Ready for delensing use lensing maps 02.00

From CMB-S4 wiki
Revision as of 15:20, 30 September 2018 by Jcarron (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

J. Carron Sept 30 2018.

This page describe the 300 lensing tracers maps (healpy alm fits files, with lmax=3000) available at


These maps were directly built from the quadratic estimated raw lensing maps Lensing reconstructions 02.00, after Wiener-filtering and mean-field (contamination of the lensing estimate by mask and noise anisotropies) subtraction, as described below. They can be used as is for delensing. The lensing reconstructions were themselves built from this data: Sim map sets to demonstrate "real delensing" (02.00_and_02.09)

There are 3 types of maps:

  • sim_ptt_????_lmax3000.fits

built from temperature data only, with ???? from 0000 to 0099

  • sim_p_p_????_lmax3000.fits

built from polarization data only,

  • sim_p_????_lmax3000.fits

built from combination and temperature and polarization data (MV).


The Wiener-filter C / C + N0 is isotropic, and is given by the last column of the files

   clwf_{ptt, p_p, p}.dat

in the same folder. The noise spectra are obtained empirically, by taking (twice) the spectrum of a lensing reconstruction where the two legs of the quadratic estimator are independent, and averaging over 100 such reconstructions.


To subtract the mean-field from one lensing estimate, we have subtracted the average of the lensing reconstructions, excluding that very estimate (resulting in the average of 99 lensing maps). This adds a small, ~1% contribution to the reconstruction noise which is neglected in the Wiener-filter. This procedure also causes the simulations to be not exactly independent anymore, even if this is a small effect.