not a minor correction but can shift the inferred values by non-trivial amounts.
  • SFB Formalism is Powerful: The SFB formalism is well-suited for incorporating wide-angle and relativistic effects, making it a valuable tool for analyzing future large-scale surveys.
  • Cross-Bin Correlations are Key: Analyzing correlations between different redshift bins is essential for isolating the signal of integrated GR effects, such as lensing convergence, which are otherwise masked by local effects in auto-correlations.
  • Limber Approximation Fails on Large Scales: The commonly used Limber approximation breaks down for lower angular scales (larger angular separations), where integrated GR effects are most prominent. A full, non-Limber calculation is necessary for accurate analysis in these regimes.

Key Results

  • The paper demonstrates that neglecting GR effects can lead to a systematic shift in the best-fit value of $f_\mathrm{NL}$. For a SPHEREx-like survey, this shift can be as large as 0.4 times the predicted precision for $f_\mathrm{NL}$ for the high-number-count sample.
  • The magnitude of this bias depends on survey specifications, such as magnification bias and galaxy population properties.
  • The cross-bin correlation analysis is shown to be a clean way to search for relativistic correction signals, particularly for integrated effects like lensing.
  • The Limber approximation systematically underestimates the spherical spectrum in cross-bin correlations, highlighting the need for the full, non-flat-sky formalism.

Limitations and Future Work

  • Computational Cost: The full, non-diagonal covariance matrix calculation is computationally expensive, requiring efficient numerical methods.
  • Model Dependencies: The quantitative impact of the Best-Fit Shift (BFS) is dependent on specific survey properties (e.g., magnification bias, galaxy bias models). While the BFS is a general result, its precise value requires detailed survey information.
  • Future Work: The authors suggest further work on a more extensive detectability analysis, exploring the impact of uncertainties in bias and magnification bias modeling, and applying this formalism to actual observational data. They also mention the need for more efficient numerical methods for integrating the highly oscillatory spherical Bessel functions.

Summary updated

Abstract

We present a formalism for analyzing galaxy clustering on the lightcone with the 2-point correlation in the Spherical Fourier-Bessel (SFB) formalism, which is a natural choice to account for all wide-angle and relativistic (GR) effects. We extend previous studies by including all projection and GR effects, developing an efficient numerical implementation that avoids the use of the Limber approximation, includes multi-bins correlations and a full non-diagonal covariance. Using this formalism, we investigate the impact of neglecting GR corrections, and in particular how much this could bias measurements of the non-Gaussianity parameter $f_\mathrm{NL}$. Our results show that not including relativistic projection terms can systematically and non-negligibly bias estimates of $f_\mathrm{NL}$. The exact results depend on survey specifications and galaxy population properties, but we stress that a bias will generally be present. Finally, we develop a novel prescription for cross-bin correlations that allow to search for a clean signal of relativistic corrections, and show that this requires the use of the 3D full-sky formalism.
BibTeX
@article{semenzato2025full,
  author = {Federico Semenzato and Daniele Bertacca and Alvise Raccanelli},
  title = {{The full-sky Spherical Fourier-Bessel power spectrum in general relativity}},
  journal = {Physical review. D/Physical review. D.},
  year = {2025},
  doi = {10.1103/physrevd.111.023522},
}

Loading related content...

Figure