The clustering of galaxies in the SDSS-III Baryon Oscillation Spectroscopic Survey: Signs of neutrino mass in current cosmological datasets

Abstract

We investigate the cosmological implications of the latest growth of structure measurement from the Baryon Oscillation Spectroscopic Survey (BOSS) CMASS Data Release 11 with particular focus on the sum of the neutrino masses, $\sum m_ν$. We examine the robustness of the cosmological constraints from the Baryon Acoustic Oscillation (BAO) scale, the Alcock-Paczynski effect and redshift-space distortions ($D_V/r_s$, $F_{\rm AP}$, $fσ_8$) of \citet{Beutler:2013b}, when introducing a neutrino mass in the power spectrum template. We then discuss how the neutrino mass relaxes discrepancies between the Cosmic Microwave Background (CMB) and other low-redshift measurements within $Λ$CDM. Combining our cosmological constraints with WMAP9 yields $\sum m_ν = 0.36\pm0.14\,$eV ($68\%$ c.l.), which represents a $2.6σ$ preference for non-zero neutrino mass. The significance can be increased to $3.3σ$ when including weak lensing results and other BAO constraints, yielding $\sum m_ν = 0.35\pm0.10\,$eV ($68\%$ c.l.). However, combining CMASS with Planck data reduces the preference for neutrino mass to $\sim 2σ$. When removing the CMB lensing effect in the Planck temperature power spectrum (by marginalising over $A_{\rm L}$), we see shifts of $\sim 1σ$ in $σ_8$ and $Ω_m$, which have a significant effect on the neutrino mass constraints. In case of CMASS plus Planck without the $A_{\rm L}$-lensing signal, we find a preference for a neutrino mass of $\sum m_ν = 0.34\pm 0.14\,$eV ($68\%$ c.l.), in excellent agreement with the WMAP9+CMASS value. The constraint can be tightened to $3.4σ$ yielding $\sum m_ν = 0.36\pm 0.10\,$eV ($68\%$ c.l.) when weak lensing data and other BAO constraints are included.

Figures (11)

• Figure 1: Comparison between the likelihood distributions in $\Omega_m$-$\sigma_8$ within $\Lambda$CDM. We show Planck Ade:2013zuv (brown contours), Planck SZ clusters Ade:2013lmv (magenta contours), CFHTLenS Kilbinger:2012qz (grey contours), galaxy-galaxy lensing Mandelbaum:2012ay (green contours) and CMASS-DR11 Beutler:2013b (orange contours). The Planck contours in this plot assume $\Lambda$CDM and $\sum m_{\nu} = 0.06\,$eV.
• Figure 2: Relative amplitude difference between a linear power spectrum monopole (top) and quadrupole (bottom) with $\sum m_{\nu} = 0\,$eV (black lines) and $\sum m_{\nu} = 0.4\,$eV (red lines). We keep $\Omega_ch^2$ fixed when including the neutrino mass, so that the total physical matter density increases as $\Omega_mh^2 = \Omega_ch^2 + \Omega_bh^2 + \Omega_{\nu}h^2$. The black dashed lines show the fitting range for the CMASS-DR11 results of Beutler:2013b. We subtract $0.5$ from the quadrupole for plotting purposes.
• Figure 3: Scale-dependence of the growth rate for different values of the neutrino mass parameter. The black line shows the commonly used linear assumption, while all other lines are obtained as derivatives of the growth factor $D(k, a)$ using a camb power spectrum. In this figure we fix $\Omega_mh^2$ when increasing the neutrino mass.
• Figure 4: Results of the reliability tests for the CMASS and CFHTLenS constraints. The red contours include a neutrino mass of $\sum m_{\nu} = 0.4\,$eV in the modelling, while the black contours assume $\sum m_{\nu} = 0\,$eV. (left) Here we show the Alcock-Paczynski parameter $F_{\rm AP}$ and the growth rate $f\sigma_8$ from CMASS-DR11, which are the two parameters most affected by the change in the neutrino mass parameter (in the analysis we also include the BAO scale via $D_V/r_s$). The crosses mark the maximum likelihood values. (right) Here we show the $\Omega_m$-$\sigma_8$ constraints of CFHTLenS including the degeneracy line used in our analysis and reported in Kilbinger:2012qz (black dashed line). The black contours show the original fitting results using the CosmoPMC implementation of Kilbinger:2011bu, while the blue contours use a modified code with the biggest difference being the new halofit implementation of Bird:2011rb (see text for details).
• Figure 5: Two-dimensional likelihood for $\Omega_m$-$\sigma_8$ (left) and $\sum m_{\nu}$-$\sigma_8$ (right) when combining the WMAP9 MCMC chain within $\Lambda$CDM and free $\sum m_{\nu}$ with different low redshift growth of structure constraints. The orange contours show WMAP9+Beutler2013, where Beutler2013 stands for the constraints on $D_V/r_s$, $F_{\rm AP}$ and $f\sigma_8$ reported in Beutler:2013b. The green contours show WMAP9+Beutler2013+CFHTLenS. The results are summarised in Table \ref{['tab:para2']}.