Cosmology of neutrinos and extra light particles after WMAP3

TL;DR

This study uses a comprehensive cosmological data set to constrain standard and non-standard neutrino properties and possible extra light particles. By implementing a flexible, Gaussian-approximate analysis framework and bespoke cosmological tools, the authors quantify how freely-streaming versus interacting relativistic species, and neutrino–boson couplings, shape the cosmic perturbation evolution. The key results show a bound of $\sum m_\nu < 0.40\,\mathrm{eV}$ (99.9% C.L.) and an effective relativistic density of $N_\nu = 5 \pm 1$ for freely-streaming extras, while interacting extra massless species are limited to $\triangle N_\nu = 0 \pm 1.3$, with scenarios featuring three interacting neutrinos strongly disfavored at about $4\sigma$; Ly$\alpha$ data critically influence several bounds. Overall, the work highlights the data-driven tension with LSND-like sterile neutrino interpretations and demonstrates the sensitivity of cosmology to the microphysics of light relics and neutrino interactions.

Abstract

We study how present data probe standard and non-standard properties of neutrinos and the possible existence of new light particles, freely-streaming or interacting, among themselves or with neutrinos. Our results include: sum m_nu < 0.40 eV at 99.9% C.L.; that extra massless particles have abundance Delta N_nu = 2 pm 1 if freely-streaming and Delta N_nu = 0 pm 1.3 if interacting; that 3 interacting neutrinos are disfavored at about 4 sigma. We investigate the robustness of our results by fitting to different sub-sets of data. We developed our own cosmological computational tools, somewhat different from the standard ones.

Figures (8)

• Figure 1: Our computation of CMB and LSS spectra in standard $\Lambda$CDM cosmology, compared with data.
• Figure 2: Difference between our code and CAMB, at the standard-cosmology best-fit point for $m_\nu=0$ (red solid line) and for $m_\nu=0.5\,{\rm eV}$ (blue dashed line). Our code does not employ any approximation specific for standard cosmology. In both codes various parameters allow the user to increase the accuracy; this plot holds for the choice employed in the present paper. The dotted line shows the $1\sigma$ accuracy obtained by WMAP3 results (binned data), indicating that we have a good enough accuracy (as confirmed by other tests). A similar $\%$-level accuracy is found for the TE and EE CMB spectra, that presently are measured with much larger uncertainties than the TT spectrum.
• Figure 3: Fit of cosmological data at 68, 90 and 99% C.L. The shaded areas show our global fit without Lyman-$\alpha$, and the dotted lines our WMAP3-only fit, such that this figure can be directly compared with the analogous WMAP Science Team plots in fig. 10 of WMAPparam.
• Figure 4: Fit as function of the energy density in freely-streaming slow particles, parameterized by the sum of active neutrino masses, assumed to have the standard abundance $N_\nu=3.04$. We studied different combinations of data-sets, as indicated by the legend.
• Figure 5: Fig. \ref{['fig:NNu']}a): fit as function of the energy density in freely-streaming relativistic particles, parametrized by the usual 'number of neutrinos' $N_\nu$. Fig. \ref{['fig:NNu']}b): fit as function of the energy density in extra interacting relativistic particles, with abundance parametrized by $\Delta N_\nu$. We studied different combinations of data-sets, as indicated by the legend.