ACT (DR6.02) Needlet ILC Maps

Needlet ILC maps of the Compton Y and CMB blackbody signal made from Planck (LFI+HFI) and ACT DR6v4 maps. Made using the method of Coulton et al., Phys. Rev. D 109, 063530 (2024) but with the updated ACT DR6v4 maps.

The files are formatted as follows:

act-planck_dr6.02_nilc_{signal_type}_{deproj_opts}.fits
where signal_type can be either:
   ComptonY: Compton-y maps (in dimensionless units)
   blackbody_T: CMB blackbody temperature components. uK units.
   blackbody_E: CMB E mode polarization map. uK units

The deproj_opts denotes which sky components have been deprojected with a constrained ILC. For example: act-planck_dr6.02_nilc_blackbody_T_deproj_SZ_cib_cibdBeta_1.4_24.0.fits denotes that the tSZ, CIB and the first term in a Taylor expansion around the spectral index beta of the CIB SED ("cibdBeta") are deprojected. The CIB SED that is deprojected and around which the Taylor expansion is performed has a spectral index beta=1.4 and temperature T=24 K.

In addition to the maps, there also three types of masks:

footprint_mask.fits: a slightly more restricted version of the footprint mask used to pre-process the maps. This denotes the well tested sky region ilc_inpaint_mask.fits: Binary mask that denotes the inpainted sources. ilc_subtracted_mask.fits: binary mask that denotes the locations of the sources subtracted in the preprocessing steps.

The maps have a 1.6 arcmin Gaussian beam applied to them.

Note, most of these maps have an effective lmax of ~17000. Maps with 2 deprojections have an lmax of ~11000 and maps with three have an lmax of ~4000 (as we require at least #deprojections+1 frequency channels).

The tSZ map has a lot of residual CIB that can bias some analyses. Here is a suggested framework for testing for this:

Compare a basic ILC to a CIB deprojected ILC: see if there is a systematic shift in your signal. (The noise changes so you expect some scatter). If there is a negligible impact that probably means that there isnt much CIB! If there is a big change try using different spectral indices. If that is stable then you may not need the derivative deproj. If there are big changes with the different indices, then try dBeta. Repeat for different temperatures. If that shows variations add in dT.