Link to main foregrounds telecons page
Link to logbook
Attendance: Ragnhild, Mathieu, Jacques, Shaul
Regrets:
Notes: Shaul
-
She is working with 90.91, r=0 and r=1e-3.
Slide 3: r=0.003 (note typo), CMB map is much cleaner compared to last week. This is due to tweaking parameters of underlying foreground model. (compare to slide 3 of two weeks ago.)
comparison of Slides 4,5 shows that dust temperature is reconstructed well. PICO is very good in extracting dust temperature. Not clear what's happening with dust amplitude. Is it a different color scale or really amplitude not reconstructed well?
Slides 6-9 comparison of reconstruction of other foreground parameters
Slide 10: chi^2. JD: chi^2 of what? A: of residual maps; Q: what units? MR says unitless.
Slide 11: power spectrum. Commander is orange. Excess power from last week for \el~10 disappears. Now somewhat lower power at low \ell.
Slide 12: power spectrum for r=0, which looks very good.
Discussion about MR's work, next steps, number of simulations, and points to highlight in a future paper.
Conclusions: concentrate for now on 90.91 and 90.92 with r=0 and r=0.003 (this would be new for MR).
JD: analyze 90.92 sky with 90.91 best model
Use various cuts of data (bands, sky coverage) to show robustness of results, and absence of such robustness with less available bands, or higher noise that prevents mission cuts.
For realizations: use N=10 for now. The expectation is that variance due to statistics would be small compared to variance coming from different sky models. No point getting the variance due to statistics determined to too high precision.