2022-10-17 DM-SST vF2F Agenda and Meeting Notes

Short : ls.st/dmsst 

Zoom: https://noirlab-edu.zoom.us/j/97839121776?pwd=K1JPeUpSMXFvSXJSa2xORGkyVk5zdz09

Zoom Meeting ID: 978 3912 1776 
Zoom Password: 512314

Dial closest IP: 162.255.36.11 (east coast) and 162.255.37.11 (west coast) then use the Zoom meeting ID 978 3912 1776  as the dialing extension.

For example: 978 3912 1776 @162.255.37.11 or: 162.255.37.11##978 3912 1776  Password: 512314

Dial-in numbers:

+1 346 248 7799 (US Toll)

+1 669 900 6833 (US Toll) Meeting ID: 936 2540 1560 International numbers available: https://gemini.zoom.us/u/adcUNrbXzS

No PST meeting last week. Next PST meeting 2022-10-19

• SR interested in algorithms for CR removal. 
• JB: Assumption that DI will remove all CRs 
• EB: AP did a big push on this a few months ago - existing stack algorithm performs well. Slides from the cosmic ray analysis earlier this year
• GPDF: Presumption of the same answer for every band? No
• SR: 1x30 per u band and 2x15 in the other band. Operational complexity will remain if no decision to abandon the complexity of 2x15
• JB: Could stop a lot of work if we make this decision to abandon 2x15.  

• Flow chart presented is the order of operations agreed at CalibPalooza and will form the outline for the construction paper for calibration. 
• Red - work that is not done yet, Black - something is in place.
• Earlier we can combine snaps, the less expensive it is. 
• JB: Most steps before BF are fast - as long as we do snap combine < BF we are fine. BF is not fast
• XT correction is per detector - no evidence for inter-detector XT
• SR: make clear after each step what the units are 
• ER: Yes, some are still under debate. Will be electron-like of ADU-like Awaiting a writeup from Pierre
• GPDF: Figure out what processing is needed in the RT system - rapid analysis producing a post-ISR output. 
• ER: Will be rewriting ISR task  - for QL rapid analysis various steps can be turned off  - expect most to be fast except BF. e.g. variance plane doubles the size of every image 
• CTS: Is the purpose of diving by gain primarily to get normalization btw amps or to get electrons
• ER: 100% primary purpose is normalization btw amps - this cannot be captured by a FF  its a model that depends on 2 temperatures.  And - it would be nice to normalize the gains at the same point and bring them all to 1.
• ER: Critical issue learnt from AT is that if calibration products re run in a diff. sequence to what there are used there are problems.  
• LPG: What is the con of normalization at the same time as gain normalization? 
• ER: Does not understand RHL's concern. 

Post ISR - Now new plans for FFing. The dream of doing FFing better

• Need to be careful with flats to no t introduce unintended effects. 
• Subtract fringes was not discussed at CP - ER's thinks it goes aft Subtract BKG. 
• Do we need a bespoke sly SED when processing or is a reference SED OK? TBD - complicated.
• fCBP encompasses a lot of work
• KB: How do we know in a data driven way, which are lateral e-field effects or quantum efficiency effects 
• ER: Most ate lateral e-field  effects. The 3 listed things we have physical models - develop a template for each and remove it. Not using astrometric residuals. 
• SR: QE dominated by variations in the anti reflective coating - agree to make evaluation of residual leftovers data driven. 
• KB: I hope that we will have quasi-independent methods for verification and validation based on on-sky data. This will take some thought.
• SR: mentioning techniques to be used to assess residuals. Can use on-sky data to evaluate or verify residuals
• ER: No box for amp-to-am matching - hope that GN does it correctly and no bogus aa-to-a matching is needed. If this is messed up it affects background measurement
• YAS: amp-to-amp matching could be done in  PCA box. Cannot assume that GN will be done perfectly. 
• LPG: Withouth RFF without CBP version we are still beyond state of the art due to the different use of MDF
• DF may fail  -most likely place for dust to change is on the filters - filter change operation will probably move the dust around so once at the beginnign of the night will not work. 
• RSN: Is there an existing one that is sufficient?? No - what are we missing without this?
• ER: If Gaia SP gets to the point of being better than our internal - GDR3 is worse than internal. But if in a few years and are down to 17th mag and ave synth ugrizy to mmag level then we can use it and get better calibration.
• KB: ETA of cell based coadds? 
• YAS: next summer
• ER: Cell size is 250x250 incl. boundary 
• CTS: 
• LPG: Important to have a CBP/non-CBP route + prioritization scheme
• ER: CP follow up to discuss prioritiation

#refcats channel to help out on the monster