This is the repo for the "Optical Depth, Reionization of the Universe and Neutrino Masses" projet paper.
We used two models as mock data (see LB_fiducials_mock)
-
Planck: Planck data are spread into three different likelihoods following the original version of the Planck likelihoods.-
Planck lowT: likelihood based on the PR3 low-ℓ TT posterior distributions of measured$C_\ell$ fitted with a Wishart ($\alpha_\ell, \beta_\ell$ ). We then rescaled to the mock fiducial model with$\beta_\ell \longrightarrow \beta_\ell^{\rm mock} = (\alpha_\ell+1)C_{\ell,{\rm mock}}^{TT}$ . Multipole range available$\ell \in [2,29]$ . -
Planck lowE: Hamimeche&Lewis approximation using the Planck PR4 covariance matrix (LoLLiPoP) evaluated at the mock fiducial spectrum. Multipole range available$\ell \in [2,29]$ . -
Planck high-ℓ (TT,TE,EE): Gaussian likelihood for TT , TE and EE power spectra, with the full covariance matrix from the PR3 Planck-lite likelihood (marginalised over foreground parameters). Multipole range available$\ell \in [30,2508]$ .
-
-
LiteBIRD: Wishart likelihood for TT , TE, and EE spectra, using beam-deconvolved noise spectra including foreground residuals after ILC and an effective sky fraction fsky ∈ {0.7, 0.8} from E-mode group. Multipole range available (depending on the noise file). -
Simons-Observatory: Wishart likelihood with noise curves including foreground residuals. Multipole range available (depending on the noise file). -
CMB-S4: Wishart likelihood with noise curves including foreground residuals. Multipole range available (depending on the noise file).
Noises are described in LB_fiducials_noises
We fit the data with 4 different models:
-
tanh: tanh model with$z_{reio}$ ($\Delta z$ fixed) -
tanhdz: tanh model with$z_{reio}$ and$\Delta z$ -
flex1: FlexKnot model with 3 knots ( (z1=0,xe1=1.08), (z2,xe2=1.08), (z3,xe3=0) ) -
flex2: FlexKnot model with 4 knots ( (z1=0,xe1=1.08), (z2,xe2=1.08), (z3,xe3), (z4,xe4=0) )