On the call: Brendan, Karl, Shaul, Jacques, Maurizio, Kris

Regrets: Eric, Julian

Notes:

• GRASP sims → toast-generated timelines to estimate impact on BB
  • alm's generated for GRASP beams for Andrea
  • Andrea to set up a simulation. Do we have a sky model?
  • Brendan to follow up to see how this fits in with the other efforts.
  • What to compare this to: the noise level and the expected signal level.
• GRASP sims → moon/sun/galaxy avoidance
  • Jacques will attempt to do a 4pi beam convolution check to cross-check his more straightforward results.
• Polarization angle calibration / pointing reconstruction
  • Eric posted a bug fix
  • with 1 arcsecond pointing reconstruction, leakage signals is typically about a factor 10 better than the instrumental noise. noise shown here is for single multipoles. Aiming for r~1e-4 we would want to bin the data and reduce the instrument noise. Would the pointing errror noise also integrate down by binning? In this simulation, it probably would since the pointing errors are assumed to be Gaussian. But reality is another story.
• NERSC sims, esp. related to the calibration simulation Simulation results here
  • initial results: look at the pointings only to see how well the dipole is measured vs. time.
  • There are some time periods where PICO only scans along the dipole equator thus seeing no dipole signal but these only last a few days (compare with Planck where these dead periods lasted many weeks).
  • 6 hour bins were chosen, but other time periods could be chosen. These will be compared to assumptions about the PICO instrument stability and bound
  • DaCapo run on 2-year mission for 1 detector: calibrating in 6 hour signal chunks. Error is 1%, calibration is biased by 0.6% by the Galaxy. Maurizio changing sky masks: these jobs are submitted.
  • Maurizio would like to do naive component separation, i.e. be a little bit fancier than masking the sky.
  • noise: white + 1/f noise (as described in the optimizing scan strategy and the baseline imager definition), but DaCapo implements destriping.
  • input sky: using pysm at 90 GHz. no kinetic dipole.
  • What is the actual deliverable here? These results set limits on how well dipole calibration on short time scales can mitigate systematic errors: perhaps revisit the systematics list in light of these results to see which are actually mitigated.
  • Correlated noise post re-calibration
• next week: