Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-17846

Understand PSF issues in deblending

    Details

    • Type: Story
    • Status: Done
    • Resolution: Done
    • Fix Version/s: None
    • Component/s: meas_deblender
    • Labels:
      None
    • Templates:
    • Story Points:
      7
    • Sprint:
      DRP S19-3, DRP S19-4, DRP S19-5
    • Team:
      Data Release Production

      Description

      This ticket has been expanded to investigate two effects that we are seeing in blends while using PSF matching.

      1. We commonly see color gradients in the residuals when running scarlet on real HSC data. This is the main issue with running scarlet on real science data.

      Peter Melchior and I believe that this is due to a misalignment of the PSF kernels given by the DM pipeline as opposed to their offsets in the image. In other words, it looks like the PSF kernel gives a PSF that is centered on the central pixel while the actual PSF in the image may have offsets between the bands. This needs to be verified, and regardless we need to make sure there is agreement between where the stack says the PSF should be centered and where scarlet applies it.

      So part of this ticket is to determine the cause of the color fluxes and the rest is to fix the issue.

      2. We also see rings around the centers of bright sources, with centers underestimated by the model and a ring around the center that is overestimated by the model.

        Attachments

          Activity

            People

            • Assignee:
              fred3m Fred Moolekamp
              Reporter:
              fred3m Fred Moolekamp
              Reviewers:
              Peter Melchior
              Watchers:
              Fred Moolekamp, Peter Melchior
            • Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved:

                Summary Panel