This PR will homogeneize the synchrotron models to the latest updates implemented on dust :
input templates the same as before: HASLAM (rescaled to 23GHz and WMAP K-band, smoothed at 2deg)
two pivot scales ell1=36,ell2=400 for small scale injection
spectral indices for EE, BB coming from literature , Planck 2018 :
i. alpha_TT=alpha_TE= -1
ii. alpha_EE = -0.84
iii. alpha_BB= -0.78
Notice that for TT and TE the value is justified by what actually see in the spectra of input maps :
small scales have with non-zero TE
Modulation of qu maps with a single map, p:
i. to avoid modulation w/ negative values,
ii. to preserve non-zero TE
we propose to modulate polarization small scales as it has been done in pysm2(https://arxiv.org/pdf/1608.02841.pdf) , with a couple of differences:
i. small scales are expected to be injected with non-gaussian content (thanks to the logpoltens formalism);
ii. split the sky with high reso pixels (nside=8) ;
iii. use amplitude of E-mode spectra to derive the modulation template
Validation
TT, TE, EE, BB Power Spectra at different GAL masks
E-to- B ratio at different GAL masks
EE BB spectra in small patches
Polarization Fraction pixel distribution
Notebook and final synchrotron maps
Notebook can be found here
Maps available at NERSC :/global/cscratch1/sd/giuspugl/cmbs4/pysm_validation/synchro_pysmmod.fits
Updates on Synchrotron
This PR will homogeneize the synchrotron models to the latest updates implemented on dust :
input templates the same as before: HASLAM (rescaled to 23GHz and WMAP K-band, smoothed at 2deg)
two pivot scales
ell1=36
,ell2=400
for small scale injectionspectral indices for EE, BB coming from literature , Planck 2018 : i. alpha_TT=alpha_TE= -1 ii. alpha_EE = -0.84 iii. alpha_BB= -0.78 Notice that for TT and TE the value is justified by what actually see in the spectra of input maps :
small scales have with non-zero TE
Modulation of qu maps with a single map, p: i. to avoid modulation w/ negative values, ii. to preserve non-zero TE
we propose to modulate polarization small scales as it has been done in
pysm2
(https://arxiv.org/pdf/1608.02841.pdf) , with a couple of differences: i. small scales are expected to be injected with non-gaussian content (thanks to the logpoltens formalism); ii. split the sky with high reso pixels (nside=8) ; iii. use amplitude of E-mode spectra to derive the modulation templateValidation
TT, TE, EE, BB Power Spectra at different GAL masks
E-to- B ratio at different GAL masks
EE BB spectra in small patches
Polarization Fraction pixel distribution
Notebook and final synchrotron maps
Notebook can be found here Maps available at NERSC :
/global/cscratch1/sd/giuspugl/cmbs4/pysm_validation/synchro_pysmmod.fits