[Paper Review] Unveiling Neutrino Masses: Insights from Robust (e)BOSS Data Analysis and Prospects for DESI and Beyond

H. E. Noriega, Alejandro Avilés|arXiv (Cornell University)|Jul 8, 2024

Neutrino Physics Research4 citations

TL;DR

This paper reveals that projection effects in full-shape galaxy power spectrum analyses significantly bias neutrino mass constraints toward zero or negative values, even when the true mass is positive. By decomposing the power spectrum into broadband and BAO wiggles, the study shows that neutrino mass information is primarily extracted from the suppressed wiggles, not broadband evolution, suggesting that future robust constraints should focus on wiggle amplitudes rather than background expansion models.

ABSTRACT

Recent findings from DESI BAO combined with Planck CMB data have set an upper limit on the total neutrino mass of $\sum m_ν< 0.072\, ext{eV}$ (95% confidence level), ruling out the inverted hierarchy. Indeed, methods that rely on the background expansion of the Universe tend to suggest negative neutrino masses. In this work, we contribute to the quest for accurately constraining neutrino mass using cosmological probes. By conducting a full-shape analysis on data from BOSS, eBOSS, and synthetic power spectra, we showed that projection effects can significantly influence constraints on neutrino mass, rendering these measurements largely unreliable. Our results highlight the need for better techniques to measure the neutrino mass accurately. Based on the large-scale structure suppression, we identified a critical blind spot in the full-shape analysis. By splitting the galaxy power spectrum into broadband and wiggles, we noticed that information on neutrino mass is primarily extracted from the suppressed wiggles rather than broadband suppression. This opens the possibility of developing alternative methods based only on the wiggles of the power spectrum that can be more robust than those heavily reliant on background evolution.

Motivation & Objective

• To investigate why neutrino mass constraints from (e)BOSS full-shape analyses peak at zero or negative values despite physical positivity.
• To assess whether projection effects—arising from parameter degeneracies and prior volume—distort neutrino mass posteriors in large-scale structure data.
• To determine if splitting the power spectrum into broadband and BAO wiggles can isolate more reliable neutrino mass information.
• To propose an alternative method focused solely on the relative amplitude of BAO wiggles as a more robust path to neutrino mass constraints.
• To evaluate the impact of external priors and bias parameter scaling on the stability of neutrino mass constraints.

Proposed method

• Conducted full-shape modeling of the galaxy power spectrum using (e)BOSS data from BOSS DR12 galaxies and eBOSS quasars across multiple redshift bins.
• Split the power spectrum into broadband (smooth) and BAO wiggles (oscillatory) components to isolate distinct sources of information on neutrino mass.
• Applied both Bayesian and frequentist inference to detect and verify the presence of projection effects via posterior profile analysis.
• Generated noiseless synthetic power spectra using Planck 2018 fiducial parameters, including $M_{ u} = 0.06$ eV, to test the robustness of constraints under controlled conditions.
• Used Jeffreys prior to mitigate projection effects and evaluated its impact on posterior distributions.
• Compared constraints from direct bias sampling versus scaled bias parameters ($b_n \sigma^m_8$) to reduce degeneracy-driven biases in $A_s$ and $\sigma_8$.

Figure 1: Comparison of 1-dimensional credible intervals (means and 68 % c.l.) from full-shape analyses of the BOSS power spectrum using different pipelines [ 18 , 19 , 45 ] , assuming a flat $\Lambda$ CDM model with $M_{\nu}=0.06\,\text{eV}$ . Points (stars) with error bars represent results from s

Experimental results

Research questions

• RQ1 Does the peak at $M_{\nu} = 0$ in the posterior distribution of neutrino mass persist in full-shape analyses of (e)BOSS data?
• RQ2 To what extent do projection effects—due to parameter degeneracies and prior volume—distort neutrino mass constraints in full-shape LSS analyses?
• RQ3 Is the information on neutrino mass primarily encoded in the broadband suppression or in the relative amplitude of BAO wiggles?
• RQ4 Can a method based solely on the amplitude of BAO wiggles provide more robust neutrino mass constraints than those relying on background expansion history?
• RQ5 How do external priors and bias parameter scaling affect the stability and reliability of neutrino mass constraints?

Key findings

• The peak at $M_{\nu} = 0$ in the posterior distribution is driven primarily by eBOSS data, which imposes tighter constraints than the combined BOSS redshift bins.
• Synthetic data with $M_{\nu} = 0.06$ eV produced identical posterior profiles to real data, confirming that projection effects—not data systematics—are the root cause of the $M_{\nu} = 0$ peak.
• Neutrino mass information is predominantly extracted from the suppression of BAO wiggles, not from broadband power spectrum suppression, contrary to common assumptions.
• Full-shape analyses relying on background evolution are highly sensitive to degeneracies with cosmological parameters like $\sigma_8$ and $A_s$, leading to unreliable constraints.
• Scaling bias parameters with $\sigma_8$ improves constraints on $A_s$ and $\sigma_8$ but fails to eliminate projection effects on $M_{\nu}$, indicating deeper structural issues.
• Despite weaker constraints, methods based solely on BAO wiggle amplitudes could be more robust than current full-shape approaches, especially under model uncertainties.

Figure 2: Constraints from BOSS $z_{1}$ ( $z_{\text{eff}}=0.31$ ) and $z_{3}$ ( $z_{\text{eff}}=0.61$ ) galaxies, and eBOSS quasars at redshift $z_{\text{eff}}=1.52$ . The upper panel shows the results from fitting the real data, while the bottom panel shows the results from fitting synthetic noisel