Differences

This shows you the differences between two versions of the page.

--- foregroundstelecon20210204 [2021/02/17 21:31] – hanany
+++ foregroundstelecon20210204 [2021/02/17 22:41] (current) – hanany
@@ Line 15: / Line 15: @@
   * Discussion of analysis of small patches of the sky. What are we doing about E/B separation? **Action: Shaul will check with Clem what has been done for the S4 paper**
-  * {{ :20210203_commander_pico.pdf |Commander Update}}
+  * Discussing {{ :nilc_gnilc_pico_20210204.pdf |MR's NILC Update}}.
+    * slides 2,3: completed 90.91 and 90.92, today discussing removing bands.
+    * slides 5,6: model 90.92, foreground removal is 'better' than 90.91, but we are suspicious because according to Brandon PySM underestimates the variability across the sky. But for now we are proceeding with this model ...
+    * slide 7: MR removed both high and low frequencies -> there is larger residual. It is factor of ~5 at the lowest \ell.
+    * slide 12,13: Comparing these two slides we see that there is more bias with the restricted frequency range, but even so, it is an upper limit. SH and HK say that "removing the high/low frequency bands would not have led to the wrong conclusion".
+    * Also, this conclusion ("would not have led to bias") is true for the PICO noise level. May not be true at other noise levels.
+    * **Action Item: Redo analysis removing only high or only low frequency bands**.
+  *  {{ :20210203_commander_pico.pdf |Discussing Ragnhild's report}}
+    * In trying to optimize the mask RA finds varying level's of r depending on the mask (model 90.91, with r=0)
+    * Generally finding aliasing of power: the resolution of the maps is coarser than the PICO beam.
+    * **Action: Uni is producing in-house simulations to check that this is indeed the issue**

User Tools

Differences

Page Tools