Relative Photometric Calibration Resource Needs¶
Number/name of Requirement or Goal: Cal-R1, R4, R5, G4, G5, G6¶
These goals are about creating a flat-fielding and sky-subtraction regimen that makes reproducible relative stellar photometry possible. The strategy will be to (a) use diffuse-light flats (domes, twilights, DECal, night-sky) to calibrate small-scale responsivity of pixels, which we hope to be independent of color across a given filter band (except for fringing). (b) Create solutions for large-scale behavior of the stellar photometric response in each filter by solving for internal consistency in a series of dithered exposures of a rich stellar field. (c) Test these solutions by applying the star flat to observations with DES-style full-depth tilings, and measuring consistency of relative magnitudes on distinct exposures.
For Cal-R4, consistency of small-scale flat structure: will take these flats and then measure variance of high-pass-filtered ratios of flats of different kinds:
- Field(s) to observe: Dome flats, twilight flats, DECal flats taken at red and blue ends of each filter
- Number, duration of exposures per filter: Sufficient for creation of mean flats of each type with statistical errors <0.5% per pixel
- Dithering / tiling pattern: Twilights dithered to remove stars in median stacks
- Sky conditions needed (seeing, photometric, moon phase, etc:): no clouds for twilights
- Other timing constraints: Clearly some will be taken during Commissioning but during SV should check at least one other complete set to see if small-scale structure is reproducible in time as well as across different types of flats.
For Cal-R5, G6, checking variability of fringing: will look at variability of structure in night-sky flats that does not go away after application of dome flats and high-pass filters.
- Field(s) to observe: Series of >10 exposures in each filter, any field that is not crowded or containing large objects. Must repeat several times in Y and z
- Number, duration of exposures per filter: Sufficient for creation of median sky flats of each type with statistical errors <0.5% per pixel
- Dithering / tiling pattern: Dithering larger than any objects in the field
- Sky conditions needed (seeing, photometric, moon phase, etc:): not thick clouds. Y and z should be done at varying levels of sky brightness (airglow)
- Other timing constraints: Clearly some will be taken during Commissioning but during SV should repeat z and Y to check fringing variation.
- Field(s) to observe: For solution construction: field with high, but not crowded, stellar density. For validation: any DES field
- Number, duration of exposures per filter: Dither sequence for solutions is TBD (10-20 exposures per filter). For validation: Can use any fields with substantial dithers, preferably something showing full-depth DES tiling pattern.
- Dithering / tiling pattern: TBD for calibration field, standard DES tilings for validation fields
- Sky conditions needed (seeing, photometric, moon phase, etc:): clear (But see below)
- Other timing constraints: A photometric solution will be taken during Commissioning but during SV should repeat to check stability.
For Cal-G5: should take overlapping dithered exposures during non-photometric conditions, (probably can use just about any data being taken on non-phot nights) in order to test what RASICam outputs signify inhomogeneous clouds that will cause relative photometry errors across DECam FOV.
Data processing needs¶
Which of the following products would you need to conduct the test:
- Raw images available at some off-mountain site? Probably, for experimentation with different flattening techniques
- Images with cross-talk, debias, trim, dome-flat correction? Yes
- Images with an "illumination" or "star" flat applied for better relative photometry? Yes, will be examining cosmetics of flattening / sky subtraction
- Single-exposure object catalogs? Yes
- with astrometric solution accurate to: 1 arcsec (for matching)
- and photometric zeropoint accurate to: (N/A - will be deriving photometric solutions here)
- Only need bright stars, or also want galaxy magnitudes? Bright stars only - depend critically on stellar photometry algorithm from images
Telescope or other engineering data¶
External data sets¶
Codes and scripts for analysing the test data¶
- Code for combining dome and sky flat exposures into deep, clean flats
- Code for high-pass filtering of flats and their ratios
- Code for robust quantification of flat variance
- Code for deriving sky flat from internal agreement of dithered exposures
- Must give this dome/sky flat to be used for flattening of SV exposures
- Code for comparing relative photometry on dithered exposures of same field
- Code for subtracting / manipulating fringe frames.
- Need to bin photometry residuals by position on focal plane, look for patterns
Are there real-time decisions to be made during the observations? Maybe to take new exposures for fringe testing when Y/z sky brightness is seen to vary during night
How quickly can / must the data be available and analysed? Images and stellar photometry catalogs available within 24 hrs.
Who will do it, and where? Could be off-site