Do we have simulated main beams? Eric wants to take a look at the cross-polar beam.
Eric recently learned that in Litebird, the peak of the U beam for GRASP sims of the main beams is something like 10% of the peak of the I beam map. This is much larger than simulations for Planck HFI and LFI (which was < 1% peak) and was thus determined to not be a problem. Being very far off-axis might be the issue.
There are two effects: 1. Calibration will correct any “constant” effect. so how accurately do we need to calibrate? 2. to what effect I/Q/U beams maps are the same.
Can we find a main lobe GRASP simulation for PICO that we can look at? Eric can determine whether the cross-polar response can be corrected by a simple calibration factor. If it's a simple rotation, then no problem. But if there are extra terms, then we might need to have a different approach.
action to Eric: to describe this in more detail next week (building on his previous post).
action to Shaul: talk to Karl to look at main beam GRASP sim for two orthogonal detectors.
Maurizio worked with Andrea. Andrea using PICO as a test case for some “large focal plane” development for toast.
Maps are available for 2 year simulation requested by Kris: /scratch1/scratchdirs/tomasi/PICO_simulations/pico_4detectors_2yrs.fits input map is /global/cscratch1/sd/zonca/pico/cal_sims/pico_cmb_nodip.fits
Andrea was able to patch toast to bring in dacapo input. Allows gain estimation sims: 1. a map of the full sky reconstructed over 2 years and 4 detectors, 2. gain factors vs. time, and 3. a map that has the effects of residual gain errors.
Maurizio can use these products to compute angular power spectra and show the science impact of the residual gain errors. He will do this in the next few days.