UCSD-2019: Analysis/Pipeline Working Group: Maps to C ell

From CMB-S4 wiki
Jump to: navigation, search

Link back to agenda

Charge

Note that "Maps to Cl" is actually shorthand for "Maps to Cl to Parameters" or "Maps to Parameters from (non-low-ell-BB) Power Spectra." This is important for understanding the charge and defining the responsibilities of this group.

The charge from the meeting SOC is twofold (see https://docs.google.com/document/d/1KtRif7_BPi3gACMkIA7__AWK63iUJiljN6UlKI41MS0 for the exact charge). One part is a set of general instructions for all analysis working groups:

  • Identify key decisions that must be made (and justified) prior to CD-1,
  • Make progress on (or actually make) those decisions,
  • Lay out a timeline and process for making each decision, consistent with the post-decision work and internal reviews that will be needed to complete preparations for CD-1,
  • Ensure that those timelines and processes are understood and supported by the collaboration, and that we (together) believe we can follow them.

The other part is a set of specific questions to support other working groups / WBSs as we move to the next level of design. Our questions are:

  • How are we calibrating beams to meet high-ell science requirements? Can we use high S/N point sources alone?
  • How does the galaxy impact Neff inference?
  • Does this drive frequency coverage?
  • Is there a path to realistically achieve both the necessary cadence for transients and the necessary sky coverage for light relics goals?
  • What are the necessary analysis tools to answer these questions?

Agenda

Reminder: What are the science targets, and what was previously identified as important decisions / questions by the forecasters? (30 minutes total)

1. Light relics summary (Ben Wallisch, on behalf of Dan Green and Joel Meyers and others) slides

2. Dark matter science summary (Vera Gluscevic) slides


General charge (45 minutes total)

1. Identify key decisions that must be made (and justified) prior to CD-1 (open discussion with some structure and example questions from organizers)

Background/clarifying questions:

  • What does “by CD-1” mean, and what are the implications for when tools need to be in place and working?
    • According to APC white paper (https://arxiv.org/abs/1908.01062), CD-1 is in Q3 of FY2021 (so June 2021?).
      • But according to project office, "Plan [must be] finalized by start of 2020 for delivering...CD-1"
    • Working backward from there, any tool that could reasonably influence a CD-1 decision needs to be in place and working by ... ?
    • Give an example timeline for an example decision?

Some possible key decisions/questions:

  • Start with known high-priority things
    • Which instrumental effects are important, what requirements can we put on them, and are they relevant for decisions being made on the timescale of CD-1?
      • Beams
      • T -> P leakage
      • Time constants
      • Gain / gain drift / elevation-dependent gain
    • What limits how much sky we can observe?
    • What science goals beyond Neff potentially drive measurement and instrument requirements for non-BB power spectra? Are we missing something by concentrating on Neff?
    • What level of sims to maps to Cell to parameters pipeline do we need for CD-1? Should a tool be developed to be used by the whole collaboration?
      • Do we need to include delensing in near-term forecasting tools?
    • Should Maps to Cell and low-ell BB groups have a common set of simulations and analysis tools to be addressing “non-idealities” by CD-1?
      • This is happening in DM / Simulations, though other groups will have input on the nature of these tools.
    • Should we be setting up data challenges (e.g., make a CMB sky with beam properties varying across the field and have people analyze it and get unbiased Neff)?

2. Make progress on (or actually make) those decisions

3. Lay out a timeline and process for making each decision, consistent with the post-decision work and internal reviews that will be needed to complete preparations for CD-1

  • Summary of decisions identified as most important, possible tasking and timelines.

4. Ensure that those timelines and processes are understood and supported by the collaboration, and that we (together) believe we can follow them.

  • We can talk in the session about the best ways to ensure this.


Individual charge questions (45 minutes total)

Questions regarding how much sky we could possibly observe without hitting some systematic or noise floor? (what is limiting fsky?)

1. How does the galaxy impact Neff inference and does this drive frequency coverage? (Colin Hill) File:JCH Gal Neff v2.pdf

  • also think about ground pickup?

2. Is there a path to realistically achieve both the necessary cadence for transients and the necessary sky coverage for light relics goals? (Reijo Keskitalo) slides animation wiki post WAFFT results part 3


Questions regarding systematics:

3. How are we calibrating beams to meet high-ell science requirements? Can we use high S/N point sources alone? (Tom Crawford slides)

  • What about T -> P leakage for Neff? (incl higher-order terms)


General questions'

4. What are the necessary analysis tools to answer these questions? (all, discussion)

Remote attendance

Zoom link

Notes

Early notes on planning/agenda are here: https://docs.google.com/document/d/1uxn1AedOzWAObeG59aS7WPfctV_GC6kDbDw0KgdR_Ow


Note further "charge" from Lloyd's flowdown parallel summary:

  • Working groups need to develop (or gather) software needed for reduction of simulated maps to science results.
  • Working groups need to think about development of analyses for rapid exploration of different measurement properties (e.g., frequency distribution, angular resolution, levels of systematic effects).
  • Flowdown from Science to Meas will be done with a hybrid approach of semi-analytic (or otherwise “fast”) methods, map-based sims, and occasional time-domain sims.


Notes on Science Goal mini-talks

  • Need to confirm that main ACT beam eigenmode identified as problematic does really correspond closely to beam width.
  • Are Dark Matter constraints really in the purview of this group? DM-baryon scattering constraints really come from lensing, so probably not.
    • For DM annihilation, do we get everything we need if we have already satisfied N_eff constraints, or are there independent drivers?
      • Consensus is "yes" and "no."


Brainstorming on key decisions on the path to CD-1

Some possible key decisions/questions:

Start with known high-priority things

  • Which instrumental effects are important, what requirements can we put on them, and are they relevant for decisions being made on the timescale of CD-1?
    • Beams
    • T -> P leakage
      • What *kind* of T->P leakage? How does this actually bias N_eff?
    • Time constants
    • Gain / gain drift / elevation-dependent gain
    • cross-talk?
      • if you have xtalk in pixel partners, it is just a reduction in polarization efficiency, and there's nothing you can do about it. someone should just declare a limit on that.
    • polarization-dependent beams?
      • is this different from just beam mismatches between polarization partners?
      • can we measure polarized beams? how?
      • is it hopeless to do this off of astronomical sources? do we absolutely need a terrestrial source? how to set this requirement?

What limits how much sky we can observe?

What science goals beyond Neff potentially drive measurement and instrument requirements for non-BB power spectra? Are we missing something by concentrating on Neff?

What level of sims to maps to Cell to parameters pipeline do we need for CD-1? Should a tool be developed to be used by the whole collaboration?

Do we need to include delensing in near-term forecasting tools?

  • we think this was included in Science Book forecasts.
    • and in DSR
      • if you just include 7 parameters (LCDM+Neff), the difference between doing forecasts on unlensed spectra and real delensing is not that important, but if you include Yp, it starts to matter (like 0.3 sigma)
      • confirm with Dan & Joel, but probably already done

Should Maps to Cell and low-ell BB groups have a common set of simulations and analysis tools to be addressing “non-idealities” by CD-1?

  • This is happening in DM / Simulations, though other groups will have input on the nature of these tools.
  • well, the simulation side is definitely happening, but what about the tools? Do we need to produce a unified set of maps to C_ell tools?
    • general consensus is yes.
    • Who maintains and develops these tools?
      • this working group (not Data Management, not TOASTers)
      • needs to be a schedule for delivery of these tools, and they need to be ready for the first Mock Data Challenge

Should we be setting up data challenges (e.g., make a CMB sky with beam properties varying across the field and have people analyze it and get unbiased Neff)?


Individual charge questions

1. How does the galaxy impact Neff inference and does this drive frequency coverage?

  • How certain are we that CO, free-free, and AME are unpolarized?
    • AME is probably most uncertain (could be 1%); others are constrained to << 1%
  • ACTION ITEM: Was polarized Galactic emission included in DSR N_eff constraints?
    • Colin tracks this down later, answer is yes but DSR text does not accurately reflect this.
  • Main result: For 62% of sky, Dl_dust / Dl_CMB ~= 6% in TE at ell ~= 2500. So probably not a huge deal but need to include in fits.
  • also think about ground pickup?
    • Did not talk about ground pickup

2. Is there a path to realistically achieve both the necessary cadence for transients and the necessary sky coverage for light relics goals?

  • For detector people: How long does it take to retune (to go from observing at el=30 to el=60, for instance)?
  • ACTION ITEM: Reconcile high-cadence-strategy N_eff constraints with Reijo's (and everyone's) intuition.
    • If we can really tolerate a 50% noise hit to go to more sky (this has been said out loud at meetings), what's limiting N_eff in high-cadence strategies?


Questions regarding systematics:

3. How are we calibrating beams to meet high-ell science requirements? Can we use high S/N point sources alone?

  • In terms of beam width and not considering polarization-specific beam systematics, yes, with plenty of margin.
  • What about T -> P leakage (incl higher-order terms) and "polarized beams"?
    • Needs better definition and some study.