Photo-z Resource Needs

Each of the SV Requirements / Goals will need a page describing what is needed to conduct the test or meet the goal. Here is a suggested breakdown. In the near term, we are most concerned with the data processing needs (i.e. what will DESDM or some alternative pipeline need to do).

Of course the answers to many of these questions will not be known until we proceed further in our planning, but we should fill in what we know.

Number/name of Requirement or Goal: Photo-z goals PZ-G1 through PZ-G4 combined

DECam observations

  • Field(s) to observe:
    • VVDS-Deep 02hr field (RA = 02:26, Dec = -04:30)
    • VVDS-Deep CDFS field (RA = 03:32, Dec = -27:48) can serve as an alternative, but isn't as ideal as it has fewer redshifts
  • Number, duration of exposures per filter:
    • Can be deeper, but at least main survey depth (800 sec each of g,r, 1000 sec i,z, 500 sec Y), under main survey observing conditions
  • Dithering / tiling pattern:
    • Both standard main survey tiling pattern and supernova dithered exposures are ok, just need to cover full VVDS-Deep field to above depths
  • Sky conditions needed (seeing, photometric, moon phase, etc:):
    • Standard main survey observing conditions to above exposure depths
  • Other timing constraints:

Data processing needs

Which of the following products would you need to conduct the test:

  • Raw images available at some off-mountain site? no
  • Images with cross-talk, debias, trim, dome-flat correction? not needed separately from DESDM processing of the data
  • Images with an "illumination" or "star" flat applied for better relative photometry? desirable, but not needed separately from DESDM processing of the data
  • Single-exposure object catalogs? no
    • with astrometric solution accurate to:
    • and photometric zeropoint accurate to:
    • Only need bright stars, or also want galaxy magnitudes?
  • Registered stacked images (co-adds)? yes
    • PSF homogenization needed? not required, but desirable
  • Catalogs from coadds? yes, with mag_auto, galaxy model fitting photometry, and detmodel colors
    • with photometric zeropoint (and/or colors) accurate to: 0.05 mag
  • Weight maps? standard DESDM coadd output weight map
    • Bad & saturated pixels? yes
    • Bright-star exclusion zones? yes
    • other?
  • Polygon masks?
  • Difference images?
  • Further SN data products:
  • Other:

Telescope or other engineering data

External data sets

  • existing VVDS-Deep spectroscopic redshift catalog

Codes and scripts for analysing the test data

  • existing DESDM neural network photo-z module
  • existing DC6B Photo-z Challenge analysis scripts in IDL


Are there real-time decisions to be made during the observations? no

How quickly can / must the data be available and analysed? no requirement

Who will do it, and where?

  • Given DESDM coadd catalogs of VVDS-Deep field, Photo-z WG can derive neural network photo-z solution and then carry out remaining tests
  • Ideally, DESDM will run photo-z module inside DESDM framework to apply solution and ingest photo-z results into DESDM database, but this is not required to carry out the tests, which only need the DESDM coadd SExtractor catalogs to be available

Other important resources or things to note