Abstract: When neutrino masses arise from the exchange of neutral heavy leptons, as in most seesaw schemes, the effective lepton mixing matrix N describing neutrino propagation is non-unitary, hence neutrinos are not exactly orthonormal. New CP violation phases appear in N that could be confused with the standard phase delta(CP) characterizing the three neutrino paradigm. We study the potential of the long-baseline neutrino experiment DUNE in probing CP violation induced by the standard CP phase in the presence of non-unitarity. In order to accomplish this we develop our previous formalism, so as to take into account the neutrino interactions with the medium, important in long baseline experiments such as DUNE. We find that the expected CP sensitivity of DUNE is somewhat degraded with respect to that characterizing the standard unitary case. However the effect is weaker than might have been expected thanks mainly to the wide neutrino beam. We also investigate the sensitivity of DUNE to the parameters characterizing non-unitarity. In this case we find that there is no improvement expected with respect to the current situation, unless the near detector setup is revamped.

Abstract: We study the impact of stochastic noise on the generation of primordial black hole (PBH) seeds in ultra-slow-roll (USR) inflation with numerical simulations. We consider the non-linearity of the system by consistently taking into account the noise dependence on the inflaton perturbations, while evolving the perturbations on the coarse-grained background affected by the noise. We capture in this way the non-Markovian nature of the dynamics, and demonstrate that non-Markovian effects are subleading. Using the Delta N formalism, we find the probability distribution P(R) of the comoving curvature perturbation R. We consider inflationary potentials that fit the CMB and lead to PBH dark matter with i) asteroid, ii) solar, or iii) Planck mass, as well as iv) PBHs that form the seeds of supermassive black holes. We find that stochastic effects enhance the PBH abundance by a factor of O(10)-O(10(8)), depending on the PBH mass. We also show that the usual approximation, where stochastic kicks depend only on the Hubble rate, either underestimates or overestimates the abundance by orders of magnitude, depending on the potential. We evaluate the gauge dependence of the results, discuss the quantum-to-classical transition, and highlight open issues of the application of the stochastic formalism to USR inflation.

Abstract: We revisit the joint constraints in the mixed hot dark matter scenario in which both thermally produced QCD axions and relic neutrinos are present. Upon recomputing the cosmological axion abundance via recent advances in the literature, we improve the state-of-the-art analyses and provide updated bounds on axion and neutrino masses. By avoiding approximate methods, such as the instantaneous decoupling approximation, and limitations due to the limited validity of the perturbative approach in QCD that forced to artificially divide the constraints from the axion-pion and the axion-gluon production channels, we find robust and self-consistent limits. We investigate the two most popular axion frameworks: KSVZ and DFSZ. From Big Bang Nucleosynthesis (BBN) light element abundances data we find for the KSVZ axion Delta N-eff < 0.31 and an axion mass bound m(a) < 0.53 eV (i.e., a bound on the axion decay constant f(a) > 1.07 x 10(7) GeV) both at 95% CL. These BBN bounds are improved to Delta N-eff < 0.14 and m(a) < 0.16 eV (f(a) > 3.56 x 10(7) GeV) if a prior on the baryon energy density from Cosmic Microwave Background (CMB) data is assumed. When instead considering cosmological observations from the CMB temperature, polarization and lensing from the Planck satellite combined with large scale structure data we find Delta N-eff < 0.23, m(a) < 0.28 eV (f(a) > 2.02 x 10(7) GeV) and Sigma m(nu) < 0.16 eV at 95% CL. This corresponds approximately to a factor of 5 improvement in the axion mass bound with respect to the existing limits. Very similar results are obtained for the DFSZ axion. We also forecast upcoming observations from future CMB and galaxy surveys, showing that they could reach percent level errors for m(a) similar to 1 eV.