status of beams l. page & c. barnes et al.. why do we need to model the beams? computing solid...
TRANSCRIPT
Status of Beams
L. Page & C. Barnes et al.
Why do we need to model the beams?
• Computing solid angles (T_jup) & windows from the raw maps depends on the cut-off. How one treats the transition to noise dominates the error on the answer.
• Fitted beams are conditioned by correlation length of deformations. Calculations are easier and errors can be assessed. They also allow one to understand the deformations in the beam.
• Fits give SA, amplitude, and beam shape.
Contours: 0.09, 0.3, 0.6, 0.9
Contours: 0.06, 0.09, 0.3, 0.6, 0.9
Contours: 0.03, 0.06, 0.09, 0.3, 0.6, 0.9
K
Ka, Q
V, W
Solid angle most difficult thing to get right.
Baseline subtraction critical.
A/B integrals of beam as a function of angle
Integrated offset
Beams “clean” beyond 1.1-1.4 deg in agreement with measured correlation length and Ruze theory.
Can restrict all fits to inside 1.1 to 1.4 deg --based on physical arguments
AP_INT_V5 Status
• All W-band beams processed.
• 17 parameter fit, 14 for surface.
• Complicated covariance structure.
• Different surface for each beam.
• Can be misguided by near-field current distribution.
Compare symmetrized beams
Model vs “Raw” pixelized beams
0.017
0.014
Rms
Lambda/7
All 32 W detectorsEvery two should be the same
Expect a slope in the spectrum but the maps are not quite there.
Need fits to individual channels
JanetRaw sym
Raw
Model
W-band window comparison
0.23 deg
Symmetry conjecture….
Other Issues & Next steps• Pointing: A/B sides are not back to back
symmetric. Almost like TRS is split open.
• Bin on 0.3 deg and use pixel win to check the low l parts of the windows.
• AP_INT routines on smaller target region.
• Polarization/3/4 consistency.
• Get better beams. Rt 2 matters here! Offset subtraction is critical.
• Full surface modeling with DADRA (Chris)