owned this note
owned this note
Published
Linked with GitHub
# Cross-Cut: Simulations for Measurement Requirements
[Link to wiki](https://cmb-s4.org/wiki/index.php/UCSD-2019:_Cross-Cut:_Simulations_for_Measurement_Requirements)
> [name=Andrea Zonca] Please help me out taking notes, it is very important to produce a complete description of the discussion during the parallel session. Don't worry if there is duplication. I'll cleanup and post to wiki afterwards.
## Agenda and notes
* 5 min - Andrea Zonca: PySM 3, simulation tools developed for Simons Observatory
* 5 min - Marcelo Alvarez: Extragalactic modeling integration in PySM 3, Simulations requirements for cluster finding
* 5 min - Reijo Keskitalo: Scanning strategy for S4
* 5 min - Ben Racine: Noise simulations for PGW analysis
* 25 min - Discussion:
### Unified sky model
#### Galactic
* PySM (`so_pysm_models`) NSIDE 8192 Planck + Gaussian
* MHD: Dust + Synch small patches
* d7 - implement in PySM 3 and increase resolution
* Currently used model up until DSR by the r-forecasting group:
** 00 - Simple Gaussian Sync/Dust with uniform amplitude across the sky
** 01 - PySM base (@zonca has higher nside than 512, to 4096? will get to 8192)
** 02 - PySM extended (@zonca has higher nside than 512, to 4096? will get to 8192)
** 03 - Same as 02 but Hensley/Draine dust model
** 04 - HiDPol dust model (by Tuhin Ghosh)
** 05 - Strong decorrelation model
** 06 - MHD based model from Flauger/Hensley (should swap in "08": one with more correct synchrotron)
** 07 - Amplitude modulated Gaussian
** 08 - MKD multilayer model from arxiv/1706.04162
** 09 - Vansyngel model as described in arxiv/1611.02577 with multi-frequency extension
* For delensing, we need small scale polarized non-Gaussian Galactic foregrounds. Currently, only 09 has small scale NG.
* discuss: 00 should be replaced with 07? 00 won't be meaningfully used by the large area survey. But it is useful for the deep survey for sanity checks.
* To generalize 08 and 09 into PySM, @zonca has to ask the creator of these models for code/info.
* Q: are there other Galactic models used in S4 outside of the r-forecasting group?
* Check consistency with Planck
* CO lines?
### How to gather requirements from AWG
### Extragalactic + Clusters
* Websky, 1 realization
* validated against planck
* lensing: 1 realization
* radio galaxy sources, catalog input, PySM component could generate map on the fly
* tSZ from Planck, CIB from Herschel and Planck, dusty galaxy count compared to Herschel, stack clusters
### Noise
#### Scanning strategy
* <https://cmb-s4.org/wiki/index.php/Deeper_SAT_from_Chile_II>
* <https://cmb-s4.org/wiki/index.php/High_cadence_LAT_from_Chile>
High cadence LAT gives lot of f_sky and daily scans for transients
* <https://cmb-s4.org/wiki/index.php/Deep_SAT_from_the_Pole>
Below only done in case of r detection from Deep SAT:
* <https://cmb-s4.org/wiki/index.php/Wider_SAT_from_the_Pole_II>
Jan & Feb we could remove
Clem has already noise pipeline
Delensing LAT from Pole??
#### Noise simulations for PWG
* SAT: start from BK data, fit noise model, just scaling NET and observing time, only data from Pole
* BK NET calculator does not agree with data (in absolute terms), but only relative scaling matters
* SO Noise calculator is just used for LAT
* SAT: Extrapolating 220 to 270, 95GHz to lower freqs. Will need to be checked when data is available.
* Check POLARBEAR new released performance, use that to scale?
Delensing **LAT** at Pole needs a noise model, maybe scaling SPT??