23 Oct 2012 BICEP2 weekly telecon
On the line:
Stanford: Jamie T., Walt
San Diego: Angiola, Jon, Brian
Harvard: Colin, Abby
Caltech: Randol, Jeff, Jamie B.
Minnesota: Clem, Chris, Stefan
1) Weekly report from Pole [ Steff ]
- Thermal test Oct. 16
-- how did it look?
2) Data quality / reduc czar report [ Jon ]
* tags 20121014C10 through 20121022F10
* general: bad weather
* tags of interest:
- Bad WX, but not cut: 20121017D03
change in el nod gain; elnod_fracdel is a loose cut
- Looks good, but gets cut: 20121020F09
stationarity_ab removes everything
stationarity_ab is on p3 filtered a, b;
not shown in reducplots
- slanted el nod: 20121014H05
* Next week’s czar [ Jon ]
3) Testing and deployment planning
* Thermal testing [ Jon ]
- Recent data looks great
- Sent new PID testing schedule to Steff
- One additional test pending, waiting approval from John
- Diagnose crosstalk: thermal or electrical?
- limits precision of thermal filter suppression
for freq > 10 mHz
- that’s OK, gives us precision we need
* Deployment / measurement plan [ Colin, Randol ]
https://docs.google.com/spreadsheet/ccc?key=0Ao4UxezLygx5dExiM2U4RHFyZmVFLU…
- Follow up for far sidelobe testing [ Randol, Abby ]
- want meas at two attenuation settings
- ideas for alignment of rotating polarized source
- use open-ended waveguide instead of a horn?
* FTS plan [ Zak ]
- Coming together nicely at Caltech
- Taking data today, pack up tomorrow for shipping
4) Analysis status reports
Baseline 2yr data reduction updates [ Walt ] brief!
- El nod goodness of fit cut: changes in progress [ Martin, Angiola ]
- offline NTD cal ready [ Walt ]
- tweak_elnods changes, need to reprocess a few tags [ Walt ]
http://bicep0.caltech.edu/~spuder/analysis_logbook/analysis/20121019_tweak_…
- new cutparams files, plot cuts pending [ Angiola ]
Characterizing non-beam systematics [ Walt / Angiola ] brief!
- relgain mismatch from per-channel abscal still in progress [ Walt ]
Addendum: first pass finished.
revealed indexing problem in A/B maps.
Code has been patched, now re-running.
- crosstalk
- thermal
- magnetic: dark scan analysis [ Angiola ]
- applied basic cut; many chans removed entirely
- may need additional data
Characterizing beam systematics [ Abby (/Randol) ]
- Abby and Randol working out details of test plan
- Testing plan for additional far sidelobes
Deprojection / beamsim algorithm coding status report [ Chris, Clem ]
* deproj sim test of linearity: add timestream noise
- Done. Shown to be linear (posting from last week)
http://bicep0.caltech.edu/~spuder/analysis_logbook/analysis/20121015_sim123…
* deproj sim test of beam width / ellipticity
- gen_templates mod in CVS to make templates for this
- Stefan has run sims: posting now up
Data consistency check status report [ Chris (/Randol) ]
* B2 2-year deck jack maps cleaned up dramatically with deproj
http://bicep0.caltech.edu/~spuder/analysis_logbook/analysis/20121015_sim123…
- improvement in real data jack maps is good match to sims
* Comparison of recovered deproj coeffs w/ measured beams
- Randol has deproj coefficients
- In progress
* full jack chi2's, sum_bp_devs, etc.
- 50-realization simset now running on B2 2010+11 data.
- current code and TOD’s, accumulating deproj templates
- sims include diff point in full simset
- bandpowers available for next week’s telecon
Foregrounds status report: [ Jamie T ]
* PSM model for polarized dust and synchrotron
- ready for re-observing with B2 pipeline.
- Emailed Jacques about whether PSM will include
Planck ERCSC...haven't heard back.
- will probably have to run as separate catalogue
and compare to the PSM prediction.
- Will try this out first for B1.
Statistics for reporting bandpowers, r constraints [ Colin / John ]
* Current work: Posting in B1 analysis logbook:
http://bicep0.caltech.edu/~bicep/analysis_logbook_north/20121019_toy_model
- compare method in toy model: direct calc, quadratic estimator,
offset-lognormal, Hamimeche-Lewis
- offset-lognormal method not a good match
* Planned work:
- Immanual to extend “toy model” to include:
- freq cross spectra, TE cross spectrum, E->B leakage.
- Colin refining rho statistic (quadratic estimator)
* Eventual goal
- Use Hamimeche-Lewis for bandpower likelihoods
(CosmoMC compatible)
- Use quadratic estimator rho for limits on r
- Use H-L bandpower likelihoods for 95% CL BB band powers