PySM pysm3 v3.4.0b6 dust models validation

[erussier]

I did the comparison between the PySM bandpass integrated dust models d9, d10, d11 and d12 with Planck's 3rd data release at 353 GHz and nside 2048.

Bandpass used:
HFI_RIMO_R3.00.fits

Observations used:
Noise simulation: ffp10_noise_353_psb_full_map_mc_00000.fits
PR3 data: HFI_SkyMap_353-psb-field-IQU_2048_R3.00_full.fits
PR3 Half mission 1: HFI_SkyMap_353_2048_R3.01_halfmission-1.fits
PR3 Half mission 2: HFI_SkyMap_353_2048_R3.01_halfmission-2.fits

You can see the summary of the results of this analysis in these slides.
The code where I do the bandpass integration is here, which you would need to run first if you want to launch my analysis notebook and the notebook where I did my analysis is here.

For the power spectra analysis, I used two different apodized masks for the intensity and for the polarization that you can see in slide 33.
This is what I get for the EE, BB power spectra for d9, d10, d11, d12, PR3 and PR3 HM1xHM2:

[erussier]

I also did the same analysis at 217GHz and 857GHz between PR3 data, d9 and d10 models at nside 2048.
I did the bandpass integration of the d9 and d10 models, you can find the python scripts here for 217GHz and here for 857GHz.

I changed the resolution of the d9, d10 and PR3 maps to 30 arcmin for the intensity, polarized intensity plots in these scripts, here for 217GHz and here for 857GHz.

Here are the notebooks where I did the comparison between PR3 data, d9 and d10 models:
for 217GHz
for 857GHz.

You can find the summary of the results and detail of my analysis in these slides for 217GHz and 857GHz.

This is what I get for the EE/BB power spectra at 217GHz for d9, d10, d9+noise, PR3 and PR3 HM1xHM2:

and this is what I get for the TT power spectra at 857GHz for d9, d10, d9+noise, PR3 and PR3 HM1xHM2:

Note: The y axis unit is wrong in the last plot, I did my analysis in MJy/sr for 857GHz and uKcmb for 217GHz.