Delensing LAT relative Nhits v0p1

From CMB-S4 wiki
Revision as of 23:34, 14 December 2020 by Wlwu (talk | contribs) (→‎Results)
Jump to navigationJump to search

This is a 2nd in a series of postings that document the process to answer the question: how should the Nhits map of the Delensing LAT be optimized in order to minimize ?

Previous posting: Delensing_LAT_relative_Nhits_v0p0


This posting is written in response to a discussion from the scan-strategy call: if we have tophat-like hit patterns for the LAT, given fixed total N-hits, how would the delensed A_L change if the tophat pattern covers all the pixels hit by SATs vs the smaller tophat hit pattern given in the [DM reference-design sims ].

Note that in the DM reference-design sims, the LAT (Pole-LAT-MFL1) N-hit pattern doesn't cover all the pixels hit by the Pole SAT-MF(HS1/HS2/LS1/LS2). So the regions outside the LAT hit would not have delensing.

Key takeaways:

  • In the reference-design sims, the LAT hits cover about 50% of the SAT hit pattern; within that region, it gets 85% of the total number of hits.
  • Delensed A_L is ~0.09 in the reference-design LAT region.
  • Delensed A_L is ~0.12 if spreading the LAT hits across all the pixels seen by the SAT.


  • I used information from one LAT and one SAT (both Pole) for extracting all the numbers (using just white noise)
    • LAT: Pole-LAT-MFL1 (93GHz, 2.2 FWHM); the N-hits for the MFs are the same.
    • SAT: Pole-SAT-MFLS1 (85GHz); the N-hits for all the MFs are the same.
  • I calculate the noise spectra from the LAT-MFL1 noise map (/global/cfs/cdirs/cmbs4/dm/dstool/output/s4_reference_design_noise_atmo_7splits/LAT-MFL1_pole/cmbs4_KCMB_LAT-MFL1_pole_nside4096_1_of_1.fits) and estimate the equivalent white noise level in uK-arcmin to be ~0.7uK-arcmin.
  • For the effective white noise level when spreading all the hits uniformly, I scale the uK-arcmin number by sqrt(new_area/old_area).
  • The delensed A_L is from DSR fig.68.


Fig.1 Pole LAT-MFL1 EE/BB noise

1207 LAT-MFL1 nlev ukam.png

A TF corrected N_ell included here (also include comparison with Low-ell BB DC06 sim Pole LAT 95GHz noise).

Fig.2 Pole LAT-MFL1 Nhits

1207 LAT-MFL1 nhits.png

Fig.3 Pole SAT-MFLS1 Nhits cut with LAT-MFL1 tophat

1207 SAT-MFLS1 Nhits bound.png

Fig.4 Pole SAT-MFLS1 map noise uK-arcmin

1207 SAT-MFLS1 map ukam.png

Fig.5 Pole LAT-MFL1 map noise uK-arcmin

1207 LAT-MFL1 map nlev ukam.png