SLAC-2017:FSM Small angular scales
Back to SLAC-2017 main page
Foregrounds, Systematics and Modeling: Small Angular Scales
Post session talks here.
Questions To be addressed for each small-scale science ‘probe’ (TT/TE/EE, lensing, tSZ/kSZ) towards defining measurement requirements - to start discussion/for context
How large do the large telescopes need to be? (NB - size of telescope is not a measurement requirement, but it is a simple proxy for resolution that varies with frequency in a realistic way)
- Current status: studies indicate 5m (1.8' at 150 GHz) is sufficient for lensing (mnu and delensing) and high-ell TE/EE science (Neff), but may not be enough for cluster-derived science. Extra resolution doesn’t seem needed for cleaning TE/EE and lensing.
- To-do: push forward cluster forecasts (see lunchtime discussion).
- To-do: Baryonic physics of clusters is interesting -- can we define a good science metric?
- To-do: Beam/other systematics -- plan to do more sophisticated treatment.
How many/which wavelengths do we need in the large telescopes?
(1) for TE/EE 2-pt for Neff
- Current status: indicates that three is sufficient, based on new power spectrum simulations.
- To-do: check impact of Galactic contamination
(2) for lensing (mnu/delensing) to avoid EB lensing bias
- Current status: determining whether only 90/150/220 needed: work underway to test effect on lensing of expected non-Gaussian dust foregrounds. Might need synchrotron channels.
- To-do: continue this work, test other sims. Compute delensing effect for various sims. Does delensing or mnu drive requirements?
- Can look at real stage-3 data to avoid modelling uncertainies.
(3) for tSZ/kSZ (mnu/w/patchy-tau/growth/feedback)
- Current status: at least 90, 150, 220 for tSZ. Suspect 270 GHz will be beneficial but assess expected noise levels. More frequencies would help separate kSZ, especially in power spectrum for reionization.
- To-do: run cluster forecasts (for counts, tSZ power spectrum, y-maps, kSZ signal) on extragalactic sims
What are our systematic requirements for the 2pt and 4pt functions?
Can we achieve science requirements given realistic atmospheric noise?
- Current status: realistic atmospheric noise consistent with Chilean observations without HWP has now been modeled and included in fisher forecasts for high-ell and lensing science. Neff/mnu parameters are not degraded.
- To-do: include multi-freq realistic atmospheric noise in tSZ/kSZ forecasts, include South Pole noise curves. Establish agreement on atmospheric noise model being used by everyone.
Some people who have things to say on these topics:
- Matthew Hasselfield - atmospheric noise model for large telescopes
Neff (TE/EE)-Discussion of Systematics and Foregrounds
- Dan Green - Overview + Beams/Pointing - slides
- Alex van Engelen - Impact of polarized dust on lensing
- Mat M. / Colin H. - Multigrequency cleaning and extra spatial filtering
- Yuuki Omori - Biases from extragalatic sources in temperature lensing
- Marcelo Alvarez - extragalactic sims
- Kevin Huffenberger (remote, 5 min) - Foregrounds and Atm noise for tSZ science slides
- Simone Ferraro - foregrounds for kSZ slides
- Lindsey Bleem - more on sims slides
Notes from session
- Most info is from TE, l ~ 2000 - 4000
- Beam effect: taking beam error as a single parameter on its width: sigma(10^-2) arcmin^2 to get sigma(Neff) < .03
- Question: What is the best way to model the beam uncertainties -- if low-order polynomials, need uncertinties on each one
- CIB / radio -- adding noise in TE
- How well can this be cleaned with multi-frequency
- Will finding sources to reduce the power from Poisson sources drive resolution requirements?
- Dust models - Alex showed ongoing work characterizing bias. Also plan for computing delensing bias. Both come from <EBEB>
- What measurement requirements are more stringent -- Lensing auto or delensing?
- What frequency, angular resolution, and sky area do we need?
- Discussion of map-level simulation analysis. To what extent do we need to do this for high-ell targets?
- Baryonic physics of clusters is interesting -- can we define a good science metric?
Action items/Next steps
Summarize action items here