Agenda And Minutes for 4/28 10am CDT¶
1. Dial Toll-Free Number: 866-740-1260 (U.S. & Canada)
2. International participants dial:
Toll Number: 303-248-0285
Or International Toll-Free Number: http://www.readytalk.com/intl
3. Enter your 7-digit access code, followed by “#”
Your access code: 7691719
- More Testing Completeness With UFIG. And with SNX3 as well!
- Improved Gold 1.0 Depth Maps
- Plans for collaboration meeting
Attending: Eli, Angelo, Gary, Nacho, Aurelio, Chihway, Diego, Steve K, Huan, Robert G, others...
We discussed the collaboration meeting plans, including the sessions on "Value Added Catalog" generation. We decided it would make sense to have a separate session on SVA1 Gold specifically, as well as VAC sessions on how best to handle value-added catalogs in the future. In particular, SVA1 Gold is somewhat of a "replacement" catalog for SVA1, while value-added catalogs in the future will not necessary be such a full replacement.
We discussed the completeness plots (linked to above). First: The science requirement R-4 from Science Requirements states:
The DES wide-area survey will be carried out in 5 optical passbands grizY and reach g = 24.6, r = 24.1, i = 24.3, z = 23.8, and Y=21.5 AB at ≥ 10σ S/N in 1.5” apertures (apertures of this size are appropriate for faint galaxies in 0.9” seeing), with ≤100 mas accuracy positions, over at least 90% of the survey area. Inside this area there should be ≥ 97.5% completeness (excluding area occulted by stars) and 95% purity.
So the numbers and the completeness do refer to a 1.5" aperture magnitude and not a total magnitude, as was surmised by Diego. However, when moving to total rather than aperture magnitude the targets do shift to (brighter) magnitudes. It is an assumption that the completeness in total magnitude will shift in the same way as the completeness in aperture magnitude, but this needs to be explicitly checked.
Diego also pointed out that it would be useful to have a map of completeness in magnitude--surface-brightness space. The current UFIG run cannot produce this, but the next UFIG run will have all the profiles correctly generated and recorded, so these maps can be made from UFIG in the near future.
Diego also had questions about the way that DETMODEL magnitudes are measured, and noted that the documentation is somewhat sparse. The current description on the DESDM wiki is very sparse.
The key is that DETMODEL is not a total magnitude, but rather a flux ratio between the measurement image and the detection image, using the same galaxy model (which is convolved with the local psf). On its own, it doesn't make much sense. However, when you have two bands, each with DETMODEL magnitudes which are flux ratios compared to the same reference, then a (relatively) low-scatter color can be computed by taking the difference of DETMODEL magnitudes (which is the flux ratio between the two bands, since the common reference flux cancels out).
Huan had some comments on the SNX3 coadds that Eli made. In particular, Eli did not do any checks on the ellipticity quality. Now that things are set up at SLAC, Eli will make more test coadds in the SN fields, expanding beyond 90s images, using the best images to add up to ~900 s (full depth) to compare to the deep coadds.