Neutrino mass and dark energy constraints from redshift-space distortions

TL;DR

The study tackles the challenge of constraining the sum of neutrino masses $\\sum m_\\nu$ and the evolution of dark energy using redshift-space distortions in the BOSS DR11 galaxy power spectrum, complemented by Planck CMB and JLA supernova data. It advances a perturbative framework based on Time-Renormalization Group theory, augmented by FAST-PT FFT techniques, to model scale-dependent growth and redshift-space effects with redshift-space multipoles, while marginalizing over scale-dependent galaxy bias. The main result is a 95% CL upper bound of $\\sum m_\\nu < 183$ meV in the $\\nu\\Lambda$CDM context, which degrades to $\\sum m_\\nu < 540$ meV when allowing a time-varying dark energy equation of state $(w_0,w_a)$; the neutrino constraint is thus strongly entangled with dark energy. The work also shows that bias-model variations have limited impact on cosmological parameters, though larger high-k data sets and future tomographic surveys are needed to break the degeneracy and sharpen neutrino-mass bounds, highlighting the critical role of dark-energy knowledge in precision cosmology.

Abstract

Cosmology in the near future promises a measurement of the sum of neutrino masses, a fundamental Standard Model parameter, as well as substantially-improved constraints on the dark energy. We use the shape of the BOSS redshift-space galaxy power spectrum, in combination with CMB and supernova data, to constrain the neutrino masses and the dark energy. Essential to this calculation are several recent advances in non-linear cosmological perturbation theory, including FFT methods, redshift space distortions, and scale-dependent growth. Our 95% confidence upper bound of 180 meV on the sum of masses degrades substantially to 540 meV when the dark energy equation of state and its first derivative are also allowed to vary, representing a significant challenge to current constraints. We also study the impact of additional galaxy bias parameters, finding that a greater allowed range of scale-dependent bias only slightly shifts the preferred neutrino mass value, weakens its upper bound by about 20%, and has a negligible effect on the other cosmological parameters.

Figures (10)

• Figure 1: Perturbation theory with top-town bias models compared with HOD models. (a) MR with $3$ parameters, using linear or Time-RG perturbation theory for $P_{\delta\delta}$. Switching from linear theory to Time-RG reduces errors by factors of $2$-$3$. (b) Time-RG with a simple $3$-parameter bias model $b(k) = (b_0 + b_1 k)/(1 + b_2 k^2)$, which has substantially larger errors than the MR model with $3$ parameters. (c) Time-RG with $5$-parameter MR model. Systematic errors are approximately as large as with the $3$-parameter MR model. (d) Time-RG with $5$-parameter MR model fit in the range $0.005~h/\mathrm{Mpc} < k < 1~h/\mathrm{Mpc}$.
• Figure 2: Time-RG perturbation theory with the MR($3$-param) bias model compared with mock HOD power spectrum monopoles for a model with $\omega_\nu=0.0009$ ($\sum m_\nu = 84$ meV). Yellow bands show BOSS DR11 error bars.
• Figure 3: Same as Fig. \ref{['f:rsd_hod_mock_mono']} but for the quadrupoles.
• Figure 4: Binned, windowed model power spectra vs. BOSS data, for the $\nu\Lambda$CDM model with MR($3$-param) bias and the Planck + BOSS data combination. One hundred power spectra were randomly chosen from the converged portion of the Markov chains and plotted against the data from the northern sky patch. The lower panels show the residuals.
• Figure 5: Similar to Fig. \ref{['f:bestfit_N']} but for the southern sky data.