First steps to sim input maps
Dec 1 2016, Clem Pryke
Updated Dec 14 - fix E/B in dust map
Here are some plots showing first steps towards making signal and noise maps to be used for forecasting sims. The initial goal is to make maps which ought to reproduce the results in the Science Book when run through power spectrum estimation and likelihood analysis. An ellmin of 30 is now imposed in all cases to simulate filtering of ground based experiments.
NOISE: I have setup code to output arbitrary N_l spectra into a fits file, run synfast to make a sets of alms up to l=4096, and then run synfast again to render these for a given Heaplix nside. Here is the first realization rendered on nside=512 (with lmax=2048). I am now waiting on Victor and Colin to provide sensible N_l spectra.
LCDM: Jullian provided sets of alms for unlensed and lensed-LCDM. Rendering the first of these for nside=512 (lmax=2048) we get:
GAUSSIAN DUST: To estimate the effect of dust on r constraints we need multiple realizations of dust. This is a problem as we don't really know how to generate that. Up to now in BK analysis we have used Gaussian dust realizations. We can use the same method as for the noise sims to generate Gaussian realizations which have the PIPXXX spectral shape, and the A_d normalization which is the best fit in the BK field:
APODIZATION MASK: For a start we can use the apodization mask as used in the BK14 analysis. Interpolating this to Healpix, writing to file and re-plotting we get:
MASKED NOISE: In a real experiment with limited sky coverage the observed noise of course blows up around the edge of the field. We can simulate this by taking the full sky noise map from above and dividing by the square root of the apodization mask.