Telecon Meeting Notes July 12, 2018

On the call: Brendan, Shaul, Ranajoy, Colin, Maurizio, Eric, Joy

Actions for next time:

• Brendan to review Planck low-ell paper at next call
• Colin to review S4 tools at next call
• Ranajoy to brief on the CORE systematics paper (though next week he's not available)
• Ranajoy to begin a Systematics List on the wiki
• Joy to review Ranajoy's list

- see Brendan's intro/context slides here: systematics_telecon_1.pdf

Shaul's comments: baseline notional design looks like EPIC-IM. For now, let's pretend that the inflation probe IS EPIC-IM. This may change in the future, but it provides a concrete starting point.

Brendan to post link to EPIC-IM report: epic-im-report.pdf

Colin: what frequency bands are going to be assumed?

Shaul's baseline is coming out soon - should be finalized within days. For now, see the spreadsheet on the wiki: https://zzz.physics.umn.edu/ipsig/imageroptions

Shaul suggests not focusing on spectrometer systematics in order to create more focus: this is an option only but it's not settled.

Shaul emphasizes Systematics are an essential portion of this study for creating a convincing case.

Try to leverage very soon: explore S4 ways to inject systematics into simulations

Colin: what is the baseline analysis pipeline? Like a Fischer analysis. For S4, copied likelihood analysis from Bicep/Keck. But is that translatable to a space mission?

Ranajoy: With CORE, systematics were evaluated in terms of impact on measurements of r, so he has ways to do that.

Eric: But Fischer matrices might be too optimistic; correlated noise is a big issue. We may need a two-stage approach, start with simple (white) noise model but then make things more complicated.

Joy: should we assume there is a half-wave plate or not? Shaul: Let's start with assuming NO HWP. But we might want to keep that in mind and possibly consider how much better we can do with a HWP or other fast modulator at some point in the future.

Ranajoy: HWP has a huge impact but CORE spent a lot of time convincing themselves they didn't need one. But, what's the scan strategy? Bandpass leakage (and other effects) directly relate to how well we have angular coverage across the sky

To figure out: 1. what tools to use? S4-like tools? 2. What are a realistic noise model that should be used? - maybe start optimistically, and use a S4-like Fischer tool. - Colin has written down a knox-style bandpower covariance matrix but assumes a dust amplitude - how foregrounds are handled will have to be explored later. - Colin has worked on a generic “additive” systematic error for S4 and its effects depend on how common-mode the systematic is.