2019-10-21 DM SST Agenda and Meeting notes

Chuck submitted (under protest) LCR-2045: Commissioning De-Scope LSSTCam Observing and SV Mini-Surveys. The motivating reason is to improve the contingency ratio on both budget and schedule relative to remaining planned work. 

This will: 

• Eliminate  the planned 2-months of observing for science pipeline testing from the commissioning plan. 
• Compress the time for science validation surveys from 3 months to 2 months. 
• Alert Production and Data Release Production each will be shortened by 2 weeks. Science pipelines will not be a thoroughly vetted prior to start of operations. Smaller sky footprint, increased risk of poor weather affecting commissioning. 
• 8 months only now for on-sky commissioning 

Will has created a response page: Response to LCR-2045 - Shorter Commissioning

Recap from last week's meeting & PST discussion

Who will participate in the provenance WG? Timeline is to convene and wrap-up by end-2019.  Goal is to review the DM provenance based on what is now available in butler Gen3 and the SDM.  

Scope:

• Assess available provenance information in Gen3 metadata registry
• Identify provenance gaps in the data model and how they should be filled.
• Classify the resulting provenance needs as critical (needed for commissioning), important (need for full ops) or nice to have.
• Not in scope to solve how missing  provenance information should be added. 

Meeting notes

• Investigate whether anyone has run photo-z codes on HSC (or DES) data. We should try to get an estimate of the effort involved from these groups. Jim Bosch recommends contacting  Atsushi Nishizawa  (head of the HSC photo-z group)  and/or Masayuki Tanaka.  The HSC group are already dealing with data of fairly comparable qualities to those of  LSST, e.g. depth, though with somewhat less confusion than LSST will have. 
• Other projects such as DES are also producing photo-z and could be interested. 
• Michael Wood-Vasey  will be at the Subaru meeting 20th anniversary meeting  - Nov 16 (meeting website). 
• Based on this experience,  estimate the effort to build the MVP described in DMTN-049. Is this a 1-month or 1-year effort?
• Gregory Dubois-Felsmann points out that  ingesting and federating user catalogs with the LSST data products is a fundamental use case for the LSP. 
• Option 2 is an operations project task, this option is in the baseline and is important for science regardless of whether DM computes Photo-z or not. 
• Option 1 could be done in collaboration with the community or an international partner, and could form an in-kind contribution. 
• Gregory Dubois-Felsmann suggests the SST write an RFP describing how we would do work with an international group to produce photo-z, incl. interaction with the project project, what should be delivered, formats, documentation, etc.