BlueJeans: https://bluejeans.com/426716450

attending: Eric Bellm John Swinbank Meredith Rawls Krzysztof Findeisen Unknown User (cmorrison) Unknown User (gkovacs)

Topics for discussion

  • HiTS reprocessing (Meredith Rawls):
    • re-ran ap_pipe as part of the regular ~monthly rerun.  did not remake the templates this time, just kept the ones in the dataset.  only subtle changes.
  • DES processing (Unknown User (emorganson)):
  • ap_association (Unknown User (cmorrison)):
    • has dipole inputs in review with Eric; got stuck on SNR of detection, which we agreed needs a ticket–not necessarily by Chris, as it may be in the C++ layer
  • Metrics & ap_verify (Krzysztof Findeisen ):
    • nothing to report
  • image differencing (Unknown User (gkovacs) ):
    • got bogged down with some debugging and in a diversion into a ticket from John Parejko 
    • open discussion: other examples of good difference imaging on DECAM: DMTN-006, DMTN-021, DMTN-061
    • Eric Bellm to confirm that our detection rates are actually "bad" relative to e.g. DMTN-006/21
    • Meredith finds in "known issues" DM-11993 - Getting issue details... STATUS , which alludes to possibility of artifacts due to "poorly calibrated" data–is there a correlation with our bad photo calibration that we already know about?
  • DCR (Ian Sullivan ):
  • Review CI (https://chronograf-demo.lsst.codes/):
  • AOB
    • John Swinbank to report on DRP plans/progress on fake insertion
      • Notes from John Swinbank, 2019-04-01 (ie, before I forget):
        • The existing Synpipe system is already capable of injecting a model sky into single frame processing (and hence into calexps that we could use in differencing). There is no need to wait for any further DRP work there.
        • The work being carried out by DRP extends this in two directions:
          • It makes it possible to inject simulated objects using a Jointcal (or meas_mosaic) derived WCS, which means they can be propagated into coadds (otherwise, they get smeared out during warping). It follows that this work is necessary if you want to simulate sources in the template, but not in the science image.
          • It makes it possible to have a dynamic sky model (so you can inject variables and recover their lightcurves, for example).
        • The new DRP work has produced InsertFakeSourcesTask, which is currently sitting on a branch in pipe_tasks and elsewhere.
        • we'll discuss this in the next cycle plan
    • Unknown User (cmorrison) will be working on proposal for ra/dec covariances