wiki:DC3SimulationProducts
Last modified 12 years ago Last modified on 12/19/2007 09:43:13 PM

DC3 Simulation Products

from: DC3 Management.

I wanted to start the discussion of what we want from simulations for DM at the start of DC3. This is meant to be the minimal set of properties that we would need such that the simulations are useful in testing the image subtraction and coaddition or stack processing (ie not the final suite with cosmology through to ray tracing). Running through the image subtraction and stack processing use cases I'd propose the following properties for the simulations:

Properties common to all simulated image sets:

  • FITS files with header keywords sufficient to be processed through the pipelines
  • An accurate WCS defined within the FITS header (at the accuracy of the stretch goals for the LSST 10 mas)
  • Images will contain a population of stars and galaxies (with a density comparable to 200K/deg2 for galaxies and 10K/deg2 for stars at a depth of S/N=5)
  • Galaxies are a combination of images defined by shapelets (or associated) basis functions plus scaled observations of nearby galaxies
  • Source populations are input to the images to a limiting signal-to-noise S/N=1(for point sources)
  • A variable background for the images (low order polynomial for the background sky)
  • Sources can be distributed randomly on an image (enhanced goal a realistic 2pt function for the galaxies)
  • For each pointing we deliver an associated catalog containing the input parameters describing the sources, the underlying PSF and the functional form for its variation, the background model, the noise model.
  • For pointing on the sky we define 100 realizations of the images with a variable PSF (see below)

From this I would assume at least 3 sets of simulations:

  1. A series of images (100 CCD pointings with 100 realizations per pointing) with a variable PSF but with no spatial variation in the PSF
  1. A series of images (100 CCD pointings with 100 realizations per pointing) variable PSF with a spatial variation that can be characterized by a low order polynomial
  1. A series of images (100 CCD pointings with 100 realizations per pointing) variable PSF with a spatial variation that can be characterized by a low order polynomial that are rotated relative to one another.