Description
DMTN-049 was recently updated by the DM System Science Team to be a community-facing roadmap towards DR Object catalog photo-z.
This DMTN is not subject to the CCB, but since this is a big change and a big initiative, we are posting this RFC to solicit feedback in advance of starting to advertise to the community (which we plan to begin 2021-01-29).
Briefly, this roadmap includes a series of monthly drop-in virtual forums to ingest community feedback about the evaluation criteria for photo-z estimators, a call for "Letters of Recommendation" (due 2021-09-30) for the science community to advocate for certain estimators or types of estimators, and (during commissioning, dates TBD) a "Photo-z Validation Cooperative" process for community participation in the validation and implementation of photo-z estimators shortlisted by the "LoR" process.
Attachments
Activity
Working in branch tickets/DM-28526, so far have made the following big changes (and a bunch of smaller ones):
Rewrote Section 2.2 to attempt to make participation from the in-kind teams in this roadmap clearer (thanks Phil Marshall).
In the abstract, replaced "This roadmap includes possible in-kind contributions from international partnerships to {\tt Object} catalog photo-$z$ for DR1 and beyond, as well as potential..." with "This roadmap includes the role of proposed in-kind contributions from international partnerships as well as potential..."
In Section 3, added:
\textbf{External Data Sets:}
The assembly of vetted training sets (e.g., compilations of spectroscopic and many-band photometric redshifts) is a separate but related aspect with its own technical considerations.
Any such external data sets that would be required for a photo-$z$ estimator to run on the LSST data must exist by the time of commissioning so that the algorithms can be trained, run, and validated.
These external data sets should also be in the public domain and formatted to be ingested by the estimator.
Thanks all for your comments, they have been incorporated and branch tickets/DM-28526 of DMTN-049 has been merged to master.
Thanks Melissa! Good stuff. I wonder if there is a simplification in organization that you could make: the original intent with the photo-z coordination group was to make a forum where all teams interested in photo-z estimation could interact, not just the international in kind contribution teams. (I think we should be encouraging US groups to join in with the discussion - and even to become in-kind contributors themselves, if they can get funding to provide development effort to a science collaboration or Rubin team, as the international teams will be.) With that generalization, would the photo-z coordination group then look sufficiently similar to the "Photo-z Validation Co-op" to think about doing all this within the same structure? In any case, it might be good to show, in this Roadmap, how you see the various actors and groups interacting during various phases (preparing Letters of Recommendation, validating estimators, ongoing through the survey).