Mass-varying neutrino in light of cosmic microwave background and weak lensing

TL;DR

This paper constrains mass-varying neutrino models using large scale structure observations, forecasts Euclid survey capabilities, and finds preferences for higher coupling parameters, with combined data reducing uncertainties.

Contribution

It provides new constraints on mass-varying neutrino models with different potentials and couplings, and forecasts the Euclid survey's ability to improve these constraints.

Findings

Upper limit on $ u$ density parameter: $oxed{ ext{Ω}_ u h^2<0.004}$ at 2-$oxed{ ext{σ}}$ for exponential potential.

Null coupling excluded at >2-$oxed{ ext{σ}}$ for inverse power law potential.

Combined Euclid-like lensing and Planck-like CMB data reduce parameter errors by a factor of ~2.

Abstract

We aim to constrain mass-varying neutrino models using large scale structure observations and produce forecast for the Euclid survey. We investigate two models with different scalar field potential and both positive and negative coupling parameters \beta. These parameters correspond to growing or decreasing neutrino mass, respectively. We explore couplings up to |\beta|<5. In the case of the exponential potential, we find an upper limit on $\Omega_\nu h^2$<0.004 at 2-$\sigma$ level. In the case of the inverse power law potential the null coupling can be excluded with more than 2-\sigma significance; the limits on the coupling are \beta>3 for the growing neutrino mass and \beta<-1.5 for the decreasing mass case. This is a clear sign for a preference of higher couplings. When including a prior on the present neutrino mass the upper limit on the coupling becomes |\beta|<3 at 2-$\sigma$ level for the exponential potential. Finally, we present a Fisher forecast using the tomographic weak lensing from an Euclid-like experiment and we also consider the combination with the cosmic microwave background (CMB) temperature and polarisation spectra from a Planck-like mission. If considered alone, lensing data is more efficient in constraining $\Omega_\nu$ with respect to CMB data alone. There is, however, a strong degeneracy in the \beta-$\Omega_\nu h^2$ plane. When the two data sets are combined, the latter degeneracy remains, but the errors are reduced by a factor ~2 for both parameters.