the GNILC large-scale template used in the PySM: unires or varres?

[yaojian95]

Hi Andrea,

I'm calculating the full-sky power spectra of map from last November /global/cfs/cdirs/cmb/www/pysm-data/dust_gnilc/![image](https://user-images.githubusercontent.com/14847517/216408071-033685cd-e921-4c59-b6f8-698b53ddbe99.png) and it's shown in the plot (ignore the titles)

and also this map but transformed into small iqu quantities then calculated the full-sky power spectra

Then I also use the latest map from Giuseppe /global/cscratch1/sd/giuspugl/cmbs4/pysm_validation/dust_gnilc_varres_pysm_circles_uKcmb.fits to calculate the full-sky power spectra

also transformed into small iqu quantities

From what I know, besides the modulation to the small scales, there should not be other differences between the gnilc_dust_template_nside2048.fits and dust_gnilc_varres_pysm_circles_uKcmb.fits. But the power spectra from gnilc_dust_template_nside2048.fits seems to show that this map is based on GNILC-unires large-scale templates, which can be shown if we have a look at the power spectra the Peppe had in his slides:

Can you check that if you use varres or unires large-scale template to generate the maps?

[zonca]

see https://pysm3.readthedocs.io/en/latest/preprocess-templates/small_scale_dust_pysm3.html

cell 11 and below.

For temperature we are using a custom map at 21.8’, for polarization we were supposed to use varres.

However, there is a bug in cell 13:

    !wget -O $dust_varresP http://pla.esac.esa.int/pla/aio/product-action?MAP.MAP_ID=COM_CompMap_IQU-thermaldust-gnilc-unires_2048_R3.00.fits

it looks like it is downloading unires even for the varres case. Thank you very much for catching this bug, I'll make sure I fix it when I regenerate the templates with the new small scales.

[zonca]

This fix is included in https://github.com/galsci/pysm/pull/150