Publications

We develop a generic framework within CLASS that allows for one dark matter species to have simultaneous interactions with photons, baryons, and dark radiation. We use this framework to show that the effect of these interactions on the CMB is largely additive, with no clear degeneracies in the parameters. We also show that the combination of dark matter – photons and dark matter – dark radiation interactions can reduce both the S₈ tension and the H₀ tension. This is the release paper of CLASS v3.2.

We consider a scenario where dark matter has cannibalistic number changing processes (3->2), which suppress the growth of structure on small scales. We implement this model in CLASS to study its cosmological implications. We show that this model can accommodate low S₈ values while remaining compatible with CMB+BAO data, providing a solution to the S₈ tension.

Using the spectral distortions pipeline we previously developed, we study how constraints and degeneracies between cosmological parameters evolve while varying the proposed sensitivity of spectral distortion experiments. This allows us to address the question of which sensitivity we need for such an experiment in order to improve upon CMB anisotropy constraints.

We consider non-cold dark matter produced by the evaporation of primordial black holes, which would only interact gravitationally but could be warm enough to be subject to structure formation constraints. We perform a detailed calculation of the dark matter phase-space distribution, which we then interface with CLASS to extract the corresponding matter power spectrum. By remapping warm dark matter bounds, we are able to derive constraints from Lyman-α data without needing to perform new N-body simulations.

We study a model of dark matter – dark energy interactions with a coupling term that is linear in the energy density of the dark energy, which has been proposed to solve the H₀ tension. We show that when including BAO and Pantheon data, the model does not solve the tension, but can alleviate it. We also find no preference in the data for this model over ΛCDM.

We incorporate the calculation of spectral distortions into the Boltzmann code CLASS, and develop accompanying mock likelihoods for future spectral distortion missions for MontePython. We use this framework to forecast the achievable constraints with such a mission on the ΛCDM parameters, on decaying dark matter, and on evaporating primordial black holes. This is the release paper of CLASS v3.0.

The combination of BAO and BBN data can probe the background cosmological history independently of CMB and supernovae data. We use this combination to show that when assuming ΛCDM these two datasets are in 3.2σ tension with the SH0ES measurement of H₀. We further show that increasing the radiation density parameter N_eff can only reduce the tension down to the 2.6σ level, therefore not fully resolving the tension.

We develop a new method to extract information from Lyman-α forests which does not require new N-body simulations, but rather relies on interpolating in a grid of pre-existing simulations. We use this method to constrain several different dark matter – dark radiation interactions, placing stringent bounds on the interaction strength and amount of dark radiation. This is the release paper of CLASS v2.9.

We forecast the expected sensitivity to the sum of neutrino masses for seven CMB missions or combinations thereupon, five large scale structure experiments, and using four fiducial cosmological models, resulting in a total of 140 different scenarios. We show that a neutrino mass detection from cosmology in the next decade is very likely when using a combination of ground-based and satellite missions.