We forecast the sensitivity of thirty-five different combinations of future Cosmic Microwave Background and Large Scale Structure data sets to cosmological parameters and to the total neutrino mass. We work under conservative assumptions accounting for uncertainties in the modelling of systematics. In particular, for galaxy redshift surveys, we remove the information coming from non-linear scales. We use Bayesian parameter extraction from mock likelihoods to avoid Fisher matrix uncertainties. Our grid of results allows for a direct comparison between the sensitivity of different data sets. We find that future surveys will measure the neutrino mass with high significance and will not be substantially affected by potential parameter degeneracies between neutrino masses, the density of relativistic relics, and a possible time-varying equation of state of Dark Energy.

Theoretical uncertainties on non-linear scales are among the main obstacles to exploit the sensitivity of forthcoming galaxy and hydrogen surveys like Euclid or the Square Kilometre Array (SKA). Here, we devise a new method to model the theoretical error that goes beyond the usual cut-off on small scales. The advantage of this more efficient implementation of the non-linear uncertainties is tested through a Markov-Chain-Monte-Carlo (MCMC) forecast of the sensitivity of Euclid and SKA to the parameters of the standard $\Lambda$CDM model, including massive neutrinos with total mass $M_\nu$, and to 3 extended scenarios, including 1) additional relativistic degrees of freedom ($\Lambda$CDM + $M_\nu$ + $N_\mathrm{eff}$), 2) a deviation from the cosmological constant ($\Lambda$CDM + $M_\nu$ + $w_0$), and 3) a time-varying dark energy equation of state parameter ($\Lambda$CDM + $M_\nu$ + $\left(w_0,w_a \right)$). We compare the sensitivity of 14 different combinations of cosmological probes and experimental configurations. For Euclid combined with Planck, assuming a plain cosmological constant, our method gives robust predictions for a high sensitivity to the primordial spectral index $n_{\rm s}$ ($\sigma(n_s)=0.00085$), the Hubble constant $H_0$ ($\sigma(H_0)=0.141 \, {\rm km/s/Mpc}$), the total neutrino mass $M_\nu$ ($\sigma(M_\nu)=0.020 \, {\rm eV}$). Assuming dynamical dark energy we get $\sigma(M_\nu)=0.030 \, {\rm eV}$ for the mass and $(\sigma(w_0), \sigma(w_a)) = (0.0214, 0.071)$ for the equation of state parameters. The predicted sensitivity to $M_\nu$ is mostly stable against the extensions of the cosmological model considered here. Interestingly, a significant improvement of the constraints on the extended model parameters is also obtained when combining Euclid with a low redshift HI intensity mapping survey by SKA1, demonstrating the importance of the synergy of Euclid and SKA.

In cosmologies with massive neutrinos, the galaxy bias defined with respect to the total matter field (cold dark matter, baryons, and non-relativistic neutrinos) depends on the sum of the neutrino masses $M_{\nu}$, and becomes scale-dependent even on large scales. This effect has been usually neglected given the sensitivity of current surveys, but becomes a severe systematic for future surveys aiming to provide the first detection of non-zero $M_{\nu}$. The effect can be corrected for by defining the bias with respect to the density field of cold dark matter and baryons instead of the total matter field. In this work, we provide a simple prescription for correctly mitigating the neutrino-induced scale-dependent bias effect in a practical way. We clarify a number of subtleties regarding how to properly implement this correction in the presence of redshift-space distortions and non-linear evolution of perturbations. We perform a MCMC analysis on simulated galaxy clustering data that match the expected sensitivity of the \textit{Euclid} survey. We find that the neutrino-induced scale-dependent bias can lead to important shifts in both the inferred mean value of $M_{\nu}$, as well as its uncertainty. We show how these shifts propagate to other cosmological parameters correlated with $M_{\nu}$, such as the cold dark matter physical density $\Omega_{cdm} h^2$ and the scalar spectral index $n_s$. In conclusion, we find that correctly accounting for the neutrino-induced scale-dependent bias will be of crucial importance for future galaxy clustering analyses. We encourage the cosmology community to correctly account for this effect using the simple prescription we present in our work. The tools necessary to easily correct for the neutrino-induced scale-dependent bias will be made publicly available in an upcoming release of the Boltzmann solver \texttt{CLASS}.