Uploaded image for project: 'OBSOLETE DM Long-range Planning OBSOLETE'
  1. OBSOLETE DM Long-range Planning OBSOLETE
  2. DLP-264

Framework for representing astrometric pixel distortions

    Details

    • Type: Milestone
    • Status: Unplanned
    • Priority: Major
    • Resolution: Unresolved
    • Affects Version/s: None
    • Fix Version/s: F16
    • Component/s: Planning
    • Labels:
      None
    • Milestone Level:
      4

      Description

      The current mechanism is to do a two fold process where smooth perturbations caused by the atmosphere and optics are represented as a single polynomial solution over the focalplane or chip. The higher frequency distortions caused by sensor effects are modeled in a different, unrelated way.

      Astrometry and all measurements need a way to represent these different distortions in a single self-consistent framework. This leads to the notion that the general concept of a "WCS" needs to be redefined to be more general.

        Attachments

          Activity

          Hide
          jbosch Jim Bosch added a comment -

          I just wanted to comment that I think that this probably can't be the final WCS interface, as I think that will ultimately need to include chromaticity, and I don't think we're in a position to do that yet. DRP's work on PSF/Kernel interfaces that are chromatically dependent (DLP-85) is blocked by DLP-489, which adds utility classes for chromaticity that I think we'll want to use for both PSF and WCS evaluation - things like a real filter curve classes, better color terms, and code for manipulating SEDs. I'm not by any means suggesting this issue should be blocked by DLP-489, but we probably need a later WCS redesign milestone that is.

          Show
          jbosch Jim Bosch added a comment - I just wanted to comment that I think that this probably can't be the final WCS interface, as I think that will ultimately need to include chromaticity, and I don't think we're in a position to do that yet. DRP's work on PSF/Kernel interfaces that are chromatically dependent (DLP-85) is blocked by DLP-489, which adds utility classes for chromaticity that I think we'll want to use for both PSF and WCS evaluation - things like a real filter curve classes, better color terms, and code for manipulating SEDs. I'm not by any means suggesting this issue should be blocked by DLP-489, but we probably need a later WCS redesign milestone that is.
          Hide
          krughoff Simon Krughoff added a comment - - edited

          This was started in W16, but we under estimated how hard this was, and how much ramp-up time there would be for the people working on this. Much of the precursor work has been done to figure out what we need to implement (see DMTN-005.lsst.io and DMTN-010.lsst.io). We will have a design for the implementation by the end of X16, but I do not think we will have an implementation until F16.

          Impact of delay:
          This is one of the major enhancements needed to remove the false positives cause by astrometric residuals in the difference images. Not having this does not make things worse, but when we deliver it, we hope to do much better than we have to date. The largest impact on other components is that the jointcal task (astrometric fit on a stack of images) depends on the LSST representation of coordinate transformations.

          Show
          krughoff Simon Krughoff added a comment - - edited This was started in W16, but we under estimated how hard this was, and how much ramp-up time there would be for the people working on this. Much of the precursor work has been done to figure out what we need to implement (see DMTN-005.lsst.io and DMTN-010.lsst.io). We will have a design for the implementation by the end of X16, but I do not think we will have an implementation until F16. Impact of delay: This is one of the major enhancements needed to remove the false positives cause by astrometric residuals in the difference images. Not having this does not make things worse, but when we deliver it, we hope to do much better than we have to date. The largest impact on other components is that the jointcal task (astrometric fit on a stack of images) depends on the LSST representation of coordinate transformations.

            People

            • Assignee:
              krughoff Simon Krughoff
              Reporter:
              krughoff Simon Krughoff
              Watchers:
              Jim Bosch, John Parejko, Simon Krughoff
            • Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

              • Created:
                Updated:

                Summary Panel