Chicago-2016: Blue Team

From CMB-S4 wiki
Jump to: navigation, search

BLUE TEAM: Mixed array of small (~1m) and large (>~6m) segmented dishes

Moderated by Akito Kusaka. Notes from Suzanne Staggs.

Also present in the room: Sheehy, Devlin, Benson, Holzapfel, Ruhl, Carlstrom, Shirikoff, Thompson, Cho, Barron, Ahmed, Kernokowsky, Sehgal, Padin, deBrouille, Lawrence, Wollack, Arnold, Snow, Suzuki, Chang, Appel, Henning, Hall, Crites, Madhavacheril, Newburgh, 3 people I couldn't quite see in the back of the room, Papitashvili, Lee, Stebbins

  • Are we thinking of a configuration that works for both surveys? Answer in the room is yes.
  • Are we assuming the large aperture cannot go to low ell for this? Answer in the room is yes.
  • Begin by defining a strawperson -- what elements need to be considered?
    • Large: D(f), lmin, segmented or not, pol modulation or not, design
    • Small: D(f), lmin, lmax, pol modulator in front of the whole system or not, reflector or refractor, single-frequency (or two frequency including dichroic) per aperture or not
    • Resolution range per frequency band, and also Npix per frequency band
    • How many telescope sizes?
  • What issues do we need to roll in?
    • complexity -- this could drive us to wanting to only do things we have already demonstrated -- ie, not going to much larger receivers than we have done, or trying to have more bandwidth than we have demonstrated -- but R & D soon might mitigate this, and could make scaling up to large N more feasible and/or make the systems more robust
    • size of cryostat
    • do we need the same resolution for all frequencies for foregrounds?
  • For the configuration how much does small-aperture lmax and large-aperture lmin depend on calibration issue, for comparing them both?
  • For this configuration we could put the 30 GHz on another instance of a large aperture.
  • Pros and cons
    • Pro of this configuration is can have more flexibility for smaller apertures, and perhaps can scan faster too
    • Con is complexity of more than one telescope
    • Con if there are only two sizes of dishes is that the small aperture size for 30-280 GHz maybe be extra limiting for the angular resolution issue
    • Pro is could address more science because we could get to higher resolution and get all the cluster science as well as what the 5m array could do
    • Pro is can have boresight rotation only for the small apertures and not have to have that complexity with the large apertures
    • Con is dividing up the detectors so you don't get all the sensitivity over the l's that don't overlap (could argue that you only need the extra detectors needed for the delensing survey over what the homogeneous 5m array would need)
  • What is the limit on the aperture size for a small telescope that works? B3: 0.7 m gives 76 arcmin at 30 GHz. Existence proof for that size window/optics from SPT-3G. It takes 7-8 of those (2x the size of the B3 array which has 0.5 m). You get 5k pixels so 20k TESes in that area.
  • Possibility: could have a range of telescope sizes so that you only have a factor of two in the resolution for the entire frequency range
  • What is the criterion for deciding if you need a pol modulator in front of a small aperture? (Could also be boresight rotation, though doesn't do as much for systematics.) How to decide how fast you need to modulate; if atmosphere is OK (as for B2) then you don't need rapid modulation. (Atmosphere may require modulation in Chile.)
  • Homework:
    • SPT, ACT and SA folks to come up with a means of defining lmin for large apertures
    • Define criterion for needing polzn modulation (including eg HWP and boresight rotation) -- systematics and atmosphere 1/f both need to be considered
    • Systematics comparison between eg B2 and ABS
    • Figure out if we need 30 and 40 GHz for delensing?
    • Need to get started on how to decide on telescope architecture for large aperture -- segmented versus not, CD versus not, shielding
    • Should we use more than 2 sizes of telescopes? Or consider a monolithic center and degraded outer panels to try to get same resolution at all.